Reducing herbicide use in spring cereal production Jukka Salonen Agricultural Research Centre of Finland Institute of Plant Protection FIN-31600 Jokioinen, Finland Academic dissertation To he presented, with the permission of the Faculty of Agriculture and Forestry of the University of Helsinki, for public criticism in Viikki, Auditorium 82, on December 17th, 1993, at 12 o’clocknoon. https://www.c-info.fi/en/info/?token=cKj97U7GR-Ta7-v1.gzfUz4LAzbTx5WwxT36lIw.MuiJ12S9luD98eN4Z1WLUD7flzN4dRHRlIsc48wVNb9_jUAOWPgPWPvMx3ZLuFmoVozgKb1Ju1s4HETFXAs5D_Sefu80Zu75WevkbU24h1aO6nCFNmEbEzkt_wLh_vQbHMmgnFW6N9CBFCLrHHoa5PVzr4bLoOF_lO3E6uAFUkSeYwDOWgM-kFt71HvZXdM_34REMm7ir2nVDMnJ0i8Z62oJJmMK-YGongHBMh-2l2CEId6L-KDqvTHVyQy9lpERO6Lsqw 3 PREFACE The research projects summarized in my thesis were conducted at the Agricultural Research Centre of Finland (ARC) during 1982-1992. I am most grateful to Dr. Leila-Riitta Erviö (Head of the Institute of Plant Protection), to Professor Emeritus Jaakko Mukula, and to the heads of the regional research stations of the ARC for providing the possibilities and facilities required to undertake the experiments. Dr. Erviö and Professor Mukula are especially acknowledged for being my teachers and supervisors in Weed Science. I am indebted to Professor Emeritus Eero Vans, my teacher in Crop Science, for valuable instructions on my research subject and for encouragement during my post-gradu- ate studies. I express my sincere gratitude to Professor Eija Pehu for her enthusiastic guidance and support during the last stages of the compilation ofmy thesis. I am grateful to Professor Haldor Fykse and to Dr. Jari Peltonen, referees of the thesis, for their valuable advice and constructive criticism on the manuscript. I sincerely thankDr. Alan Courtney, Mr. George Cussans and Mr. Per Kudsk whose relevant comments and suggestions greatly improved the manuscript of the introductory chapter. I also wish to thank my Nordic colleagues Professor Sigurd Håkansson and Professor Jens Streibig for fruitful discussions and guidance in Weed Science and statistics. I give my warm thanks to the technical staff of the Weed Science Section of the ARC. Under the leadership of Mr. Kauko Aunola, Mrs. Eiramaija Tanni and Mr. Jari Poikulainen they collected a huge amount of data. I sincerely thank Mrs. Tarja Aurén-Kamaattu, Mr. Heikki Jalli and Mrs. Inkeri Tähkävuori for being the co-leaders of the field trials. 1 am indebted to my colleague Mrs. Sanni Junnila for sharing my duties at the Institute. I wish to express my thanks to Ms. Elise Ketoja, Dr. Caspar Looman, Mrs.Liisa Mattila and Dr. Jukka Öfversten for guidance in statistics. I thank the staff of the Data and InformationServices of the ARC for efficient co-operation. Especially, pleasant team work with Mrs. Sari Torkko, Co-Editor of this journal, is warmly acknowledged as it speeded up the compilation of my thesis. I am pleased to get my thesis published in this journal. Linguistic revision by Mrs. Sevastiana Ruusamo (original papers) and by Dr. Jonathan Robinson (introductory chapter) is sincerely acknowledged. I am greatful for the financial support to our research projects from the Ministry of Agriculture and Forestry and from the Academy of Finland. I am pleased to acknowledge the special funds from the ARC, and personal scholarships from the Academy ofFinland, the Agricultural Reseach Foundation of Tiura, the Finnish Association of Academic Agronomists, the Kemira Research Foundation and the Research and Science Foundation of Farmos. My dearest thanks I owe to my wifeMerja, M.Sc. in Soil Science, for being a demanding reviewer of my manuscripts and for her cheerful attitude and understanding while I was wrapped in my scientific thoughts. Both my wifeand the KVARKA circle of friends have made a great effort in optimizing my life between work and leisure. Jokioinen,November 1993 Jukka Salonen 4 LIST OF ORIGINAL ARTICLES The thesis is a summary and discussion of the following articles, which in the introductory chapter are referred to by their Roman numerals: I Salonen, J. 1993. Weed infestation and factors affecting weed incidence in spring cereals in Finland - a multivariate approach. Agricultural Science in Finland 2: (in press). II Salonen, J. & Erviö, L.-R. 1988. Efficacy ofchemical weed control in spring cereals in Finland. Weed Research 28: 231-235. 11l Salonen, J. 1992. Efficacy of reduced herbicide doses in spring cereals of different competitive ability. Weed Research 32: 483-491. IV Salonen, J. 1992. Yieldresponses ofspring cereals to reduced herbicide doses. Weed Research 32: 493-499. V Salonen, J. 1993. Performance of reduced herbicide doses in spring cereals. Agricul- tural Science in Finland 2: (in press). Reprints of the original articles 11-IV are published with the kind permission ofBlackwell Scientific Publications. 5 CONTENTS PREFACE 3 LIST OF ORIGINAL ARTICLES 4 ABSTRACT 7 INTRODUCTION 8 I Weed incidence in arable lands 8 1,1 Species composition of weed floras 8 1.2 Changes in weed floras 8 1.3 Seasonal dynamics of annual weeds 10 2 Trends in chemical weed control 10 2.1 Proposals for reducing herbicide use 10 2.2 Strategies for reducing herbicide use 11 2.3 Use of herbicides in Finland 13 2.4 Performance of herbicides 14 3 Objectives of the study 15 MATERIALS AND METHODS 15 1 Weed survey in spring cereals 15 2 Dose reduction of herbicides in field experiments 15 2.1 Herbicides applied 17 2.2 Weed assessment 18 3 Statistical methods 18 3.1 Data transformation 18 3.2 Ordination analysis 18 3.3 Linear models 19 RESULTS AND DISCUSSION 19 1 Weed vegetation in spring cereals 19 1.1 Species composition of weed flora 19 1.2 Weed infestation levels in cereal fields 21 2 Chemical weed control in spring cereals 21 2.1 Use of herbicides 21 2.2 Efficacy of herbicides 22 2.3 Impact of dose reduction on herbicide efficacy 23 2.4 Yield responses to chemical weed control 25 2.5 Basing the herbicide use on crop-weed interactions 26 2.6 Impact of herbicides and crop rotation on weed infestation 28 3 Economic impact and practical implications of herbicide dose reduction 30 SUMMARY AND CONCLUSIONS 32 REFERENCES 34 SELOSTUS 41 7 Reducing herbicide use in spring cereal production Jukka Salonen Salonen, J. 1993. Reducing herbicide use in spring cereal production. Agric. Sci. Finl. 2: Supplement No. 2.42 p. Academic dissertation. (Agric. Res. Centre of Finland, Inst. Plant Protect., FIN-31600 Jokioinen, Finland.) A survey was conducted in southern and central Finland from 1982 to 1984 to determine the main weed species affecting spring cereal production. The weed flora was domin- ated by broad-leaved species. The most common broad-leaved weeds were Chenopo- dium album L., Galeopsis L. spp., Viola arvensis Murr. and Stellaria media (L.) Vili., and the most common grass weed was Elymus repens (L.) Gould. The density of weeds averaged 170 plants rrT 2 (median 124), and the dry weight 320 kg ha’ 1 (median 183). Ordination analyses revealed that the species composition of weed populations varied regionally and was affected by soil characteristics and crop management practices, particularly by long-term use ofherbicides. An additional aspect studied in the survey was herbicide efficacy in farmers’ fields. Phenoxy acid herbicides, MCPA, dichlorprop and mecoprop, were the most common active ingredients used in the 252 spring cereal fields surveyed. MCPA alone gave only a moderate control of 65%, determined as a reduction of weed biomass, whereas the efficacy of herbicide mixtures containing MCPA averaged 83%. Inadequate control was in most cases due to a wrong choice of active ingredient for the prevailing weed population. Reduction in the use of herbicides by applying lower doses than recommended was studied in field experiments. Herbicide formulations of MCPA/dichlorprop, MCPA/ mecoprop and MCPA/fluroxypyr were screened in spring barley (Hordeum vulgäre L.) and spring wheat (Triticum aestivum L.) fields. The efficacy of herbicides,applied at the lowest recommended dose, averaged 85%. At a 30% lower dose the efficacy still reached 79%. Even lower herbicide doses were often adequate, depending on the herbicide, weed species and the crop. The production of weed biomass was adequately suppressed with reduced doses since the most common and aggressive species, such as Chenopodium album and Galeopsis spp., were efficiently controlled with low doses. Use of reduced herbicide doses for three years in the same field caused neither an increase in the subsequent weed infestation nor changes in the species composition of weedpopulations compared with the treatments at recommended rates of application. The percentage emergence of weeds averaged 70-75% at the time of herbicide application when the crop was at the 3-4 leaf growth stage. However, spraying during the early growth stages of those weeds that emerge in the main flush is recommended since the competitive ability of the crop is normally sufficient to suppress the growth of late-emerging weed seedlings. At harvest the proportion of weed biomass in unsprayed plots, as a proportion of the total vegetative biomass, averaged 3.1% inbarley fields and 3.6% in wheat fields. The growth of weeds was more efficiently suppressed with reduced herbicide doses than by increasing the seeding rate of the crop. The mean yield gain remained below 5% at all rates of herbicide application. No reliable density-based threshold for chemical weed control was established. Instead, site-specific dose adjustment based on the composition and infestation level of the prevailing weed populations is suggested to reduce the total use of herbicides and to maintain the current low levels of weed infestation. Consequently, annual reductions of 30% in use of cereal herbicides are expected. Such a reduction corresponds to monetary savings of approximately FIM 20 million per annum at the national level. Key words: spring barley, spring wheat, broad-leaved weeds, weed survey, efficacy of herbicides, reduced doses of herbicides, yield response, CCA, CANOCO Agric. Sei. Fint. Suppl. No. 2 (1993) INTRODUCTION Optimization of crop protection measures requires adequate information on pests and weeds. Identi- fication of the principal weed species and aware- ness of their population dynamics and impact on crop production are key factors for successful weed control. Furthermore, adjustment of control meas- ures according to the prevailing weed infestation is a prerequisite for economic and sustainable crop production. The optimization of herbicide use can be defined as a reduction in the level of active ingredient used to the minimum necessary to meet a defined need (CUSSANS 1992). Appraisal of the rational use of herbicides in spring cereal produc- tion in Finland is given in this study. 1 Weed incidence in arable lands 1.1 Species composition of weed floras Worldwide, about 200 species of the 250,000plant species are classified as important weeds (Holm et al. 1977). A comprehensive review of the factors influencing the distribution of weeds in Europe is given by Holzner and Immonen (1982). They reported that the most significant alterations in weed communities have taken place since 1950. Haas and Streibig (1982) gave a detailed descrip- tion of changing patterns of weed distribution in Denmark. They concluded that although herbicides evidently have been one of the major driving forces in changing the species composition and infestation level of weed populations, several other factors included in crop production, such as crop rotation, fertilization etc., have also affected weed floras. Several studies on weed floras of arable lands in the Nordic countries have been published (Table 1). Considerable similarities in the species composi- tion of weed floras in the different countries is evident from these studies. About 50 weed species are common and of economic importance in the Nordic countries. The most frequently occurring broad-leaved weeds are Chenopodium album L., Polygonum L. spp., Stellaria media (L.) VILL., Viola arvensis MURR., and the most important grass weed is Elymus repens (L.) GOULD. Erviö and Salonen (1987) compared the weed populations of the 1960 s and the 1980 s in spring cereal fields inFinland. They found a slight decline in the frequency of C. album, GaleopsisL. spp., S. media. Erysimum cheiranthoides L., Myosotis L. spp., Spergula arvensis L. and Tripleurospermum inodorum SHULTZ BIP.. Weed species that were found more frequently in the 1980 s were e.g. V. arvensis, Fallopia convolvulus (L.) A.LÖVE, Lap- sana communis L., Polygonum aviculare L., Pu- mmia officinalis L., Galium L. spp., Lamium L. spp. and Matricaria matricarioides (LESS.) PORTER. Among the 15 most frequently occurring weed species, ten were tolerant of MCPA, in com- parison with nine in the 1960 s (Mukula et al. 1969). In general, manipulation of the environment for agricultural purposes has favoured species that can adapt to the disturbed habitats of cultivated fields, whereas sensitive species have become extinct (Young and Evans 1976, Eggers 1984). The im- pact of intensive agricultural practices on weed floras has reached the stage where even the conser- vation of endangered weed species has been sug- gested (Eggers 1987, Wilson et al. 1990, Mahn 1992). 1.2 Changes in weed floras The occurrence of weeds and the changes in weed floras are often related to crop management (e.g. Bachthaler 1969, Rademacher et al. 1970, Cussans et al. 1979. Haas and Streibig 1982, Froud-Williams et al. 1983, Post 1986). Crop management in Finland, as in Europe generally, has been intensified enormously in recent decades re- sulting in well-established crop stands and a trend of increasing yields per unit area (Mukula and Rantanen 1987). The use of herbicides and inor- ganic fertilizers have been adopted in modern cer- eal production, and crop rotations have changed 8 Agric. Sei. Finl. Sappi. No. 2 (1993) Table 1. Studies on weed floras of arable lands in the Nordic countries. Country Study 1 Crop Most common weed species 2 Reference Finland I Cereals GAESS CHEAL RAPRA SPEAR AGRRE Hilu 1948 S Grassland DECCA TAROF RUMSS AGRRE CHYLE Paatela 1953 S Spring cereals GAESS CHEAL SPEAR STEME VIOAR Mukula et al. 1969 S Grassland RANRE ACHMI RUMSS TAROF DECCA Raatikainen and Raatikainen 1975 S Winter cereals VIOAR CHEAL GAESS MATSS ERYCH Raatikainen et al, 1978 S Spring cereals CHEAL GAESS VIOAR STEME POLCO Erviö and Salonen 1987 Denmark S Winter rye STEME POAAN CHEAL VIOAR MYOAR Petersen 1943 S Cereals STEME VIOAR POLCO POAAN PLAMA Mikkelsen and Laursen 1966 S Spring barley STEME VIOAR POAAN CHEAL MYOAR Andreasen et al. 1989 S Grassland POAAN STEME TARSS CAPBP VERSS Andreasen 1990 Norway R Arable land General review on the main species Korsmo 1925 F Spring cereals CHEAT STEME GAESS VIOAR TAMPU Fykse 1993 (pers. commun.) F Grassland RUM TO RANAC TAROF RANRE RUMAC Fykse 1993 (pers. commun.)RUM TO RANAC TAROF RANRE RUMAC Fykse 1993 (pers. commun.) Sweden S Spring cereals CHEAT GAESS S Winter cereals MATMA PAPSS F Spring cereals CHEAT GAESS F Winter cereals MATMA STEME F Spring cereals CHEAT STEME F Winter cereals STEME VIOAR SPEAR STEME POTTA Granström and Almgård 1955 CENCY STEME TAMSS Granström and Almgård 1955 STEME POTSS SPEAR Gummesson 1975 VIOAR VERSS GAESS Gummesson 1975 GAESS VIOAR MYOAR Hallgren 1993a MATMA MYOAR VERSS Hallgren 1993a Type of study: F = Field trials, I = Inquiry, R = Review, S = Survey 2 Codes according to Bayer (1992): ACHMI = Achillea millefolium, AGRRE = Elymus repens, CAPBP = Capsella bursa- pasloris, CENCY = Centaurea cyanus, CHEAT = Chenopodium album, CHYTE = Chrysanthemum leucanlhemum, DECCA = Deschampsia caespitosa, ERYCH = Erysimum cheiranthoides, GAESS = Galeopsis spp., GAESS = Galium spp., TAMPU = Lamium purpureum, TAMSS = Lamium spp., MATMA = Matricaria matricarioides, MATSS = Matricaria spp., MYOAR = Myosotis arvensis, PAPSS = Papaver spp., PTAMA = Plantago major, POTCO = Fallopio convolvulus, POTTA = Polygonum lapathifolium, POTSS = Polygonum spp., POAAN = Poa annua, RANAC = Ranunculus acris, RANRE = Ranunculusrepens, RAPRA = Raphanus raphanistrum, RUMAC = Rumex acetosa, RUMTO = Rumex longifolius, RUMSS = Rumex spp., SPEAR = Spergula arvensis, STEME = Stellaria media, TAROF = Taraxacum vulgare, TARSS = Taraxacum spp., VERSS = Veronica spp., VIOAR = Viola arvensis considerably over time, often tending towards monoculture. These changes have caused both quantitative and qualitative changes in weed popu- lations (Fryer and Chancettor 1970, Rade- macher and Koch 1972, Reuss 1981, Mahn 1984). The most apparent change in weed populations in recent decades has been the decline in weed abundance (No. m ) in cereal fields (Aamisepp and Wattgren 1979, Erviö and Satonen 1987, Hattgren 1993a). Erviö and Satonen (1987) reported that the density and biomass production of weeds in Finnish spring cereal fields have de- creased to about one-third of the values recorded during the 19605. Furthermore, the number of weed species has decreased in intensive cropping systems (Fogetfors 1979, Cattauch 1981, Atbrecht and Bachthater 1988, Debaeke 1990, SPE- ranza et al. 1990). Changes in weed vegetation are not only re- stricted to alterations in species composition and their proportional abundance (interspecific), but also to changes within the population ofone species (intraspecific). A good example of the intraspecific variability ofweeds is resistance ofa weed popula- tion to a herbicide as a result of its continuous 9 Agric. Sei. Fint. Suppl. No. 2 (1993) application promoting selection in the weed popu- lation (Leßaron and Gressel 1982). Generally, a number of selective factors including light regime, soil type, biotic factors and agricultural practices, result in genetic differentiation in weed populations (Warwick 1991). 1.3 Seasonal dynamics ofannual weeds Variation in the emergence of weeds between sea- sons and within one growing season are charac- teristic to weed populations (Roberts and Potter 1980, Erviö 1981, Håkansson 1983a, An- dreasen 1990). The timing of sowing and growth period of crops are important factors determining the composition and abundance of weed popula- tions (Streibig and Haas 1979, Chancellor 1985). Annual weed species can roughly be divided into three categories according to their germination patterns (Håkansson 1992): A. Summer annuals with a germination peak in spring, decreasing towards the end of the grow- ing season. Typical species: Chenopodium album, Fumaria officinalis, Galeopsis spp., Polygonum aviculare. B. Facultative winter annuals with extensive ger- mination both in the spring and, after soil till- age, in late summer to early autumn. Typical species: Lamium spp., Myosotis arvensis, Stel- laria media, Tripleurospermum inodorum, Viola arvensis. C. Other species. Summer annuals or mainly sum- mer annual behaviour, but with extensive ger- mination both in the spring, after soil tillage, and also later in the growing season. Typical species: Brassica L. spp. (also cultivated forms), Spergula arvensis. The time of weed emergence affects the success of chemical control, as the most common post- emergence herbicides used in cereal production are foliar-active, with minor effect on the weed seed- lings emerging after herbicide application. 2 Trends in chemical weed control 2.1 Proposals for reducing herbicide use The principles of weed control have been compre- hensively documented (e.g. Hance and Holly 1990). It has been frequently shown that the judi- cious use of herbicides is characteristic of success- ful and economic crop production (Zeddies 1986, Beyer 1991). Herbicides are commonly used in cereal production, although yield benefits from chemical weed control are sometimes questionable (Gerowitt et al. 1984. Jensen 1985,Davies et al. 1989,Erviö et al. 1991). However, increase in cereal yield is not the only argument favouring the use of herbicides. It has been shown that chemical weed control also i) pre- vents weeds from interfering with cultivation, har- vesting and marketing (Elliott 1978), ii) reduces the number of host plants of pathogens and pests (Heitefuss 1986) and Hi) reduces the reservoir of weed seeds in the soil (Hurle 1974, Kees 1986, Fykse 1991a). In contrast to the negative aspects of weed interference, Heitefuss (1986) has reviewed also the benefits of weeds, including a positive effect on soil structure and soil humus, and more- over on the incidence of a beneficial fauna. Although several alternatives to herbicides, in- cluding biological, mechanicaland physical control are available (Edwards and Regnier 1989, Mor- gan 1992, Watson 1992), herbicides will prob- ably maintain their major role in weed control. However, it has been widely recognised that weed control strategies in the future are likely to require an integration of non-chemical techniques with more efficient, but restricted, use of herbicides (Combellack 1992a, Cussans 1992). In the 1980s, the Nordic countries introduced political Action Plans, which stipulate considerable reductions in the use ofpesticides (Thonke 1991). A common aim is to reduce the amount of applied active ingredients (a.i.) of pesticides by 50% of the average amount used in the early 1980s. This policy has been adopted in Denmark (Thonke 1991) and in Sweden (Bernson 1988), as well as in the Netherlands (Ministry of Agriculture 1990). In 10 Agric. Sei. Fin!. Suppl. No. 2 (1993) Norway the aim is to reduce pesticide use "as much as reasonable" (STUBSJOEN 1991). A recently published committee proposal (Ym- päristöministeriö 1992) quantified the political ex- pectations on the reduction of pesticide use in Fin- land. The target is to halve the average use of 2,000 tons a.i. (1987-1991) before 1995. Strategies pro- posed and the estimated reduction with each ap- proach were: 1) changes in crop production and land usage (15- 20%) 2) use of pesticides according to the defined need (10-15%) 3) tests and repairs of spraying equipment (10- 15%) 4) alternative control methods (15%) A co-operative effort on behalf of those involved in crop production practices, the extension services and research has to be made to meet the above- mentioned political targets. Sales statistics indicate that a desirable trend has already started (Fig. 1). Reduction in herbicide use can partly be explained by the changes in agriculture, including increased area of fallows, a more rational use of herbicides, and a shift to low-dose herbicideproducts. Fig. I. Sales of herbicides in the Nordic countries. The aver- age amount of active ingredients (tons a.i.) in 1981-1985 and the sales in 1986-1992 in Denmark (�), Sweden (A), Finland (•) and Norway (■), Data compiled from Thonke (1991), from the National Board of Agriculture in Finland, from the Kemikalieinspektionen in Sweden and from Denmark (Flak- kebjerg) and to Norway (Statens Plantevem) via personal communications . Possibilities for reducing the recommended doses of phenoxy acid herbicides by 25%, without a considerable loss in efficacy, were reported al- ready twenty years ago (Hornig 1972). Within the range of 25-50% reduction of herbicide dose, a similar trend was observed also in field experi- ments carried out in Denmark in the 1970s (Thonke 1978). Consequently, intensive research efforts in the Nordic countries were launched to optimize the use of herbicides in crop production, particularly in cereals (Aamisepp 1984, Anders- son 1984, 1986, Erviö and Hiivola 1986, Thonke 1986, Kudsk 1989,Fogelfors 1990,Lo- MAKKA 1990). 2.2 Strategies for reducing herbicide use Appraisal of the need for reduced weed control is often based on the apparent decline in weed infesta- tion levels. However, weeds vary considerably in distribution in place and time (Marshall 1988, Wilson and Brain 1990). Complexities resulting from spatial heterogeneity and the multi-species nature of weed communities make modelling of the crop-weed interactions difficult and hamper de- cisions on chemical weed control (Auld and TiS- dell 1988, Van Groenendael 1988, Kropff 1988, Thornton et al. 1990). Routine use of herbicides has been a common approach to overcome the problems ofdetermining the necessity for chemical control, often to such extent that cost-benefit considerations have been forgotten. Two different approaches, i) prophylac- tic and ii) threshold strategies, have been studied in an attempt to change the present control practices (Table 2). Gummesson and Fogelfors (1990) suggested that the annual use of herbicides should be de- creased by applying reduced herbicide doses and not by reducing the treated land area. On the other hand, the threshold approach, whether to spray or not, has been widely studied and applied, particu- larly in Germany, mainly in winter cereals (Gar- burg 1974, Heitefuss et al. 1987, Gerowitt and Heitefuss 1990, Wahmhoff 1990), but also in 11 Agric. Sei. Finl. Suppl. No. 2 (1993) Table 2. Two diverse strategies to reduce herbicide use in cereal production. Appraisals of a) threshold weed infestations to withhold herbicide application and b) the levels of reduction of the recommended herbicide doses. a) Country Crop Threshold Reference (weeds nr 2, % cover) Germany Spring barley 87 Garburg 1974 Spring wheat 50 Winter wheat 26 Winter cereals 20-30 (monocots) Gerowitt and Heitefuss 1990 40-50 (dicots) Denmark Spring barley 80-100 Streibig 1983 Spring barley 20-50 Jensen 1987 Norway Spring barley 175 Fykse 1991 b U.K. Spring barley 150 Courtney and Johnston 1986 Finland Spring cereals 52-101 Erviö et al. 1991 Germany Winter wheat 6.7-9.7% Beer 1979 Winter barley 6.1-18.7% Austria Cereals 3.2-9.3% Neururer 1976 b) Country Crop Reduction of Reference herbicide dose, % Germany Spring barley 25-50 Hornig 1972 Denmark Spring barley 25-50 Pedersen 1978 Sweden Spring cereals 33-67 Engström 1978 Spring cereals 33-67 Aamisepp 1984 Spring barley 33-67 Lomakka 1990 Cereals 50-75 Fogelfors 1990 Finland Spring cereals 50 Erviö and Hiivola 1986 U.K. Cereals 50-87.5 Davies et al. 1989 spring cereals (e.g. Fykse 1991a, Davies et. al 1993). Gerowittand Heitefuss (1990) used the fixed threshold values of 20-30 plants m for grass weeds, and 40-50 plants m for broad-leaved weeds, in winter cereals. In addition, some specific weed species such as Galium aparine L. (0.1-0.5 plants m‘ ) and Fallopia convolvulus (2.0 plants m‘ ) were considered very harmful and a substan- tially lower threshold was suggested. Cussans (1980) ranked the population densities of weeds on a logarithmic scale (Table 3) for strategic planning of weed control measures. In general, the influence of crop type and differencesbetween weed species are emphasized in the threshold approach. Thresh- old values can be based either on biological or economical considerations, as discussed by Cus- SANS et al. (1986). A more sophisticated approach, to define the effect of weed competition on yield loss, was intro- duced by Wilson (1986): different weed species were given a Crop Equivalent (CE) value based on 12 Agric. Sd. Finl. Sappi. No. 2 (1993) Table 3. Some definitions of weed densities by Cussans (1980). Population Short Notes - with special (weeds m2) description reference to cereals > 100 Very severe Certain to cause yield loss 10-100 Severe Yield loss usually greater than the cost of spraying 1-10 Serious Some competition inter- ference probable 0.1-1 Moderate Not competitive in many crops but an obvious latent threat 0.01-0.1 Light Very unlikely to have a measurable effect on yield 0.001-0.01 Very light No effect on yield or quality 0.0001-0.001 Economically Very easily rogued by (1-10 ha') unimportant hand dry weight per weed divided by dry weight per crop plant, assessed from extensive field data. A thresh- old value of 5 CEs was suggested as a Spray Decision threshold. Davies et al. (1993) compared this approach to routine use of halfdose application of herbicides and found that the threshold approach applied over some years was insufficient to keep the weed infestationat theoriginal level. Moreover, the cost savings from using the threshold option were partly absorbed by the costs ofassessing weed infestation before the decision-making. Erviö and Hiivola (1986) compared the thresh- old and prophylactic strategies in spring cereals in Finland. They found no differences in the sub- sequent weed populations whether a continuous or threshold application of herbicide was employed during the five-year study. 2.3 Use of herbicides in Finland The era ofchemical weed control started inFinland in the early 19605, when the area of cereal fields treated with MCPA reached 30% and increased rapidly (Mukula and Ruuttunen 1969). Since then MCPA has been the most common herbicide used in cereal production. In the 1960 s and 19705, MCPA alone made up over two-thirds of the total amount of all herbicides used in agriculture (MARK- KULA et al. 1990). At present, MCPA is most often used in formu- lated herbicide mixtures with dichlorprop or me- coprop (Fig. 2). Furthermore, introductionof novel low-dose herbicides such as sulfonylureas has re- sulted in decreased use of MCPA and other phe- noxy acids. In the mid 1980s, when we started the field experiments, the quantity of sulfonylureas sold in Finland was sufficient to treat approxim- ately 10% of the cultivated cereal area, but in- creased to 26% in 1992 (JUNNILA 1993). Thus, the phenoxy acid herbicides have still retained their major role in weed control in cereals in Finland. It is desirable, however, that herbicides with different mechanisms of action are available and used to avoid the selection of weed populations which may eventually become herbicide resistant (Gressel and Segel 1982). Sales statistics ofpesticides have been available since 1953; during the years 1953-1987 a total of 37,281 tons active ingredients (a.i.) of herbicides, accounting for 83% of pesticides, were applied in agricultural fields (Markkula et al. 1990). Herbi- cides are still the largest group ofpesticides used in Finland. During 1990-1992 herbicides represented 76% of the total volume of active ingredients ap- plied and 65% of the monetary value of pesticides overall (data compiled from Hynninenand Blom- qvist 1991, 1992, 1993). Herbicide sales peaked in 1980 with 2,099 tons ofactive ingredients sold that year (TIITTANEN and Blomqvist 1981). Annual use of agricultural herbicides during 1990-1992 averaged 1,320 tons a.i. per annum (Hynninen and Blomqvist 1991, 1992, 1993) representing a slightly declining trend in annual use (Fig. 2). The amount of herbicides sold in 1990-1992 was sufficient to treat 69-75% of the area under cereal production during each year. In recent years, overall, the use ofherbicides has decreased mainly due to agricultural policy (e.g. set-aside fields) aiming at reducing overproduction of cereals. In addition to the indirect decline in herbicide use, a real decline in the use of phenoxy 13 Agric. Sd. Finl. Suppl. No. 2 (1993) acids, in quantity of a.i., is expected following ap- proval in Finland in 1992 of the new isomer formu- lations of dichlorprop and mecoprop. A shift from racemic formulations to these optically active iso- mers would correspond approximately to a 50% decrease in the use of active ingredients con- cerned. 2.4 Performance of herbicides Herbicide doses shouldbe adjusted to a level that is sufficent to control a range of target weeds without damaging the crop plant. The effect of herbicides against different weed species varies and results in selective control of weed populations. Therefore, herbicide mixtures are both manufactured and made by farmers for broad-spectrum weed control. Phytotoxic effects of herbicides are normally as- sessed on a quantitative scale as a response of plant number, biomass, height etc. to the applied chem- ical, and often described relative to an untreated control. Analysis of variance is commonly applied to test treatment effects. However, the use of a dose-response curve, describing the whole dose range, from the no effect level to complete kill at high doses, is recommended for thorough herbicide bioassay (Streibig 1992). An S-shaped logistic curve fitted with non-linear regression analysis is considered appropriate to describe herbicide effi- cacy and to compare the relative potency of differ- ent herbicides (Streibig 1988). Herbicide performance is affected by environ- mental conditions before, at, and after herbicide application (Kudsk and Kristensen 1992). De- spite numerous reports on herbicide-environment interactions, the mechanisms by which herbicide activity is influenced by different environmental factors is poorly understood (DEVINE 1988). Par- ticularproblems arise in extrapolating results from controlled conditions to the field. Consequently, inconsistent weed control with herbicides is a con- tinuing problem. Agrochemical companies, which are primarily responsible for the herbicide efficacy, attempt to minimize the risk of control failures by recom- mending application rates that are expected to be sufficient even under unfavourable conditions. However, from the farmers’ economical point of view it is important to determine the potential for regulating the dose of herbicides according to the actual requirement. Fig. 2. Sales of pesticides in Finland since 1981. Data compiled from the statistics released annually by the National Board of Agriculture. 14 Agric. Sei. Fin!. Suppl. No. 2 (1993) 3 Objectives of the study This study comprised investigations into the spe- cies composition of weed flora, the level of weed infestation and the efficacy of herbicides in spring cereal fields in Finland (I, II). Secondly, possibili- ties to reduce herbicide input by applying reduced herbicide doses were studied in field experiments (111, IV, V). The objective was to establish new recommendations for the use of herbicides in spring cereal production, taking into consideration the control efficacy, yield response and the impact on subsequent weed infestation levels. The specific objectives were: (1) to identify the most important weed species affecting spring cereal production in Finland, (2) to validate the applicability of ordination ana- lyses in describing the species composition of weed flora and in relating the occurrence of weeds to environmental factors (3) to assess the efficacy of herbicides in farmers’ fields (4) to determine sufficient herbicide doses for ef- fective weed control in spring cereal production (5) to investigate the consequences of dose reduc- tion of herbicides on efficacy, yield response and subsequent weed infestation. MATERIALS AND METHODS This study comprises three research projects con- ducted at the Agricultural Research Centre ofFin- land (ARC) during 1982-1992 (Fig. 3). PROJECT 1 was a national weed survey conducted in spring cereal fields in southern and central Finland in 1982-1984 (1, II). Based on the results of the weed survey, PROJECTS 2 and 3 were designed to study the possibilities of reducing the recommended doses of commonly applied phenoxy acid herbi- cides. 1 Weed survey in spring cereals (I, II) The occurrence ofweeds, the efficacy of herbicides and the economic returns from chemical weed con- trol were investigated in 252 farmers’ fields in southern and centralFinland during 1982-1984 (see also Erviö and Salonen 1987). These fields com- prised 155 fields studied in the 1960 s by MUKULA et al. (1969) and 97 new fields. Results on the economics of weed control in the farm fields are reported elsewhere (Erviö et al. 1991), as well as the comparison of the weed incidence in the 155 fields studied both in the 1960 s and 1980 s (Sa- lonen and Erviö 1988). Basic facts about crop production in Finland are reviewed by Mukula and Rantanen (1987). The study was restricted in advance to the 35 weed species (35 weed taxa) regarded as being the most important species according to the results ofa previous survey (Mukula et al. 1969),and accord- ing to information obtained from field experiments and the extension service. Two ofthe selected weed species, Senecio vulgaris L. and Solarium nigrum L., were absent from the 252 fields included in the analysis (I) thus resulting in a data set for 33 weed species. Weed infestation was assessed in unsprayed and sprayed sample plots at the end of July, about one month after the application ofherbicides. Informa- tion concerning crop management, soil charac- teristics and climatical conditions in the fields was either estimated, measured or obtained from the farmer (I). 2 Dose reduction of herbicides in field experiments (111, IV, V) Field experiments with spring barley (Hordeum vulgäre L.) and spring wheat (Triticum aestivum 15 Agric. Sei. Finl. Suppl. No. 2 (1993) L.) were performed at the Agricultural Research Centre ofFinland (ARC) in Jokioinen and at Ylis- taro research station. Additionally, five other re- search stations of the ARC (Anjalankoski, Kokemäki, Mietoinen, Mouhijärvi and Pälkäne) in southern and central Finland were included in PROJECT 3 (V) (Fig. 4). Crop rotations in PRO- JECT 2 were barley-wheat-barley and wheat-bar- ley-wheat, whereas barley or wheat were grown continuously for three years in PROJECT 3. The variety of spring barley was ‘Arra’ in PROJECT 2 and ‘Pohto’ in PROJECT 3. The varieties of spring wheat were ‘Tapio’ and ‘Luja’, respectively. Field trials were laid out as randomized com- plete-block designs with four replicates. A split- plot arrangement of 140 plots per crop, with five Fig. 3. A flow diagram of the research projects summarized in this study. A subject of the project (□) and the main outcome (0). Fig. 4. Location ofthe regional research stations used for field experiments (111, IV, V). EPO = Ylistaro, HÄM = Pälkäne, KYM =Anjalankoski, LOU = Mietoinen, RKA = Jokioinen, SAH = Mouhijärvi, SAT = Kokemäki. 16 Agric. Sd. Finl. Suppl. No. 2 (1993) Table 4. Summary of the herbicides screened in this study (111, IV, V). The rates of applicationrecommended for spring cereals on the product label and the rates applied in the field experiments. Study Herbicide dose, g a.i. ha 1 Active ingredients (a.i.) Trade name recommended applied Reseach project 2 MCPA/fluroxypyr Starane M 400/100-600/150 200/50 300/75 600/150 MCPA/mecoprop Herbotal Plus 500/1000-800/1600 260/520 400/800 800/1600 Research project 3 MCPA/fluroxypyr Starane M 400/100-600/150 280/70 400/100 MCPA/dichlorprop-P Duplosan DP-M 451/485-716/770 331/356 464/499 MCPA/mecoprop-P Duplosan KV-M 459/519-729/824 338/381 473/534 Tribenuron-methyl Express 75 DF 6.0 9.75 5.3 crop seeding rates as the main plots, and herbicide treatments as the subplots, was used in PROJECT 2 (111, IV), and a randomized block arrangement of 24 plots in PROJECT 3. The plot size ranged from 30 to 48 nr (3-4 m x 10-12 m) and the harvested area from 10 to 30 m . The fields were ploughed every autumn to a depth of 20-25 cm. The plots were cultivated in spring with tine harrow lengthwise of the plot to avoid the weed seed movement between the plots. 2.1 Herbicides applied In the fields included in the weed survey (II) all decisions concerning weed control were made by the farmers, and herbicides were applied with their own tractor-mounted sprayers. Half of the observa- tion areas (1.8 m x 2.4 m) were covered with plastic film during herbicide application. In the field experiments (111, IV, V), commonly used herbicide formulations of MCPA/dichlorprop and MCPA/mecoprop, a new candidate MCPA/ fluroxypyr and a reference herbicide tribenuron- methyl, were screened. Herbicides and their appli- cation rates (Table 4) were determined in advance, and not in relation to the prevailing weed infesta- tion. Commercial herbicide mixtures were applied with portable van der Weij propane sprayers fitted with flat fan nozzles delivering 200 1 ha’ 1 spray solution. Herbicides were applied at the 3- to4-leaf stage of the crop (Zadoks’ scale 13-15 (Zadoks et al. 1974)). In PROJECT 2 (111, IV), the efficacy of the high- est recommended dose, and half and one-third of it were compared (Table 4). Results from PROJECT 2 (111, IV) contributed to the initiation of research PROJECT 3, in which the performance of the low- est recommended dose and a 30% lower dose (Table 4), were evaluated in various fields and under various climatic conditions (V). Moreover, new formulations of phenoxypropionic herbicides, containing only the active isomers of dichlorprop and mecoprop, were introduced at that time (SQUIRES et al. 1987), and they were screened in PROJECT 3. 17 Agric. Sei. Finl. Suppl. No. 2 (1993) 2.2 Weed assessment The emergence of weeds relative to growth stages of the crop was followed in research project 3 (V). The dates when the crop reached 1-, 2- and 3- leaf stages (stage 11, 12 and 13 on the Zadoks’ scale) were recorded and the emergence time of weeds was related to the growth stages of the crop. In all studies, the number and the above-ground biomass ofweeds was assessed about 4 weeks after herbicide application. Additionally, in field experi- ments (111, IV, V), weed infestation was assessed 0-1 day before spraying, with some exceptions of 2 to 4 days delay, and also at harvest. Weed samples were collected from sample plots of 0.25 m . Samples were taken to the laboratory, where the number and air-dry weight of weeds was recorded by species. The long-term effect of continuous use of re- duced herbicide doses was investigated by follow- ing the emergence of weeds from soil samples in a greenhouse (III) and by counting the number of weeds in the trial plots one year after the three-year trial period (V). Ten samples, 400 cm 3 each, from the top 0-20 cm soil layer were taken for green- house tests from the plots sown at the recom- mended crop seeding rate. The soil samples were kept in pots for two growing seasons, and the num- ber ofemerging weed seedlings was recorded. Scientific names of weed species (weed taxa) are according to the systematics used by Hämet-Ahti et al. (1984). BAYER codes for weeds (BAYER 1992) were used in the introductory chapter and in Chapter I. In the other original articles weed codes were adapted from scientific names but not accord- ing to the BAYER standard. 3 Statistical methods 3.1 Data transformation To define the structure ofweed data, the descriptive statistical methods of the UNIVARIATE procedure in the SAS statistical programme package (SAS Institute Inc. 1985) were used. In particular, the need and effect of data transformation were as- sessed to give homogeneity of variances and nor- mal distribution. Consequently, weed density data were square root transformed and weed biomass data logarithm transformed. The efficacy values (% scale) were transformed with arcsine (Vy~). 3.2 Ordination analysis The weed survey data (I) were subjected to ordina- tion analyses (Gauch 1982, JONGMAN et al. 1987) with the CANOCO program package (Ter Braak 1987a). In recent years great attention has been paid to the application of constrained ordination tech- niques (Birks and Austin 1992). Constrained or- dination incorporates the features of indirect or- dination methods and regression analysis to relate the species data to explanatory environmental vari- ables. Ordination analyses were used to get a com- munity level description of weed flora in Finnish spring cereal fields which was not possible with the regression techniques (Erviö and Salonen 1987). The development of Canonical Correspondence Analysis (CCA) by Ter Braak (1986) and its im- plementation in the computer program CANOCO encourages the application of ordination analyses, also in weed science. CCA is an extension of the eigenvector technique termed Reciprocal Averag- ing or Correspondence Analysis (CA) (Hill 1973, Hill and Gauch 1980). The CA procedure ordi- nates the species data only, whereas in CCA the ordination axes are constrained to linear combina- tions of environmental variables introduced into a simultaneous analysis with species data. The re- sponse model fitted by CA and CCA for the species is a unimodal bell-shaped Gaussian curve (Ter Braak 1987b): E (yik) = ckexp [-1/2 (xi-uk) 2 / tk2 ] where E(yik) denotes the expected value of species k at site i, xi, the value of environmental variable xat site i, ck , the maximum value of the curve for species k, uk , the optimum value of species k (value of x), tk , the tolerance of species k (curve breadth). 18 Agric. Sd. Fin!. Suppl. No. 2 (1993) A data set for 33 weed species and 12 environ- mental variables collected from 252 spring cereal fields was analysed with the CANOCO program (I). The significance of species-environment rela- tionships was tested with a Monte Carlo permuta- tion test. Ordination results were presented as spe- cies-environment biplots drawn with the CANO- DRAW program (Smilauer 1990). 3.3 Linear models Results from field trials were analyzed either (/) according to a standard multiple linear regression model or (ii) by basing the analyses on appropriate mixed models. Fixed effects of herbicide treat- ments were analyzed using crop densities as covari- ates in PROJECT 2. A different approach was used in PROJECT 3 (V), in which data with repeated measurements from the trial plots over three years, were analyzed in a fashion of split-plot experiment by using the following mixed model yijki = p + «i + cpi(i) + Pj + aPij + Pcpji(i) + yk + ay,k + yq)kl(i) + PYjk + «Pyijk + eijkl where. yijki is the observed response value, p is the overall mean of sample population, oti is the fixed effect of the trial site, cpi(i) is the random effect of the block nested to a site, pj is the fixed effect treatment, 7k is the fixed effect of the year, apy, oryik, Pyjk, aPYijk are the fixed interaction terms, Ptpjl(i), ytpkl(i) are the random interaction terms, Eijkl is the error term. Linear models were analyzed with the General Linear Models (GLM) procedure of the SAS statist- ical package (SAS Institute Inc. 1990). Mean sep- aration was done with Tukey’s HSD test, and pre- planned comparisons of particular treatment effects were tested with contrasts (Littell et al. 1991). RESULTS AND DISCUSSION 1 Weed vegetation in spring cereals 1.1 Species composition of weed flora Weed flora in spring cereal fields surveyed in 1982- 1984 was dominated by relatively few species (I). The number of abundant weeds per field averaged 7 species (range 2-13) of the 33 species studied. Chenopodium album and Galeopsis spp. were the most frequently occurring species (I, Erviö and Salonen 1987). These species, among others, are well adapted to the prevailing conditions in culti- vated fields. They have succeeded in maintaining viable seeds in the soil seed bank, despite the fre- quent use of herbicides. The frequency of the four most common species (>80%) was considerably higher in the Finnish fields than in cereal fields in Denmark, where the frequency of all of the species remained below 60% (Andreasen 1990). Appar- ently this is due to differences in growth conditions (soil, climate) and in crop rotation practices, but also due to a different definition of the frequency values in these two studies. All germination groups of annual weeds (see p. 10) were represented in spring cereal fields studied (I). The most common grass weed was Elymus repens. It was even more frequent than in the 1960s (Erviö and Salonen 1987). Elymus repens had, on average, the highest biomass production per unit area. Poa annua L. was the second most frequent grass weed. Avena fatuaL. is oflocal importancein coastal regions, although it was rare in the fields surveyed. Five regions of the survey area were located in southwestern Finland, three in eastern Finland and two in the western part of central Finland. Conse- quently, the frequency results of weed species 19 Agric. Sei. Fint. Suppl. No. 2 (1993) (Table 1 in I) affected by the unequal number of fields from different geographical regions. Never - theless, the data set reflects the distribution of cer- eal fields in Finland, and thus the most important weed species affecting cereal production. Connections between geographical regions and the species composition of weed populations were indicated with ordination analyses (I). Charac- teristic weed species in southern Finland were vol- unteer turnip rape (Brassica rapa L. subsp. oleifera DC. alias B. campestris L.), Fumaria officinalis, Lamium spp., Stellaria media and Tripleurosper- mum inodorum, whereas Lapsana communis, Ely- mus repens, Myosotis arvensis må Achillea L. spp. were common in central Finland. The occurrence of particular weed species was also related to management practices and to soil properties (I), as reported from other studies (Granström 1962, Bachthaler 1969, Rade- macher et al. 1970,STREißiGetal. 1984,Dale and Thomas 1987,Andreasen etal. 1991,Dale etal. 1992). Cereals dominated the crop rotations in southern Finland, where soils were mostly clay type, whereas in central Finland mixed crop rota- tions in coarse and organic soils were typical. Thus, the geographic regions and farming practices were associated. In addition to the management practices and soil types, it is evident that the cropping history of each field affects the weed incidence (Haas and STREIBIG 1982). Ordination analysis with CCA in- dicated the high relative importance of manage- ment practices, particularly continuous herbicide use, in determining the species composition of weed flora (I). However, in this study no single factor determining the composition of weed popu- lations in spring cereal fields in Finland could be identified either by regression analysis (Erviö and Salonen 1987) or by ordinationanalysis (I). Ordination techniques applied in this study were chosen from among a group of mathematically di- verse approaches. Illustrations on species-environ- ment interactions have been produced also with other numerical methods such as factor analysis (Streibig 1979) and canonical discriminant ana- lysis (LÉGÉRE et al. 1993). However, CCA and its implementation in the computer program CANOCO has received wide acceptance in ecological research (Birks and Austin 1992), pre- sumably due to its sound ecological basis and well- documented theory and practical implications re- ported by Ter Braak (1987b). The most common weed species were well rep- resented in the field trials (111-V). Weed incidence was not static during the growing season nor be- tween the seasons (V, Salonen 1993). Galeopsis spp. were typical species, emerging early in the growing season, whereas Stellaria media and Viola arvensis increased their relative proportion in the weed population towards the end of the growing season. These observations on fluctuations in weed incidence agree with results of earlier studies con- ducted in the Nordic countries (ErviÖ 1981, Håkansson 1983a, 1992, Andreasen 1990, Fykse 1993a). Long-term changes in species composition of weed populations are of great importance. Assess- ment of the diversity of weed communities provides information on the compositional changes caused by agricultural practices (TOMKINS and Grant 1977). In this study, the number of weed species surveyed in 1982-1984 was restricted to 33 (I). The remainder of the observed species were pooled in the groups of "other dicots" and "other monocots", and thus the total number of weed spe- cies occuring in spring cereal fields is not reported. Quantitative and qualitative changes in weed incid- ence between the 1960 s and 1980 s were reported in detail by Erviö and Salonen (1987). Mukula et al. (1969) found a total of 304 weed species in the 2,710 spring cereal fields surveyed in southern and centralFinland in the 19605. In 1983, Kallio-Mannila et al. (1984) returned to 73 of the spring cereal fields in central Finland studied by Mukula et al. (1969) in the 1960 s and found that the number of weed species had decreased in that particular region from 119 to 90. Weed species that had disappeared (e.g. Centaurea cyanus L., Campanula rotundifolia L., Ranunculus auricomus L. and Hieracium pilosella L.) were of minor im- portance already in the 1960 s (Mukula et al. 1969). In contrast, Andreasen (1990) found that the number of weed species occurring in Denmark had not decreased from the 19605, although the 20 Agric. Sd. Finl. Suppl. No. 2 (1993) frequency of several species has drastically de- clined. 1.2 Weed infestation levels in cereal Helds The level ofweed infestation was clearly a dynamic phenomenon as considerable differences in weed density and weed biomass were observed both within a growing season and between seasons (111, V, Salonen 1993). According to a continuous series of observations over four years (V), weed density at the time of herbicide application varied annually in the same field e.g. from 52 to 300 weeds 2 2m' in Jokioinen and from 163 to 702 weeds m in Kokemäki. There were no clear trends of increasing weed densities (No. m ) even in untreated plots. The great annual fluctuations in above-ground weed populations fully agree with the observations of Fykse (1993a) and Hallgren (1993 a). In the weed survey, the average density ofweeds was 170 plants trf (median 124), and the average weed biomass was 320 kg ha’ 1 (median 183)(I). In the 19605, the weed infestation in sping cereal 2 1fields averaged 550 plants m “ and 1000 kg ha , respectively (Mukula 1974). Observations were made, both in the 1960 s and 1980s, in unsprayed sample plots in July, i.e. at a time when most of the weeds had emerged, but only a few had withered (c.f. Mukula et al. 1969). Although the data on weed infestation levels are already ten years old, they are presumably still valid since the changes in agricultural practices have been much less during the 1980 s than during the 1960 s and 19705. More- over, Aamisepp and Wallgren (1979) and Hall- gren (1993a) have shown that the decline in weed infestation has slowed in Sweden. In the field trials (111, IV, V), the density of annual broad-leaved weeds at the time of spraying ranged from 7to 700 weeds m , averaging 70-75% of the infestation one month later (V, Salonen 1993). Crop density did not significantly affect the number of weed seedlings per unit area at the time of spraying, however, higher crop densities had an inhibitory effect on weed growth later in the grow- ing season (111, Salonen 1993). Annual broad-leaved weed species were pre- dominant in the survey fields and accounted for 77% of the total weed biomass (II). However, none of these species reached the same mean biomass production per plant as the crop plants (V). The most aggressive weed species were volunteer tur- nip rape (Brassica rapa ssp. oleifera). Galeopsis spp. and Fallopia convolvulus, which produced biomass within the range of 0.15-0.60 g DW per plant (V). The dry weight of a barley plant averaged 1.10g and that of a wheatplant 1.02gin crop stands sown at the recommended seeding rates of 450 viable seeds m for barley and 600 for wheat (V). In conclusion, based on theresults from the weed survey (I, II) and from the field experiments (111, IV, V), the most common weed species were ranked to indicate their relative importance in spring cereal production in Finland (Table 5). It should be noticed, however, that any of these spe- cies can be harmful in one particular field where the cropping history and environmental conditions have favoured their growth. Moreover, efficient removal of the most competitive weed species with herbicides may be advantageous for herbicide-tol- erant species, as discussed later in this thesis. 2 Chemical weed control in spring cereals 2.1 Use of herbicides In total, 22 different herbicide products were used by farmers in the survey fields in 1982-1984 (II). MCPA-containing herbicideformulations were the most commonly applied. MCPA alone, and in mix- tures with dichlorprop and mecoprop, constituted 91% of the total number of herbicide treatments. The observed distribution of different herbicides sprayed in the survey fields agreed with the sales statistics (Fig. 2). A recommended dose range of a herbicide is indicated on the product label.The dose recommen- dations are proposed by the manufacturer, and the herbicide is screened in official field trials and ap- proved by the authorities. In Finland, the official body releasing the herbicides onto the market is the Pesticide Commission. At present, label informa- 21 Agric. Sei. Finl. Suppl. No. 2 (1993) Table 5. Relative importance of different weed species in spring cereal production in Finland. The ranking of weeds is based on their occurrence (frequency), adverse effect on crop (interference) and their susceptib- ility to chemical weed control with special emphasis on the need for improved herbicide efficacy. Importance Ranking factor Weed species Frequency 1 Interference2 Control 5 Major Chenopodium album L, *** •** * Galeopsis spp. L. (G. speciosa, G. letrahil, G. bifida) *** *** Elymus repens (L.) GOULD ** *** ** Viola arvensis MURRAY *** * *** Slellaria media (L.) VILL. *** *• * Lapsana communis L. ** ••• �* Fallopio convolvulus (L.) Ä. LÖVE ** *** •* Brassica rapa L. ssp. oleifera DC. ** *** * Sonchus arvensis L. ** *** ** Avena fatua L. * *** ** Moderate Polygonum aviculare L. ** ** Polygonum lapaihifolium L. ** *** *• Erysimum cheiranlhoides L. ** ** * Myosotis arvensis (L.) HILL ** * ** Spergula arvensis L. ** ** * Fumaria officinalis L. ** ** ** Lamium spp. L. (L. purpureum, L. hybridum) ** ** ** Cirsium arvense (L.) SCOP. * *** ** Tripleurospermum inodorum SCHULTZ BIP. ** ** Galium spp. L. (G. spurium) ** ** ** Minor Thlaspi arvense L. * ** * Capsella bursa-pastoris (L.) MEDIK. * * * Matricaria matricarioides (LESS.) PORTER * ** Sonchus spp. L. (S. asper, S. oleraceus) Ranunculus repens L. * * ** Equisetum spp. L. (E. arvense) * ** * Poa annua L, * * *** Rumex spp. L. (R. acetosa, R. acetosella) * ** ** Gnophalium uliginosum L. * � *• Achillea spp. L. (A. millefolium, A. ptarmica) * ** ** 1 Frequency: *: <30%, **: 30-60%, ***: >60% 2 Interference (competitiveness): *: weak, **: intermediate, ***: strong 5 Control (with current herbicides): •: easy, **: moderate, ***: difficult tion for dose adjustment deals only with the recom- mended doserange. The survey conducted in farm- ers’ fields in 1982-1984 indicated that most of the farmers followed the dose recommendations. How- ever, a survey carried out in 1991 in cereal farms in southern Finland showed that reduced herbicide doses were applied in 16% ofthe 300 spring cereal fields studied (Salonen 1992a). 2.2 Efficacy of herbicides Herbicides provided good weed control, defined here as >70% reduction of weed biomass following herbicide application, in 68-84% of the fields stud- ied in weed survey (II). In general, mixtures of phenoxy acids provided more efficient and reliable control than MCPA alone (Fig. 5). On average, 22 Agric. Sei. Finl. Sappi. No. 2 (1993) MCPA alone gave only moderate control (65%), while the efficacy of the other herbicides averaged 83% (II). Only general conclusions on the effect of different herbicides can be drawn from the weed survey, since herbicides were applied in a variety of conditions. This, however, was the first extensive study on herbicide efficacy in cereal fields in Fin- land and it indicated that the efficacy of herbicides in practice corresponds with the efficacy achieved in field trials (e.g. Aamisepp 1984, Junnila 1990, HI. V). In the field experiments (V), the lowest recom- mended doses of phenoxy mixtures provided on average as high as 80-90% control, even though the choice of herbicide was not based on the prevailing weed populations. Statistically significant Year * Site * Treatment interactions indicated that the herbicide efficacy was site-specific and fluctuated between the years (V). Herbicides were applied to natural weed vegeta- tion, comprising several weed species, and resulted in selective control of the different weed species (111, V). Consequently, removing some of the com- petitive but susceptible species (e.g. Cltenopodium album, Galeopsis spp.) provided more space for less competitive but herbicide-tolerant species (e.g. Fumaria officinalis, Viola arvensis). As a result, biomass production of the tolerant species was sometimes higher in treated plots than in untreated plots. Hallgren (1993b) reported similar results for V. arvensis in cereal experiments in Sweden. The incidences of low efficacy (%-scale) of herbicides were mainly explained by the applica- tion ofan inappropriate herbicidefor the prevailing weed population, but also by the fact that weed infestation was sometimes low even in unsprayed plots (V). Thus, the difference in weed biomass between sprayed and unsprayed plots remained low. Consequently, efficacy results, given on a rel- ative scale (% control), hid the actual level of weed infestation and the need for control in low infesta- tion situations e.g. at Jokioinen in 1986 (III) and several research stations during 1989-1991 (V). Therefore, the results were reported as changes in 2 2weed density (No. m )or weed biomass (g m )on a quantitative scale, to give a realistic account of the weed infestation in cereal stands of the field experi- ments (V). The new active ingredient, fluroxypyr, showed no particular advantage over the other herbicides, since it is a narrow-spectrum herbicide active against particular species including Galium spp., which were not abundant in our field experiments (111, V). On the other hand, the low dose herbicide tribenuron-methyl provided broad-spectrum con- trol (V) and is likely to become one of the standard herbicides used in cereal production in Finland. In general, all herbicide products included in our experiments were effective for weed control in spring cereals providing that species composition of weed population was considered as a determin- ing factor in choosing the herbicide. No commer- cial herbicide product covers the entire range of the most common weed species. To overcome this problem, mixtures of different herbicide formula- tions can be used when necessary. 2.3 Impact ofdose reduction on herbicide efficacy Reduction of the herbicide dose did not cause a corresponding reduction in efficacy against broad- leaved weeds. Halving the highest recommended Fig. 5. Comparison of the herbicide efficacy (% reduction of weed biomass) achieved with MCPA (left bar) and with the mixtures of MCPA/dichlorprop or MCPA/mecoprop (right bar). Distribution of the 252 spring cereal fields into efficacy classes. Data compiled from theTable 3 in 11, 23 Agric. Sei. Finl. Suppl. No. 2 (1993) dose of MCPA/mecoprop and MCPA/fluroxypyr reduced the average efficacy (% control) of 88% with recommended doses to 76% with reduced doses (Fig. 3 in III). Further dose reduction to one- third of that recommended resulted in highly vari- able levels of control, mainly because the examined herbicides were either not effective or their efficacy against some weed species (e.g. Viola arvensis) declined considerably when the dose was reduced. The efficacy of MCPA/dichlorprop-P, MCPA/ fluroxypyr and MCPA/mecoprop-P, applied with the lowest recommended doses, averaged 85%, and 79% with 30% lower dose, respectively (V). A considerable (>l5 %units) difference in the effi- cacy between the plots treated with recommended and reduced doses was observed in 8-29% of the cases depending on the herbicide, weed species and the crop. Moreover, to summarize the results of the field experiments during 1989-1991 on a quantitat- ive scale, weed densities from 7 to 700 weeds m'2 were observed in spring cereal experiments at the time of spraying in June (V). The weed biomass. recorded one month later, averaged 12.8 (SE 1.4) g DW m’“ in unsprayed plots, 1.8 (SE 0.2) gDW m~2 in the plots treated with the lowest recommended herbicide doses and 2.2 (SE 0.2) g DW m'2 in the plots treated with 30% lower doses, respectively. As an example ofother studies, JENNÉUS (1992) showed on demonstration trials on farmers’ fields in Sweden that an adequate reduction of weed biomass (below 50 g fresh weight m' 2 ) in spring cereals was achieved in 55% of cases with 1/4 of the normal herbicide dose, and in 70% and 91 % of cases with 1/2 and 1/1 of the normal dose, respect- ively, using herbicides chosen by the farmer. The efficacy of MCPA/fluroxypyr decreased more rapidly than that of MCPA/mecoprop when reduced doses were applied (III). Halving the high- est recommended dose reduced the efficacy of MCPA/mecoprop on average by 7 %-units and the efficacy of MCPA/fluroxypyr by 14 %-units. This can be interpreted as either i) dose recommenda- tions for new products (MCPA/fluroxypyr) are closer to the optimum than recommendations for old products (e.g. MCPA/mecoprop) or it), more likely, that fluroxypyr was, to a lesser extent than mecoprop, supplementing the efficacy of MCPA against the weed species that predominated in our field experiments. Relatively slight reduction in efficacy with re- duced doses is in accordance with the results of other experiments (e.g. Aamisepp 1984, Pallutt 1988, Proven et al. 1991). In the case of inad- equate control, split applications at reduced rates have been suggested if the first application is not successful (Pallutt 1988, Meinlschmidt and Karch 1992). Such an approach could be more useful in practice if the second application coin- cided with e.g. spraying of growth regulators or insecticides against cereal aphids. Otherwise split applications are hardly cost-effective due to relat- ively high costs (machinery, labour) of herbicide application. Dose reduction of herbicides caused differential changes in efficacy against weed species (111, V). Typically, the level of control of Chenopodium album was over 95% even at the lowest doses, while the efficacy against e.g. Viola arvensis rap- idly decreased when the herbicide dose was re- duced. As an example, the efficacy of MCPA/ fluroxypyr against V. arvensis decreased from 78% at the full dose to 52% and 44% when the dose was reduced to one-half and one-third, respectively (III). Within the range from one-third to the highest recommended dose screened in our experiments (Table 4), a linear response model between herbi- cide dose and the remaining weed biomass ap- peared to be a feasible approximation (III). On the other hand, the dose range screened in the experi- ments was insufficiently broad to determine the shape of the response curve within the entire range from zero to the recommended dose. A linear re- sponse model, including two explanatory factors, herbicide doseand crop density, was fitted to field data (Table 3 in III). The results indicated the higher relative power of herbicides to suppress weed growth compared with crop density. As an example, the "weed biomass - herbicide dose - crop density" relationship is graphed (Fig. 6) by fitting a response surface on the weed data from a field experiment, in spring wheat in 1988, when the greatest weed infestation was recorded during the three-year field experiment (III). 24 Agric. Sd. Pint. Suppl. No. 2 (1993) Regarding competition, crop seeding rate is in many respects an analogous management practice to herbicide dose in attempting to suppress the growth of weeds (Håkansson 1986). In our ex- periments, weed growth was more efficiently sup- pressed by herbicides, even at low application rates, than by increasing crop seeding rates above the recommendations (111, Fig. 6). However, a dense crop was advantageous in maintaining the weed growth at low levels towards harvest. This was detected particularly when the herbicide effect was only moderate, as in our field experiments in Ylis- taro (see Table 3 in III). Nevertheless, weed control through increased seeding rates above the optimum for yield production is not economically meaning- ful compared with the use ofherbicides even at very low rates of application (III). Substantial reduction in the recommended dose of herbicides is possible only if a herbicide is chosen according to the weed species and is applied at early growth stages of weeds under favourable conditions (Kudsk 1989). Satisfactory control of weeds in our experiments is apparently explained by the fact that most of the weed seedlings were at an early growth stage (cotyledon stage - first true leaves) at the time of herbicide application (V). Evidently, recommended herbicide doses are still needed in cases of delayed application, under ad- verse growth conditions, particularly in uneven crop stands, and against specific weed species (see also Kudsk 1989). 2.4 Yield responses to chemical weed control Yield differences between untreatedplots and plots treated with herbicides was, in most cases, below 5% in crop stands sown at the normal seeding rate (IV, V). Moreover, yield response to herbicide dose was low, particularly in spring barley, irrespective of the yield levels, which variedbetween the sites and years (IV). Only in one trial out of the 21 spring wheat trials wheat yield was significantly (P<0.01) lower in the plots treated with reduced doses of Fig, 6. Response surface of the effect of herbicide dose (MCPA/mecoprop formula- tion) and crop density on the weed biomass fitted to the data from the spring wheat experiment at Jokioinen in 1988 (III). 25 Agric. Sei. Finl. Suppl. No. 2 (1993) MCPA/mecoprop-P and MCPA/fluroxypyr than in the plots treated with recommended doses. Small yield responses from chemical weed control clearly indicate the relevance of minimized herbicide in- puts in cost-effective cereal production. Responses ofcrop yields to herbicide application were even negative in 32% of the wheat plots, and in 43% of the barley plots (IV). In the spring cereal fields of our weed survey in 1982-1984 chemical weed control was profitable in 60% of fields (Erviö et al. 1991). Gerowitt et al. (1984) re- ported that only 50% of herbicide treatments in spring barley were profitable in Germany. The crop yield was often higher in the plots treated with reduced herbicide doses than in those which received the highest recommended dose (IV). Although phenoxy herbicides caused neither visual phytotoxic symptoms in crop plants nor de- creased their dry weight, cereal plants may on occa- sion have been at a sensitive development stage (from ’mid-vegetative’ to ’double ridge’ stage), in terms of ear distortions, as discussed by Thomson and Stokes (1985) and Kirby et al. (1989). Nevetheless, statistically significant differences in yield parameters (1000 kernel weight, bulk weight etc.) between untreated and treated plots were de- tected only in a few cases in our field experiments (IV). Similarly, the yield quality was not affected by chemical weed control in the fields of our weed survey (Erviö et al. 1991). The impact ofseeding rate on crop-weed interac- tions, and consequently on yield response, was more significant in wheat than in barley (Fig. 1 in IV). Nevertheless, even in sparse wheat stands the same yield level was reached both at reduced as well as at recommended herbicide doses. On the other hand, high rates of application were profitable in terms of yield if less susceptible weed species, such as Tripleurospermum inodorum, were preval- ent in the experimental site (IV, V). The recommended sowing densities for spring 2barley (450 seeds m ) and for spring wheat (600 seeds m ) are higher inFinland than in more south- erly situated countries due to the strong dominance of the main stem of cereals under our long-day conditions (Mela and Paatela 1974, Peltonen- Sainio and Järvinen 1993). Apparently, the re- commended seeding rates for spring cereals are high enough to assure the competitive advantage of the crop stand against weeds (III). Increase of crop density above the recommended seeding rate gave an insignificant benefit, particularly in barley, in limiting the growth of weeds. This has also been reported by Håkansson (1975) and Erviö (1983). The highest crop density (900 and 1000 viable seeds m ~ for barley and wheat, respectively) used in our experiments (111, IV) was too dense, often yielding lower than the less dense crop stands (IV). It is likely that the crop was suffering from intra- specific competition and lack of nutrients. Although yield responses were sometimes erratic (IV), the general conclusion was that in most cases reduced herbicide doses were sufficient to guaran- tee the same yield level as therecommended doses. The benefit of efficient weed removal was partly lost when unnecessarily high levels of herbicide were applied, since higher yield were achieved from the plots treated with lower doses (Fig. 4 in IV), as discussed also by Thonke (1986) and by Courtney (1991). The results from experimental fields on low yield responses to herbicide application are in agreement with the observations in the weed survey; the crop yield in treated areas was, on average, only 3% higher than in untreated ones (Erviö et al. 1991). Hughes (1966) and Evans (1968) reported that the control of broad-leaved weeds in the U.K. had little impact on cereal yields already in the 19605. Sim- ilar results were reported in the 1980 s by Court- ney and Johnston (1986), by Davies et al. (1989) and in Scandinavia by Jensen (1985). Significant yield increases tend to be detected only when high densities of competitive weed species are treated with herbicides (SCRAGG 1980, Wilson 1982, LOTZet al. 1990). 2.5 Basing the herbicide use on crop-weed interactions Decisions on the type and application rate of herbi- cides have to be based on observations and predic- tions which are made at the beginning of the grow- ing season. However, in our field trials there was a 26 Agric. Sei. Fint. Suppl. No. 2 (1993) poor correlation between the weed density at the time of spraying and the biomass production of weeds later in the summer (111, V). As discussed 2earlier, several hundreds of weeds m at the time of spraying had little effect on crop yield in some trials, whereas in others only a few weeds reduced the crop yield significantly. This is reflected in the weak correlation(1-0.48) between the total number of annual weeds per unit area (m ) and weed biomass in untreatedplots. In general, the density-based thresholds were found unreliable (IV, V), as discussed also by Be- HRENDT (1986) and by Bleiholder and Nuyken (1986). The results from field plots with multispe- cies weed infestation are in accordance with the results of field experiments with single weed spe- cies (Lutman 1992). Thus, predictions on yield loss according solely to the weed density (Table 3) should be interpreted with caution in spring cereal production in the Nordic regions, and regarded only as indicative of trends. In spite of critical reviews against the threshold approach to weed control, the strategy is applied in practice e.g. in Norway and in Germany (Heite- FUSS et al. 1987, Fykse 1993b). In Norway, farm- ers are advised through an integrated telefone-com- puter system TELEVIS (Fykse 1993c). It is based on a threshold model which takes into account the weed species that occur in the field, the number of weeds per unit area and moreover the % cover of crop and weeds (Fykse 1991b). In Germany, the threshold approach is applied particularly in winter cereals, resulting in approximately 65% correct control decisions in terms of profit margin (GE- rowitt et al. 1986). By applying the threshold approach in that study, herbicide use was withheld in 26% of cases in winter barley and winter wheat, and in 36% of cases in winter rye. A more recent validation of that decision model resulted in with- holding chemical weed control in 45% of cases in winter wheat(Gerowitt 1992). Weed biomass produced by the broad-leaved weed species in the field experiments was relatively low compared with crop biomass even in untreated crop stands; the proportion of weeds was seldom higher than 5% of the total vegetative biomass in crop stands sown at the normal seeding rates (V). Thus, the effect of weeds on crop yield often re- mainedbelow the statistically significant level. On the other hand, in crop stands sown at low seeding rates (100 and 200 seeds nT of barley and wheat, respectively) the proportion of weeds often in- creased, reaching approximately 20% of the total biomass (111, Salonen 1992b), and yield increase achieved with chemical weed control in wheat was in some cases more than 30% (IV). Growthhabitand the ability of a weed species to grow under low light intensity are important factors in competition between plants (Fogelfors 1974). Typically, only a few broad-leaved weed species can successfully compete with cereal plants: WIL- SON (1986) ranked only Galium aparine L. more aggressive than winter wheatand Dock Gustavs- son (1986) reported that only Sinapis arvensis L. out-competed barley. In the experiments of Jensen (1991a), in Denmark, Brassica napus L., Sinapis an>ensis L. and Galeopsis spp. were the most effective biomass producers in spring-sown crops. In addition to botanical factors, edaphic factors can affect crop-weed interactions. Benefits from weed control are expected to be higher in soil types other than clay, because the production of weed biomass per plant was higher in coarse and organic soils than in clay soils (I), which were predominant in the field experiments. Indeed, yield responses of spring barley to chemical weed control have been found to be lowest in clay soils (Jensen 1985, Hallgren 1988). Some explanations for low yield responses can be given. The relative time of emergence is con- siderd to be of great importance in crop-weed com- petition (c.f. Håkansson 1983b, Cousens et al. 1987,Håkansson 1991). In the field trials of PRO- JECT 3, crop plants normally reached the second leafstage (stage 12-13 on the Zadoks’ scale) before the flush of annual weeds (V). Apparently, this was an advantage for crops in competition with weeds. In addition to direct control methods discussed above, some additional considerations on indirect methods can be given. The barley variety ‘Arra’, used in PROJECT 2 (III), is characterized by rapid initial growth (Jokinen 1991). The good competit- ive ability of this variety apparently partly explains the low production of weed biomass in our trials. In 27 Agric. Sei. Fin!. Suppl. No. 2 (1993) fact, now that indirect methods for weed control are emphasized, the use of competitive varieties has been suggested as a potential means ofreducing the use of herbicides. (Moss 1985, Richards 1989. Richards and Davies 1991, Whiting et al. 1991, Christensen 1993). Christensen et al. (1990) reported that the same level of weed control was achieved using the most competitive barley variety with only one-thirdof the herbicidedose neededfor weed control in fields of the least competitive bar- ley variety. In addition, the current management practices in Finland, with placement of fertilizers, obviously favour the nutrient uptake and growth of the crop particularly in the early stages ofgrowth. This gives an advantage to crops in competition with weeds (Espeby 1989, Salonen 1992b). Extensive research has been focused on the na- ture and effects of competition between the crop and weeds. Quantification of yield responses is often based on growth densities of the crop and the weeds (e.g. Håkansson 1983b, Spitters and Van Den Bergh 1982). It seems, however, that the status of competition between the crop and weeds at the time of herbicide application should be de- fined more precisely than has so far been done in threshold models (Pallutt and Roder 1992). One recent approach is to estimate the required weed control with a relative leaf cover model (Kropff and Spitters 1991). In this approach, predictions of yield loss caused by weeds are based on relative leaf area of weeds compared with leaf area of the crop. Several working groups within the European Weed Research Society, including one at the In- stitute of Plant Protection in Jokioinen, are vali- dating the performance of that particular model under different environmental conditions (Lotz et al. 1993). In addition, Assémat (1992) has sug- gested that the species-environment interaction in- fluencing the growth rates of weeds (and crops) shouldbe studied in more detail instead ofintroduc- ing average growth parameters into the prediction models. In conclusion, regarding the estimation of yield losses caused by weeds in spring cereals in Finland, the most difficultproblem is accurate assessment of relatively low levels ofweed infestation and predic- tion of their growth potential during the growing season. 2.6 Impact of herbicides and crop rotation on weed infestation When herbicide use is drastically decreased or completely stopped, weed control has been re- ported to create major problems, particularly in low-input farming systems (Edwards and Reg- NIER 1989). Subsequent effects of crop rotations and different herbicides and their application rates on weed infestation were followed in PROJECT 2 in Jokioinen. Differences in subsequent weed in- festation, one year after the three-year trial period, were detected between the crop rotations, and between untreated and treated plots, but not sig- nificantly between different herbicides and their rates of application. The subsequent weed infesta- tion in untreated plots of wheat-barley-wheat rota- tion averaged 715(±88) weeds m , whereas the weed density in barley-wheat-barley rotation re- mained at the level of 336(±50) m'“ (III). On the other hand, in treated plots of both rotations the mean density was 265(±7) weeds m’ 2 . No signific- ant differences in weed infestation between herbi- cides (MCPA/fluroxypyr, MCPA/mecoprop) and their rates of application were detected (III). Thus, if herbicides are continuously applied, even with reduced doses only, farmers can more freely choose their crop rotations without causing any significant adverse effect on the subsequent weed infestation. Similarly in PROJECT 3, the subsequent weed infestations were recorded from plots treated with different herbicides (MCPA/dichlorprop, MCPA/ fluroxypyr, MCPA/mecoprop, tribenuron-methyl) and at different rates of application (the lowest recommended and 30% lower dose). No significant differences in weed infestation between the treat- ments were detected after continuous use of herbi- cides during the three-year trial period (V). Jensen (1991b) suggested that the level of weed infestation can be kept stable either by applying half recom- mended herbicide doses every year, or alternatively by applying full doses every two years out of three. Production of weed biomass was significantly 28 Agric. Sei. Finl. Suppl. No. 2 (1993) higher in wheat than in barley stands (Fig. 1 in 111, Salonen 1993).Evidently, the production ofweed seeds was also higher in the wheat plots than in barley plots since the production of weed seeds has been found to be linearly related to the weed biomass (Wilson et al. 1988, Debaeke 1988, Pe- dersen and Rasmussen 1990). Moreover, it is likely that a 20- to 30-day longer growing season required by spring wheat in comparison with spring barley was effectively used by the annual weeds for seed production, maturation and shedding. Although ploughing the fields every autumn partly buffered the changes in weed infestation, the increase in weed infestation was detected in the wheat-dominated rotation within three to four years. Thus, competitive ability of the crops in- cluded in the crop rotation is of great importance in long-term weed management, as emphasized in other studies (Zwerger et al. 1990, Flintzsche 1990). In addition to the choice of crop, the level of weed infestation is obviously also influenced by crop management practices such as tillage (Knab and Hurle 1986, LÉGÉre et al. 1990). At present, the most severe threat of increasing weed infesta- tion in Finnish fields is neglected weed manage- ment in the obligatory set-aside fields. Changes in weed infestation were detected not only as the number ofweeds emerging in spring but also in the seed bank of weeds in the soil. In PRO- JECT 2 (III), observations from soil samples revealed that the weed seed reservoir was signific- antly (PcO.001) higher after a wheat-dominated ro- tation than after a barley-dominated one (Fig. 5 in III). However, even in wheat, reduced rates of herbicide application were sufficient to keep the subsequent weed infestation at the same level as normal rates. Also Lawson et al. (1992) found that treatment with 50% normal dose incurred no penalty in terms of weed seed banks in soil. Although herbicides evidently exert a selection pressure on weed communities (STRYCKERS 1979), the response in weed communities to continuous use of herbicides was, in our trials, in the first place quantitative and not so much qualitative. This is in agreement with results from earlier studies by Mahn and Helmecke (1979) and by Hume (1987). Similarly, ErviÖ and Hiivola (1986) found no changes in the species composition of weed populations in spring wheat experiments after different control regimes (recommended herbicide dose, half dose and threshold treatment) imposed for five years in Finland. In contrast, Kees (1986) found changes in the species composition of weed populations after four years of different weed con- trol regimes ("maximum", "threshold" and "every 2nd year control") in a four-year rotation (sugarbeet - winter wheat - spring barley - spring oats) in Germany. CATIZONE et al. (1990) found during a five-year study that herbicides had a marked influ- ence on the species composition of weed popula- tions in a continuous winter wheat experiment in Italy. Slow alterations in weed populations in relation to control regimes in Finnish conditions may be due to a short growing season in relation to the winter period. Factors related to our growing conditions may affect both weed seed production and mortal- ity of weed seeds in the soil. Even so, long-term selection pressure in weed populations, caused by management practices and herbicide use, affects the species composition of weed populations as detected in the weed survey (I). However, the res- ults based on the above-ground observations on changes in species composition should be inter- preted with caution, since the above-ground weed infestation has been shown to correlate only weakly with the level and species composition of weeds in the seed bank in the soil (Fykse 1993b). The results ofchanges in species composition of weed populations in PROJECT 2 in Jokioinenhave been reported elsewhere (Salonen 1993). How- ever, in summary: the occurrence of Chenopodium album increased particularly in untreated plots in the wheat-dominated rotation, and Viola arvensis proportionally increased in the barley-dominated rotation, even in herbicide-treated plots. Chenopo- dium album is well-known for its high seed produc- tion capacity; a recently published figure is 7,000- 30,000 seeds m in an untreated crop stand of spring barley (Rasmussen 1993). The proportion of V. an’ensis increased in three years from approximately zero to almost 50% ofthe total number of weeds (SALONEN 1993). Viola ar- vensis was relatively tolerant of herbicides applied 29 Agric. Sei. Finl. Sappi. No. 2 (1993) in our experiments, particularly of low rates of application. Furthermore, as mentioned earlier, the proportion of weed seedlings that emerged before herbicide application averaged 70-75% of the peak infestation determined one month later. Appar- ently, some V. arvensis plants emerged after herbi- cide application in the beginning of June, since Erviö (1981) has observed that e.g. V. arvensis and Lamium spp. reach their emergence peak during mid-summer. The late emerging plants are able to produce seeds, particularly if there are gaps in the crop stand (Niemann 1990). 3 Economic impact and practical implications of herbicide dose reduction Appraisals of the economic return from weed con- trol should be based on i) the density of each weed species, ii) their relative competitiveness, in) the estimated weed free yield, iv) the level of weed control expected with the herbicide, v) the antici- pated net value of the crop and vi) the cost of selected treatment (COMBELLACK 1992b). Thus, site-specific circumstances determine the profits derived from chemical weed control. In this study, the possibilities for reducing the recommended doses were validated only withphenoxy acid herbi- cides, but the same approach has been shown to apply also to sulfonylureas (Fogelfors 1992, JEN- NÉUS 1992, Junnila, pers. commun.). In fact, tribe- nuron-methyl applied at a rate of7 g product ha 1 in our field experiments (V) provided good weed con- trol, taking into account that 10g ha" 1 is regarded as a "normal" dose for spring cereals. The results achieved in the field experiments (IV, V), both in terms of weed control and yield response, hopefully support farmers’ considera- tions towards cost-effective weed control with re- duced herbicide doses. Relatively low yield re- sponses to chemical weed control of annual broad- leaved weeds were in accordance with results from the weed survey Erviö et al. (1991). The weed infestation levels in the field experiments (111, IV, V) were equal to the infestation in farmers’ fields (I, 11, Erviö and Salonen 1987) in terms of weed density, but not always in terms of weed biomass. Erviö et ai. (1991) reported that chemical weed control in spring cereals provided, on average, a 3% higher crop yield than without herbicides. At the yield level of 3,500 kg ha 1 this corresponds to a yield gain of 105 kg ha' 1 with a monetary value of FIM 163 for barley (reference net price FIM 1.55 kg' 1) and FIM 226 for wheat (FIM 2.15 kg' 1). The profit wouldbe sufficient to cover the present aver- age herbicide cost of FIM 100 ha" 1 and even the other costs (machinery, labour) of weed control at least in wheat production. Based on the results of this study and additional field experiments at the ARC, the Institute of Plant Protection of the ARC in autumn 1993 issued a statement to the Pesticide Commission in Finland, urging revision of the product labels of phenoxy- acid herbicides.The reduction of the lowest recom- mended doses by approximately 30% was pro- posed, and moreover, some recommendations for dose-adjustment were suggested to be included on labels. As herbicides applied in cereal production con- stitute the majority of herbicides used in Finland, even a slight reduction in their use on individual farms, if practiced by significant numbers of farm- ers, would lead to considerable savings at the na- tional level. The amount of cereal herbicides sold in 1992 was sufficient to treat 635,000 ha, accounting for 69% of the total cereal acreage in that year (Hynninen and Blomqvist 1993), corresponding to a monetary value of approximately FIM 63.5 million in herbicide sales. If the use of these herbi- cides on a national scale could be reduced by 30%, by adjusting the application rates according to the actual needs, the savings would be approximately FIM 20 million per annum. A prerequisite for successful long-term reduction in pesticide use is to introduce reliable recommen- dations and control measures for farmers. The good weed control achieved in the field experiments, with 30% lower doses than those recommended (V), should encourage farmers to test even lower doses in their own fields. However, the complex factors (management practices, crop rotation, pre- vious weed control) influencing the dynamics of weed populations and survival rates create a com- plicated environment for decision making. There- 30 Agric. Sei. Finl. Suppl. No. 2 (1993) Table 6. Towards a computerized decision support system for chemical weed control in spring cereal pro- duction. Appraisal of the current status of available information in Finland by ranking different factors with a special emphasis on reducing herbicide use. Factor Relative Availability importance l of data 2 a) Factors related to weeds Identification of weed species *** *** Competitive ability of weed species *** ** Effect of growth stage of weed on susceptibility to herbicides ** * Effect of physiological stage of weed on susceptibility to herbicides ** * Production of weed seeds in cereals * * Selection in weed population ** ** b) Factors related to herbicides Efficacy of individual herbicide against individual weed species *** ** Effect of dose on efficacy (dose response) *** ** Effect of weather on efficacy ** ** Flerbicide mixtures ** ** Adjuvants ** ** Integration to other control practices * * Sequential application * * Price ** *** Persistence of herbicides ** ** Side effects of herbicides ** * Toxicology of herbicides * *** c) Factors related to crop Competitiveness of crop *** *** Competitiveness of crop cultivar ** ** Prediction of yield loss (quantitative & qualitative) ** * Crops included in crop rotation * ** d) Factors related to technology Software for PC *** *** Valid weather forecast ** ** Detection of weed infestation (level, patchiness) ** Sprayer technology (differences between nozzle types etc.) ** Replacement of herbicides with other control practices * ** 1 Importance: ***: Essential, **: Intermediate, *: Useful but not necessary to commence 2 Availability: �**: Good data exist, **: Adequate data to commence but more research required, *: Finnish (or relevant imported) data inadequate fore, computerized advisory systems have already been introduced to assist farmers in weed manage- ment (e.g. Baandrup and Ballegaard 1989, Garvert et al. 1990, Fykse 1993c). Introducing the recently obtained results into computerized decision support systems will be the next step in the process of optimizing herbicide use in Finland. The status of available information to develop a decision support system in Finland is summarized in Table 6. Some data (e.g. impact of weatherparameters, efficacy, doseresponse) can in part be adopted from other countries, but further research in Finland is still needed to improve the validity of system. In Denmark one of the most sophisticated PC-based decision support systems is already in use (Baandrup and Ballegaard 1989, 31 Agric. Sei. Fint. Suppl. No. 2 (1993) Murali and Secher 1991), and could be used as a source ofdata if applicable to Finnish conditions. Another prospect for applying reduced doses of herbicides is associated with herbicide resistance. Worldwide almost 100 weed species are known to have evolved herbicide resistance, particularly against triazine herbicides used e.g. in maize pro- duction (Putwain and Mortimer 1989). Reports on herbicide (mecoprop and chlorsulfuron) resist- ant biotypes of Stellaria media (LUTMAN and Snow 1987,Reed et al. 1989, Kudsk et al. 1992) indicate that herbicide resistance is a potential problem for cereal production, too. Since the development ofresistance is suggested to evolve primarily against heavily used, repeatedly used, highly persistent and high-kill herbicides (Gressel and Segel 1982), the use of low herbi- cide doses could be an indirect way to delay the development of resistant weed populations. As shown in our experiments, low herbicide doses de- creased the selection pressure of herbicides against many weed species (111, V). Reduced herbicide doses could be intergrated into a strategy together with careful selection of /) crop rotation, ii) use of herbicide mixtures or a rotation of herbicides and Hi) other management practices known to affect weed infestation (Shaner et al. 1992). Although the overall recommendation of this study promotes continuous use of herbicides in spring cereals with adjusted doses, one-year breaks from herbicide application in circumstances of low weed infestation do not necessarily lead to harmful consequences in terms of yield and the subsequent weed infestation (111, IV, V), particularly in spring barley production. In practice, control measures should primarily be based on the prevailing weed infestation to protect the current crop, as the sea- sonal changes in weed infestation are not easy to manage or predict (c.f. Firbank 1989, Fykse 1993a). SUMMARY AND CONCLUSIONS Since the 1960 s herbicides, together with improved management practices, have made a considerable contribution to profitable cereal production in Fin- land. Studies detailed in this thesis on the pos- sibilities of reducing herbicide input in crop pro- duction were focused on spring cereals as i) spring cereals comprise over 50% of the cultivated field area in Finland, ii) cereals are known to be relat- ively competitive against weeds and finally, Hi) the majority of herbicides in Finland are applied to cereal fields. Therefore, an overall reduction of herbicide use in cereal production would have a significant economic and environmental impact on Finnish cereal production. Based on the results of the weed survey con- ducted in 252 fields in southern and central Finland, target weeds for chemical control in spring cereal fields are mainly annual broad-leaved species. The only frequent grass weed was Elymus repens. The weed infestation in unsprayed spring cereal fields 2averaged 170 plants m (median 124), and air-dry weight 320 kg ha 1 (median 183). Weed vegetation in individual cereal fields was often dominatedby a few species. Regular weed surveys are needed to follow the responses of weed floras to changes in agricultural practices. Ordination methods proved feasible for analyzing the associations between multispecies weed populations and interrelated environmental factors. Ordination analyses of these datarevealed that species composition of weed populations varied regionally and was affected by soil char- acteristics and crop management practices, par- ticularly by long-term use of herbicides. Phenoxy acid herbicides are commonly used to control broad-leaved weeds in Finland, and they were found to give sufficient control of the the prevailing weed vegetation, both in terms of yield response and long-term weed infestation. The re- duction of weed biomass with MCPA application 132 Agric. Sei. Finl. Suppl. No. 2 (1993) averaged 65%, and with other herbicides (mainly MCPA/dichlorprop or MCPA/mecoprop) 83%. In the field experiments the weed infestation (No. of plants m"2) at the time of herbicide applica- tion averaged 72% of the peak infestation recorded later in the growing season in unsprayed trial plots. However, spraying during the early growth stages of those weeds that emerge in the main flush is recommended, since the competitive ability of the crop is normally sufficient to suppress the growth of late-emerging weed seedlings. Reduction of therecommended rates of applica- tion seldom resulted in a considerable loss in efficacy, detectable yield loss or adverse effect on the subsequent weed infestation. This applied also to the new herbicide mixtures containing optically active isomers of dichlorprop and mecoprop, as well as to sulfonylurea herbicides. Considerable reductions up to 30-50% of the recommended her- bicide doses were often feasible, depending on the herbicide, weed species and the crop. If reduced herbicide doses are to be applied, a precise selec- tion of herbicides according to the weed species composition will be required. As a consequence of dose reduction, the dif- ferential sensitivity of weed species to a particular herbicide became evident. Even so, no drastic changes in weed composition were detected after continous use of reduced herbicide application rates for three years. The species composition of weed populations was affected more by crop rotation thanby herbicide type and dose. Weed biomass frequently comprised less than 5% of the total vegetative biomass of the crop stands. Consequently, the impact of weeds on crop yield was often very minor even in untreated plots, and statistically significant yield responses to herbicide treatments were seldom detected. There were a number of fields where the yield benefit did not cover the total cost of chemical weed control. Yieldresponses to herbicide doseproved to be low and this suggests that herbicide inputs in spring cereal production could be minimized. Since the dose adjustment proved to be site- specific, the responsibility for the final control re- sult should be transferred from the agrochemical companies and authorities to farmers. After all, farmers are responsible for timing and accuracy of herbicide application and for maintaining their spray-equipment. Computerized advisory systems provide a novel way to assist farmers in decision making. The present study provided valuable data for development of this type of advisory system suited to Finnish growing conditions. Threshold strategies based on weed densities proved to be inadequate for describing crop-weed interactions in Finnish spring cereal fields. There- fore, before recommending seasonal breaks in the use of herbicides, precision of thresholds has to be improved to minimize decision failures in weed control. Meanwhile, continuous use of herbicides with site-adjusted doses is suggested for spring cereal production. According to the results of this study, a realistic target should be 30% reduction in use ofherbicides at the national level. Implementation of such a reduction in Finland would correspond, at the current level of herbicide sales, to savings of ap- proximately FIM 20 million per annum. The level and the means of reduction suggested may not alleviate the public disquiet over the use of pesti- cides, but rather are meant to represent a reliable recommendation for sustainable weed manage- ment. 33 Agric. Sei. Finl. Suppl. No. 2 (1993) REFERENCES 0 Aamisepp, A. 1984. Behovsprövad ogräsbekämpning i vårsäd. Slutrapport. Proc. 25:e svenska ogräskonfer- ensen, Ogräs och ogräsbekämpning. Rapp. p. 33-47. & Wallgren, B. 1979. Ogräs i stråsäd. Verkan av kemisk ogräsbekämpning och andra odlingsåtgärder, 1950-1978. Akt. frän Lantbr.univ. 280. Uppsala. 15 p. Albrecht, H. & Bachthaler, G. 1988. Die Segetalflora zweier bayerischer Ackerstandorte 1986/87im Vergleich zu Untersuchungsergebnissen von 1955/56bzw. 1965.Z. Pfl.krankh. Pfl.schutz, Sonderh. 11: 163-174. Andersson, B. 1984. Utsädesmängder och MCPA-doser i vårkom. Proc. 25:e svenska ogräskonf.. Ogräs och ogräs- bekämpning. Rapp. p. 49-58. 1986. Influence of crop density and spacing on weed competition and grain yield in wheat and barley. Proc. EWRS Symp. Economic Weed Control, p, 121-128. Andreasen, C. 1990. Ukrudtsartemes forekomst på danske sasdskiftemarker. Summary: The occurrence of weed species in Danish arable fields. The Royal Vet. and Agric. Univ., Copenhagen. 53 p. (Diss.). —, Haas, H, & Streibig, J.C. 1989. Floraaendringer, fore- Ipbig status. Proc. 6. Danske Plantev.konf. Ukrudt. p. 125-133. —, Streibig, J.C. & Haas, H. 1991. Soil properties affecting the weed distribution of37 weed species in Danish fields. Weed Res, 31: 181-187. Assémat, L. 1992. Stability of crop and weed growth rates. Proc. IXerne Colloque Int. sur la Biologic des Mauvaises Herbes, Dijon, p. 361-367. Auld, B.A. & Tisdell, C.A. 1988. Influence of spatial dis- tribution of weeds on crop yield loss. Plant Prot. Quart. 3: 81. Baandrup, M. & Ballegaard, T, 1989. Three years field experience with an advisory computer system applying factor-adjusted doses. Proc. Brighton Crop Prot. Conf. - Weeds. 2: 555-560. Bachthaler, G. 1969. Entwicklung der Unkrautflora in Deutschland in Abhängigkeit von den veränderten Kul- turmethoden, Angew, Bot. 43: 59-69. BAYER 1992.Important crops of the world and their weeds. 2nd ed. BAYER AG. Leverkusen. 1681 p. Beer, E. 1979. Ermittlung der Bekämpfungsschwellen und wirtschaftlichen Schadensschwellen von monokotylen und dikotylen Unkräutem in Winterweizen und Winter- gerste anhand von Daten aus der amtlichen Mittel- priifung. Univ. Gottingen. 139 p. (Diss.). Behrendt, S. 1986. Einige Gedanken zur Unkrautbekämp- fung in der Zukunft. Proc. EWRS Symp. Economic Weed Control, p. 407-417. 11 Several papers referred to in the introductory chapter and in the original articles have been printed in the sympo- sium proceedings of the European Weed Research Society (EWRS). Proceedings are available from: CABO/ DLO, Dept. Weed Sci., Bornsesteeg 65, NL-6700 Wageningen, The Netherlands. Bernson, V. 1988. Regulation of pesticides in Sweden. Proc. Brighton Crop Prot. Conf. - Pests and Diseases. 3: 1059- 1064. Beyer, E.M. 1991. Crop protection - meeting the challenge. Proc. Brighton Crop Prot. Conf. - Weeds. 1: 3-22. Birrs, H.J.B. & Austin, H.A. 1992. An annotated biblio- graphy of canonical correspondence analysis and related constrained ordination methods 1986-1991. Univ, Ber- gen, Norway. 29 p. Bleiholder, H. & Nuyken, W. 1986. Neue Ansätze zur Darstellung und Interpretation des Zusammenhanges zwischen dem Deckungsgrad der Unkräuter und dem Ertrag von Getreide. Proc. EWRS Symp. Economic Weed Control, p. 61-68. Callauch, R. 1981. Vergleich der Segetalvegetation auf ’konventionell’ und ’biologisch’ bewirtschafteten Ack- em in SO-Niedersachsen. Z. Pfl.krankh. Pfl.schutz, Son- derh. 9: 85-95. Catizone, P., Tedeschi, M. & Saldoni, G. 1990. Influence of crop management on weed populations and wheat yield. Proc. EWRS Symp. Integrated Weed Management in Cereals, p. 119-125. Chancellor, R.J. 1985. Changes in the weed flora of an arable field cultivated for 20 years. J, Appi. Ecol. 22: 491-501, Christensen, S. 1993, Weed suppression in cereal varieties. Statens Planteavlforspk. SP rapport Nr. 1/93. 104 p. (Diss.). —, Streibig, J.C. & Haas, H. 1990. Interaction between herbicide activity and weed suppression by spring barley varieties. Proc. EWRS Symp. Integrated Weed Manage- ment in Cereals, p. 367-374. Combellack, J.H. 1992a. The importance of weeds and ways of reducing concerns about methods for their control. Proc. First Int. Weed Control Congr., Melbourne. 1: 43- 63. —1992 b. An appraisal of opportunities to reduce herbicide use. Crop Prot. Quart. 7; 66-69. Courtney, A.D. 1991. The role of competition in developing an appropriate rate strategy for weed control in spring barley. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 1217-1224. & Johnston, R.T. 1986. An assessment of weed popula- tions and yield responses on spring barley subjected to a programme of reduced herbicide usage. Proc. EWRS Symp. Economic Weed Control, p. 301-308. Cousens, R., Brain, P., O’Donovan, J.T. & O’Sullivan, A. 1987. The use of biologically realistic equations to de- scribe the effects of weed density and relative time of emergence on crop yield. Weed Sci. 35: 720-725. Cussans, G.W. 1980. Strategic planning forweed control - a researcher’s view. Proc. 1980 Brit. Crop Prot. Conf. - Weeds. 3: 823-831. 1992. Identifying ways of optimizing herbicide use in crops. Proc. First Int, Weed Control Congr., Melbourne. 1:208-214. 34 Agric. Sei. Fint. Sappi. No. 2 (1993) -, Cousens, R.D. & Wilson, B.J. 1986. Thresholds for weed control - the concepts and their application. Proc. EWRS Symp. Economic Weed Control, p. 253-260. -, Moss, S.R., Pollard, F. & Wilson, B.J. 1979. Studies on the effects of tillage on annual weed populations. Proc. EWRS Symp. Influence of Different Factors on the De- velopment and Control of Weeds, p. 115-122. Dale, M.R.T. & Thomas, A.G. 1987. The structure of weed communities in Saskatchewan fields. Weed Sci. 35: 348- 355. -, Thomas, A.G. & John,E.A. 1992. Environmental factors including management practices as correlates of weed community composition in spring seeded crops. Can. J. Botany. 70: 1931-1939. Davies, D.H.K., Proven, M.J., Courtney, A.D. & Lawson, H.M, 1993. Comparison of the use of weed thresholds and routine herbicide use at reduced rate on the econom- ics of cereal production in the rotation. Proc. Bth EWRS Symp. Quantitative Approaches in Weed and Herbicide Research. 2: 747-754. -, Whiting, A.J. & Whytock, G.M. 1989. Yield responses to herbicide use and weed levels in winter wheat and spring barley in Scottish trials and consequences for economic models. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 955-960. Debaeke, P. 1988. Dynamique de quelques dicotylédones adventices en culture de céréale. 11. Survie, floraison et fructification. Weed Res. 28: 265-279. 1990. Effets de systemes de culture diversement intensi- fies sur la composition et la dynamique de la flore adven- tice des cereales d’hiver. Proc. EWRS Symp. Integrated Weed Management in Cereals, p. 143-152. Devine, M.D. 1988. Environmental influences in herbicide performance: A critical evaluation of experimental tech- niques. Proc. EWRS Symp. Factors Affecting Herbicidal Activity and Selectivity, p. 219-226. Dock Gustavsson A.-M. 1986. Relative growth rate and competitive ability of annual weeds. Proc. EWRS Symp. Economic Weed Control, p. 105-112. Edwards, C.A. & Regnier, E.E. 1989. Designing integrated low-input farming systems to achieve effective weed contol. Proc. Brighton Crop Prot. Conf. - Weeds. 2: 585-590. Eggers, T. 1984. Wandel der Unkrautvegetation der Äcker. Proc. 44. Deutsche Pfl.sch.-Tagung. Mitt. Biol. Bundes- anst. Land- u. Forstw. Heft 223: 62. 1987. Environmental impact of chemical weed control in arable fields in the Federal Republic of Germany. Proc. 1987 Brit. Crop Prot. Conf. - Weeds. 1: 267-275. Elliott, J.G. 1978.The economic objectives of weed control in cereals. Proc. 1978 Brit. Crop. Prot. Conf. - Weeds. 3: 829-841. Engström, E. 1978. Control of dicots in spring cereals with MCPA, MCPA+dichlorprop and different doses of Oxy- tril 4. Proc. 19th Swed. Weed Conf. - Weeds and Weed Control, p. 858-868. Erviö, L.-R. 1981. The emergence of weeds in the field. Ann. Agric. Fenn. 20: 292-303. 1983. Competition between barley and annual weeds at different sowing densities. Ann. Agric. Perm. 22: 232- 239. & Hiivola, S.-L. 1986. Herbisidien käytön vähentäminen viljakasvustoissa. Maatalouden tutkimuskeskus. Tiedote 8/86. p. 28-42. & Salonen, J. 1987. Changes in the weed population of spring cereals inFinland. Ann. Agric. Fenn. 26:201-226. —, Tanskanen, T. & Salonen, J. 1991. Profitability of chemical weed control in spring cereals. Ann. Agric. Fenn. 30: 199-206. Espeby, L. 1989. Germination of weed seeds and competition in stands of weed and barley. Crop Prod. Sci. 6. Uppsala, Sweden. 172 p. (Diss.). Evans, S.A. 1968. Project N.A.E. 30 - Results and conclu- sions. Proc. 9th Brit. Weed Contr. Conf. 3: 1165-1166. Firbank, L.G. 1989. Forecasting weed infestations - the desirable and the possible. Proc. Brighton Crop Prot. Conf. - Weeds, p. 567-572. Fogelfors, H. 1974. Några ogräsarters utveckling under skilda ljusföhällanden och konkurrensförmåga i kom- beständ. Lantbr.högsk. Inst. Växtodl., Uppsala. 74 p. 1979. Floraförändringar i odlingslandskapet. Sveriges Lantbr.univ. Inst. Ekol. and Miljöv., Rapp. 5. 66 p. 1990. Different doses of herbicide forcontrol of weeds in cereals - final report from the long-term series. Proc. 31st Swedish Crop Prot. Conf. p. 139-151. 1992. Erfarenheter från försöksverksamheten med dos- nycklar. Proc. 33:e svenska växtskyddskonferensen. Ogräs och ogräsbekämpning. Rapp. p. 13-20. Froud-Williams, R.J., Drennan, D.S. & Chancellor, R.J. 1983. Influence of cultivation regime on weed floras of arable cropping systems. J. Appi. Ecol. 20: 187-197. Fryer, J.D. & Chancellor, R.J. 1970. Evidence of changing weed populations in arable land. Proc. 10th British Weed Control Conf. 3: 958-964, Fykse, H. 1991a. Skadetersklar for ugras. Norsk Landbr.forskn., Suppl. 10: 40-43. —1991 b. Skadetersklar for ugras i värkorn. Informa- sjonsmpte i plantevern 1991. Faginfo 2/93. p. 165-173. —1993 a. Dynamics of weeds in long-term experiments in spring cereals. Proc. EWRS Symp. Quantitative Ap- proaches in Weed and Herbicide Research. 2: 689-696. —1993 b. Ugrassituasjonen i vårkorn ved ulike bekjemp- ningsstrategier. Informasjonsmpte i plantevern 1993. Faginfo 3/93. p. 172-180. —1993 c. Råd via TELEVIS om sprpyting mot ugras. Infor- masjonsmote i plantevern 1993. Faginfo 3/93. p. 143- 146. Garburo, W. 1974. Untersuchungen zur Ermittlung der ökonomischen Schadensschwellen und der Bekämp- fungsschwellen von Unkräutem in Getreide. Univ. Göt- tingen. 133p. (Diss.). Garvert, U., Wagner, J., Kochs, H.J., Heyland, K.U. & Steets, R, 1990. HERBEXPERT - ein computergestiltz- tes System zur schlagspezifischen Herbizidberatung im Getreidebau. Z. Pfl.krankh. Pfl.schutz, Sonderh. 12: 191- 195. Gauch, H.G. 1982. Multivariate analysis in community eco- logy. Cambridge University Press. Cambridge. 298 p. 35 Agric. Sei. Finl. Suppl. No. 2 (1993) Gerowitt, B. 1992. Dreijährige Versuche zur Anwendung eines computergestiitzten Entscheidungsmodells zur Un- krautbekämpfung nach Schadensschwellen im Winter- weizen. Z, Pfl.krankh, Pfl.schutz, Sonderh. 13: 301-310. —, Bodendörfer, H. & Heitefuss, R. 1984. Zur Wirt- schaftlichkeit des Herbizideinsatzes im Getreide - Auswertung von Versuchen des Pflanzenschutzdienstes den Jahren 1977-81. Z. Pfl.krankh. Pfl.schutz, Sonderh. 10: 127-135. & Heitefuss, R. 1990. Weed economic thresholds in cereals in the Federal Republic of Germany. Crop Prot. 9: 323-331. —, Heitefuss, R., Fischer, A. & Reiner, L. 1986. Versuche zur Erarbeitung von Kriterien fiir EDV-gestiitzte En- scheidungshilfen zur Unkrautbekämpfung nach Schadensschwellen im Wintergetreide. Proc. EWRS Symp. Economic Weed Control, p. 351-360. Granström, B. 1962. Studier över ogräs i vårsådda grödor. Kungl. Lantbr.högsk. och St. Lantbr.försök, Medd. Nr. 130, Uppsala. 188 p. & AlmoArd, G. 1955. Studier över den svenska ogräs- floran. Stat. Jordbr.försök, Meddel, nr 56, 22 p. Gressel, J. & Segel,L.A. 1982. Interrelating factors control- ling the rate of appearance of resistance: the outlook for the future. In: Leßaron, H.M. & Gressel, J. (eds.). Herbi- cide resistance in plants, J. Wiley & Sons. p. 325-347. Gummesson, G. 1975. Ogräsförekomsten i stråsäd. Proc. 16:e svenska ogräskonf. Ogräs och ogräsbekämpn. 1. Före- drag. p, K2l-K37. & Fooelfors, H. 1990. Möjligheter till minskad bekämpning av ogräs i stråsäd. Proc. 31:a svenska växtskyddskonf., Skadedj. och Växtsjukd. p. 60-68. Haas, H. & Streibig, J.C. 1982. Changing patterns of weed distribution as a result of herbicide use and other agronomic factors. In: Leßaron, H.M. & Gressel, J. (eds.). Herbicide resistance in plants. J. Wiley & Sons. p. 57-79. Håkansson, S. 1975. Grundläggande växtodlingsfrågor. 1. Inflytande av utsädesmängden och utsädets horisontella fördelning pä utveckling och produktion i kortvariga växtbestånd. Sveriges Lantbr.univ, Rapp. o. Avhandl. 33. Uppsala. 192 p. 1983a. Seasonal variation in the emergence of annual weeds - an introductory investigation in Sweden. Weed Res. 23: 313-324. —1983 b. Competition and production in short-lived crop- weed stands. Swed. Univ. Agric. Sci., Dept. Plant Husb., Report 127. Uppsala. 85 p. 1986. Competitionbetween crops and weeds - influencing factors, experimental methods and research needs. Proc. EWRS Symp. Economic Weed Control, p. 49-60. 1991. Growth and competition in plant stands. Swedish Univ. Agric. Sci., Crop Prod. Sci. No 12. Uppsala. 241 p. 1992. Germination of weed seeds in different seasons. Proc. IXe Colloque Int. sur la Biologie des Mauvaises Herbes, Dijon, p. 45-54. Hallgren, E. 1988. Influence of different factors on the effect of spraying cereals in the spring with Oxitril 4 as regards weeds and grain yield. 2. Influence of crop, de- velopmental stage, prevailing conditions, geographical and climatic situation, soil type, organic content and N-rate on weed stand and effect on weeds. Composition of weed stand at different relative yields. Proc. 29th Swedish Weed Contr. Conf., Weeds and Weed Control, p. 39-74. —1993 a. Förändras ogräsfloran och verkan av ett ogräs- medel (Oxitril 4) med tiden. Proc. 34:e svenska växt- skyddskonferensen. Ogräs och ogräsbekämpn. Rapp. p. 15-38. —1993 b. Verkan av några ogräsmedel mot olika tvåhjärt- bladiga ogräsarter vid skilda doser och behandlingstid- punkter, Summary: Effects of some herbicides or mix- tures of herbicides on individual dicot weed species at different doses and developmental stages. Sver. Lantbr.univ. Växtodling 44. Uppsala. 85 p. Hämet-Ahti, L., Suominen, J., Ulvinen, T., Uotila, P. & Vuokko, S. (eds.). 1984. Retkeilykasvio. Suomen Luon- nonsuojelun Tuki, Helsinki. 544 p. Hance, R.J. & Holly, K. (eds.). 1990. Weed control hand- book: Principles. Brit. Crop Prot. Council, Blackwell Sci. Pubi. 582 p. Heitefuss, R. 1986. Auswirkungen von Unkräutern und Massnahmen der Unkrautbekämpfung auf andere Kriterien als den Ertrag. Proc. EWRS Symp. Economic Weed Control, p. 189-199. —, Gerowitt, B. & Wahmhoff, W. 1987. Development and implementation ofweed economic thresholds in the F.R. Germany. Proc. Brit. Crop Prot. Conf. - Weeds. 3: 1025- 1034. Hill, M.O. 1973. Reciprocal averaging: an eigenvector method of ordination. J. Ecol. 61: 237-249. & Gauch, H.G. 1980. Detrended correspondence ana- lysis, an improved ordination technique. Vegetatio 42:47-58. Hilli, A. 1948. Rikkaruohomme ja niiden taloudellinen merkitys. Maatal. ja Koetoim. 3: 154-168. Hintzsche, E. 1990. Zur Wirkung acker-und pflanzenbau- licher Massnahmen auf einzelne Unkrautarten bzw. Unkrautartengruppen. Z. Pfl.krankh. Pfl.schutz, Sonderh. 12: 211-218. Holm, L.G., Plucknett, D.L., Pancho, J.V. & Herberger, J.P. 1977. The world’s worst weeds, distribution and biology. The Univ. Press of Hawaii, Honolulu. 609 p. Holzner, W. & Immonen, R. 1982. The agrestal weed flora and vegetation of the world: examples and aspects, Europe: an overview. In: Holzner, W. & Numata, M. (eds.). Biology and ecology ofweeds. Dr. W. Junk Pubi., The Hague, p. 203-226. Hornig, H. 1972. Unkrautbekämpfung mit reduzierten Auf- wandmengen in Getreide. Gesunde Pfl. 24: 145-150. Hughes, R.G. 1966. Project N.A.E. 30 "The efficiency of weed control in cereals on commercial farms". Proc. Bth Brit. Weed Control Conf. p. 200-205. Hume, L. 1987. Long-term effects of 2,4-D application on plants. I. Effects on the weed community in a wheat crop. Can. J. Botany 65: 2530-2536. Hurle, K. 1974.Effects of long-term weed control measures on viable weed seeds in the soil. Proc. British Crop Prot. Conf.-Weeds. 3: 1145-1152. 36 Agric. Sd. Finl. Suppl. No. 2 (1993) Hynninen, E.-L. & Blomqvist, H. 1991. Torjunta-aineiden myynti Suomessa. [Summary: Sales of pesticides in Fin- land in 1990], KemiaKemi 18: 506-509, & Blomqvist, H. 1992.Pesticide sales in Finland in 1991. Kemia-Kemi 19: 563-565. & Blomqvist, H. 1993. Pesticide sales in Finland 1992. Kemia-Kemi 20: 535-537. Jennéus, B, 1992. Erfarenheter från ogräsbekämpning i de- monstrationsodlingar med hjälp av dosnyckhtr. Proc. 33:e svenska växtskyddskonferensen, Ogräs och ogräs- bekämpning. Rapp. p. 21-32. Jensen, P.K. 1985. A review of yield responses to weed control in one thousand spring barley experiments. Proc. British Crop Prot. Conf. - Weeds. 2: 687-692. 1987. Skadetasrskler for tokimbladet ukrudt i värbyg. Tidsskr. Planteavl. 90: 369-376. —1991 a. Weed size hierarchies in Denmark. Weed Res. 31: 1-7. —1991 b. Behov og ekonomi ved ukrudtsbektempelse i landbrugsafgrpder. Bilag til Moder om plantevem 1991, Landbrugsafgroder. p. 14-16. Jokinen, K. 1991. Competition and yield advantage in bar- ley-barley and barley-oats mixtures. J. Agric. Sci. Finl. 63: 255-284. (Diss.). Jonoman, R.H.G., Ter Braak, C.J.F. & Van Tonoeren, O.F.R. 1987. Data analysis in community and landscape ecology. Pudoc, Wageningen. 299 p. Junnila, S. 1990. Influence of spraying time on herbicide efficacy in spring cereals. Proc. EWRS Symp. Integrated Weed Management in Cereals, p. 375-382. 1993. Över 10-års försöksarbete med S.U.-herbicidema i Finland, Proc. DuPont-seminarium i Stockholm. Mimeogr. 8 p. (Available at the Agric. Res. Centre of Finland, Jokioinen). Kallio-Mannila, K., Raatikainen, T. & Raatikainen, M. 1984. Rikkaruohot vähentyneet - torjuntaa muutettava. Koetoim. jaKäyt. 41:53. Kees, H. 1986. Einfluss zehnjähriger Unkrautbekämpfung mit 4 unterschiedlichen Intensitätsstufen unter Beriick- sichtigung der wirtschaftlichen Schadensschwelle auf Unkrautflora und Unkrautsamenvorrat im Boden. Proc. EWRS Symp. Economic Weed Control, p. 399-406. Kirby, E.J.M., Siddique, K.H.M., Perry, M.W., Kaeseha- gen, D. & Stern, W.R. 1989. Variation in spikelet initia- tion and ear development of old and modern Australian wheat varieties. Field Crops Res. 20: 113-128. Knab, W, & Hurle, K. 1986. Einfluss derGrundbodenarbei- tung auf die Verunkrautung - ein Beitrag zur Prognose der Verunkrautung. Proc. EWRS Symp. Economic Weed Control, p. 309-316. Korsmo, E. 1925. Ugress i nutidens jordbruk. J.W. Cappe- lens forlag. Oslo. 694 p. Kropff, M.J. 1988. Modelling the effects of weeds on crop production. Weed Res. 28; 465-471. & Spitters, C.J.T. 1991. A simple model of crop loss by weed competition from early observations on relative leaf area of the weed. Weed Res. 31: 97-105. Kudsk, P. 1989.Experiences with reduced herbicide doses in Denmark and the development of the concept of factor- adjusted doses. Proc. Brighton Crop Prot. Conf. - Weeds. 2: 545-554. & Kristensen, J.L. 1992. Effect of environmental factors on herbicide performance. Proc. First Int. Weed Contr. Congr., Melbourne. 1: 173-186. —, Mathiasen, S.K. & Petersen, E.F. 1992. Sulfonylureare- sistens i fuglegraes. Proc. 9. Danske Plantev.konf., - Ukrudt.p. 147-156. Lawson, H.M., Wright, G.McN., Davies, D.H.K. & Whit- ing, A.J. 1992. The effects of reduced herbicide strategies on the weed flora in cereals in Scotland: an interim report. Proc. IXe Colloque Int. sur la Biologie des Mauvaises Herbes, Dijon, p. 269-278. Leßaron, H.M. & Gressel, J. (eds.). 1982. Herbicide resist- ance in plants. John Wiley & Sons, Inc. New York. 401 p. Légére, A., Lemieux, C., Simard, R.R. & Lapierre, C. 1993. Response of weed communities to fertility and tillage. Proc. Bth EWRS Symp. Quantitative Aspects of Weed and Herbicide Research, p. 41-48. —-, Samson, N., Lemieux, C. & Rioux, R. 1990. Effects of weed management and reduced tillage on weed popula- tions and barley yields. Proc. EWRS Symp, Integrated Weed Management in Cereals, p. 111-118. Littell, R.C., Freund, R.J. & Spector, P.C. 1991, SAS® System for Linear Models, 3rd Edit. SAS Institute Inc. Cary, NC. 329 p. Lomakka, L. 1990. Minimerad kemisk bekämpning av ogräs vid ensidig kornodling i Norra Sverige. Proc. 31 :a svens- ka växtskyddskonferensen. Ogräs och ogräsbekämpning. Rapp. p. 27-38. Lotz, L.A.P., Kropff, M.J. & Groeneveld, R.M.W. 1990. Modelling weed competition and yield losses to study the effect of omission of herbicides in winter wheat. Neth. J. Agric. Sci. 38; 711-718. —, Kropff, M.J. & Groeneveld, R.M.W. 1993.The relative leaf cover model tested for practice. Proc. Bth EWRS Symp. Quantitative Aspects of Weed and Herbicide Re- search. 2: 793-798. Lutman, P.J.W. 1992. Prediction of the competitive effects of weeds in the yields of several spring-sown arable crops. Proc. IXe Colloque Int. sur la Biologie des Mauv. Herbes, Dijon, p. 337-344. - & Snow, H.S, 1987.Further investigations into the resist- ance of Chickweed (Stellaria media) to mecoprop. Proc. 1987 British Crop Prot. Conf. - Weeds. 3: 901- 908. Mahn, E.G. 1984. The influence of different nitrogen levels on the productivity and structural changes of weed com- munities in agroecosystems. Proc. 7th Int. Symp. on Weed Biology and Systematics, Paris, p. 421-429. 1992. Ackerunkräuter - ihre bedeutung im Agro-Okosys- tem aus ökologischer Sicht. Z. Pfl.krankh. Pfl.schutz, Sonderh. 12: 21-30. & Helmecke, K. 1979. Effects of herbicide treatment on the structure and functioning of agro-ecosystems; 11. Structural changes in the plant community after the appli- cation of herbicides over several years. AgroEcosyst. 5: 159-179. 37 Agric. Sei. Fin!. Suppl. No. 2 (1993) Markkula, M., Tiittanen, K. & Vasarainen, A. 1990. Tor- junta-aineet maa-ja metsätaloudessa 1953-1987. Maata- louden tutkimuskeskus, Tiedote 2/90, 58 p. Marshall, E.J.P. 1988.Field-scale estimates of grass weed populations in arable land. Weed Res. 28: 191-198. Meinlschmidt,E. & Karch, K. 1992.Weed growth suppres- sion by means of repeated applications of reduced herbi- cide doses. Proc. IXe Colloque Int. sur la Biologic des Mauvaises Herbes, Dijon, p. 369-376. Mela, T. & Paatela, J. 1974. Grain yield of spring wheat and oats as affected by population density. Ann. Agric. Fenn. 13: 161-167. Mikkelsen, V.M. & Laursen, F. 1966. Markukrudtet i Dan- mark omkring 1960. Botan. Tidsskr. 62: 1-26. Ministry of Agriculture 1990. Multi-year crop protection plan. Essentials, Min. Agric., Dept. Nature Management and Fisheries, Amsterdam. 14 p. Morgan, W. 1992. Strategies to reduce dependence on herbicides. Proc. First Int. Weed Contr. Congr., Mel- bourne. 1: 289-294. Moss, S.R. 1985. The influence of crop variety and seed rate on Alopecurus myosuroides competition in winter cer- eals. Proc. Brit. Crop. Prot. Conf. - Weeds. 2: 701-708. Mukula, J. 1974. Weed competition in spring cereals in Finland. Forskn. o. Fors. Landbr. 25: 585-592. —, Raatikainen, M., Lallukka, R. & Raatikainen, T. 1969. Composition of weed flora in spring cereals in Finland. Ann. Agric. Fenn. 8: 59-109. & Rantanen, O. 1987. Climatic risks to the yield and quality of field crops in Finland. 1. Basic facts about Finnish field crops production. Ann. Agric. Fenn. 26: 1-18. & Ruuttunen, E. 1969. Chemical weed control in Finland in 1887-1965.Ann. Agric. Fenn. 8, Suppl. 1: 1-45. Murali, N.S. & Secher, B.J.M. 1991. Status on the com- puter-based plant protection systems in Denmark. Proc. Workshop on computer-based plant prot. advisory sys- tems. Danish J. PI. and Soil Sci. - Spec, series. Report no. S 2161. p. 7-10. Neururer, H. 1976. Ökonomische Schadensschwelle und tolerierbare Verunkrautungsstärke in der Unkrauts- bekämpfung. Land u. Forstw. Forsch. in Österreich, Band 7: 143-152. Niemann, P. 1990. Zur Häufigkeit von Bestandesliicken und deren Bedeutung fur die Verunkrautung. Z. Pfl.krankh. Pfl.schutz, Sonderh. 12:59-69. Paatela, J. 1953, Maammeheinänurmien botaanisesta koos- tumuksesta. Summary: On the botanical composition of the tame-hayfields in Finland. Acta Agr. Fenn. 79: 1-128. Pallutt, B. 1988. Beträge zur integrierten Unkrautbekämp- fung im Getreideanbau. Akad. Landw, der DDR, Berlin. 129p. (Diss.). & Roder, W. 1992. Zur Verbesserung der Vorher- sagegenauigkeit von unkrautbedingten Komertragsver- lusten bei Wintergetreide. Z. Pfl.krankh. Pfl.schutz, Son- derh. 13; 129-137. Pedersen, H.E. 1978. Use of reduced doses of herbicides in barley. Proc. 19th Swedish Weed Control Conf., Weeds and Weed Control, p. 843-849. Pedersen, J.O. & Rasmussen, LA. 1990. Herbiciders in- flydelse pä ukrudets fröproduktion. Proc. 7, Danske Plan- tev.konf, Ukrudt. p. 73-83. Peltonen-Sainio, P. & Järvinen, P. 1993.Effects of seeding rate on growth duration and accumulation and partition- ing ofdry-matter in oats. Mimeogr. (Available at the Inst. Crop Sci., Univ. of Helsinki). Petersen, H.I. 1943. Undersdgelserover Ukrudstarter i Rug- marken. Ugeskr. forLandm. 88: 492-497. Post, B.J. 1986. Factors of influence on the development of an arable weed vegetation. Proc. EWRS Symp. Economic Weed Control, p. 317-325. Proven, M.J., Courtney, A., Picton, J., Davies, D.H.K. & Whiting, A.J. 1991. Cost-effectiveness of weed control in cereals - systems based on thresholds and reduced rates. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 1201- 1208. Putwain, P.D. & Mortimer, A.M. 1989, The resistance of weeds toherbicides: rational approaches for containment of a growing problem. Proc. Brighton Crop Prot. Conf. - Weeds. 1: 285-294. Raatikainen, M. & Raatikainen, T. 1975. Heinänurmien sato, kasvilajikoostumus ja sen muutokset. Summary: Yield, composition and dynamics of flora in grassland for hay in Finland. Ann. Agric. Fenn. 14: 57-191. —, Raatikainen, T. & Mukula, J. 1978. Weed species, frequencies and densities in winter cereals in Finland. Ann. Agric. Fenn. 17: 115-142. Rademacher, B. & Koch, W. 1972. Kulturanbedingte Veränderungen in der Unkrautflora eines Feldes von 1956 bis 1971. Z. Pfl.krankh. Pfl.schutz, Sonderh. 6: 149-160. —, Koch, W. & Hurle, K. 1970. Changes in the weed flora as the result of continuous cropping of cereals and the annual use of the same weed control measures since 1956. Proc. 10th Brighton Weed Conf. p. 1-6. Rasmussen, LA. 1993. Seed production of Chenopodium album in spring barley sprayed with different herbicides in normal to very low doses. Proc. EWRS Symp. Quant- itative Approach in Weed and Herbicide Research. 2: 639-646. Reed, W.T., Saladini, J.L., Cotterman, J.C.,Primiani, M.M. & Saari, L.L. 1989. Resistance in weeds to sulfonylurea herbicides. Proc. Brighton Crop Prot. Conf. - Weeds. 1: 295-300. Reuss, H.-U. 1981. Untersuchung des Einflusses produk- tionstechnischer und ökologischer Faktoren auf die quan- titative und qualitative Veränderung der standörtlichen Unkrautflora auf Ackerland. Technischen Universitäl Miinchen. 116p. (Diss.). Richards, M.C. 1989. Crop competitiveness as an aid to weed control. Proc. Brighton Crop Prot. Conf. - Weeds. 2: 573-578. & Davies, D.H.K. 1991. Potential for reducing herbicide inputs/rates with more competitive cereal cultivars. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 1233-1240. Roberts, H.A. & Potter, E.M. 1980. Emergence patterns of weed seedlings in relation to cultivation and rainfall. Weed Res. 20: 377-386. 38 Agric. Sei. Fin!. Suppl. No. 2 (1993) Salonen, J. 1992a. Viljanviljelyn kannattava rikkakasvintor- junta. Maataloustieteen päivät. Suomen Maatal.tiet. seur. tied. No 16. p. 60-67. —1992 b. Distribution of nitrogen between crop and weeds in spring cereals. Acta Agric. Scand., Sect B, Soil and Plant Sci. 42:218-223. —1993. Dynamics ofweed infestation inspring cereals with special reference to the use of reduced herbicide doses. Proc. Bth EWRS Symp. Quantitative Approaches in Weed and Herbicide Research. 2: 715-722. & Erviö, L.-R. 1988. Changes in the weed infestation of spring cereals in Finland between the 1960’s and 1980’s. Proc. Ville Colloque Int. sur la biologie, Ideologic et la systématique des Mauvaises Herbes, Dijon, p. 439-444. SAS Institute Inc. 1985. SAS® User’s Guide: Basics, Vers. 5 ed. SAS Inst. Inc. Cary, NC. 1290p. 1990. SAS/STAT® User’s Guide, Vers. 6,4th ed., Vol 2. SAS Inst. Inc. Cary, NC. 848 p. Scragg, E.B, 1980. Cost-effective weed control in spring barley in the north of Scotland. Proc. British Crop Prot. Conf. - Weeds. 1:69-76. Shaner, A., Sinha, A. & Braddock, R. 1992. Designing strategies to delay development of resistance to herbi- cides. Proc. First Int. Weed Control Congr., Melbourne. 1:236-239. Smilauer, P. 1990. CANODRAW- a companion program to CANOCO for publication-quality graphical output. Mi- crocomputer Power, Ithaca, New York. 33 p. Speranza, M., Govi, G. & Catizone, P. 1990.Crop manage- ment and weed flora of cereals in Italy. Proc. EWRS Symp. Integrated Weed Management in Cereals, p, 163- 171. Spitters, C.J.T. & Van Den Beroh, J.P. 1982. Competition between crop and weeds: A system approach. In: Holz- ner, W. & Numata, M. (eds.). Biology and ecology of weeds. Dr. W. Junk Pubi., The Hague, p. 137-148. Squires, N.R.W., Radtke, M. & Hunt, S. 1987. New formu- lations of phenoxypropionic herbicides containing only the herbicidally active isomer for the control of broad- leaved weeds in cereals. Proc. British Crop Prot. Conf. - Weeds. 1: 225-231. Streibio, J.C. 1979. Numerical methods illustrating the phy- tosociology of crops in relation to weed flora. J. Appi. Ecol. 16: 577-587. 1983. Ukrudtssprojtning og merudbytte i korn. Ugeskrift forJordbr. 40: 811-816. 1988.Herbicide bioassay. Weed Res. 28: 479-484. 1992.Quantitative assessment of herbicide phytotoxicity with dilution assay. Royal Vet. and Agric. Univ. Copen- hagen. 98 p. (Diss.). , Gottschau, A., Dennis, 8., Haas, H. & Moloaard, P. 1984. Soil properties affecting weed distribution. Proc. 7th International Symposium on Weed Biology, Ecology and Systematics, Paris, p. 147-154. & Haas, H. 1979. Zusammensetzung der dänischen Un- krautflora und deren Veränderung in den letzten 60 Jahren. Proc. EWRS Symp. Influence ofDifferent Factors on the Development and Control of Weeds, p. 273-280. Stryckers, J.M.T. 1979. Veränderungen in der (Un)kraut- flora durch Herbizidanwendung. Proc. EWRS Symp. In- fluence of Different Factors on the Development and Control of Weeds, p. 25-38. Stubsjoen, R. 1991. Ny landbrugspolitikk - Handlingsplan for reduceret bruk av plantevemsmidler. Informas- jonsmpte i Plantevem 1991. Faginfo 2/91. p. 23-28. Ter Braak, C.J. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradi- ent analysis. Ecology. 67: 1167-1179. —1987 a. CANOCO - a FORTRAN program for Canonical Community Ordination. Microcomputer Power, Ithaca, New York. 95 p. —1987 b. Unimodal models to relate species to environ- ment. Agric. Univ. Wageningen. 151 p. (Diss.). Thomson, W.J. & Stokes, D.T. 1985. Cereal apex develop- ment in relation to crop management. Aspects Appi. Biol. 10: 415-429. Thonke, K.E. 1978. Influence of growth factors when using reduced doses of hormone herbicides in barley. Proc. 19th Swedish Weed Conf. Weeds and Weed Control, p. 851-857. 1986. Muligheder for anvendelse af reducerede dose- ringer af herbicider. Proc. 3. Danske Plantev.konf. - Ukrudt. p. 117-124. 1991. Political and practical approaches in Scandinavia to reduce herbicide inputs. Proc. Brighton Crop Prot. Conf. -Weeds. 3: 1183-1190. Thornton, P.K., Fawcett, R.H., Dent, J.B. & Perkins, T.J. 1990. Spatial weed distributions and economic thresh- olds for weed control. Crop Prot. 9: 337-342. Tiittanen, K. & Blomqvist, H. 1981. Sales of pesticides in Finland in 1980. Kemia-Kemi 8: 504-506. Tomkins, D.J. & Grant, W.F. 1977.Effects of herbicides on species diversity of two plant communities. Ecology. 58: 398-406. Van Groenendael, J.M. 1988. Patchy distribution ofweeds and some implications for modelling population dynam- ics: a short literature review. Weed Res. 28: 437-441. Wahmhoff, W. 1990. The use of economic thresholds over a three year period in cereal crop rotations and the effects on weed infestations two years later. Proc. EWRS Symp. Integrated Weed Management in Cereals, p, 323-330. Warwick, S. 1991. The influence of intraspecific variation on the biology and control of agricultural weeds. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 997-1006. Watson, A.K. 1992. Biological and other alternative control measures. Proc. First Int. Weed Contr. Congr., Mel- bourne. I: 64-73. Whiting, A.J., Davies, D.H.K., Brown, H. & Whytock, G.P. 1991. The field use of reduced doses of broad-leaved weed herbicides in cereals. Proc. Brighton Crop Prot. Conf. - Weeds. 3: 1209-1216. Wilson, B.J. 1982. The yield response of winter cereals to autumn or spring control ofbroad-leaved weeds. Aspects of Appi. Biol. 1. Broad-leaved weeds and their control in cereals, p. 53-61. 1986, Yield responses of winter cereals to the control of broad-leaved weeds. Proc. EWRS Symp. Economic Weed Control, p. 75-82. 39 Agric. Sei. Fin!. Soppi. No. 2 (1993) & Brain, P. 1990. Weed monitoring on a whole farm - patchiness and the stability of distribution of Alopecurus myosuroides over ten year period. Proc. EWRS Symp, Integrated Weed Management in Cereals, p. 45-51. —, Peters, N.C.8., Wright, K.J. & Atkins, H.A. 1988. The influence of crop competition on the seed production of Lamium purpureum, Viola arvensis and Papaver rhoeas in winter wheat. Aspects Appi. Biol. 18: 71-80. Wilson, P.J., Boatman, N.D. & Edwards, PJ. 1990. Strat- egies for the conservation of the endangered arable weeds in Great Britain. Proc. EWRS Symp. Integrated Weed Management in Cereals, p. 93-101. Ympäristöministeriö 1992. Ehdotus maaseudun ympäristö- ohjelmaksi Ympäristönsuoj.os., Työryhmämiet. 68/92. Helsinki. 49 p. Young, J.A. & Evans, R.A. 1976. Responses of weed popu- lations to human manipulations of the natural environ- ment. Weed Sci. 24: 186-190. Zadoks, J.C., Chang, T.T. & Konzak, C.F. 1974. A decimal code for growth stages of cereals. Weed Res. 14; 415- 421. Zeddies, J. 1986. Wirtschaftlichen Bedeutung der Unkraut- bekämpfung unter zukunftigen Rahmenbindungen. Proc. EWRS Symp. Economic Weed Control, p. 39-45, Zwerger, P., Hurle, K. & Kemmer, A. 1990. Untersuchun- gen zum Einfluss von Fruchtfolge und Abbauintensität auf die Entwicklung des Unkrautsamenhaltes im Boden. Proc. EWRS Symp. Integrated Weed Management in Cereals, p. 127-134. 40 Agric. Sei. Finl. Suppl. No. 2 (1993) SELOSTUS Herbisidien käytön vähentäminen kevätviljan viljelyssä Jukka Salonen Maatalouden tutkimuskeskus Rikkakasvien kemiallinen torjunta yleistyi Suomessa 1960- luvun puolivälissä. Herbisidien käyttö on muun viljelytekni- sen kehityksen ohella lisännyt viljelykasvien satoa javiljelyn kannattavuutta. Herbisidien käytön tarkentamista on kuiten- kin perusteltu mm. tuotantokustannusten optimoinnilla ja tor- junta-aineiden haitallisilla ympäristövaikutuksilla. Tässä tut- kimuksessa selvitettiin herbisidien merkitystä ja käyttöä ke- vätviljan viljelyssä, joka on suurin herbisidien käyttökohde. Tutkimus jakautui i) peltojen rikkakasvilajiston, rikkakas- vien runsauden ja herbisidien tehokkuuden selvittämiseen Etelä- jaKeski-Suomen viljelyksillä, sekä ii) kenttäkokeisiin, joissa tutkittiin suositeltua pienempien herbisidiannosten käyttökelpoisuutta. Tavoitteena oli tarkentaakevätviljan vilje- lyssä yleisimmin käytettyjen fenoksihappoherbisidienkäyttö- suosituksia nykyistä tarvetta vastaaviksi. Kevätviljapeltojen rikkakasvilajisto Etelä-ja Keski-Suomen kevätviljapelloillakasvoi 1980-luvun alussa keskimäärin 170 rikkakasvia/m2 (mediaani 124) janii- den tuottama ilmakuiva biomassa oli keskimäärin 320 kg/ha (mediaani 183) ruiskuttamattomilla näytealoilla. Rikkakasvi- en lukumäärä japaino pinta-alayksikköä kohden olivat vähen- tyneet kolmannekseen 1960-luvun alun tilanteeseen verrat- tuna. Peltojen rikkakasvillisuuskoostui pääosin kaksisirkkaisista lajeista. Yleisimpiä ja runsaimmin esiintyneitä rikkakasveja olivat jauhosavikka, pillikkeet, pelto-orvokki, pihatähtimö ja kiertotatar. Yleisin rikkaheinä oli juolavehnä. Yksittäisen pellon rikkakasvillisuus koostui keskimäärin seitsemästä (2-13) torjunnan kannalta merkittävästä lajista. Kenttäkokeissa todettiin, että herbisidien suositeltuna ruisku- tushetkenä viljan 3-4 -lehtiasteella rikkakasveista oli taimettu- nut keskimäärin 70-75 %. Hyvässä kasvukunnossa oleva vilja kuitenkin ehkäisi herbisidiruiskutuksen jälkeen taimettunei- den rikkakasvien kasvun varsin tehokkaasti. Kevätviljapeltojen rikkakasvillisuuteen vaikuttavia tekijöi- tä havainnollistettiin ordinaatioanalyysillä. Rikkakasvien esiintymiseen vaikuttivat maaperätekijät sekä viljelytekniset toimet kuten viljelykierto ja torjunta-aineiden käytön ylei- syys. Yksittäisiä rikkakasvien esiintymiseen vaikuttavia teki- jöitä ei havaintoaineistosta voitu erottaa, sillä maaperään ja viljelytoimiin liittyvät tekijät kytkeytyivät toisiinsa. Peipit, matarat, peltoemäkki ja saunakukka olivat yleisiä Etelä- ja Lounais-Suomen viljanviljelyalueella, jossa maalaji oli pää- asiassa savea. Keski-Suomessa, jossa viljelykierto on moni- puolisempaa ja maalajit kevyempiä, tyypillisiä lajeja olivat hierakal, tatarlajit jaleinikit. Jauhosavikkaa ja pillikettä esiin- tyi kaikkialla. Herbisidien tehokkuus MCPA-valmisteet ja yleisimmät seosvalmisteet (MCPA/ diklorproppi, MCPA/mekoproppi) muodostivat yhdessä 91 % viljelyksillä käytetyistä herbisideistä. Ruiskutus MCPAdIa vähensi rikkakasvien biomassaa keskimäärin 65 % ja seos val- misteilla 83 % käsittelemättömään havaintoruutuun verrattu- na. Kenttäkokeissa oli suositellulla annoksella ruiskutettujen herbisidiseosten teho vastaavasti 85-90 %. Alimman nykyisin suositellun herbisidiannoksen pienentä- minen 30 %:lla heikensi MCPA/diklorproppi-, MCPA/flu- roksipyyri- ja MCPA/mekoproppi-valmisteiden tehoa keski- määrin alle 10prosenttiyksikköä. Vieläkin pienemmät herbi- sidimäärät olivat riittäviä, jos torjunta tehtiin hyvissä oloissa ja herbisidi valittiin rikkakasvilajiston mukaan. Riski torjun- nan epäonnistumisesta kasvoi kuitenkin annosta pienennet- täessä. Tulosten perusteella voidaan suositella nykyisin voimassa olevien annossuositusten pienentämistä 30 prosentilla. Ny- kyistä pienempien annosten käyttö edellyttää viljelijöiltä tai- toa valita herbisidi rikkakasvilajiston mukaan. Tietokonepoh- jaiset asiantuntijajärjestelmät ovat tulossa viljelijöiden avuksi herbisidiä ja tarpeenmukaista annosta valittaessa. Tästä tutki- muksesta saadut tulokset antavat arvokasta tietoa asiantuntija- järjestelmien kehittämiseen. Herbisidien käyttötarve Rikkakasvien osuus viljakasvuston kokonaisbiomassasta oli alle 5 % ruiskuttamattomillakin näytealoilla. Nykyistä viljelytekniikkaa käyttäen jakylvötiheyssuosituksia noudatta- en viljakasvustot kehittyvät erittäin kilpailukykyisiksi ylei- simpiä rikkakasvejamme vastaan. Rikkakasvit vähensivätkin viljan satoakeskimäärin alle 5 %. Herbisidiruiskutus suositel- tua pienemmillä annoksilla tuotti saman satotuloksen kuin suositellulla ainemäärällä ruiskutettuna. Viljan kylvösiemenmäärän lisääminen ei osoittautunut ta- loudelliseksi vaihtoehdoksi rikkakasvien torjunnassa. Hyvin- kin pienillä herbisidiannoksilla ehkäistiin rikkakasvien kas- vua tehokkaammin kuin viljan suositeltua kasvutiheyttä lisää- mällä. Viljan kasvutiheys ei vaikuttanut ruiskutushetkeen mennessä taimettuneiden rikkakasvien lukumäärään, mutta tiheä viljakasvusto ehkäisi rikkakasvien biomassan kasvua myöhemmin kesällä. 41 Agric. Sei. Fin!. Suppl. No. 2 (1993) Luotettavia kynnysarvoja herbisidin käytöstä luopumiselle ei löydetty. Herbisidien käyttötarvetta ei voitu arvioida pel- kästään taimettuneiden rikkakasvien lukumäärän perusteella, sillä lukumäärä ei yksinään kuvannut kasvustoon myöhem- min kesän aikana kehittyvää rikkakasvimassaa eikä viljalle aiheutuvaa sadon menetystä. Muutama rikkakasvi neliömet- rillä harvassa viljelykasvustossa saattoi tuottaa enemmän bio- massaa ja rikkasiemeniä kuin yli sata yksilöä tasaisesti orastu- neessa hyvässä kasvukunnossa olevassa kasvustossa. Herbisidien käyttö ei kaikissa tilanteissa ollut taloudellises- ti kannattavaa, jos toijunnan hyötyä arvioidaan pelkästään sadon määränä. Herbisidien käytöstä luopuminen lisää kui- lenkin maahan kertyvien rikkasiementen määrää. Herbisidien jatkuvakäyttö pienillä annosmäärilläkin osoittautui tässä suh- teessa turvallisemmaksi vaihtoehdoksi kuin välivuodet tor- junnassa. Rikkakasvien määrän ja lajiston suuri vuosittainen japelto- kohtainen vaihtelu edellyttää rikkakasvillisuuden kartoitusta joka vuosi ennen herbisidivalintaa ja -ruiskutusta. Ruiskutuk- sen ajoitus rikkakasvien varhaiselle taimivaiheelle tuo mah- dollisuuden käyttää nykyistä pienempiä herbisidiannoksia. Näin menetellen, realistisena tavoitteena voidaan pitää vilja- herbisidien käytön vähenemistä Suomessa 30 prosentilla. 42 Agric. Sd. Fint. Suppl. No. 2 (1993) ERRATA FOR THE ORIGINAL ARTICLES Chapter 111. Position p. 485. Table 2. Ylistaro 10986 1987 1989 Correction Ylistaro 1986 1987 1988 Chapter IV. Position p. 495. Fig. 1. Fig. 1. Dose-response ofcrop yield (open symbols) and weed biomass (closed symbols) to the MCPA/mecoprop treatment in three growth densities of(a) spring barley and (b) spring wheat in Jokioinen in 1988. (O) indicates the normal, (□ ) 200 seeds lower and (0) 200 seeds higher sowing densities. Correction (of symbols) Fig. 1. Dose-response ofcrop yield (open symbols) and weed biomass (closed symbols) to the MCPA/mecoprop treatment in three growth densities of (a) spring barley and (b) spring wheat in Jokioinen in 1988. (□) indicates the normal, (O) 200 seeds lower and (O) 200 seeds higher sowing densities. Agric. Sei. Fin!. Sappi. No. 2 (1993) Weed infestation and factors affecting weed incidence in spring cereals in Finland - a multivariate approach Jukka Salonen Salonen, J. 1993. Weed infestation and factors affecting weed incidence in spring cereals in Finland - a multivariate approach. Agric. Sci. Finl. 2: 000-000. (Agric. Res. Centre of Finland, Inst. PI. Prot., FIN-31600 Jokioinen,Finland.) Weed vegetation of spring cereal fields in southern and central Finland was analyzed by ordination methods to provide a community level description of weed populations. Attention was paid particularly to the relative importance of environmental factors affecting weed incidence such as crop management, soil properties and weather condi- tions. A data set of 33 weed taxa from 252 fields was subjected to both indirect and direct gradient analysis. Indirect ordination was obtained with correspondence analysis (CA), and direct gradient analyses were performed with redundancy analysis (RDA) and with canonical correspondence analysis (CCA) relating environmental factors to the occurrence of weeds. Among several management factors, continuous herbicide use explained best the variation in the species composition of weed flora. Weed vegetation was also associated with soil type, moisture conditions and soil pHh 2o- Ordination diagrams visualized the species-environment interactions and detected characteristic weed species for different geographical regions. In addition to ordination analyses of weed flora, the level and structure of weed infestation are described. The density of weeds averaged 170 plants m 2(median=l24) and the air-dry weight of weeds 320 kg ha '(median=lB3). The average weed density was the same in different soil types, but the weed biomass was lower in clay soils than in coarse mineral and organic soils. Key words: broad-leaved weeds, ordination, barley, oats, wheat,canonical correspond- ence analysis, CA, CCA, RDA, CANOCO Introduction Arable fields are continuously subjected to differ- ent agricultural measures particularly in annual crops. Although many weed species are adapted to the prevailing conditions, the constantly changing habitat selectively affects weed communities and, consequently, changes the weed flora (Rade- MACHER et al. 1970, REUSS 1981, Mahn 1984, Chancellor 1985,Légére et al. 1993). Weed flora in spring cereals was investigated during 1982-84 in Finland (Erviö and Salonen 1987). Attention was paid particularly to the changes in weed infestation by comparing the data with the previous study from the 1960 s (Mukula et al. 1969). The occurrence of individual weed species was related to several explanatory variables by the analysis of variance and regression tech- niques. These methods are appropriate if detailed responses of particular weed species to explanatory factors are studied. The problem was, however, to give a summary of the relative importance of fac- tors affecting the weed incidence. Therefore, the data from weed survey was subjected to ordination 1 Agric. Sei. Finl. 2 (1993)2ndProof analyses which have proved to be appropriate for community level description of weed vegetation (Ter Braak 1987a). Multivariate analysis of community data is fre- quently applied in ecological studies to summarize the information in samples-by-species data matri- ces (Gauch 1982). In weed science, the multivari- ate approach is feasible to describe and predict the response of weed vegetation to farming practices (POST 1988). Multivariate methods in ecology can be divided into three groups (JONGMAN et al. 1987): direct gradient analysis (regression), indirect gradi- ent analysis (ordination) and classification (cluster analysis). Indirect methods analyze the species data only, whereas species-environment interactions can be analyzed simultaneously by direct methods. In this paper, the weed survey data from 1982- 1984 was subjected to ordination analyses to give a community level descriptionof weedflora in spring cereal fields. The objective was to find charac- teristic weed species in different geographical re- gions and to illustrate responses ofweed vegetation to environmental factors. Furthermore, the level of weed infestation, proportion of the most abundant weed species and the occurrence of weeds in differ- ent soil types are reported. Material and methods A total of 267 spring cereal fields (barley, oats or wheat) in southern and centralFinland were studied during 1982-1984. In each field there were 4 to 5 9sample plots of 0.25 m" in size from which the above-ground occurrence of 33 weed species (Table 1) or, in fact, weed taxa was assessed in late July by counting the number and weighing the air- dry biomass of weeds. The sample plots were not sprayed with herbicides. Frequency of weeds (Table 1) denotes the proportion of the fields where the particular weed species was observed out of the all fields studied. Detailed information of the sur- vey and the occurrence of weed species has been given by Erviö and Salonen (1987). Data on factors involved in each field was col- lected either by observing, measuring or by inter- viewing the farmer. Twelve factors describing either the current crop, crop rotation, soil properties or climate (Table 2) were used as environmental variables in the CCA. The factors were chosen from among the 21 factors studied in the regression ana- lysis and considered the most important (Erviö and Salonen 1987). The survey localities were grouped into three regions based on their geo- graphical locations: South-western Finland (SW), eastern part of central Finland (CE) and western part of central Finland (CW). Features of regression analysis and ordination are integrated in canonical ordination techniques (Jongman et al. 1987). These techniques provide a direct analysis of species-envir- onment interactions which was earlier possible only by re- gression methods. ’Canonical correspondence analysis’ (CCA) by Ter Braak (1986) is probably the most common canonical ordination technique currently applied in various ecological studies (Birks and Austin 1992). CCA and the related indirect technique ’correspondence analysis’ (CA) (Gauch 1982) have been applied also in agricultural research (Jukola-Sulonen 1983,Wentworth et al. 1984, Post 1986, Siepel et al. 1989, Pysek and LepS 1991, Dale et al. 1992). CA and CCA fit the unimodal curve to the species-environ- ment data, whereas a linear response model between species data and environmental variables can be fitted by the ’redundancy analysis’ (RDA). The ordination techniques mentioned above are all available in the computer program CANOCO (Ter Braak 1987b). Environmental variables were either qualitative (nominal scale) or quantitative (interval scale) (Table 2). The crop rotation was considered cereal dominant if a cereal crop had been grown at least for three years of theprevious four years. Otherwise it was classified as mixed rotation. The use of herbicides indicates only the intensity of chemical weed control, not the type of herbicides applied during the last nine years. The soil pHh2 0 was measured from the top 0-20 cm layer. The soil type of fields was classified into three categories: clay (clay content >30%), organic (>20% organic mat- ter) and coarse mineral soils. The subjective assess- ment of soil moisture was primarily based on the soil type and the drainage of the field. Nominal type environmentalfactors were transformed into binary dummy variables. Due to missing values of ex- planatory factors, some sample fields had to be excluded, since missing data are not accepted in the CANOCO run. Thus, a final data set consisted of 2 2nd ProofAgric. Sei. Finl. 2(1993) 3 Table 1. Frequency, the effective number of occurrences (N2) and average biomass production of the 33 weed species studied in 252 spring cereal fields. Frequency denotes the proportion of the fields where the species was found. The N 2 value obtained from the CANOCO run is based on the weighted averages of weed densities and it indicates the number of fields where the species was abundant. Air-dry biomass indic- ates the average infestation of the species in those fields it was found. Weed taxa Codell Frequency N 2 Biomass % g m 2 Chenopodium album L. CHEAT 87 163 5.0 Galeopsis spp. L. GAESS 85 166 6.1 Viola arvensis MURRAY VIOAR 85 146 1.0 Slellaria media (L.) VILL. STEME 81 155 2.8 Fallopio convolvulus (L.) Ä. LÖVE POLCO 61 112 1.3 Erysimum cheiranthoides L. ERYCH 58 95 1.5 Lapsana communis L. LAPCO 54 94 4.0 Polygonum aviculare L. POLAV 52 71 0.5 Myosotis arvensis (L.) HILL MYOAR 52 66 0.5 Elymus repens (L.) GOULD AGRRE 51 92 13.0 Spergula arvensis L. SPRAR 46 68 2.9 Fumaria officinalis L. FUMOF 43 74 1.4 Galium spp. L. GALSS 35 57 1.0 Tripleurospermum inodorum SCHULTZ BIP. MATIN 32 34 0.7 Polygonum lapalhifolium L. POLLA 30 45 1.7 Sonchus arvensis L. SONAR 27 43 2.8 Lamium spp. L. LAMSS 25 39 1.9 Matricaria matricarioides (LESS.) PORTER MATMT 18 23 1.9 Gnaphalium uliginosum L. GNAUL 18 15 0.1 Capsella bursa-pastoris (L.) MEDIK, CAPBP 17 23 0.3 Ranunculus repens L. RANRE 17 13 0.2 Thlaspi arvense L. THLAR 16 21 0.8 Equisetum spp. L. EQUSS 13 26 2.1 Brassica rapa L. ssp. oleifera DC. (volunt.) BRSRO 13 25 4.0 Poa annua L. POAAN 13 14 0.5 Brassica spp. L. BRSSS 12 15 4.2 Rumex spp. L. (Sorrels) RUMSS 12 14 0.7 Achillea spp. L. ACHSS 5 11 2.1 Cirsium arvense (L.) SCOP. CIRAR 5 7 2.0 Sonchus spp. L. (S. asper, S. oleraceus) SONSS 4 5 9.4 Urtica spp. L. URTSS 2 1 0.5 A vena fatua L. AVEFA 1 2 9.4 Slachys palustris L. STAPA 1 1 3.0 11 Weed codes are according to the BAYER standard (BAYER 1992). 252 fields. The geographical regions were used as environmental variables in RDA, and as covari- ables in partial CCA. Ordination analyses were performed with the CANOCO program (Ter Braak 1987b) applying CA, CCA and RDA. Ordination diagrams (species- environment biplots) were drawn with the CANO- DRAW program (Smilauer 1990). The relation- ship between the weed communities and environ- mental variables is displayed with the first two ordination axes. Only the central area of the dia- gram is shown to improve the visibility of species near the origin. Consequently, some species and environmental variables lie outside the drawn area (Figs. 4 and 5). Due to the skewed distribution of the response values (weed density and weed biomass) the weed data was log-transformed (ln(y+l)) in the CANOCO run. Species diversity was described by the N 2 value from the CANOCO output. The N2 Table 2. Environmental variables subjected to the canonical correspondence analysis (CCA). Variable (scale) Code Range or No. of fields CROP VARIABLES Cover, % (interval) COVER 13-100 Yield, kg ha 1 (interval) YIELD 520-7300 MANAGEMENT VARIABLES Cereal dominance (nominal) Cereals CER 144 Mixed rotation MIX 108 Herbicide use during 9 previous years (interval) HERB 0 years 6 1 8 2 14 3 26 4 20 5 30 6 17 7 14 8 20 9 97 SOIL VARIABLES Soil type (nominal) Coarse COARSE 110 Clay CLAY 112 Organic ORGANIC 30 Moisture type (nominal) Dry DRY 40 Normal NORMAL 199 Wet WET 13 Soil pH H2o (interval) PH 4.85-7.65 CLIMATIC VARIABLES (between sowing and sampling) Effective temperature sum, DD (base 5°C) (interval) ETS 281-857 Precipitation, mm (interval) PREC 40-222 where X is Simpson’s diversity index, m is the number of ith species in the population and N is the total number of all S species in the population. Results Occurrence of weeds 2The weed density averaged 170 plants m (SE= 10, median=l24) and the biomass production 320 kg ha" 1 (SE=23, 183).The total weed biomass correlated weakly P90% MCPA II 21 30 38 Dichlorprop/MCPA 7 14 33 45 Mecoprop/MCPA 4 14 32 50 Other herbicides 0 17 38 46 All fields 8 17 32 44 Efficacy ofherbicides on dicotyledonous weeds In total, 22 different products were used to control broad-leaved weeds. These were classified into four groups on the basis of their active ingredients (Table 2). The fourth group of ‘other herbicides’ included products containing mix- tures of phenoxy acids/ioxynil/bromoxynil, phe- noxy acids/clopyralid or chlorsulfuron as active ingredients. Due to the small number of fields in which these products were applied, a division of this group was not meaningful. The average dose ofactive ingredient of MCPA was 1-2 kg ha-1 when sprayed alone. The most common products were herbicides containing either dichlorprop or mecoprop (400 g l~ l together with 200 gf MCPA. The average dose of these mixtures in spring cereals was 2-5-4 0 I ha-1 , and the average amount of water was 250 I ha The average time interval between sowing and herbicide spraying was 26 days (12-43) and the weed samples were collected on average 38 days (12-46) after herbicide treatment. The average total dry weight ofdicotyledonous weeds was 243 kg ha' 1 in untreated, and 61 kg ha 1 in treated sample plots, respectively. Thus, the average control with herbicide treatment was 75% (Table 2). Differences in the dry weights ofweeds between untreated and treated plots were statistically significant (/*< 0-001) with all herbicides, but differences between herbicide groups were not. The efficacy of chemical weed control varied considerably in different fields and with different herbicides. The most unreliable control was obtained with products containing MCPA alone. The performance of herbicide mixtures was good or excellent in about 80% of fields assessed (Table 3). The best effect was achieved against Chenopo- dium album and Erysimum cheiranthoides, which were easy to control with all herbicides. The weakest effect was observed against Lapsana communis with MCPA significantly (/ > <0 05) lower than other herbicides (Table 4). Discussion According to the present survey the weed species, producing the most biomass in spring cereal fields in Finland were Chenopodium album, Galeopsis spp., Stellaria media, Lapsana communis and Elymus repens (Table 1). Although Viola arvensis was common, its bio- mass production was usually quite low. In our study V. arvensis was controlled fairlywell (Table 4), even though it is considered to be quite resistant to phenoxy herbicides (Fryer & Make- peace, 1978). In Swedish field trials V. arvensis was observed to recover from herbicide treat- ment, and to be able to utilize the space obtained from other species controlled by spraying (Gum- messon, 1983). Results indicating the efficacy of control of the less frequent weed species (Table 4) should be interpreted with caution due to their random occurrence in sample plots. These weed species may have been absent either in untreated or treated sample plots. The fifteen most common weed species were affected to different degrees by the herbicides 233 J. Salonen and L.-R. Erviä Table 4 Effect ofdifferent herbicides on the biomass ofdicotyledonous weed species % ofcontrol Weed species MCPA Dichlorprop/MCPA Mecoprop/MCPA Other herbicides Chenopodiumalbum 87 92 93 99 Galeopsis spp. 71 71 83 95 Violaarvensis 79 84 91 79 Stellaria media 82 69 87 92 Fallopio convolvulus 69 89 55 83 Erysimum cheiranthoides 84 98 98 97 Impsana communis 9 49 90 89 Polygonum aviculare 40 89 99 78 Myosotis spp. 28 79 32 59 Spergula arvensis 53 76 95 99 Fumaria officinalis 49 87 92 58 Galium spp. 32 91 94 66 Tripleurospermum inodorum 70 71 80 97 Polygonum lapathifolium 30 62 75 84 iMmium spp. 55 77 55 89 (Table 4). In most of the cases herbicide mixtures gave a better control than MCPA alone (see also Mukula & Köylijärvi, 1965), but in many fields the dense crop stand already caused so much competition that MCPA alone was sufficient. This is in agreement with results by Fogelfors (1977). In spring cereals herbicides can be applied from the three-leaf stage to shooting, but the right timing may be difficult underFinnish conditions where the development of plants is often very rapid (Pessala, 1976).Because of the weather the number of days suitable for spraying during that time may be restricted. Environmental factors are also recognized as a major cause of inconsistency in herbicide activity (Gerber, Nyffler & Green, 1983; Caseley, 1984). In a recent study (Luoma & Lavonen, 1987) on the tractor sprayers ofFinnish farms it was shown that 63% of sprayers had defects resulting in uneveness in application, which might have caused poor weed control also in some fields of our study. The main reason for inadequate control in a number offields was undoubtedly a wrong choice of herbicide (Table 2). Weeds like Galium spp., Lapsana communis and Polygonum spp. had been sprayed with MCPA, but they are known to be tolerant to MCPA (Fryer & Makepeace, 1978). According to our study MCPA should be rep- laced by mecoprop/MCPA for the control of L. communis, and by dichlorprop/MCPA when Fal- lopia convolvulus or Myosotis spp. occurs (Table as. '' More than one-third of the fieldssurveyed were treated with MCPA alone. Sales statistics of herbicides in Finland also indicate that a great proportion of herbicides used against broad- leaved weeds still consists of products containing MCPA alone (Hynninen & Blomqvist, 1987). The low rate herbicides like chlorsulfuron and metsulfuron-methyl are nowavailable inFinland, and they are likely to replace phenoxy acids to some extent in the future. In general, these low rate herbicides have a wider spectrum ofactivity but they have also their weak points like Fumaria officinalis, Galium spp. and Viola arvensis (Aamis- epp, 1986). Thus, their use may lead to botanical changes in cereal fields. In conclusion, to obtain a satisfactory and economic weed control, the weed spectrum of each field should be considered more carefully before selecting a herbicide, Acknowledgments We wish to thank Mr Juha Mustonen and Mrs Katri Pahkala for their excellent work as leaders of the field studies during the first 2 years of this survey. Planning of the data collection and analysis by Mrs Liisa Mattila is also warmly acknowledged. The survey was financed by the Finnish Ministry ofAgriculture and Forestry. References Aamisepp A. (1986) Undersökningar av sulfonylurea-herbicider i stråsäd i svenska försök. Aktuellt fra Statens Fagtjenestefor Landhrukei, 8, 229 237. 2 234 Aamisepp A. & Nilsson H. (1987) De kemiska ogräsmedlens användningsområde och verkan mot ogräs. Ogräsnyckeln. Aktuellt från Lantbruksuniversitet , 354, 20-52. Aamisepp A. & Wallgren B. (1979)Ogräs i stråsäd. Verkan av kemisk ogräsbekämpning och andra odlingsåtgärder, 1950- 1978. Aktuellt från Lantbruksuniversitet, 280, 15. Caseley J. (1984) The influence ofenvironmental factors on foliage-applied herbicides. NJF-seminarium Nr 58. ‘Klima- faktörers inverkan på herbicidernas effekt. Erviö L.-R. & Salonen J. (1987) Changes in the weed population of springcereals in Finland. Annales Agriculturae Fenniae . 26, 201-226. Fogelfors H. (1977) The competition between barley and five weed species as influenced by MCPA treatment. Swedish Journal of Agricultural Research, 7, 147-151. Fryer J.D. & Makepeace R.J. (1978) Weed Control Handbook. Volume 11. Bth edn. Blackwell Scientific Publications, Oxford. Gerber H R.. Nyffler A. & Green D.H. (1983) The influence of rainfall, temperature, humidityand lighton soil and foliar- applied herbicides. Aspects of Applied Biology, 4, 1-14. Gummesson G. (1983) Chemical and non-chemical control changes in weed stand following different control measures. Weedsand Weed Control. 24th Swedish WeedConference, pp. 234-245. Hynninen E.-L. & Blomqvist H. (1986) Torjunta-aineiden myynti Suomessa 1985. Summary: Sales of pesticides in Finland in 1985. Kemia-Kemi, 13, 725-728. Hynninen E.-L. 8c Blomqvist H. (1987) Torjunta-aineiden myynti Suomessa 1986. Summary: Sales of pesticides in Finland in 1986. Kemia-Kemi, 14, 569-572. Kolbe W. (1983) Kulturpflanzenbau und Unkrautbekämp- fung. Pflanzenschutz-Nachrichten Bayer, 36, 209-381. Luoma T. & Lavonen A. (1987) Maatiloilla käytettävien kasvinsuojeluruiskujen tekninen kunto. Kasvinsuojeluseuran 8. Kasvitauti- ja Tuhoeläinpäivä, pp. 6-10. Maas G. (1979) Einfluss von Umweltfaktoren und Spritztech- nik auf die Wirkung von Herbiziden. Proceedings EWRS Symposium on The Influence of Different Factors on the Development and Control of Weeds, pp. 39-47. Mukula J. (1965) Tilastoa rikkakasvihävitteiden käytöstä Suomessa. Pellervo, 66, 448-449. Mukula J. & Köylijärvi J. (1965) Comparative studies with three herbicides, MCPA, MCPA plus 2,3,6-TBA and mcco- prop forweed control in spring cereals. Annales Agriculturae Fenniae, 4, 256-276. Mukula J. & Rantanen O. (1987) Climatic risks to the yield and quality of field crops in Finland. I. Basic facts about Finnish field cropsproduction. Annales AgriculturaeFenniae, 26, I-18. Mukula J. & Ruuttunen E. (1969) Chemical weed control in Finland in 1887-1965. Annales Agriculturae Fenniae, 8, Supplement 1, 59-110. Pessala B. (1976) Time of application of herbicides for Avena fatua control in spring wheat and barley. Proceedings 1976 British Crop Protection Conference— Weeds, pp. 39-46. Steel R.G.D. & Torrie J.H.(1960) Principles andProcedures of Statistics. McGraw-Hill, New York. 235 Efficacy of reduced herbicide doses in spring cereals of different competitive ability J. SALONEN Agricultural Research Centre of Finland, Institute of Plant Protection, SF-31600 Jokioinen, Finland Received 11 October 1991 Revised version accepted 6 April 1992 Summary: Résumé: Zusammenfassung The effect of dose reduction on the efficacy of MCPA/mecoprop and MCPA/fluroxypyr mix- tures in the control of broad-leaved weeds was studied in spring barley (Hordeum vulgäre L.) and spring wheat (Triticum aestivum L.). Five crop densities were sown to obtain crop stands of different competitive ability. Halving the highest recommended dose decreased the con- trol efficacy by 5-15 percentage units, but still provided a good weed control. Further re- duction of the dose resulted in high variation in efficacy. The level of weed suppression was mainly determined by the competitiveness of the crop, environmental conditions and weed spectrum with differential sensitivity of species to the herbicides. No changes in the weed infes- tation were detected after continuous use of low herbicide doses for 3 years. The seed bank of weeds in soil was higher after a wheat-domin- ated rotation than after a barley-dominated rotation. Efficacité de doses réduites d'herbicides chez des céréales deprintemps d’aptitudes ä la competitive différentes L’impact de la réduction des doses sur I’effica- cité de mélange MCPA/mecoprop et MCPA/ fluroxypyr en matiére de desherbage des dicotyledones a été étudié chez I’orge de prin- temps (Hordeum vulgäre L.) et le blé de prin- temps (Triticum aestivum L.) 5 densités de culture ont été semées pour obtenir des cultures d’aptitudes ä la compétitivité différentes. La re- duction de moitié de la dose recommandée la plus élevée a réduit I’efficacité herbicide de 5 ä 15% mais a encore fourni un bon résultat. Des réductions plus élevées de la dose ont conduit ä de fortes variations dans I’efficacité. Le niveau de suppression des adventices était principale- ment déterminé par la compétitivité de la cul- ture, les conditions environnementales et le spectre des adventices avec une sensibilité dif- férentes des espéces aux herbicides. Aucun changement dans I’infestation en adventices n’a été observé apres une utilisation continue de faibles doses d’herbicidependant 3 ans. Le stock grainier d’adventices du sol était plus élevé apres une rotation ä dominante blé, qu’apres une rotation ä dominante orge. Wirksamkeit reduzierten A ufwands von Herbi- ziden in Sommergetreide mil unterschiedlicher Konkurrenzkraft In Sommergerste (Hordeum vulgäre L.) und Sommerweizen (Triticum aestivum L.) wurde untersucht, wie die Reduzierung des Aufwands die Wirksamkeit von MCPA-Mecoprop- und MCPA-Fluroxypyr-Mischungen beeinflußt. Die Getreide wurden in 5 Dichten gesät, um Pflanzenbestände unterschiedlicher Konkur- renzkraft zu haben. Die Halbierung des höchsten empfohlenen Aufwands verminderte die Wirksamkeit um 5 bis 15 Prozentpunkte, ergab dennoch eine gute Unkrautbekämpfung. Eine weitere Reduzierung des Aufwands brachte sehr unterschiedliche Wirksamkeit. Der Grad der Unkrautunterdriickung wurde iiberwiegend durch die Konkurrenzkraft des Getreides, die Umweltbedingungen und das Unkrautspektrum mit gegeniiber den Herbi- ziden verschieden empfindlichen Arten Weed Research , 1992, Volume 32, 483-491 bestimmt. Nach 3 Jahren niedrigen Herbizidaufwands ließen sich Änderungen der Verunkrautung nicht erkennen. Der Unkraut- samenvorrat im Boden war nach einer weizen- dominierten Fruchtfolge größer als nach einer gerstendominierten. Introduction The encouraging results regarding the use of reduced herbicide doses in spring barley (Elbek Pedersen, 1978; Thonke, 1978), prompted many research projects in Scandinavia in order to study the possibilities of decreasing the total use of herbicides. As a result of decreased weed infestation in spring cereal fields (Erviö & Salonen, 1987) and of inconsistent economic return from chemical weed control (Erviö etal. , 1991), similar research was also considered necessary in Finland. Weed management by applying reduced herbicide doses every year differs from the alternative approach of weed control thres- holds. Thresholds have been extensivelystudied in Germany (Heitefuss et al., 1987). Gummesson & Fogelfors (1990) suggested that in the long-term the annual use of reduced doses prevents the increase of the weed seed bank in soil more reliably than spraying only if certain thresholds are exceeded. Herbicides for the control of broad-leaved weeds in cereals represent the most important group of pesticides used in Finland (Hynninen & Blomqvist, 1990). Therefore, the reduction of chemical weed control in spring cereals would be economically and environmentally attractive. Thonke (1986) reviewed several promising results on the use of reduced herbicide doses. On the other hand, little is known about the long-term effects of successive applications of reduced doses, particularly on the soil seed bank of weeds and the effect on future weed infesta- tions (Kudsk, 1989). The objective of the present study on cereal herbicides was to investigate the efficacy levels achieved with lower dose rates than recommen- ded in spring cereals. The long-term effects of continuous applications of reduced herbicide rates in crop stands of varying competitive ability were also studied. Materials and methods Field trials Field trials with spring barley (cv. Arra) and spring wheat (cv. Luja) were performed at the Agricultural Research Centre ofFinland during 1986-1988. The experiments were conducted in Jokioinen in Southern Finland (61° N) and in Ylistaro in Central Finland (63° N). The trials lasted 3 years with a barley-wheat-barley rota- tion at both sites and, additionally, with a wheat-barley-wheat rotation in Jokioinen. All three trials were established on clay soil, and 90 kg N ha -1 was applied every year. The pre- ceding crop at Jokioinen was grassland for cattle and spring barley in Ylistaro. Field trials were laid out in a randomized complete-block design with a split-plot arrange- ment and four replicates. The main plots had five crop seed rates. The subplots, 3 x 10 m in size, were subjected to treatments with two herbicide mixtures and three ascending dose rates up to the highest recommended commer- cial dose (Table 1). At Jokioinen the trials with barley and wheat were placed adjacent to each other in order to compare the competitive differences in the crops. Each trial consisted of 140 plots. Commercial herbicide mixtures of MCPA 200 g a.i. r'/mecoprop 400 g a.i. I- ' (Flerbotal Plus) and MCPA 400 g a.i. C'/fluroxypyr 100 g a.i. 1“' (Starane M) contained salt-formulated phenoxy acids and ester-formulated fluroxypyr. Herbicides were applied with a portable 3-m- -wide sprayer fitted with Hardi 4110-10 flat fan nozzles which gave 200 I ha~' spray solution at a pressure of 300 kPa. The treatment time (Table 2) was at the 3- to 4-leaf stage of the crop when Table 1. Treatments in the factorial designed field trials in spring barley and spring wheat in Jokioinen and Ylistaro in 1986-1988 MAIN PLOT: SEED RATE (viable kernels m" 2) Barley: 100, 300, 500 (normal), 700 and 900 Wheat: 200. 400, 600 (normal). 800 and 1000 SUBPLOT: HERBICIDE AND RATE Treatment Iha-1 kga.i.ha-1 Untreated 0 0 MCPA/mecoprop 1-30 0-26/0-52 MCPA/mecoprop 2 00 0-40/0-80 MCPA/mecoprop 4-00 0-80/1-60 MCPA/fluroxypyr 0-50 0-20/0-05 MCPA/fluroxypyr 0-75 0-30/0-08 MCPA/fluroxypyr 1-50 0-60/0-15 484 Table 2. Application data from the field trials with spring barley and spring wheat at two locations in 1986-1988 Jokioinen Ylistaro 1986 1987 1988 10986 1987 1989 Application date 18 JUN 24 JUN IOJUN 19 JUN 23 JUN 3 JUN Temperature (°C) 21 22 18 28 21 15 Relative humidity (%) 53 50 53 40 30 40 Days between sowing/spraying 19 28 25 30 33 30 Spraying/assessment barley 29 31 31 33 31 wheat 29 35 32 29 spraying/harvest barley 70 63 53 68 52 wheat 99 94 83 91 Precipitation (mm) between sowing/spraying 7 75 27 21 64 53 spraying/assessment barley 55 29 22 73 49 wheat 55 44 22 17 spraying/harvest barley 160 153 154 157 92 wheat 272 227 223 137 Effective temperature sum (> 5°C) between sowing/spraying 203 182 226 279 195 154 spraying/assessment barley 326 297 406 331 367 wheat 326 329 420 270 spraying/harvest barley 717 524 684 641 663 wheat 790 635 956 611 the weed seedlings were between the cotyledon and the four-leaf stage. Weed assessments The number of weed seedlings that emerged from the natural seed bank of the field was counted before spraying. Furthermore, the weed infestation was assessed 4 weeks after spraying and at harvest by counting seedling numbers and recording the air-dry biomass of weeds in a 0-25 m 2 circle per plot. The herbicide efficacy against all broad-leaved weeds and separately against the most common weed species was determined by comparing the biomass ofweeds in untreated and treatedplots. The subsequent effect of different treatments on the number of weeds which emerged by the 3- to 4-leaf stage of barley was studied at Jokioinen in 1989, 1 year after the three year trial period. At the same time, soil samples were taken from the plots sown at normal seed rate in order to compare the effects of different control intensities on the seed bank of weeds. Soil samples were taken from a depth of 0-20 cm, and all weed seedlings that emerged during the following two growing seasons (1989-1990) were counted and removed from the sample pots which were kept in open-air greenhouses. Statistical analysis The biomass of weedswas transformed with the common logarithm log'°(y+l) to achieve normal distribution and homogeneity of var- iances. The exponential pattern between weed infestation and crop density and herbicide rate was thus linearized. Similarly, the relative efficacy values were transformed with arcsin (Vy) and the number ofweeds with square root (Vy) transformation before the data were subjected to analysis of variance and analysis of covariance using the density of crop as a covariate. The sum of squares waspartitioned to orthogonal comparisons in order to test in- dividual hypotheses. A linear regression model relating the weed infestation to the crop density and herbicide dose was used. Statistical analyses 485 were performed with the Generalized Linear Models procedure of the SAS statistical package. Results and Discussion Occurrence of weeds The beginning of the growing season was extremely dry in 1986. On the other hand, the moisture conditions favoured the emergence of weeds in 1987 and in 1988, resulting in higher weed densities. At the time of treatment the density of weeds in untreated barley plots in Jokioinen was 48 weeds m“ 2 in 1986, 335 weeds m~2 in 1987 and 353 weeds m~2 in 1988. In spring wheat the corresponding weed densities were 43, 322 and 560 weeds m-2. In Ylistaro there were 111 weeds m~2 in barley in 1986,192 weeds m-2 in wheat in 1987, and 415 weeds m~ 2 in barley in 1988. Accordingly, in most of the trials the density of weeds was higher than the average of 173 weeds m“2 in spring cereal fields inFinland (Erviö & Salonen, 1987). The most abundant weed species at Jokioinen were Chenopodium album L. and Galeopsis speciosa L., as is also the case in spring cereal fields in Finland (Erviö & Salonen 1987). Stellaria media (L.) Vili. and Matricaria spp. L. were the main species at Ylistaro. Effect of crop density The crop density did not significantly affect the number of weed seedlings that emerged prior to spraying, but affected the weed infestation later in the growing season. In untreated plots the in- crease in crop seed rate improved the competi- tive ability of wheat more than that of barley (Fig. 1). However, the effect was usually marginal above the normal seed rates, as has also been shown by Håkansson (1975) and Erviö (1983). Particularly in treated plots, the increase in crop seed rate above the recommended crop density gave an insignificant benefit in stunting the growth of the weeds. The suppression of weed growth with subnormal dose rates was clearly more efficient than the use of higher than recommended crop seed rates, as has also been suggested by Andersson (1984). In each trial the above-ground biomass of weeds varied between crops (P<0.05) in Jokioinen, and was also explained with crop densities (P<0.01), both 1 month after applica- tion (data not shown) and at harvest (Fig. 2). Decreased weed suppression with subnormal dose rates was to some extent compensated for by the increased crop seed rate. Herbicide efficacy The herbicide efficacy (Fig. 3), measured as a relative reduction in weed biomass compared to untreated plots within each crop density, did not vary significantly between crop densities. There- fore, the efficacy values were pooled across the five densities. The response curves shown in Fig. 3 represent three different control situations; in 1986, the weed infestation was low and dominated by susceptible species, whereas in 3 Fig. I. Relationship between weed biomass and crop density of (a) barley and (b) wheat in 1986-88. Comparisons between the untreated plots and plots treated with either one-third dose of MCPA/mecoprop (MM) or one-third dose of MCPA/ flyroxypyr (MF) are presented. 486 1988 there was a high weed infestation with a more tolerant weed spectrum resulting in lower efficacy. An intermediate situation occurred in 1987. A parallel-line assay technique (Finney, 1979) with sigmoid dose-response curves, has been applied to desciibe the effects of herbicides (Streibig, 1988). The non-linear response was also found in our experiments (Figs 3 and 4), although our data were insufficient to cover the whole response curve. Within the dose interval used here, the linear model appeared, in some cases, to be a feasible approximation for des- cribing the effect of dose rate on the biomass production of a mixed weed population. Significant differences in efficacy between herbicides and their rates varied between trials. On average, higher control efficacy was achieved inJokioinen, where the most abundant weed species were Chenopodium album L. and Galeopsis speciosa L., compared to Ylistaro, where Stellaria media (L.) Vili., Matricaria spp. L. and Polygonum spp. L. were the main species. Effect ofdose reduction The best weed control was achieved with the highest herbicide rate, but lower doses often provided good weed control. The dose reduc- Fig. 2. Relative dry weight of broad-leaved weeds affected by crop density and herbicide rate. Relative value 100 equals the untreated plot at normal sowing density. Assessments were made at harvest time in Jokioinen.(0= -400, • = +4OO, + = -200, ■ = +2OO, � = normal [viable seeds of crop m“ 2].) 487 488 tion decreased the control efficacy of MCPA/ mecoprop less than that of MCPA/fluroxypyr (Fig. 3). The average efficacy of herbicide mixtures at recommended rates in Finnish spring cereal fields is 80-85% (Salonen & Erviö, 1987). With regard to the two products used in our trials, the intermediate dose rates of 2 and 0-75 1 (Figs 3 and 4) corresponded to a 25-30% reduction of the lowest recommended dose. Halving the highest recommended dose rate decreased the efficacy by 5-15 percentage units on average. With the one-third dose, which corresponded to a 50% reduction in the lowest recommended dose, the reduction of efficacy was 10-25 percentage units compared to the full dose. These results are in agreement with Swedish (Hagsand, 1983; Aamisepp, 1984; Lomakka, 1990) and German trials (Pallutt, 1991). The susceptibility of weeds to reduced herbicide doses varied among species (Fig. 4). MCPA/fluroxypyr showed a linear response with all the five species studied, but the slope of the line varied between species. Differences in the control efficacy between the two herbicide mixtures were most clearly demonstrated against V. arvensis and Polygonum spp L. Consequently, weed monitoring is a prerequi- site for successful adjustment of dose rate, since the dose responses are species-specific (e.g. Elba;k Jenson el al., 1989; Kudsk, 1989). The herbicides used in our experiment were Fig. 3. Percentge efficacy of (a) MCPA/mecoprop and (b) MCPA/fluroxypyr applied at three rates in different control situations in 1986 (■), 1987 (A) and 1988 (•) (see text for explanation). Data express the reduction in weed biomass compared with untreated plots 4 weeks after spraying. Fig. 4. Dose responses of (a) MCPA/mecoprop and (b) MCPA/fluroxypyr against diferent weed species. Figures in parentheses denote number of observations. Table 3. Regression coefficients (standard error S.E.) for the linear regression log (Y+l) = a+log bx+log cz relating the biomass of weeds (Y) at harvest to the crop density (x plants m-2) and the dose rate of herbicide (z 1 ha-1 ) in crop standsof spring barley and spring wheat treated with either MCPA/mecoprop or MCPA/fluroxypyr MCPA/mecoprop MCPA/fluroxypyr b(S.E.) b(S.E.) Crop/trial •10E-2 c(S.E.) •10E-2 c(S.E.) Spring barley JOKB6 -0-121(0-019) -0-117(0-039) -0-114(0-020) -0-303(0-106) JOKB7 -0-087(0-014) -0-136(0-032) -0-055(0-018) -0-127(0-109) JOKBB -0-112(0-014) -0-194(0-032) -0-098(0-013) -0.285(0-078) YLIB6 -0-134(0-017) -0-153(0-033) -0-125(0-020) -0-142(0-099) YLIBB -0-169(0-014) -0-086(0-034) -0-164(0-011) -0-194(0-073) Spring wheat JOKB6 -0 049(0 038) -0 130(0-059) -0-096 (0 035) -0-345(0-148) JOKB7 -0-129(0-015) -0-099(0-030) -0-158(0-017) -0-380(0-097) JOKBB -0-099(0-023) -0-232(0-052) -0-120(0-020) -0-807(0-120) YLIB7 -0-202(0-025) -0-096(0-050) -0-193(0-023) -0-163(0-119) common broad-spectrum mixtures for cereals, and they were not selected on the basis of the weed species present at spraying. Regression models of dose-responses should be developed for each herbicide against impor- tant weed species. However, the concept of effective dose (Streibig, 1989) is a function of many factors associated with the herbicide, the target weeds and environmental conditions, as well as interactions between these factors. This was also clearly demonstrated in our trials repli- cated in time and under different growth conditions. The linear regression model (Table 3), in- cluding both the rate of herbicide and the crop density as independent variables, explained 13- 80% of the variation in weed biomass at harvest in treated plots. The poor fit of the model was usually caused by the wide variation in data at the lower end of the dose range. In practice, the use of lower dose rates than those employed in our trials seems unrealistic and unreliable due to the increasing variation in control efficacy against mixed weed popula- tions. The effect of environmental factors, growth conditions and spray conditions on control also increases with decreasing dose rate. Subsequent effect on weed infestation Considerable differences in weed density were detected in untreated plots between the two crop rotations 1 year after the 3 year trial period. In the wheat-barley-wheat rotation the number of weed seedlings averaged 715 m“2 (S.E. 88, n =20), whereas in the barley-wheat-barley rotation it remained at the level of 336 m“ 2 (S.E. 50, n = 20). On the other hand, in treated plots the mean density of weeds was 265 plants m~ 2 (S.E. 7, n = 240). No significant differences between herbicides and their rates were detected. The weed biomass correlates closely with the seed production of weeds (Debeake, 1988; Wilson et al., 1988). Subnormal herbicide doses may not prevent weed seed production (Fogelfors, 1977). In our trials the measurement of the weed seed bank (Fig. 5) revealed no in- crease in the number of weed seeds in the soil Fig. 5. Relative number of germinated weed seeds (untreated barley = 100) in soil samples from crop stands of barley and wheat sown at normal seed rate. Samples were taken after 3 years of successive application of MCPA/mecoprop or MCPA/fluroxypyr with three dose rates (data pooled across herbicides. 489 after continuous use of reduced herbicide doses for 3 years. Pedersen & Rasmussen (1990) also found that a quarter of the normal dose of MCPA or chlorsulfuron adequately inhibited the seed production of weeds in barley stands compared with the normal dose. The observations from the soil samples (Fig. 5) revealed that the seed reservoir of weeds in the field was significantly (P< 0.001) higher in the wheat-dominated rotation than in the barley-wheat-barley rotation. The highest seed bank was found in the untreated plots of wheat rotation. In addition to lower competitiveness, spring wheat has a 20- to 30-day longer growth period than spring barley. Weeds effectively use this time for seed maturation and shedding before harvest. Therefore, weed control should be more efficient in wheat than in barley. Conclusions Chenopodium album and Galeopsis spp. are the most frequent and important weed species in spring cereals in Finland. Since they were effec- tively controlled even with the lowest dose rates, the possibility of making new detailed dose recommendations for cereal herbicides seems feasible. Lowering the dose reveals differentialsuscep- tibility between weed species, and may favour more tolerant species within one growing season. The long-term changes in weedpopula- tion are most efficiently prevented by choosing a herbicide according to the dominating weed species. The reduction of herbicide dose to half or even less is often possible in well-defined weed/ herbicide situations in spring cereals. The complexity of affecting factors creates a decision problem which is most efficiently solved with computerized expert systems. However, additional data on the competitive ability of the crop and enviromental factors are needed to develop a reliable advisory system for weed control, and particularly for the adjust- ment of adequate herbicide dose. References Aamisepp A. (1984) Behovsprövad ogräsbekämpning ivårsäd. Slutrapport. Ogräs och Ogräsbekämpning, 25:e svenska orgräskonferensen, Uppsala, pp. 33—47. Andersson B. (1984) Utsädesmängder och MCPA-doser i vårkorn. Ogräs och Ogräsbekämpning, 25:e svenska ogräskonferensen, Uppsala, pp. 49-58. Debaeke P. (1988) Dynamique de quelques dicotyledones adventices en culture de cereale. 11. Survie, floraison et fructification. Weed Research, 28, 265-279. Elbaek Jensen P., Juhl Petersen E. & Jorgensen H. (1989) Herbicider og vaekstreguleringsmidler under afprovning. Proceedings 6. Danske Plantevaernkonfer- ence/Ukrudt, pp. 62-93. Elbek Pedersen H. (1978) Use of reduced doses of herbi- cides in barley. Weeds and Weed Control, I9th Swedish Weed Conference, pp. 843-849. ErviO L.-R. (1983) Competition between barley and annual weeds at different sowing densities. Annales Agriculturae Fenniae, 22, 232-239. ErviO L.-R. & Salonen J. (1987) Changes in the weed popu- lation of spring cereals in Finland. Annales Agriculturae Fenniae, 26, 201-226. ErviO L.-R., Tanskanen T. & Salonen J. (1991) Profitabil- ity of chemical weed control in spring cereals. Annales Agriculturae Fenniae, 30, 199-206. Finney D.J. (1979) Bioassay and the practice of statistical inference. International Statistical Review, 47, 1-12. Fogelfors H. (1977) The competition between barley and five weed species as influenced by MCPA treatment. Swedish JournalofAgriculturalResearch, 7, 147-151. Gummesson G. & Fogelfors H. (1990) Möjligheter till minskad bekämpning av ogräs i stråsäd. 31 :a Svenska Växtskyddskonferensen, Uppsala, pp. 60-68. Hagsand E. (1983) Utsädesmängder och kemisk ogräs- bekämpning i sexradskorn i Norra Sverige. Röbäcksdalen meddelar, 1983:6, Umeå, 19 pp. Håkansson S. (1975) Grundläggande växtodlingsfrågor. I. Inflytande av utsädesmängden och utsädets horisontella fördelning på utveckling- och produktion i kortvariga växtbestånd. Sveriges Lantbruksuniversitet, Rapporter och Avhandlangar33, Uppsala, 192 pp. Heitefuss R., Gerowitt B. & Wahmhoff W. (1987) Development and implementation of weed economic thres- holds in the F.R. Germany. Proceedings of the 1987 British Crop Protection Conference Weeds, pp. 1025-1034. Hynninen E.-L. & Blomqvist H. (1990) Torjunta-aineiden myynti Suomessa 1989. Summary: Sales of pesticides in Finland in 1989. Kemia-Kemi, 17, 530-533. Kudsk P. (1989) Experienceswith reduced herbicide doses in Denmark and the development of the concept of factor- adjusted doses. Proceedings of the Brighton Crop Protec- tion Conference — Weeds, pp. 545-554. Lomakka L. (1990) Minimerad kemisk bekämpning av ogräs vid ensidig kornodling i Norra Sverige. 31 :a svenska växtskydds-konferensen, Ogräs och ogräsbekämpning, Uppsala, pp. 27-38. Pallutt B. (1991) Beiträge zur integrierten Unkrautbekämp- fund im Getreideanbau. Dissertation, Akademie der Landwirtschafts-wissenschaften der DDR, Berlin, 129 pp. Pedersen J.O. & Rasmussen I.A. (1990) Herbiciders in- flydelse på ukrudets fröproduktion. Proceedings 7. Danske Plantevaernskonference 1990, Ukrudt, pp. 73-83. Salonen J. & ErviO L.-R. (1988) Efficacy ofchemical weed control in spring cereals in Finland. Weed Research, 28, 231-235. Streibig J.C. (1988) Herbicide bioassay. Weed Research, 28, 479-484. Streibig J.C. (1989) The herbicide dose-response curve and the economics of weed control. Proceedings of theBrighton Crop Protection Conference Weeds, pp. 927-935. Thonke K.E. (1978) Influence of growth factors when using reduced doses of hormone herbicides in barley. Weedsand Weed Control, 19th Swedish Weed Conference, pp. 851-857. 490 Thonke K. (1986) Muligheder for anvendelse af reducerede doseringer af herbicider. Proceedings 3. Danske Plante- vaernkonferense/Ukrudt, pp. 117-124. Wilson 8.J., Peters N.C.8., Wright K.J. & Atkins H.A. Reduced herbicide doses in spring cereals 491 (1988) The influence of crop competition on the seed pro- duction of Lamium purpureum, Viola arvensisand Papaver rhoeas in winter wheat. Aspects of Applied Biology, 18, 71-80. Weed Research. 1992, Volume 32, 493-499 Yield responses of spring cereals to reduced herbicide doses J. SALONEN Agricultural Research Centre of Finland, Institute of Plant Protection, SF-31600 Jokioinen, Finland Received 11 October 1991 Revised version accepted 6 April 1992 Summary: Résumé: Zusammenfassung Reduction of the dose of MCPA/mecoprop and MCPA/fluroxypyr mixtures to half or one-third of the recommended rates still provided good weed control efficacy in spring wheat (Triticum aestivum L.) and particularly in spring barley {Hordeum vulgäre L). The average yield of treated plots was 8% higher in wheat and 1% higher in barley compared with untreated plots. However, yield reductions were observed in 32% of wheat plots and in 43% of barley plots treated with herbicides. The yield responses were poorly correlated with the weed density of mixed weed flora at the time of spraying. No reliable threshold density for chemical weed control was found. Réponses au rendement de céréales deprintemps, å des doses d’herbicides réduites La réduction de la dose de MCPA/mecoprop et MCPA/fluroxypyr å la moitié ou au tiers des taux recommandés a encore assure une bonne efficacité herbicide chez le blé de printemps (Triticum aestivum L.) et spécialement chez I’orge de printemps (Hordeum vulgäre L.). Le rendement moyen des parcelles traitées était de 8% plus élevé chez le blé et de 1% élevé chez I’orge que dans les parcelles non traitées. Cependant, des reductions de rendement ont été observées dans 32% des parcelles de blé et dans 43% des parcelles d’orge traitées aux herbicides. Les réponses du rendement étaient faiblement liées ä la densité en adventices d’une flore mixte au moment du traitement. Aucune densité seuil-sérieuse pour le desher- bage n’a été trouvée. Ertragsbildung von Sommergetreide bei redu- ziertem Aufwand von Herbiziden Eine Reduzierung des Aufwands von MCPA- Mecoprop- und MCPA-Fluroxypyr-Mischun- gen auf die Hälfte oder ein Drittel des empfohlenen Aufwands ergab in Sommerwei- zen (Triticum aestivum L.) und besonders in Sommergerste (Hordeum vulgäre L.) noch eine gute Unkrautbekämpfung. Im Mittel war der Ertrag der behandelten Parzellen beim Weizen um 8 % und bei der Gerste um 1 % höher als in den unbehandelten, aber in 32 % der Herbizidparzellen mit Weizen und in 43 % der mit Gerste wurden Ertragsminderungen beob- achtet. Die Ertragsbildung stand kaum in Beziehung zur Unkrautdichte zur Zeit der Behandlung. Es wurde keine verläßliche Schadensschwelle gefunden. Introduction Modern herbicides provide efficient control of weeds in cereals at a reasonable price compared with other economic inputs of cereal cultivation. Herbicides are primarily used to avoid yield loss by preventing weeds from interfering with cultivation, harvesting and marketing of grain (Elliott, 1978, 1980). However, environmental, economic and even political factors compel farmers to minimize their dependence on the use zf herbicides. Benefits from chemical weed control decrease with increasing yield level (Beer, 1979; Gummesson, 1987). The increasing crop yields (Mukula & Rantanen, 1987) and the decreasing infestation of weeds (Erviö & Salonen, 1987)in Finland gave an impetus to the evaluation of the present need for herbicides, particularly in spring cereals, which have a relatively high competitive ability. Spring cereals are grown on 53% of the cultivated field area in Finland (National Board of Agriculture, 1990). The use of herbicides in spring cereals is not always profitable (e.g. Evans, 1969; Courtney & Johnston, 1982; Aamisepp, 1984; Erviö et al., 1991). Substantial cost savings in chemical weed control are aimed at, either by using control thresholds (Heitefuss et al. , 1987)or by applying reduced herbicide doses (Kudsk, 1989). The objective of this study was to determine the yield responses of spring barley and spring wheat to chemical weed control, particularly with lower than recommended herbicide rates. The herbicidal efficacy in these field experi- ments has been describedby Salonen (1992). Materials and methods A total of nine field trials were conducted in 1986-1988 with a six-row spring barley (cv. Arra) and spring wheat (cv. Luja) in Southern (Jokioinen) and Central (Ylistaro) Finland. Cereals were sown on clay soil using five seed rates; 100, 300, 500, 700 and 900 viable kernels of barley m~2 and 200, 400, 600, 800 and 1000 kernels of wheat m~2 . Trials were sown with a combined drill and fertilizer applicator which placed the NPK fertilizer (90 kg N ha-1) between the crop rows under the seed bed at a depth of 7-9 cm. The distance between crop rows was 12-5 cm. The experimental plots were ploughed to a depth of 20 cm every autumn. All these measures followed the common farming practice in Finland, Two herbicide mixtures at three dose rates were applied at the 3- to 4-leaf stage of the crop (Zadoks’ scale 13-14) in order to control the broad-leaved weeds that emerged from the natural seed bank of the soil. The highest re- commended rate of commercial herbicide mixture containing MCPA (200 g a.i. 1~') and mecprop (400 g a.i. I -1) (Herbotal Plus) was 4-0 1 ha~' and that of MCPA (400 g a.i. I-1 ) and fluroxypyr (100 g a.i. 1~') (Starane M) mixture was 1-5 1 ha -1 . In addition half and one-third rates were sprayed. The experimental design was a split-plot with crop seed rate and herbicide treatments comprising main and sub-plot factors, respec- tively. There were four replicates. Grain yields were combined from an area of 21 m 2 from the centre of each 3 x 10cm plot. The crop yield was adjusted to 15% moisture content. Moreover, the following yield components and quality parameters were determined: number of ears, number of kernels per ear, thousand kernel weight, bulk weight, moisture content at harvest and Hagberg’s falling number of wheat. Results In 1987, the growing season was prolonged due to the poor weather conditions during the whole summer. Consequently, the spring wheat did not ripen sufficiently to be harvested, so that yield results from that year are not available. In other trials the crop yielded the average level for spring cereals in Finland. The weed infestation varied considerably between years and sites (Table 1). The domin- ant weed species in the field trials were Cheno- podium album L., Galeopsis spp. L., Stellaria media (L.) Vili. and Viola arvensis Murr. Chemical weed control efficiently reduced the biomass of weeds at all herbicide rates applied (Fig. 1). The results from 1988 are shown as an example of good efficacy which, however, resulted in only slightly positive or even erratic yield response compared with the untreated plots. The yield level varied between the years and the trial sites. However, the yield response to weed control was similar at all yield levels (Fig. 2). Herbicides neither caused visible phytotoxicity to the crop nor affected to the dry weight of crop plants. Table I. Occurrence of weeds at harvest in unsprayed plots sown with normal seed rates Weed infestation Density Biomass Year Site (m-1 ) (gm'2 ) Spring barley 1986 JOK 44 4 YLI 102 61 1987 JOK 322 10 1988 JOK 306 17 YLI 305 32 Spring wheat 1986 JOK 45 54 1988 JOK 301 112 JOK = Jokioinen, YLI = Ylistaro. 494 The proportional yield response to herbicide application compared with untreated crops varied between growth densities and trial sites (Fig. 3). Results were pooled across the years, since no significant interaction was found. The use of herbicides was naturally more profitable in crop stands of low growth density, and in Ylistaro, where the weed infestation was higher than in Jokioinen. The reduced herbiciderates of Jokioinen gave higher yield increases than the recommended dose. However, the latter was the most economic at Ylistaro, where the weed popula- tion was more tolerant for the herbicides used in our trials. When the lowest and highest crop densities were excluded from the calculations as in- appropriate growth densities in practical farming, the yield response to chemical weed control in spring wheat averaged 202 kg ha-1 , corresponding to a yield increase of 8%, and in spring barley 30 kg ha~* (1%). However, the yield response was negative in 32% of treated wheat plots and in 43% of barley plots. The graphical plotting of results from separ- ate trials (data not shown) revealed that yield responses to chemical weed control were poorly related to the weed density at the time of spraying. Yield responses to biomass produc- tion of weeds also remained fairly low even in 1988 (Fig. 1), when the weed infestation was highest. Spring barley in particular managed well, even without herbicides. Even at a rather high weed infestation, as in 1988,the benefit of usingreduced herbicide doses was clearly demonstrated in spring wheat (Fig. 4). The highest yield increases were achieved with subnormal rates of MCPA/fluroxypyr, des- pite the fact that the lowest weed biomass was in plots treated with the recommended rate. Fig. 1. Dose-response of crop yield (open symbols) and weed biomass (closed symbols) to the MCPA/mecoprop treatment in three growth densities of (a) spring barley and (b) spring wheatin Jokioinen in 1988. (O) indicates the nor- mal, (□) 200 seeds lower and (O) 200 seeds higher sowing densities. 495 496 Statistically significant differences in yield parameters between untreated and treated plots were only detected in some cases (Table 2). No differences were found between herbicides and their dose rates. Discussion The density of weeds was rather low in 1986 due to the long dry period after sowing (Table 1). Otherwise, the weed infestation was higher than the average density of 173 weeds m“2 in Finnish spring cereals, but the main weed species in our trials corresponded well with the prevailing weed flora of cereal fields in Finland (Erviö & Salonen, 1987). In the presentation of results, preference is given to data from 1988, when the weed infestation was highest. The yield benefits for barley resulting from chemical weed control remained rather low. Similar results have been noted elsewhere (Courtney & Johnston, 1986; Davies et al., 1989). Erviö et al. (1991) reported that the average yield increase with chemical weed control in spring cereal fields inFinland is 123 kg ha-1 , and that 60% of treatments are profitable. In Norway, herbicides have decreased the crop yield in 25% of cereal fields (Fykse, 1991), and in Germany more than half of the herbicide applications in spring cereals have been un- profitable (Gerowitt etal., 1984). Our trials were situated on clay soil where the harmful effects of weeds tend to be less than on other soil types (Kryger, 1985; Hallgren, 1989). In a comparison between soil types Jensen (1985) found that herbicide treatment resulted in yield decreases in 27% of all trials, but in 38% of the trials conducted on clay soils. Relationships between crop yield and weed infestation have been described with different models (e.g. Cousens, 1985;Håkansson, 1991). However, when the weed infestation is manip- ulatedby chemical control (Fig. 4), the effect of herbicide is also an important factor affecting the yield response of the crop (Brain & Cousens, 1989; Streibig et al., 1989). Yield responses to chemical weed control may be erratic, as in our trials with barley (Fig. 1), where the crop managed well in competition against weeds even without herbicide appli- cation, Obviously, the number of weeds assessed at the time of spraying is insufficient to describe the thresholds for chemical weed control in spring cereals (Bleiholder & Nuyken, 1986). On the other hand, the use of thresholds seems to be more reliable in winter cereals (Heitefuss et al., 1987). Yield results for spring wheat from 1988(Fig. 4) showed that even at relatively high levels of weed infestation the benefit of efficient weed removal can be partly lost if unnecessarily high herbicide rates are applied, as was also shown by Aamisepp (1984), Andersson (1984), Gummesson (1988) and Davies & Whiting (1990). The results from 1988 can be partly explained by the interaction of herbicide and long-lasting drought, which effectively sup- pressed the growth of weeds between herbicide application and weed assessment 1 month later. At Jokioinen, the reduced herbicide doses provided a control efficacy of 70-90% and, on Fig. 2. Yield responses of spring barley ( ) and spring wheat ( ) in Jokioinen and spring barley ( ) in Ylistaro to the control of broad-leaved weeds with different doses of (a) MCPA/mecoprop and (b) MCPA/fluroxypyr in crop stands sown with normal seed rate (500 and 600 seeds m" 2 ofbarleyand wheat(■) = 1986, (A) = 1987, (■) = 1988. average, gave higher yield increases than the highest dose, which suppressed weed infestation by more than 90%. This is in agreement with the conclusions of Thonke (1986) who reviewed several Scandinavian field trials. In Finland, the initial growth of spring cereals and weeds is rapid, and the outcome of their competition is difficult to predict. The dose reduction appears to be a more reliable weed control strategy than thresholds. In conclusion, reduced rates of herbicides provided adequate control of broad-leaved weeds in terms of crop yield, although the efficacy was lower than that of recommended Fig. 3. Percentage yield response of spring cereals to weed control with different rates of MCPA/mecoprop and MCPA/ fluroxypyr. Mean values from 1986-1988 in different sowing densities. 497 Table 2. Effect of herbicide treatment on yield parameters of crop sown with normal seed rate at Jokioinen Parameter Year Crop Untreated Treated P-value Grain moisture 1986 Barley 25 0 23*6 0 009 at harvest (%) 1988 Wheat 24-7 2M 0 001 1000kernel weight(g) 1986 Wheat 31-9 33-4 0 000 1988 Wheat 30 1 32 1 0 000 Bulk weight (kg) 1988 Wheat 76-2 77-5 0 000 Numberof ears (nr 2 ) 1987 Wheat 490 584 0 016 Numberof kernels/ear 1987 Barley 23-9 28-3 0 037 doses. High dose rates were profitable only if less susceptible weed species occurred or in crop stands of low competitiveness. Particularly in barley sown at recommended seed rates, low dose rates seem to be a feasible way of reducing the use of herbicides by at least 25-30%. Under favourable conditions a dose reduction of 50% or more may be possible. References Aamisepp A. (1984) Behovsprövad ogräsbekämpning i vårsäd. Slutrapport. Ogräs och Ogräsbekämpning, 25:e svenska ogräskonferensen, Uppsala, pp. 33-47. Andersson B. (1984) Utsädesmängder och MCPA-doser i vårkorn. Ogräs och Ogräsbekämpning, 25:e svenska ogräskonferensen, Uppsala, pp. 49-58. Beer E. (1979) Ermittlung der Bekämpfungsschwellen und wirtschaftlichen Schadensschwellen von monokotylen und dikotylen Unkräutern in Winterweizen und Wintergerste anhand von Daten aus der amtlichen Mittelpriifung. Dissertation. University of Göttingen. Bleiholder H. & Nuyken W. (1986) Neue Ansätze zur Darstellung und interpretation des Zusammenhanges zwischen den Deckungsgrad derUnkräuter unddem Ertrag von Getreide. Proceedings of the EWRS Symposium 1986, Economic Weed Control, pp. 61-68. Brain &. & Cousens R. (1989) An equation to describe dose responses where there is stimulation of growth at low doses. Weed Research, 29, 93-96. Courtney A.D. & Johnston R.T. (1982) The influence of competitive stress and the application of a herbicide, based on 2,4-DP/MCPA, on the components of yield in spring barley. Aspects ofApplied Biology I, 239-246. Courtney A.D. & Johnston R.T. (1986) An assessment of weed populations and yield responses on spring barley subjected to a programme of reduced herbicide usage. Proceedings of the EWRS Sympsoium 1986, Economic Weed Control, pp. 301-308. Cousens R. (1985) A simple model relatingyield loss to weed density. Annals ofApplied Biology, 107, 239-252. Davies D.H.K. & Whiting A.J. (1990) Effect of reduced herbicide dose on weed growth and crop safety in cereals and consequences for grain quality and harvesting. Proceedings of the EWRS Symposium 1990, Integrated Weed Management in Cereals, pp. 331-336. Davies D.H.K., Whiting A.J. & Whytock G.M. (1989). Yield responses to herbicide use and weed levels in winter wheat andspring barley in Scottish trials and consequences for economic models. Proceedings of the 1989 Brighton Crop Protection Conference Weeds, pp. 955-960. Elliott J.G. (1978) The economic objectives of weedcontrol in cereals. Proceedings of the 1987 British Crop Protection Coference Weeds, pp. 829-841. Elliott J.G. (1980) The economic significance of weeds in the harvestingof grain. Proceedings ofthe 1980 British Crop Protection Conference Weeds, pp. 787-797. Erviö L.-R. & Salonen J. (1987) Changes in the weed population of spring cereals in Finland. Annales Agricul- ture Fenniae, 26, 201-226. ErviO L.-R., Tanskanen T. & Salonen J. (1991) Profitabil- ity of chemical weed control in spring cereals. Annales Agriculturae Fenniae, 30, 199-206. Evans S.A. (1969) Spraying of cereals for the control of weeds. Experimental Husbandry, 18, 102-109. Fykse H. (1991) Skadetersklar for ugras. Norsk Land- bruksforsking, Suppl. 10, 40-43. Gerowitt 8., BodendörferH. & Heitefuss R. (1984) Zur Wirtschaftlichkeit des Herbizideinsatzes im Getrcide Auswertung von Versuchen des Pflanzenschutzdienstes den Jahren 1977-81. Zeitschrift fur Pflanzenkrankheitcn und Pflanzenschulz, Sonderheft X: pp. 127-135. Gummesson G. (1987) Kan kemisk bekämpning mot ogräs halveras med bibehållen lönsamhet. Växskyddsrapporter, Jordbruk, 42, 19-30. Gummesson G. (1988) Mängden ogräsmedel kan minskas genom bättre anpassning av dosen. Växskyddsrapporter, Jordbruk, 49, 13-20. Hallgren E. (1989) Influence of different factors on the effect of spraying cereals in the spring with Oxitril 4 as regards weeds and grain yield. 2. Influence of crop, developmental stage, prevailing conditions, geographical and climatic situation, soil type,organic content and N-rate on weed stand and effect on weeds. Composition ofweed stand at different relative yields. Weeds and Weed Control. 29th Swedish Weed Control Conference , pp. 39-74. Håkansson S. (1991) Growth and competition in plant Fig. 4. Relationship between the wheat yield and the biomass of weeds at harvest in three crop densities(400 to800). The in- festation levels of weeds were achieved with MCPA/ fluroxypyr applied at the recommended(1), half (2) and one- third dose (3). 0 indicates untreated plots. 498 stands. Swedish University of Agricultural Sciences, Crop Production Science 12, Uppsala, 241 p. Heitefuss R., Gerowitt B. & Wahmhoff W. (1987) Development and implementation of weed economic thresholds in the F.R. Germany. Proceedings of the 1987 British Crop Protection Conference Weeds, pp. 1025- 1034. Jensen P.K. (1985) A review of yield responses to weed control in one thousand spring barley experiments. Proceedings of the 1985 British Crop Protection Conference Weeds, pp. 687-692. Kryger J. (1985) Muligheder for fastsaettelse af skadetaerskler i vårbyg. 2. Danske Plantevaernkon- konference! Ukrudt, pp. 203-216. Kudsk P. (1989) Experiences with reduced herbicide doses in Denmark and the development of the concept of factor- adjusteddoses. Proceedings ofthe BrightonCrop Protection Conference Weeds, pp. 545-554. Mukula J. & Rantanen O. (1987) Climatic risks to the yield and quality of field crops in Finland. I. Basic facts about Finnish field crops production. Annales Agriculturae Fenniae, 26, 1-18. National Board of Agriculture (1990) Monthly review of AgriculturalStatistics, n:o 7:23. Helsinki. Salonen J. (1992) Efficacy of reduced herbicide doses in spring cereals of different competitive ability. Weed Research, 32, 000-000. Streibig J. (1983) Ukrudtssprojtning og merudbytte i korn. Ugeskrift for Jordbrug, 128, 811-816. Streibig J.C., Combellack J.H., Pritchard G.H. & Richardson R.G. (1989) Estimation of thresholds for weed control in Australian cereals. Weed Research, 29, 117-126. Thonke K.E. (1986) Muligheder for anvendelse af re- ducerede doseringer af herbicider. 3. Danske Plante- vaernkonference/Ukrudtpp. 2117-124. 499 Performance of reduced herbicide doses in spring cereals Jukka Salonen Salonen, J. Performance of reduced herbicide doses in spring cereals. Agric. Sci. Finl. 2: 000-000. (Agric. Res. Centre of Finland, Inst. PI. Prot., FIN-31600 Jokioinen, Finland.) The consequences of dose reduction of three new herbicide formulations were studied for the control ofannual broad-leaved weeds in fields of spring barley (Hordeum vulgare L.) and spring wheat (Triticum aestivum L.). The herbicide formulations were MCPA/mecoprop-P, MCPA/dichlorprop-P and MCPA/fluroxypyr. The efficacy of the lowest recommended dose and a 30% lower rate were tested and compared with the reference herbicide tribenuron-methyl. Trials were conducted at seven sites for three years. Considerable annual fluctuations in weed infestation were recorded. Although the dose reduction occasionally caused considerable decline in control (on %-scale), sup- pression of weed biomass was still satisfactory in most of the trials. On average, a 75% reduction of weed biomass in spring barley and an 83% reduction in spring wheat were achieved with reduced herbicide doses. Use of reduced herbicide doses for three years in the same fields caused neither a significant increase in weed infestation nor changes in the species composition of weed populations compared with treatments at recom- mended rates of application. There was a significant difference in biomass production between weed species. Consequently, the total biomass production of annual dicotyle- donous weeds correlated only weakly (r=0.48) with the total weed density. Even in untreated plots the weed biomass at harvest constituted, on average, only 3.1-3.6% of the total vegetative biomass of crop stands. Thus, the crop yield responses to chemical weed control remained low. Key words: spring barley, spring wheat, broad-leaved weeds, MCPA/mecoprop-P, MCPA/dichlorprop-P, MCPA/fluroxypyr, tribenuron-methyl Introduction Reduced herbicide doses have provided adequate control of broad-leaved weeds in many recent cer- eal experiments (e.g. Baandrup and Ballegaard 1989, Davies et al. 1989, Fogelfors 1990, Kem- mer and Hurle 1990, Proven et al. 1991, Sa- lonen 1992a). At present, political Action Plans stipulate thereduction of pesticide use in the Nordic countries (Thonke 1991, Ympäristöministeriö 1992). Reduction of herbicide doses is one of the measures suggested and studied to achieve this ob- jective. In the Nordic countries (Denmark, Finland, Nor- way, Sweden), herbicides represent 60-80% oftotal pesticide use (Thonke 1991, Markkula et al. 1990). Herbicides are commonly used to control broad-leaved weeds in fields of small-grain cereals, which represent the most widely cultivated crops. Therefore, special attention is paid to optimization of herbicide use in cereal fields as cereals are prob- ably able to out-compete weeds even at low rates of herbicide application. 1 Agric. Sei. Fint. 2 (1993)2nd Proof The recommended herbicide doses given on the product labels are normally suggested by chemical companies and then officially tested and approved by the relevant national authorities. The recom- mended "normal" dose implicitly ensures reliable weed control in most situations. The use of factor- adjusted doses is, however, emphasized by the ex- tension service and computer-based advisory sys- tems (Kudsk 1989, Baandrup and Ballegaard 1989,Jennéus 1991). Formulated mixtures of MCPA/dichlorprop and MCPA/mecoprop are commonly used in spring cer- eal crops in Finland (Hynninen and Blomqvist 1993). To date, the commercial formulations of phenoxypropionic acids, dichlorprop and meco- prop, have been mixtures of two optical isomers, R <+) and S* *. However, only the R(+) isomer is an active part of herbicide. Recently, these isomers have been separated, and formulations containing only the active isomer have been developed (Squires et al. 1987). Replacement of conven- tional racemic isomers by the new active isomers, dichlorprop-P and mecoprop-P, results in approx- imately 50% reduction in the use of the active in- gredients, dichlorprop and mecoprop. The first commercial products containing active isomers were registered in Finland in 1992. The purpose of this study was to investigate pos- sibilities for reducing the lowest recommended ap- plication rates of the new cereal herbicidesby 30%. The risk of failure was assessed, and the con- sequences of continuous use of reduced herbicide doses on weed infestation were studied. Further- more, crop yield responses to chemical weed con- trol were measured. Material and methods Field experiments were conducted at seven experi- mental stations of the Agricultural Research Centre. Four stations (Anjalankoski (KYM), Jokioinen (RKA), Kokemäki (SAT) and Mietoinen (LOU)) are located in southern Finland and three stations (Mouhijärvi (SAH), Pälkäne (HÄM) and Ylistaro (EPO)) in central Finland. The same trial protocol was used for three years, 1989-1991, in spring bar- ley and spring wheat monocultures in the same field. At each site there was one spring wheat (cv. ’Luja’) trial and at four sites (EPO, KYM, LOU, RKA) there was a spring barley (cv. ’Pohto’) trial. Thus, during the 3-years of experimentation there were in total 21 spring wheat trials and 11 spring barley trials. The experiments were established in 1989 in fields where spring cereals were sown in 1988. The crops were sown at the recommended seed rates: 450 viable seeds of barley and 600 seeds of wheat m 2. Various soil types from ranging from sandy clay to heavy clay were represented. The experi- mental plots (4.0/5.0 m x 12 m) were ploughed to a depth of 20-25 cm every autumn. Commercial herbicide formulations of MCPA/mecoprop-P (270/305 g a.i. 11,I 1, 'Duplosan KV-M’) for use in wheat fields and MCPA/dichlor- prop-P (265/285 g a.i. f l , 'Duplosan DP-M’) in barley fields were applied at their lowest recom- mended rates and at 30% lower rates. MCPA/fluroxypyr (400/100 g a.i. 11,I 1, 'Starane M’) was applied to both crops. In addition, tribenuron- methyl (750 g kg' 1 granular formulation, ’Express 75 DE’) (Ferguson et al. 1985) was used as a reference herbicide (Table 1). New formulations of phenoxy acid herbicides containing only the optically active isomers of di- chlorprop and mecoprop (Squires et al. 1987) were Table 1. Treatments in the field experiments in spring barley and spring wheat fields in 1989-1991.MCPA/dichlorprop-P was applied only in spring barley and MCPA/mecoprop-P only in spring wheat. Treatment Herbicide dose lha' 1 g a.i. ha' 1 Unsprayed 0 MCPA/fluroxypyr 0.70 280/ 70 MCPA/fluroxypyr 1.00 400/100 MCPA/dichlorprop-P 1.25 331/356 MCPA/dichlorprop-P 1.75 464/499 MCPA/mecoprop-P 1.25 338/381 MCPA/mecoprop-P 1.75 473/534 Tribenuron-methyl 11 7 g 5.3 11 Non-ionic surfactant (’Citowett’) 0.05% was added to the spray solution (water volume 200 I ha’ 1). 2 2nd ProofAgric. Sei. Finl. 2 (1993) included in the experiments. The objective was to investigate whether the positive results of reducing the recommended doses of racemic mixtures (SA- LONEN 1992a) would apply also to new formula- tions. Fluroxypyr was introduced into the official screening trials in Finland in 1982 as a new herbi- cide for weed control in cereal crops, with particu- lar effect on Galium aparine L. (Paul et al. 1985). Treatments were arranged as a randomized com- plete block design with four replicates. Herbicides were applied at the 3- to 4- leaf stage of the crop (Zadoks’ scale 13-15 (Zadoks et al. 1974)) with a portable van der Weij propane sprayer that deliv- ered 200 1 ha 1 spray solution at a pressure of 300 kPa. Herbicides were applied between the end of May and mid-June, about one month after sowing. The temperature at the time of application ranged from 10 to 25°C, and the relative humidity from 33 to 77%. The emergence of crops and weeds were moni- tored before the herbicide application. Crop devel- opment (growth stages) and weed emergence were recorded. Weeds were assessed in 0.25 m‘ 2 sample plots. Annual dicot weeds were counted 0-1 days before spraying (with some exceptions of 2-4 days delay). Furthermore, the weed infestation (number and air- dry weight per unit area) was assessed one month after spraying and at harvest. The relative number ofemerged weeds at the time of spraying herbicides was calculated by comparing the number of weeds (No. m'2 ) at spraying and one month later. Crop yield results are given at 15% moisture content. The impact of different control regimes on the subsequent weed infestation was assessed one year after the 3-year trial period in 1992. Weeds were counted at the timeof spraying herbicides in spring cereal fields. Statistical analysis Analysis of variance was applied to weed and crop data by introducing Year as a within-subject factor and Site, Treatment and Block as between-subject factors. The random factor Block was nested in the site. The biomass ofweeds, as a dependent variable, was transformed with the common logarithm log(y+l) to achieve normal distribution and homo- geneity of variances. Weed density (No. m 2) was transformed with square root. The data from un- sprayed plots were excluded from the final statisti- cal analyses. The effect ofherbicide dosereduction was tested with single degree-of-freedom contrasts. The effect ofweed infestation (density, biomass) on crop yield was tested withregression analysis. Stat- istical analyses were done with the General Linear Models procedure of the SAS statistical package (SAS Institute Inc. 1990). Results Occurrence of weeds Weed density at the time of herbicide application varied within therange of7-702 weeds m 2 (Fig. 1). Also, the relative number of weeds which emerged before herbicide application, compared with the number of weeds per unit area one month later, varied considerably (Fig. 2). On average, 72% of the annual dicotyledonous weeds emerged before spraying. Crop plants usually reached at least the second leaf stage (Zadoks’ scale 12-13) before the first flush of weeds. Most weed seedlings were between the cotyledon stage and the first true-leaf stage at the time of herbicide application. The predominant weed species in the experimen- tal fields were typical of Finnish cereal fields (c.f. ErviÖ and SALONEN 1987). The weed populations varied between sites (Table 2) and, to some extent, between years at the same site. The most frequent and abundant weed species were Chenopodium al- bum L., Fumaria officinalis L., Lamium L. spp., Stellaria media (L.) Vili., Matricaria L. spp.(in- cluding Tripleurospermum inodorum Schultz Bip.) and Viola arvensis Murray. Volunteer turnip rape (Brassica rapa L. subsp. oleifera DC.) occurred in those fields (LOU, SAT) where there were trials with turnip rape some years before the experiment. The weed biomass (air-dry weight, DW) in the untreatedplots ranged from 0.4 (SE 0.2) to 61.5 (SE 3 Agric. Sei. Fin I. 2 (1993)2nd Proof Table 2. Predominant weed species in the experimental fields. Site Weed species 1 * Anjalan- CHEAT FUMOF GALSS POLCO koski (KYM) Jokioinen CHEAT LAMSS STEME VIOAR (RKA) Kokemäki BRSRO CHEAT THLAR VIOAR (SAT) Mietoinen BRSRO FUMOF LAMSS STEME (LOU) Mouhijärvi CHEAT MYOAR TRFPR VIOAR (SAH) Pälkäne CHEAT MATSS STEME VIOAR (HÄM) Ylistaro LAMSS MATSS POLCO STEME (EPO) 0 BAYER codes for weeds (BAYER 1992): BRSRO = Brassica rapa spp. oleifera (volunteer), CHEAT =Chenopo- dium album, FUMOF = Fumaria officinalis, GAESS = Gaieopsis spp., GALS = Galium spp., LAMSS = Lamium spp., MATSS =Matricaria spp., POLCO = Fallopio convol- vulus, STEME = Stellaria media. THLAR = Thtapsi ar- vense, TRFPR =Trifolium pralense, VIOAR = Viola arven- sis. (LOU) Fig. 1.Weed infestation in the experimental fields at the time of herbicide application in a) spring wheat and b) spring barley. The experimental sites are: EPO =Ylistaro, KYM = Anjalankoski, LOU =Mietoinen, RKA = Jokioinen, HÄM = Pälkäne, SAFI = Mouhijärvi, SAT =Kokemäki. Fig. 2. Weed emergence in unsprayed plots at the time of herbicide application given as a percentage (five classes) of the weed density (No, m"2) one month later in 33 spring cereal experiments in 1989-1991. 4 2nd ProofAgric. Sei. Fint. 2 (1993) 2nd Proof 12.2) gDW m 2 one month after herbicide applica- tion, and from 1.8 (SE 0.9) to 116.3 (SE 7.7) g DW m 2 at harvest. Biomass production varied consider- ably between weed species. Consequently, the total weed biomass in unsprayed plots, one month after spraying, weakly correlated (r=0.48) with the total weed density at spraying. The most competitive weed species producing the highest biomass per plant were volunteer turnip rape (0.60 g DW plant '), Galeopsis L. spp. (0.27 gDW plant 1) and Fallopio convolvulus (L.) A. Love (0.15 g DW plant 1). The biomass production ofbarley averaged 1.10 g plant 1 and that of wheat 1.02 gplant 1 at the four sites whereboth crops were grown in the same field. Herbicide efficacy Generally, all herbicides were effective when ap- plied at the lowest recommended dose, and 30% dosereduction reduced the efficacy, on average, by less than 10 percentage units (Fig. 3). However, a significant (PcO.001) Year*Site*Treatment inter- action was detected (Table 3). This indicates that there were differences in the effectiveness of weed control between sites and between years within a site. Tribenuron-methyl was the most effective herbicide in most trials (Fig. 4), particularly when Matricaria spp. and Lamium spp. (EPO, LOU) were the predominant weed species. The reduction in herbicide efficacy was consid- ered significant if 30% dosereduction caused more than 15% reduction in efficacy (on %-scale) com- pared with the efficacy achieved with the recom- mended dose. In wheat trials such a reduction (>15%) in the biomass-based efficacy occurred in 29% of plots treated with MCPA/mecoprop-P, and in 14% ofplots treated with MCPA/fluroxypyr. The corresponding figures for barley trials were 8% with MCPA/dichlorprop-P and 19% with MCPA/fluroxypyr. To describe the probability ofachieving a certain level of weed control, herbicide efficacy was calcu- lated for each treatment within each replicate and theresults were ranked in four efficacy classes (Fig. 5). Accordingly, treatment with reduced herbicide doses still provided at least 70% control in 70-89% of plots monitored. At the recommended herbicide doses the 70% efficacy level was reached in 78- 91% of cases. Weed biomass in sprayed crop stands one month afterherbicide treatment was less than 15 g DW m 2 in every trial. Dry weight of crop plants in un- sprayed plots averaged 506 (SE 24) g DW m 2 in barley and 482 (SE 15) g DW m 2 in wheat. Dry weight of weeds was significantly (P<0.01) higher in unsprayed than in sprayed plots. To simplify the ANOVA analyses, data from the unsprayed plots were not included in the final analyses (Table 3). Only in some fields was the weed biomass signifi- Fig. 3. Efficacy of herbicides determined as % reduction of weed density (light bar) and dry weed biomass (dark bar). The mean efficacy and the SE of the mean in a) 12spring barley trials and b) 21 spring wheat trials during 1989-1991. 5 Agric. Sei. Finl. 2 (1993) 6 Agric. Sei. Fin!. 2 (1993) 2ndProof 2nd Proof cantly higher in the plots treated with reduced doses than in the plots treated with normal doses. In gen- eral, weeds produced more biomass in wheat stands than in barley stands (Fig. 4). The dose reduction of MCPA/mecoprop-P de- creased the effect of control particularly against Myosotis arvensis (L.) Hill, Matricaria spp.. Poly- gonum L. spp. and Viola arvensis. Similarly, the dose reduction of MCPA/fluroxypyr significantly decreased (>lO %-units) the efficacy against Pu- ntaria officinalis, Matricaria spp., Polygonum spp. and Viola arvensis. Conclusions concerning MCPA/dichlorprop-P were not drawn due to the limited number of observations. Crop-weed interactions The yield of spring barley and spring wheat aver- aged 4,900 kg ha 1 and 3,700 kg ha ', respectively. In the trial plots treated with herbicides the mean yield of wheat was 1.9% higher and barley yield was 4.0% higher than in untreated plots. The monetary value ofsuch a yield increase ranges from FIM 150 to FIM 300 which is sufficient to cover the average cost of (FIM 100) herbicides for broad- leaved weed control. Herbicide treatment did not reduce crop yield significantly (P<0.05) in any trial. There was no significant difference in the mean crop yield from plots which received a recom- mended dose and those which received a reduced dose of herbicide. Only in one trial from 21 wheat trials did the dose reduction of MCPA/mecoprop-P result in a significantly (P<0.01) lower wheat yield, and once, in the same trial, with a reduced dose of MCPA/fluroxypyr (P<0.02). The proportion of weed biomass from the total vegetative biomass of cereal fields was relatively low (Table 4). Fig. 4. Comparison of the remaining weed biomass in 1989-1991 in a) spring barley and b) spring wheat one month after treatment with different herbicide formulations and doses. The figures in paranthesis indicate the air-dry weed biomass (g m 2) in the unsprayed plots each year. The experimental sites are: EPO = Ylistaro, KYM = Anjalankoski, LOU =Mietoinen, RKA = Jokioinen, ffÄM =Pälkäne, SAff =Mouhijärvi, SAT = Kokemäki. 7 Agric. Sei. Fint. 2 (1993) Table 3. Repeated measurements analysis of variance (ANOVA) of weed biomass log(Y+l) in sprayed plots in spring barley at four sites and in spring wheat at seven sites. Trials were repeated for three years. Air-dry weight of weeds was measured one month after herbicide application. Crop Degrees Type 111 F-value F-test of Mean probability Source of variation freedom Square Spring barley Belween-suhject effect Site 3 3.64 28.83