Journal of Applied Botany and Food Quality 87, 234 - 242 (2014), DOI:10.5073/JABFQ.2014.087.033 University of Rostock, Faculty of Agricultural and Environmental Sciences, Crop Health, Rostock, Germany Important maize weeds profit in growth and reproduction from climate change conditions represented by higher temperatures and reduced humidity Kristian Peters*, Bärbel Gerowitt (Received May 27, 2014) * Corresponding author Summary Climate change is predicted to result in rising temperatures and reduced precipitation during spring and summer in Central Europe. As a consequence, crops and weeds will be affected. Our study fo- cuses on the three weed species in maize Amaranthus retroflexus, Echinochloa crus-galli and Setaria viridis. These weeds occur nu- merously in European maize fields and populations are likely to further increase. Yet, there is a lack of knowledge about particular biological strategies of the weeds. Our study focuses on how the weed species respond biologically to the climate change conditions. Experiments were conducted in two climate chambers with a 2 °C difference in temperature and the warmer one with 13 % less humid- ity. Emergence, development, biomass and seed production were de- termined of the weeds grown individually in pots and grown within maize. All tested weed species were taller during the first weeks un- der the climate change scenario. At later growth phases there was a trade-off between traits measured during vegetative growth and at the time when seeds were produced. To summarize the results, the weed species profited in the order E. crus-galli, S. viridis and A. retroflexus from the climate change conditions. Knowledge of the weeds biological responses to the predicted conditions helps to re- duce their long-term population development by targeting crop pro- tection measures at specific growth phases of the weeds. To ensure control of the tested weed species under climate change conditions various weed management strategies are necessary. Introduction Climate change will result in rising temperatures (TUBIELLO et al., 2007) and modified precipitation (ROBINSON and GROSS, 2010). Summer droughts will be more likely and will affect weeds in spring- sown crops. According to PATTERSON et al. (1999) temperature and precipitation are the most important factors for the geographical distribution, the growth and the competitive abilities of weeds. Our study focuses these two climate variables. We have chosen three important Central European C4 weeds in maize: Amaranthus retro- flexus, Echinochloa crus-galli and Setaria viridis, since currently a shift of the local weed flora is occurring in favour of weeds of the C4 photosynthesis type. Most of these weeds are late germinators and emerge from early summer to early autumn. Thus, they are con- sidered thermophilous in Central Europe and these weeds are ex- pected to migrate further north with changes in climate conditions (WALTHER et al., 2002). These weeds are already most numerous and most competitive in maize fields of Southern Europe (NOVÁK et al., 2009). They may also extend their damage potential in spring- sown crops, such as maize, under the predicted future conditions in Central Europe. Therefore, our study focuses on three thermophilous C4 weeds, whose future status and biological strategies are not well understood for Central European populations. Amaranthus retroflexus L. (redroot pigweed) is a successful weed in maize (OVEISI et al., 2013). Fast growth and indeterminate flowering enable a single plant to produce up to 500,000 seeds (STECKEL et al., 2004). The reaction of European populations to warming and dif- ferent humidity is not well studied so far (HYVÖNEN, 2011). Never- theless, because the species produces more biomass and more seeds in a warmer climate in North America and Canada (SCHIMPF, 1977), we expect the species to develop better with warmer conditions in Europe as well (KIGEL et al., 1977; OVEISI et al., 2013). Setaria viridis (L.) P. Beauv. (green foxtail) is currently the most wide-spread species of the Setaria genus in Europe (DEKKER, 2003). It generally emerges after the last herbicide treatment in maize fields. Thus, plants are often not affected by herbicides (MEHRTENS et al., 2005; BECKIE and TARDIF, 2012). A high genetic variability enables the species to grow under a great range of temperatures (DEKKER, 2003). Plants of North-American populations produce less bio- mass but increase generative reproduction under warmer conditions (SWANTON et al., 1999). Echinochloa crus-galli (L.) Beauv. (barnyard grass) exhibits high phenotypic plasticity (BARRETT and WILSON, 1981; MAUN and BAR- RETT, 1986). Central European populations are most competitive under high temperatures, high nutrient availability and mid level humidity (OTTE et al., 2006). The species benefits from higher tem- peratures but not from dry conditions (BARRETT and WILSON, 1981; CHAUHAN and JOHNSON, 2011). Most studies focus on the development of weeds under various ag- ricultural practices or on their resistance to herbicides (BARRETT and WILSON, 1981; POTVIN, 1986; CHAUHAN and JOHNSON, 2011). With a special focus on European populations, we ascertain a lack of biological knowledge concerning the vegetative development, the development speed and the generative reproduction for the studied weeds under climate change conditions. The aim of our study is to explore how the three species perform under the predicted future conditions and which distinct biological strategies they realise. Fur- thermore, this study attempts to assess which biological weed prop- erties are important with climatic changes. Based on these biological data, a better framework for weed management and crop protection can be devised. To achieve this, the effects of the raised temperature and less humidity on the weeds were studied combined in the experi- ment. Materials and methods The three weed species Amaranthus retroflexus, Setaria viridis and Echinochloa crus-galli used in the experiments were collected near Göttingen, Germany in 2007. Seeds originated from different field populations in that area and were stored for two years until the ex- periments started in 2009. Preliminary germination tests showed high germination rates (approx. 50-75 % germination percentage). Thus, the seeds were not subjected to any special treatment. The experiment was conducted in two climate chambers, 2.15 m in width, 3.45 m in length and 2.10 m in height. Temperature, humid- ity and light were independently adjustable. The first climate cham- ber had temperatures that represent the conditions of the current Climate change and three maize weed species 235 Tab. 1: Conditions in the two climate chambers. Values given after ± are the mean variations between replications. period factor chamber with current climate chamber with predicted future climate 0-2 weeks night/day temperature 7 °C/13 °C ±0.3 % 9 °C/15 °C ± 0.6 % night/day length 12/12 12/12 night/day humidity 70 %/66 % ± 1 % 58 %/52 % ± 3 % 2-6 weeks night/day temperature 9 °C/18 °C ± 0.3 % 11 °C/20 °C ± 0.6 % night/day length 14/10 14/10 night/day humidity 70 %/66 % ± 1 % 58 %/52 % ± 3 % 6-12 weeks night/day temperature 11 °C/20 °C ± 0.3 % 13 °C/22 °C ± 0.6 % night/day length 14,5/9,5 14,5/9,5 night/day humidity 70 %/66 % ± 1 % 58 %/52 % ± 3 % 12-15 weeks night/day temperature 11 °C/20 °C ± 0.3 % 13 °C/22 °C ± 0.6 % night/day length 14/10 14/10 night/day humidity 70 %/66 % ± 1 % 58 %/52 % ± 3 % 15-21 weeks night/day temperature 9 °C/18 °C ± 0.3 % 11 °C/20 °C ± 0.6 % night/day length 13/11 13/11 night/day humidity 70 %/66 % ±1 % 58 %/52 % ± 3 % climate of Northern Germany. The humidity (air moisture content) was raised slightly with water nozzles when compared to the sec- ond chamber that represented predicted future climate conditions (Tab. 1). Mean humidity was set to 68 % in the climate chamber with current conditions and was 13 % lower in the other chamber due to the disabled water nozzles. As light source ten Phillips Son-T Agro 400 were used, which were embedded in the ceiling of each climate chamber. Light levels at the centre of the chambers were 24,900 lux measured 90 cm from the ground and 120 cm below the lights. This is equal to a PPFD of approx. 400 μmol/s/m2 at the aforementioned distance. The chosen conditions were based on the A1B scenario of the IPCC (2013), which predicted an increase in temperatures of 2 °C until 2070 for Central Europe and less humidity during summer months. Thus, daily minimum and maximum temperatures were always set 2 °C higher in the second chamber (Tab. 1). The day-length was the same in both chambers, but adjusted over time to simulate an ad- vancing season. To ensure the intended climate levels (Tab. 1), tem- perate and humidity in the chambers were continuously adapted and monitored with data loggers 5 cm above ground and with additional air temperature sensors 1.60 m above ground. Each climate chamber had two large plant tubs, 1.30 m x 1.10 m in size and a depth of 0.90 m. The bottom of each tub was equipped with a 2 cm thick layer of small stones to allow excessive water to run off through small holes at the bottom-side of the tubs. Gauze was stretched above the stone layer to prevent soil from the 60 cm thick soil layer above to agglutinate at the bottom. Maize seeds were sown directly into the tubs’ soil (at a depth of 3 cm) at the begin- ning of each replication to simulate a row typical for maize fields (80 cm row distance, 10 cm plant distance in row). These tubs were already installed for preliminary studies and thus were equipped with an established soil-bed under the two climate conditions. Due to the resulting weight, they were immovable between the chambers. Therefore, the tubs rested in the same climate chamber under the same conditions in time of the experiment. The tubs were fertilized with 90 g Compo Hakaphos® Blue (equal to 135 kg N and 90 kg P per ha) before the start of each replication and again before weed seedlings were planted. For studying the emergence of the weed species and to provide seed- lings for further experiments, 130 seeds of each species were sown at a soil depth of 1 cm in five germination trays in each climate cham- ber at the same time. The number of seedlings was counted every day (Fig. 1). GDD = ( T max +T min 2 ) Tb emergence sowing early-growth 0 4 12 mid-growth final-growth 21 plant tubs + m aize crop plant pots harvest harvest time [weeks] Fig. 1: Overview of the treatments performed in each replication in both climate chambers. In order to characterize, compare and correlate processes during emergence, various parameters were calculated (GRUNDY, 2003). For predicting the cumulative emergence of seedlings, Growing De- gree Days (GDD) were calculated (DORADO et al., 2009): where Tmax and Tmin are the daily maximum and minimum tempera- tures measured by the dataloggers in the climate chamber. Tb is the base temperature for each weed. The Tb values were taken from various authors: Tb=4.0 °C for A. retroflexus from GARDARIN et al. (2009), Tb=6.2 °C for E. crus-galli from GUILLEMIN et al. (2013) and Tb=6.1 °C for S. viridis from GARDARIN et al. (2010). In order to estimate and compare the progress of seedling emer- gence, Mean Emergence Time (MET) and Emergence Rate Index (ERI) were calculated as follows (DORADO et al., 2009): 236 K. Peters, B. Gerowitt where Ni represents the newly emerged seedlings since the previous count, ti represents the GDD after sowing and n is the number of sampling occasions. For characterizing emergence speed, the emergence rate at mid emergence (v50) was additionally calculated (GARDARIN et al., 2011, modified): where m is the number of emerged seeds until mid emergence, b is a shape parameter correlated with the emergence rate at mid emer- gence, D0 represents the day after sowing on which first emergence was recorded, and D50 is a factor which represents the day on which 50 % the total emerged seedlings were recorded. The prediction curve of the cumulative emergence Gi was fitted (GARDARIN et al., 2011, modified): where Di represents the day of measurement. Seedling height of every emerged seedling was determined after 4 weeks (BBCH stage 15) (Fig. 1, early-growth phase). Ten random- ly chosen seedlings of each species were then planted in pots in each climate chamber to study further development. Pots were 30 cm in diameter and 20 cm deep. Another 10 seedlings of each weed were put in the two plant tubs between the maize rows in each climate chamber. Sampling and positioning of the weeds was random. Both maize and weeds in the large plant tubs were harvested at the end of the mid-growth phase (Fig. 1, beginning of flower onset of maize, BBCH stage 53, after approx. 12 weeks). Plant height and above ground dry mass were determined for each weed and maize plant. Height and development stage (BBCH, HESS et al., 1997) were additionally determined for the weeds in the smaller plant pots. The remaining weeds in the pots were harvested at the reproduction phase (Fig. 1, after 21 weeks) and their height and vegetative above ground dry mass (without seeds) were determined. Panicles, tillers and seeds of all weed plants were also counted. The experiment involved three replications, which were conducted consecutively. Each replication in time was set-up independently. The calculated emergence coefficients were tested for significant dif- ferences using the t-test (Fig. 3). As described above, two tubs were available within each climate chamber. Hence, the involved factor levels (chambers, tubs) were not arranged orthogonally, excluding classical ANOVA approaches. Instead, for comparing the biological parameters at the end of the early and mid-growth phases, as well as the reproduction phase (Tab. 2, 3), Linear Models with Random Effects (LMER) were used. In these models, climate (temperature and humidity combined) was chosen as fixed factor, whereas repli- cations in time were introduced as random factor. When tubs were involved (Tab. 4), they were also introduced as random factor. Each weed species was tested separately for significant differences be- tween the climate chambers. To investigate the properties of the ran- dom coefficients of the LMER, probability values for the parameters of models fitted with LMER were calculated with a Monte Carlo Markov Chain (MCMC) approach, choosing 10,000 sample simula- tions of the model (BAAYEN et al., 2008). Distribution and homoge- neity of variance were visually checked with the help of histograms and diagnostic plots. Data were log or square root transformed be- fore statistical analysis if conditions of normality were not met or to improve homogeneity of variances. The residual vs. fitted plots and the normal qq-plots were examined for each tested parameter according to FARAWAY (2006). Statistical analysis was carried out with the software R (IHAKA and GENTLEMAN, 1997). The additional packages languageR, Hmisc, agricolae and lme4 were used. Results Seedling emergence In both climate chambers maize seedlings emerged regularly at the 11th day after sowing, whereas first weed emergence was recorded between the 9th and 13th day after sowing (Fig. 2). First seedlings of Amaranthus retroflexus appeared about one day earlier in the climate chamber with future conditions. Only for Setaria viridis a significantly different D0 was measured. Seedlings emerged two days earlier under the climate change scenario. Whereas the Mean Emergence Time (MET) was significantly different for A. retroflexus and S. viridis between both climate chambers (Fig. 3), seedlings of ERI = nP i=1 Ni MET Gi = m h 1 eln(2) ⇣ D i D0 D50D0 ⌘ b i MET = nP i=1 Niti nP i=1 Ni v50 = m · b · ln(2) 2 · (D50 D0) Tab. 2: Comparison of the influence of predicted future conditions on plant height [cm] at the end of early-growth phase, mid-growth phase and reproduc- tion phase of plants grown in pots for the three replications, mean and standard error (se) are given for predicted future and current climate, significant pLMER-values are bold. plant height [cm] / phase early-growth phase mid-growth phase reproduction phase weed species measurement predicted current predicted current predicted current A. retroflexus mean 3.12 1.83 31.33 23.39 26.12 24.38 se 0.07 0.04 2.21 3.08 1.85 2.42 pLMER 0.001 0.040 0.577 E. crus-galli mean 13.84 8,44 92.50 73.22 107.77 102.15 se 0.23 0.15 2.46 1.39 2.59 3.53 pLMER 0.001 0.001 0.130 S. viridis mean 3.17 1.92 45.61 37.33 89.50 83.04 se 0.08 0.05 2.31 1.73 4.98 3.23 pLMER 0.001 0.002 0.186 Climate change and three maize weed species 237 Tab. 4: Parameters of plants grown in tubs at the end of mid-growth phase for the three replications; mean and standard error (se) are given for predicted future and current climate, significant differences LMER (pLMER) are bold. weed species A. retroflexus E. crus-galli S. viridis Maize factor measurement predicted current predicted current predicted current predicted current height [cm] mean 13.37 14.15 75.33 48.97 24.30 19.40 162.55 141.05 se 1.14 1.28 3.09 2.03 2.11 1.65 2.78 1.91 pLMER 0.650 0.001 0.034 0.001 dry mass [g] mean 1.06 1.34 8.17 3.94 0.76 0.57 78.91 62.15 se 0.23 0.16 1.51 0.62 0.11 0.06 9.96 7.71 pLMER 0.245 0.002 0.026 0.006 Tab. 3: Parameters with significant results of plants grown in pots at the end of reproduction phase for the three replications; mean and standard error (se) are given for predicted future and current climate, significant differences LMER (pLMER) are bold. weed species A. retroflexus E. crus-galli S. viridis factor measurement predicted current predicted current predicted current panicles mean 17.50 18.35 34.96 19.12 17.50 16.42 per plant se 3.18 4.92 4.20 1.44 1.30 1.61 pLMER 0.705 0.001 0.418 tillers mean 2.85 2.23 15.81 18.23 15.31 15.8 per plant se 0.43 0.40 0.81 0.76 0.91 1.39 pLMER 0.035 0.030 0.870 panicles mean 6.05 6.60 2.35 1.07 1.18 1.14 per tillers se 0.52 1.12 0.29 0.08 0.08 0.10 pLMER 0.663 0.001 0.458 number of seeds mean 2135.38 2654.46 4996.73 3553.58 2926.35 2401.12 per plant se 290.76 546.14 545.57 410.86 322.13 233.77 pLMER 0.956 0.002 0.123 seeds mean 73.28 46.03 150.97 187.84 171.94 171.71 per panicles se 7.11 4.22 7.15 17.94 18.66 19.25 pLMER 0.004 0.024 0.976 Fig. 2: Cumulated emergence of weed seeds over time for (A) Amaranthus retroflexus, (B) Echinochloa crus-galli, (C) Setaria viridis for the three replica- tions; square, diamond and circles mark readings in the climate chamber with predicted future conditions; +, x and * mark readings in the climate chamber with normal conditions; normal lines represent fitted curves under predicted future climate, dashed lines represent fitted curves under current climate. 0 5 10 15 20 25 0. 0 0. 4 0. 8 time [days] em er ge nc e [% ] (A) 0 5 10 15 20 25 0. 0 0. 4 0. 8 time [days] em er ge nc e [% ] (B) 0 5 10 15 20 25 0. 0 0. 4 0. 8 time [days] em er ge nc e [% ] (C) Echinochloa crus-galli appeared nearly at the same time in both cli- mate chambers. The time span from sowing to mid emergence (D50) differed sig- nificantly for S. viridis (Fig. 3). A D50 of 13.8 days was recorded in the chamber with the predicted future conditions compared to 16.4 days in the other chamber. The D50 of E. crus-galli was reached 1.8 days earlier and the D50 of A. retroflexus was reached 1.5 days earlier under the climate change scenario (Fig. 3). The mid emergence rate (v50) differed not significantly for all species between both climate chambers. However, Echinochloa crus-galli had the highest v50 ratio 238 K. Peters, B. Gerowitt (0.8), whereas S. viridis had the lowest v50 ratio (0.1) of the tested species (Fig. 3). The number of emerged A. retroflexus seedlings increased with each replication. Whereas in replication one, the ratio was 41 % for current climate conditions and 50 % for the chamber with predict- ed future conditions after 4 weeks, the ratio was 79 % vs. 91 % in replication three (Fig. 2). The emergence rate of E. crus-galli seed- lings was high and differences between replications were small four weeks after sowing. A rate of 75 % was reached after four weeks (Fig. 2). For Setaria viridis the rate was 58 % after four weeks in both chambers (Fig. 2). Only E. crus-galli had a significantly differ- ent Emergence Rate Index (ERI). The seedlings emerged quicker and more uniformly in the chamber with the predicted future conditions. A. retroflexus showed no significant differences (Fig. 3). Development without crop competition of plants grown individ- ually in pots Early-growth phase: Plants of all three species were significantly taller in the climate chamber with the predicted future conditions (Tab. 2). Whereas seedlings of E. crus-galli developed quite regu- larly, the variance of seedling height and development stage of S. viridis and A. retroflexus were more conspicuous between the dif- ferent replications. The mean plant height of E. crus-galli seedlings was 14 cm in the chamber with the climate change scenario and 8 cm in the chamber with current conditions. Seedlings of A. retroflexus and S. viridis were significantly smaller under current conditions: 2 cm vs. 3 cm (Tab. 2). Mid-growth phase: Amaranthus retroflexus showed a rather unde- fined growth habit. Most plants did not grow upright, were less elon- gated and branched more than plants grown under similar conditions in the field. Plants were also slightly smaller in the climate chamber with current conditions (Tab. 2). No significant differences in devel- opment stages between the two climate chambers were found, since first flower buds were developed after only 4 to 5 weeks in both chambers (BBCH stage 71). Setaria viridis grew taller under the cli- mate change scenario (Tab. 2). The average time of panicle develop- ment was the same in both climate chambers, although variance was higher in the chamber with the predicted future conditions (data not shown). Echinochloa crus-galli developed panicles significantly ear- lier (1 to 2 weeks) under the climate change scenario. Plants grown under the climate change scenario were at stage 67, whereas plants grown under current conditions were still at BBCH 55 on the mean average. Plant height was also increased under the climate change scenario (Tab. 2). Reproduction phase: We observed better growth of all species at the end of early- and mid-growth phase under the climate change sce- nario. Nevertheless, this relationship diminished at the end of the reproduction phase (Tab. 2). Amaranthus retroflexus plants still did not grow upright in both climate chambers (see above). We did not find any significant differences in plant height, panicles and seed production for A. retroflexus. The number of seeds per plant varied strongly between plants and replications. However, plants had more tillers per plant and more dry mass under the climate change scenario (Tab. 3). Echinochloa crus-galli developed quicker under predicted future conditions. First panicles were visible in the 11th week and flowering occurred over a longer period compared to the chamber with current conditions, which resulted in 550 more seeds/plant un- der the climate change scenario. The weed developed significantly more panicles with more seeds under the climate change conditions (Tab. 3). Echinochloa crus-galli also produced slightly less tillers under this scenario, which resulted also in slightly less vegetative dry matter (Tab. 3). The species was also the tallest of the three Fig. 3: Parameters characterising emergence: MET = Mean Emergence Time, ERI = Emergence Rate Index, v50 = emergence rate at time of mid emergence, D50 = the calculated day on which 50% the total emerged seeds have emerged; n = 3, triangles show values under predicted future conditions, circles show values recorded under cur- rent conditions. MET 7 8 9 10 11 12 13 ● ● ● predicted current predicted current predicted current A. retroflexus E. crus−galli S. viridis ● ● ● ● ● ● ERI 4 6 8 10 12 14 16 ● ● ● predicted current predicted current predicted current A. retroflexus E. crus−galli S. viridis ● ● ● ● ● ● v50 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 ● ● ● predicted current predicted current predicted current A. retroflexus E. crus−galli S. viridis ● ● ● ● ● ● D50 10 11 12 13 14 15 16 17 ● ● ● predicted current predicted current predicted current A. retroflexus E. crus−galli S. viridis ● ● ● ● ● ● Climate change and three maize weed species 239 weed species (Tab. 3). Setaria viridis produced first panicles after 17 weeks. They occurred at the same time in both climate chambers. Plant height, the number of panicles and tillers were rather homo- geneous after 21 weeks in both climate chambers. The weed tended to develop slightly more seeds under climate change conditions (Tab. 3). Vegetative dry matter was not affected (Tab. 3). Development with crop competition in the large plant tubs Plant height and dry mass of A. retroflexus grown in the maize rows in the large plant tubs did not differ significantly between the cli- mate chambers. Both plant height and dry mass were slightly less in the chamber with the predicted future conditions (Tab. 4). Plants of E. crus-galli were significantly taller in the chamber with the pre- dicted future conditions. This was also reflected by an increase in dry matter content (Tab. 4). Average height and dry mass of S. viridis plants differed strongly between the three replications. Plants grew taller and produced also more dry mass under the climate change sce- nario (Tab. 4). The maize plants were significantly taller when grown under the predicted future conditions. However, above ground dry mass was similar between both climate conditions (Tab. 4). All weeds were significantly smaller and had significantly less dry mass when grown in competition with maize in the large plant tubs compared to plants grown individually in smaller pots (Fig. 4). This was true for both climate chambers. Weed plants in the tubs and the pots had the same development stage. Discussion For discussing and evaluating the results, it is crucial to acknowl- edge that the two factors temperature and humidity were combined in our study. Consequently, we link the discussion mainly to these factors and conclude for climate change. Seedling emergence The growth, speed and timing of weed seedling emergence are im- portant factors for the development and seed production of weeds in maize (SARABI et al., 2011). Our results showed different emergence for all three species, which resulted in species-specific strategies to cope with the predicted future conditions. In summary, the predicted future conditions are beneficial for the tested weed species, as most of their emergence factors were enhanced under these conditions. This mainly leads to enhanced vegetative growth at later develop- ment phases (see below). The emergence results confirm that A. retroflexus seedlings typically emerge under warmer conditions later in the season (STECKEL et al., 2004). Nevertheless, Amaranthus retroflexus is also able to germi- nate in early spring with lower rates (BASKIN and BASKIN, 1985). As the emergence of seedlings is strongly affected by their parent plants, our findings also represent conditions the populations experi- enced near Göttingen, in Central Germany (KIGEL et al., 1977). The number of emerged A. retroflexus seedlings almost doubled in both climate chambers in the third replication. Such high emergence rates are untypical for A. retroflexus under low temperatures (GUILLEMIN et al., 2013). Conditions for emergence were equal in all replica- tions. Thus, we expect the dormancy status of the weeds had changed over time of storage. Seeds were stored for a total of 4 years and 2 months before the last replication started. Some other weed species show lower dormancy the longer their seeds were buried (BASKIN and BASKIN, 1985). Our findings that A. retroflexus changed its dor- mancy status over time of storage should be verified by another ex- periment, as dormancy changes were not reported by other authors (THOMPSON et al., 1997; OMAMI et al., 1999). Seedlings of E. crus-galli are able to grow faster than maize once they find sufficient conditions during the first four weeks. We found that E. crus-galli seedlings were less dependent on temperature and humidity, when compared to maize seedlings. CHAUHAN and JOHNSON (2011) reported similar results. The resulting high v50 ratio is a good indicator for the effects of higher thermal time (GARDARIN et al., 2009). The relatively low emergence rate of Setaria viridis seedlings con- firm that the species typically germinates under warmer conditions later in the season (DEKKER, 2003) and not simultaneously with the sowing of maize as it was defined by the experiment. This also ex- plains why more seedlings emerged considerably earlier under cli- mate change conditions. Current temperature conditions in North- ern parts of Central Europe are still below the considered optimum for the species (WALCK et al., 2011). Therefore, Setaria viridis will likely benefit more than the other tested weed species from higher temperatures at emergence. Development without crop competition This series of measurements was performed to study various para- meters of the weeds at the most important stages of their life and to Fig. 4: Comparison of the plant height of weeds grown in pots and tubs at the end of mid-growth phase for both climate chambers for the three replica- tions. tubs pots 10 20 30 40 pl an t h ei gh t [ cm ] Amaranthus retroflexus tubs pots 20 40 60 80 10 0 pl an t h ei gh t [ cm ] Echinochloa crus−galli tubs pots 10 20 30 40 50 60 pl an t h ei gh t [ cm ] Setaria viridis 240 K. Peters, B. Gerowitt provide answers concerning how weeds perform individually under climate change conditions. We found distinct differences between measurements taken at the time of emergence, at the end of the early and mid-growth phase and reproduction phase. The tested weeds balanced their growth and biomass allocation according to abiotic conditions and competition from neighbouring plants. Their response to the predicted future conditions started to become species-specific with the end of the mid-growth phase. Early and mid-growth phase: All tested weed species benefitted from the predicted future conditions at the end of the mid-growth phase. To conclude, greater plant heights under the climate change scenario suggest an increase in vegetative growth and thus, also en- hanced interference capabilities of the weeds during the first weeks of their development. This may give them an advantage over single maize plants. For A. retroflexus, we observed no differences in panicle onset speed between the climate chambers, although findings by KNEZEVIC et al. (2001) and STECKEL et al. (2004) suggest a different relationship. E. crus-galli had an earlier onset of flowers and a longer flowering time, which increases the opportunities for outcrossing (CLEMENTS and DITOMMASO, 2011). This is in accordance with other studies (POTVIN, 1986). These phenomena can also occur when seedlings emerge later in the season. Thus, climate change conditions may ac- celerate the life cycle as described by POTVIN (1986) for E. crus-galli and by ORYOKOT et al. (1997), KNEZEVIC et al. (2001) and HYVÖNEN (2011) for A. retroflexus. This process results in more fertile seeds at the end of the reproductive stage, but probably not in more bio- mass. By contrast, Setaria viridis develops panicles after midsummer when photoperiods are getting shorter (DEKKER, 2003). Plants grown in the climate chambers set panicles at the same time as would plants grown under natural conditions (DEKKER, 2003). We observed that the earlier seedlings emerge, the longer they delay development of reproductive parts. Therefore, later emerging arable plants have a shorter vegetative development period (FORCELLA et al., 2000; DEKKER, 2003). Similar behaviour was reported for A. retroflexus (COSTEA et al., 2004), but was not confirmed by our experiment for this species. Reproduction phase: Amaranthus retroflexus plants in the climate chambers were creeping near the ground, grew sideward and till- ered more, when compared to plants grown in the fields. This growth habit was likely a response on low red and far-red light levels in the climate chamber (COSTEA et al., 2004; GIMPLINGER and KAUL, 2009). Plants grown outside from seeds of the same population did show a normal growth habit (PETERS and GEROWITT, 2014). This could explain why A. retroflexus was less competitive in compari- son to the other two weed species. Some A. retroflexus plants also seemed to lack senescence. Unlike E. crus-galli and S. viridis, this species did not match its life cycle to the length of the season (SAUER, 1967). Even towards the end, some plants were still green and had not finished ripening. Some plants just grew and grew. As a result, we have not found differences in seed production and we can only partly confirm with HYVÖNEN (2011), who reported that raised temperatures enhance seed output but not overall growth. Echinochloa crus-galli is better suited to the predicted future con- ditions because plants had more panicles and tillers under these con- ditions. This resulted in totally more seeds per plant (BARRETT and WILSON, 1981). If arable cropping conditions allow, the species will likely extend its seed bank. This is especially the case under continu- ous maize cropping (FRIED et al., 2010). Since we have not found any differences in vegetative biomass, we assume that E. crus-galli mainly invests the possible benefits of warmer conditions into re- production. Setaria viridis is well adapted to current and future climatic condi- tions and can make use of its phenotypic plasticity (DEKKER, 2003). We observed higher variances of plant parameters throughout the replications under the climate change scenario. Since several bio- types are typical for the genus Setaria, the high variance reflects ge- netic plasticity within the used Setaria population (DEKKER, 2003). A future climate may select for certain ecotypes that are better suited for future conditions, thus enhancing fitness. Our results are in ac- cordance with SWANTON et al. (1999), who reported less biomass but increased reproductive output under warmer conditions. For S. viridis, we determined also a slight increase in seed production per plant and less dry matter content per plant in the chamber with pre- dicted future conditions. However, higher temperature difference between the climate chambers may have been needed for significant results (DOUGLAS et al., 1985; SWANTON et al., 1999). In summary, although the responses were species-specific, E. crus- galli and S. viridis had higher reproductive output under the scenario with predicted future conditions. This may lead to better long-term population development, as the weeds are able to build-up their seed banks. Development under crop competition This series of measurements was performed to study the effect of maize competition on the vegetative development of the weeds un- der climate change conditions. The broad phenotypic responses of E. crus-galli and S. viridis when grown with maize suggest that they likely will increase their success in maize when cropped under cli- mate change conditions. Plants of A. retroflexus grew even more deformed as single plants grown in the smaller pots (see above). Contrarily, OTTE et al. (2006) reported that A. retroflexus is highly competitive even when shad- ing occurs. It is also possible that temperatures in the climate cham- bers were still too low for the plants (GIMPLINGER and KAUL, 2009; HYVÖNEN, 2011; OVEISI et al., 2013). Under the maize competition, S. viridis was able to balance growth to match optimally the low light levels. Plants grown in the smaller pots showed no differences in habit and other phenological proper- ties compared to plants grown together with maize. This result con- firms studies of DEKKER (2003). However, we cannot confirm the reported result that S. viridis responds to increasing shade levels by reducing tiller production (DOUGLAS et al., 1985; DEKKER, 2003). The determined high phenotypic plasticity may enable the weed to adapt to a broad range of changing conditions. Of the three weed species investigated, E. crus-galli was the least affected in growth and development by the shading of maize. Com- pared to plants grown individually in the smaller pots, the weed balances the low light availability with growing upwards instead of tillering. The species is able to adapt fast to ecological factors such as shading, because of its broad genome, proliferation and crop mim- icking. Nevertheless, BARRETT and WILSON (1981) also showed that high phenotypic plasticity may result in differences in tiller produc- tion of E. crus-galli. Implications for crop protection From the predicted future conditions Echinochloa crus-galli prof- ited most, followed by Setaria viridis. Amaranthus retroflexus was the least benefiting species of the three tested weeds possibly due to the artificial conditions in the climate chambers. In detail, our study revealed that each weed species responded differently to changes in climate conditions. Narrow crop rotations and continuous cropping of maize are likely in Central Europe in the future (MEHRTENS et al., 2005; WEBER and GUT, 2005; FRIED et al., 2010). In subsequent years, S. viridis and Climate change and three maize weed species 241 E. crus-galli will likely build up their seed banks due to better emer- gence and higher seed output (BARRETT and WILSON, 1981; DEKKER, 2003). Better growth and faster vegetative development when grown within maize under climate change conditions will also lead to better long-term population development of the weeds in the future (DOUG- LAS et al., 1985). The results of our study indicate that under future climate conditions various weed management strategies are needed to ensure control of the tested weeds (OLESEN and BINDI, 2002; PETERS et al., 2014). For agricultural purposes, we suggest to integrate management practices such as variations in the sowing date of maize or in the choice of cultivars that result in timely harvesting, and thus reduce the chance for the late emerging weeds to fully develop and produce seeds un- der climate change conditions. A challenge for maize breeding is to select for fast growing and developing cultivars, since our study re- vealed a trade-off between the vegetative growth of the weed species and their generative reproduction when cropped together with maize (SARABI et al., 2011). Farmers themselves should avoid short crop rotations with maize alone or other late spring sown crops. Continu- ous cropping of this type of crop will select for particularly prolific biotypes of S. viridis and E. crus-galli (BARRETT and WILSON, 1981; DEKKER, 2003). Furthermore, we propose that necessary herbicide treatments should be applied late in the season to cover late emerg- ing weed cohorts with high temperature requirements, especially those of S. viridis (DEKKER, 2003; BECKIE and TARDIF, 2012). Conclusions Climate change exerts impacts on weeds during their whole life cycle. In order to get insight into underlying processes, it is neces- sary to determine biological parameters at the time of emergence, early growth and at the time of reproduction. As a result, climate mediated alterations in the measured biological and demographic at- tributes allow predictions of long-term population development of the weed species (FRIED et al., 2010; PETERS et al., 2014). Biologi- cal properties and demographic data can also be used to predict the population development under future conditions and to improve bio- climatic models (PETERS and GEROWITT, 2014). Furthermore, with future climate change costly crop protection measures at different development stages of the weeds may be needed for successful weed management. In order to continue high production under future con- ditions, crop protection has to be adapted to the biological responses of weeds to climate change. Acknowledgements This study was supported by the Ministry for Science and Culture of Lower Saxony within the network KLIFF − climate impact and adaptation research in Lower Saxony. We thank Ingolf Gliege and Martina Goltermann for assisting in the experiments. References BAAYEN, R.H., DAVIDSON, D.J., BATES, D.M., 2008: Mixed-effects modeling with crossed random effects for subjects and items. J. Mem. Lang. 59, 390-412. BARRETT, S.C.H., WILSON, B.F., 1981: Colonizing ability in the Echinochloa crus-galli complex (barnyard grass). I. Variation in life-history. Can. J. Botany 59, 1844-1860. BASKIN, J.M., BASKIN, C.C., 1985: The Annual Dormancy Cycle in Buried Weed Seeds: A Continuum. BioScience 35(8), 492-498. BECKIE, H.J., TARDIF, F.J., 2012. Herbicide cross resistance in weeds. Crop Protection 35, 15-28. CHAUHAN, B.S., JOHNSON, D.E., 2011: Ecological studies on Echinochloa crus-galli and the implications for weed management in direct-seeded rice. Crop Prot. 30, 1385-1391. CLEMENTS, D.R., DITOMMASO, A., 2011: Climate change and weed adapta- tion: Can evolution of invasive plants lead to greater range expansion than forecasted? Weed Res. 51, 227-240. COSTEA, M., WEAVER, S.E., TARDIF, F.J., 2004: The biology of Canadian weeds. 130. Amaranthus retroflexus L., A. powellii S. Watson and A. hy- bridus L.. Can. J. Plant Sci. 84(2), 631-668. DEKKER, J., 2003: Evolutionary biology of the Foxtail (Setaria) species- group. In: Inderjit (ed.), Weed Biology and Management, 65-114. Klu- wer Academic Publishers. DORADO, J., SOUSA, E., CALHA, I.M., GONZALEZ-ANDUJAR, J.L., FERNAN- DEZ-QUINTANILLA, C., 2009: Predicting weed emergence in maize crops under two contrasting climatic conditions. Weed Res. 49, 251-260. DOUGLAS, B.J., THOMAS, A.G., MORRISON, I.N., MAW, M.G., 1985: The Biology of Canadian Weeds. 70. Setaria viridis (L.) Beauv. Can. J. Plant Sci. 65(3), 669-690. FARAWAY, J.J., 2006: Extending the Linear Model with R: Generalized Lin- ear, Mixed Effects and Nonparametric Regression Models, Chapman & Hall / CRC. FORCELLA, F., COLBACH, N., KEGODE, G.O., 2000: Estimating seed produc- tion of three Setaria species in row crops. Weed Sci. 48, 436-444. FRIED, G., PETIT, S., REBOUD, X., 2010: A specialist-generalist classification of the arable flora and its response to changes in agricultural practices. BioMedCentral Ecology 10(20), 1-11. GARDARIN, A., DÜRR, C., COLBACH, N., 2009: Which model species for weed seedbank and emergence studies? A review. Weed Res. 49, 117- 130. 10.1111/j.1365-3180.2008.00683.x GARDARIN, A., GUILLEMIN, J.P., MUNIER-JOLAIN, N.M., COLBACH, N., 2010: Estimation of key parameters for weed population dynamics mod- els: Base temperature and base water potential for germination. Eur. J. Agron. 32(2), 162-168. GARDARIN, A., DÜRR, C., COLBACH, N., 2011: Prediction of germination rates of weed species: Relationships between germination speed para- meters and species traits. Ecol. Model. 222(3), 626-636. GIMPLINGER, D.M., KAUL, H.-P., 2009: Calibration and validation of the crop growth model LINTUL for grain amaranth (Amaranthus sp.). J. Appl. Bot. Food Qual. 82, 183-192. GRUNDY, A.C., 2003: Predicting weed emergence: a review of approaches and future challenges. Weed Res. 43(1), 1-11. GUILLEMIN, J.-P., GARDARIN, A., GRANGER, S., REIBEL, C., MUNIER- JOLAIN, N., COLBACH, N., 2013: Assessing potential germination period of weeds with base temperatures and base water potentials. Weed Res. 53, 76-87. HESS, M., BARRALIS, G., BLEIHOLDER, H., BUHR, L., EGGERS, T., HACK, H., STAUSS, R., 1997: Use of the extended BBCH scale – general for the descriptions of the growth stages of mono- and dicotyledonous weed species. Weed Res. 37, 433-441. HYVÖNEN, T., 2011: Impact of temperature and germination time on the suc- cess of a C4 weed in a C3 crop: Amaranthus retroflexus and spring bar- ley. Agr. Food Sci. 20, 183-190. IHAKA, R., GENTLEMAN, R., 1996: R: A Language for data analysis and graphics. J. Comput. Graph. Stat. 5(3), 299-314. IPCC, 2013: 5th Assessment Report: Climate Change 2013: The Physical Science Basis. http://www.ipcc.ch/, IPCC, Geneva, Switzerland. KIGEL, J., OFIR, M., KOLLER, D., 1977: Control of the Germination Respons- es of Amaranthus retroflexus L. Seeds by their Parental Photothermal Environment. J. Exp. Bot. 23(106), 1125-1136. KNEZEVIC, S.Z., VANDERLIP, R.L., HORAK, M.J., 2001: Relative time of re- droot pigweed emergence affects dry matter partitioning. Weed Sci 49, 617-621. MAUN, M.A., BARRETT, S.C.H., 1986: The Biology of Canadian Weeds. 77. Echinochloa crus-galli (L.) Beauv. J. Plant Sci. 66, 739-759. MEHRTENS, J., SCHULTE, M., HURLE, K., 2005: Unkrautflora in Mais – Ergeb- nisse eines Monitorings in Deutschland. Gesunde Pflanzen 57, 206-218. 242 K. Peters, B. Gerowitt NOVÁK, R., DANCZA, I., SZENTEY, L., KARAMÁN, J., 2009: Arable weeds of Hungary. Fifth national weed survey (2007-2008). Ministry of Agricul- ture and Rural Development, Hungary, 1-95. OLESEN, J.E., BINDI, M., 2002: Consequences of climate change for Euro- pean agricultural productivity, land use and policy. Eur. J. Agron. 16, 239-262. OMAMI, E.N., HAIGH, A.M., MEDD, R.W., NICOL, H.I., 1999: Changes in germinability, dormancy and viability of Amaranthus retroflexus as af- fected by depth and duration of burial. Weed Res. 39, 345-354. ORYOKOT, J.O.E., MURPHY, S.D., THOMAS, A.G., SWANTON, C.J., 1997: Temperature- and moisture-dependent models of seed germination in green and redroot pigweed (Amaranthus powellii, A. retroflexus). Weed Sci. 45, 488-496. OTTE, A., BISSELS, S., WALDHARDT, R., 2006: Samen-, Keimungs- und Habitateigenschaften: Welche Parameter erklären Veränderungstenden- zen in der Häufigkeit von Ackerwildkräutern in Deutschland? J. Plant Dis. Protect. Sp. I. XX, 507-516. OVEISI, M., RAHIMIAN MASHHADI, H., YOUSEFI, A.R., ALIZADE, H., BAGHESTANI, M.A., GONZALEZ-ANDUJAR, J.L., 2013: Predicting maize yield in a multiple species competition with Xanthium strumarium and Amaranthus retroflexus: Comparing of approaches to modeling herbicide performance. Crop Prot. 45, 15-21. PATTERSON, D.T., 1995: Weeds in a Changing Climate. Weed Sci. 43, 685- 701. PETERS, K., BREITSAMETER, L., GEROWITT, B., 2014: Impact of climate change on weeds in agriculture. A Review. Agriculture for Sustainable Development 34(4): 707-721. doi: 10.1007/s13593-014-0245-2 PETERS, K., GEROWITT, B., 2014: Weed growth properties of Amaranthus retroflexus, Echinochloa crus-galli and Setaria viridis as influenced by shifts in the maize cropping season. J. Plant Dis. Protect. In review. POTVIN, C., 1986: Biomass allocation and phenological differences among southern and northern populations of the C4 grass Echinochloa crus- galli. J. Ecol. 74, 915-923. ROBINSON, T.M.P., GROSS, K.L., 2010: The impact of altered precipitation variability on annual weed species. Am. J. Bot. 97(10), 1625-1629. doi: 10.3732/ajb.1000125 SARABI, V., MAHALLATI, M.N., NEZAMI, A., MOHASSEL, M.H.R., 2011: Effects of the relative time of emergence and the density of common lambsquarters (Chenopodium album) on corn (Zea mays) yield. Weed Biol. Manag. 11(3), 127-136. SAUER, J.D., 1967: The grain Amaranths and their relatives: A revised taxo- nomic and geographic survey. Ann. Mo Bot. Gard. 54(2), 103-137. SCHIMPF, D.J., 1977: Seed weight of Amaranthus retroflexus in relation to moisture and length of growing season. Ecology 58, 450-453. STECKEL, L.E., SPRAGUE, C.L., STOLLER, E.W., WAX, L.M., 2004: Tempera- ture effects on germination of nine Amaranthus species. Weed Sci. 52, 217-221. SWANTON, C.J., HUANG, J.Z., DEEN, W., TOLLENAAR, M., SHRESTHA, A., RAHIMIAN, H., 1999: Effects of temperature and photoperiod on Setaria viridis. Weed Sci. 47, 446-453. THOMPSON, K., BAKKER, J.P., BEKKER, R.M., 1997: The soil seed banks of North West Europe: methodology, density and longevity. Cambridge University Press, Cambridge. TUBIELLO, F.N., SOUSSANA, J.-F., HOWDEN, S.M., 2007: Crop and pasture response to climate change, Proc. Natl. Ac. Sci. 104(50), 19686-19690. WALCK, J.L., HIDAYATI, S.N., DIXON, K.W., THOMPSON, K., POSCHLOD, P., 2011: Climate change and plant regeneration from seed. Glob. Change Biol. 17, 2145-2161. WALTHER, G.-R., POST, E., CONVEY, P., MENZEL, A., PARMESAN, C., BEEBEE, T.J.C., FROMENTIN, J.-M., HOEGH-GULDBERG, O., BAIRLEIN, F., 2002: Ecological responses to recent climate change. Nature 416, 389-395. WEBER, E., GUT, D., 2005: A survey of weeds that are increasingly spreading in Europe. Agron. Sustain. Dev. 25, 109-121. Address of the corresponding author: Kristian Peters, University of Rostock, Faculty of Agricultural and Environ- mental Sciences, Crop Health, Satower Straße 48, D-18051 Rostock, Ger- many. E-mail: kristian.peters@uni-rostock.de