4018.p65 ALCES VOL. 40, 2004 MIHAJLOVICH AND BLAKE – HERBICIDE AND LICHEN 7 AN EVALUATION OF THE POTENTIAL OF GLYPHOSATE HERBICIDE FOR WOODLAND CARIBOU HABITAT MANAGEMENT Milo Mihajlovich1 and Peter Blake2 1Incremental Forest Technologies Ltd., 7327-118A Street, Edmonton, AB, Canada T6G 1V3; 2Canadian Forest Products Ltd., Bag 100, Grande Prairie, AB, Canada T8V 3A3 ABSTRACT: Two studies evaluating the effects of glyphosate used for habitat management on caribou lichens and dwarf shrubs were undertaken. Glyphosate substantially reduced blueberry cover at all rates tested in both cutover and uncut areas. Glyphosate did not affect caribou lichen cover. ALCES VOL. 40: 7-11 (2004) Key words: caribou lichen, glyphosate, habitat management, tolerance Woodland caribou (Rangifer tarandus) are listed as a threatened species, in Alberta and Canada (by Alberta Sustainable Re- source Development, Fish and Wildlife Di- vision and Committee on the Status of En- dangered Wildlife in Canada – Anonymous 2002). As a threatened species, woodland caribou are on the Blue List of species at risk of declining to non-viable levels and meriting special attention in forest manage- ment and forest management research. These efforts include developing an under- standing of the factors hindering woodland caribou population recovery when manage- ment regimes are altered in their favour. Wildlife managers suggest predation on woodland caribou young is, at present, the major factor limiting population response to more favourable management regimes. James (1999) and Chowns (2003) suggest woodland caribou in Alberta and Ontario historically used habitats that limited con- tact with other ungulate species, thus re- ducing the frequency of encounter with predators (especially wolves, Canis lupus). Reduced frequency of encounter is sug- gested as a primary means of woodland caribou avoiding predation of their young by wolves. The boreal ecotype of woodland caribou achieves this through use of habi- tats unfavourable to other ungulate species (for example, jackpine, Pinus banksiana Lamb. – caribou lichen, Cladina mitis (Sandst.) Hale & W. Culb.) sites. The mountain ecotype of woodland caribou achieves this through spending a consider- able portion (spring through late autumn) of the year in high elevation tundra site types, which are unfavourable to other high-eleva- tion species. Courtois et al. (2004) suggest joint man- agement of caribou, moose, and wolves as a management scale (3,000 – 7,000 km2) strategy. Understanding woodland caribou habitat use and the importance of predation in limiting woodland caribou recovery has led to an emphasis on managing habitat to reduce predator – prey interactions. One means of reducing predator – prey interac- tion is to provide woodland caribou with large, contiguous areas of habitat unattrac- tive to other ungulates – thereby substan- tially reducing total ungulate population den- sity and hence attractiveness to predators. To this end a number of very large cutblocks have been harvested in the boreal Lower Foothills ecoregion (Beckingham et al. 1996) caribou management zone. The area of HERBICIDE AND LICHEN – MIHAJLOVICH AND BLAKE ALCES VOL. 40, 2004 8 these large cutovers ranges from slightly larger than 100 ha to approximately 350 ha. It is presumed that the large size of these areas will reduce their appeal to large ungu- lates (moose, Alces alces andersoni) which will, in turn, reduce wolf use of these areas (Anonymous 2002). If these areas are managed for reduced appeal to browsing ungulates, management techniques to reduce browse development after harvesting may be of potential value. This would be especially appealing if major components of the diet of woodland caribou were less affected than moose browse. The authors identified the possible need for browse management in 1997 and assessed several options for longer-term browse re- duction. Broadcast herbicide application ap- peared to offer the most cost effective (Biring et al. 1996) means, and least disrup- tive to caribou calving, of managing browse – due to treatment occurring several months after calving and occurring quickly with minimal human access on the ground during treatment. Glyphosate is the most com- monly used herbicide for forest manage- ment in Canada – approximately 94% of f o r e s t m a n a g e m e n t h e r b i c i d e u s e i s glyphosate (CCFM 2004). Approximately 38% of all forest stand tending in Canada is with glyphosate herbicide. Glyphosate is chosen for its effectiveness in controlling a broad spectrum of competing species (Biring et al. 1996). Glyphosate is classed as moderately persistent material in Canadian soils (Willis and MacDowell 1983, Tortenson 1985) with a half-life in soil of 20 – 100 days depending on soil conditions. Health Canada in the 1987 decision document on glyphosate registration for forestry deemed glyphosate to be a non-leaching (i.e., not soil mobile) based on an Agriculture Canada study of potential for soil mobility of pesticides (Ag- riculture Canada 1986). Lautenshlager and Sullivan (2002) recently reviewed the ef- fects of herbicide treatments (primarily with glyphosate) on the biotic components of regenerating northern forests – they con- cluded that herbicides are a safe, effective tool for restoring conifers to previously coni- fer – dominated systems which have other- wise been replaced with hardwoods since Europeans began harvesting those systems. Oberg (2001) cites several articles that demonstrate terrestrial lichens (species not given) comprise the major component of the caribou winter diet. The case is made that both ecotypes of caribou in west-central Alberta rely on caribou lichen as their pri- mary winter food source. However, no data could be found on effects of broadcast glyphosate application, for browse manage- ment, on terrestrial lichens (caribou lichens) that form such a significant part of the woodland caribou’s diet. Therefore the authors set up two experiments to examine the impact of broadcast glyphosate herbi- cide treatment on caribou lichen. g STUDY AREA A lodgepole pine (Pinus contorta var. latifolia) stand harvested and replanted to lodgepole pine 2 years prior to treatment was selected as a study site. Located approximately 80 km southeast of Grande Prairie, Alberta the site is located on a sandy soil. Moisture regime is sub-mesic and soil nutrient levels appear to be low as key understory plants in the uncut stand were caribou lichens, bearberry (Arctosta- phylos uva-ursi), and scattered Canada wild rye (Elymus canadensis). METHODS Two small-plot herbicide treatment tri- als were installed – one in the cutover and reforested area and a second in an adjacent uncut portion of the original lodgepole pine stand. Treatments were glyphosate, applied as Vision Silviculture Herbicide, at rates of ALCES VOL. 40, 2004 MIHAJLOVICH AND BLAKE – HERBICIDE AND LICHEN 9 2, 4, and 6 L ha-1 (0, 712, 1424, and 2136 g (ae) glyphosate isopropylamine ha-1) with an untreated control (Table 1). All treat- ments were applied using a carbon dioxide propelled small plot sprayer. Application volume was 75 L ha-1, using 110015 flat fan nozzles. Applications were made on Sep- tember 9, 1997 during the typical season when herbicides might be applied for browse management – a time when browse may be controlled without negatively affecting coni- fer regeneration. Weather conditions at treatment were: wind west at 2 km h-1, temperature 23°C, and relative humidity 58 percent. Treatment plots were laid out prior to treatment. A single, fixed 0.25 m2 subplot for repeated assessment of subject plant cover was established and marked in each treatment plot. Subplots were randomly located in each treatment plot and were used to overcome variability in cover and extent of subject plants. Cover was as- sessed by ocular estimate. All estimates were by the same investigator and were made blind (without knowledge of the treat- ment being assessed). Subject species were caribou lichen (a complex of Cladina cenotea, C. rangiferana, and C. stellaris) and blueberry (Vaccinium myrtilloides); bearberry was present but was not suffi- Table 1. Browse control treatments evaluated. ciently uniform in distribution to provide an evaluation of treatment impact. Caribou lichen cover was assessed 10 months after treatment (MAT), blueberry cover was as- sessed 10 and 22 MAT. A randomized, complete block experi- mental design was used. Treatments were replicated 4 times. All comparisons were based on means of the 4 replicates. One- way analysis of variance was used to test for significant differences (P = 0.05) within a subject species, time period, and locale. If differences were found, Tukey’s Honestly Significant Difference separation of means test (University of Missouri - Rolla 2002) was used to elucidate differences between treatments at that time period and locale. RESULTS Results of cover assessments, at 10 and 22 months after treatment (MAT), are given in Table 2. Broadcast glyphosate treat- ment, regardless of application rate did not significantly affect caribou lichen cover in either area (cutover and uncut). No visual symptoms of glyphosate activity were noted on the caribou lichen. Sub-lethal glyphosate symptoms on susceptible species typically include: stunting (reduced elongation) of new stem or branch growth, dwarfing of new foliage, yellow or white color of new foliage, and clustering of foliage at branch tips. Blueberry cover was reduced signifi- cantly, at both 10 and 22 MAT by all glyphosate rates in both the cutover area and the uncut stand. Cover reduction was directly related to glyphosate application rate with the two labelled application rates (4 and 6 L of product ha-1) resulting in greater reduction in cover than the lowest rate examined (2 L of product ha-1). Reduc- tions in blueberry cover were greater in the uncut area than in the cutover area (signifi- cance not tested). Treatment Number Description Application Rate (L ha-1) Active Ingredient Rate (g ha-1) 1 Untreated control 0 0 2 ½ nominal brush rate 2 712 3 Nominal brush rate 4 1424 4 Maximum labeled rate 6 2136 HERBICIDE AND LICHEN – MIHAJLOVICH AND BLAKE ALCES VOL. 40, 2004 10 DISCUSSION Browse management to reduce use of recently reforested cutovers may be a means of limiting use of these areas by moose. It has been postulated that a reduction in moose utilization will, in turn, reduce the attractiveness of recent cutovers to preda- tors (James 1999, Courtois et al. 2004) thereby reducing predator impact on wood- land caribou populations. Herbicide use for browse reduction on large cutovers may be a feasible means of maintaining low browse densities (Biring et al. 1996). Results of this experiment sug- gest broadcast application of glyphosate herbicide for browse reduction would not negatively impact the supply of caribou lichen (and hence caribou forage) on treated areas. However, herbicide treatment would result in some reduction of blueberries thus reducing forage for species other than wood- land caribou. CONCLUSION Broadcast application of glyphosate her- bicide for browse management should not negatively impact caribou lichen extent on treated areas. Table 2. Changes in plant cover, 10 and 22 months after treatment. 1 Lichen was a complex of Cladina rangiferana, C. cenotea, and C. stellaris. 2 Cover values are averages of ocular estimates of cover in 4 – 0.25 m2 permanent sub-plots established prior to treatment in each plot. Thus each cover represents the average of a total of 16 cover assessments. 3 Percent change values shown in bold font are “honestly significantly different” from pre-treatment cover values using Tukey’s Honestly Significant Difference (HSD) test. α = 0.05. ACKNOWLEDGEMENTS The authors wish to thank Dr. Tom Sullivan of the Applied Mammal Research Institute for his review and comments; Ca- nadian Forest Products Limited for their support of this research; and two anony- mous reviewers for their comments and suggestions. REFERENCES AGRICULTURE CANADA. 1986. Pesticide priority scheme for Water Monitoring Program. Unpublished report, Pesti- cide Directorate, Food Production and Inspection Branch, Ottawa, Ontario, Canada. ANONYMOUS. 2002. Threatened Wildlife. Alberta Sustainable Resource Devel- opment, Natural Resource Services, Edmonton, Alberta, Canada. BECKINGHAM, J., I. G. W. CORNS, and J. H. ARCHIBALD. 1996. Field guide to ecosites of west-central Alberta. Natural Re- sources Canada, Canadian Forest Serv- ice, Northwest Region, Northern For- estry Centre, Special Report 9. Edmon- ton, Alberta, Canada. BIRING, B. S., P. G. COMEAU, and J. O. Locale Glyphosate (g ha-1) At Treatment 10 MAT % Change 10 MAT3 At Treatment 10 MAT % Change 10 MAT3 22 MAT % Change 22 MAT3 Cutover 0 71 77 8.5 16.7 16.7 0 18.3 9.6 712 75 89 18.7 20 8.3 -58.5 15 -25 1424 80 83 3.8 11.7 3.3 -71.8 10 -14.5 2136 78 80 2.6 13.3 0 -100 3.3 -75.2 Uncut 0 77 77 0 10 11.7 17 13.3 33 712 85 80 -5.9 11.7 4 -65.8 8.3 -29.1 1424 88 87 -1.1 16.7 2 -88 6.7 -59.9 2136 83 82 -1.2 13.3 0 -100 4 -69.9 Lichen Cover1, 2 Blueberry Cover2 ALCES VOL. 40, 2004 MIHAJLOVICH AND BLAKE – HERBICIDE AND LICHEN 11 BOATENG. 1996. Effectiveness of vegeta- tion control methods in British Co- lumbia. Canadian Forestry Service and British Columbia Ministry of Forests. Victoria, British Columbia, Canada. (CCFM) CANADIAN COUNCIL OF FOREST MIN- ISTERS. 2004. Pest control product use. http://nfdp.ccfm.org/compendium/pest/ index_e.php. CHOWNS, T. J. 2003. State of the Knowl- edge of Woodland Caribou in Ontario. Forestry Research Partnership. http:// w w w . f o r e s t r e s e a r c h . c a / product_catalogue/reports.htm. COURTOIS, R., J.-P. OUELLET, C. DUSSAULT, and A. GINGRAS. 2004. Forest manage- ment guidelines for forest-dwelling cari- bou in Québec. Forestry Chronicle 80:598-607. JAMES, A. R. C. 1999. Effects of industrial development on the predator-prey rela- tionships between wolves and caribou in northeastern Alberta. Ph.D. Thesis, Department of Renewable Resources, University of Alberta, Edmonton, Al- berta, Canada. LAUTENSCHLAGER, R. A., and T. P.SULLIVAN. 2002. Effects of herbicide treatments on biotic components of regenerating northern forests. Forestry Chronicle 78:695-730. OBERG, P. R. 2001. Responses of mountain caribou to linear features in a west- central Alberta landscape. M.Sc. The- sis, Department of Renewable Re- sources, University of Alberta, Edmon- ton, Alberta, Canada. TORTENSSON, L. 1985. Behaviour of glyphosate in soils and its degradation. Page 137 in E. Grossbard and D. Atkinson, editors. The Herbicide Glyphosate. Butterworths, London, U.K. UNIVERSITY OF MISSOURI – ROLLA. 2002. Virtual statistician website. Accessed 2002. http://web.umr.edu/~psyworld/ tukeys4mean.htm. WILLIS, G. H., and L. L. MCDOWELL. 1983. Pesticides in agricultural runoff and their effects on downstream water quality. Environmental Toxicology and Chemis- try 1:267.