Documenting habitat-related patterns in foraging behaviour at the individual level and over large temporal scales remains challenging for large herbivores. Stable isotope analysis could represent a valuable tool to quantify habitat-related foraging behaviour at the scale of individuals and over large temporal scales in forest dwelling large herbivores living in coastal environments, because the carbon (δ13C) or nitrogen (δ15N) isotopic signatures of forage can differ between open and closed habitats or between terrestrial and littoral forage, respectively. Here, we examined if we could detect isotopic differences between the different assemblages of forage taxa consumed by white-tailed deer that can be found in open, closed, supralittoral, and littoral habitats. We showed that δ13C of assemblages of forage taxa were 3.0‰ lower in closed than in open habitats, while δ15N were 2.0‰ and 7.4‰ higher in supralittoral and littoral habitats, respectively, than in terrestrial habitats. Stable isotope analysis may represent an additional technique for ecologists interested in quantifiying the consumption of terrestrial vs. marine autotrophs. Yet, given the relative isotopic proximity and the overlap between forage from open, closed, and supralittoral habitats, the next step would be to determine the potential to estimate their contribution to herbivore diet.
Citation: Giroux M-A, Valiquette É, Tremblay J-P, Côté SD (2015) Isotopic Differences between Forage Consumed by a Large Herbivore in Open, Closed, and Coastal Habitats: New Evidence from a Boreal Study System. PLoS ONE 10(11): e0142781. https://doi.org/10.1371/journal.pone.0142781
Editor: Erik Sotka, College of Charleston, UNITED STATES
Received: November 25, 2014; Accepted: October 27, 2015; Published: November 11, 2015
Copyright: © 2015 Giroux et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the Supporting Information files.
Funding: This study was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC, www.nserc-crsng.gc.ca; scholarships to MAG [Alexander Graham Bell, Michael Smith, and EnviroNorth scholarships] and grants to SDC) and by Produits forestiers Anticosti Inc. that provided matching funds for the NSERC grants to SDC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors received funding from a commercial source (Produits forestiers Anticosti Inc.), but the funder was not involved in the study design, collection, analysis, and interpretation of data, writing of the paper, and/or decision to submit for publication. This financial competing interest does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.
Determining habitat-related patterns in foraging behaviour is critical to better understand the ecology of large herbivores. For instance, the trade-off between using habitats rich in forage and those providing cover is generally recognized as a determinant of foraging behaviour in herbivores inhabiting forest ecosystems [1, 2]. Whereas forage is more abundant in open habitats due to direct exposure to sunlight [3, 4], vertical and lateral cover provide protection against extreme climatic conditions and predators, respectively [5–7]. Forage quality, however, may be higher under cover (, but see ) due to the delayed maturation of forage under shaded conditions . In coastal areas, large herbivores may also use supralittoral and littoral habitats where they can forage on coastal plants , as well as on drifted seaweed ([12, 13], reviewed in . Coastal plants can be fertilized by marine-derived nutrients [11, 15] and hence their quantity and quality may vary in parallel with the amount of drifted seaweed . The biomass of seaweed available to terrestrial herbivores in littoral habitats is highly variable in space and time, as it depends on tidal movements, storms, coast orientation, and exposure to wind and waves [11, 16].
Sampling techniques have been developed to investigate the foraging behaviour of large herbivores (telemetry: e.g. , direct observations: e.g. ), yet quantifying the proportion of resources consumed in different habitats at the individual level and over large temporal scales remains challenging . The stable isotope analysis (SIA) of animal tissues has proven to be an effective method to describe individual diet over small to large temporal scales . SIA may represent an additional tool to quantify the habitat-related foraging behaviour of forest dwelling herbivores inhabiting coastal areas, both at the scale of individuals and over large temporal scales. Indeed, carbon (δ13C) and/or nitrogen (δ15N) isotopic signatures of forage can vary between open and closed forest habitats [21–24], between plants and seaweed [25–28], and between inland and coastal plants (e.g. ).
To quantify the proportion of different types of resources in a consumer’s diet using SIA, a statistical difference must be observed between the isotopic signatures of the different types of resources . Hence, our first objective was to determine whether we could detect isotopic differences between open, closed, supralittoral, and littoral habitats using the isotopic signatures of carbon (δ13C) and nitrogen (δ15N) of forage taxa consumed by white-tailed deer (Odocoileus virginianus; hereafter “deer”). We predicted lower δ13C in closed than in open habitats [21–24], and lower δ13C and δ15N in terrestrial (open and closed) than in supralittoral  and littoral habitats [26–28]. Because different vegetation communities can occupy open and closed habitats, respectively, and isotopic signatures can vary between vegetation taxa, our second objective was to verify whether we could detect isotopic differences between open and closed forest habitats for a single species; we retained Cornus canadensis as the species of interest because this forage species was found in both open and closed habitats and is heavily used by deer in our study system [31, 32].
Materials and Methods
To conduct field work on Anticosti Island, we obtained an authorisation for scientific studies from the Ministère des Ressources Naturelles et de la Faune du Québec (Permit number: 09052701409SF). The study area spans provincially-owned lands, and no protected species were sampled.
We conducted the study from 2002 to 2010 on Anticosti Island in the Gulf of St. Lawrence, Québec, Canada (49.06–49.95°N, 61.67–64.52°W; 7943 km2; Fig 1). The climate of Anticosti Island is maritime and characterized by cool summers and long winters. The mean monthly air temperature is 16°C in July and -11°C in January . Snow precipitation averages 328 cm annually, while rainfall averages 61 cm . Anticosti Island is dominated by closed coniferous forests and open habitats covering 56% and 36% of the island, respectively . The forests of Anticosti Island belong to the boreal zone and are part of the eastern balsam fir-white birch bioclimatic region . As a result of the introduction of approximately 220 white-tailed deer on this predator-free island at the end of the 19th century, long term severe browsing induced the decline of dominant deciduous species such as white birch (Betula papyrifera) and trembling aspen (Populus tremuloides; ). Presently, Anticosti forests are dominated by balsam fir (Abies balsamea), white spruce (Picea glauca), and black spruce (P. mariana). The following open habitats are found on Anticosti Island: peatlands, canopy openings in forest stands (edaphic gaps or openings resulting from natural disturbances such as insect outbreaks or windthrowns events), clear-cuts resulting from commercial logging, and burned areas of natural origin. A supralittoral habitat dominated by graminoids and forbs is found at the periphery of the island at the ecotone between forest stands and beaches. Behavioural observations have shown that deer feed on plants in this habitat, as well as on washed-up seaweed in the adjacent littoral habitat (Giroux, Côté and Tremblay, unpublished data).
Vegetation and seaweed sampling
We sampled vegetation and seaweed in the western sector of the island in July 2008, and we collected additionnal samples of Cornus canadensis in the central and eastern sectors of the island (Fig 1). In the western sector, we randomly selected 6 sites in closed coniferous forest, as well as in the supralittoral and the littoral habitats, and 3 sites in each of the following open habitats: fens, canopy openings, and clear-cuts. We randomly selected site locations using a forest map (1:20 000) that was generated by delineating habitat polygons using 1:15,000 aerial photographs taken in 1997, to which we added recent clear-cuts. Each site was defined as a 50-m radius area around the center of the location. At each site, except in littoral habitat, we aimed to collect at least one tablespoon of leaves (l), needles (n) or whole above-ground specimens (w) of the vegetation taxa consumed by deer on Anticosti Island during the summer, as assessed by micro-histological analyses of 15 pellet groups collected on Anticosti Island in July 2004 (list of taxa in S1 Table). We also aimed to collect two additional taxa known to contribute to deer diet in specific habitats, namely Elymus arenarius (supralittoral habitat; behavioral observations) and Myrica gale (fen; evidence of deer browsing on this taxum ). At each site, we thoroughly searched the 50-m radius area for all taxa in S1 Table, sampling those that were encountered and omitting those that were not found. Because some taxa were also excluded a posteriori when they were found in only one site per habitat (n = 1 per species in those cases), we ended up with the following list of taxa (hereafter “forage taxa”): Abies balsamea (n), Carex spp. (l), Cornus canadensis (l), Dasiphora fruticosa (l), Elymus arenarius (l), Ferns (l), Fragaria spp. (l), Graminae spp. (l), Hieracium spp. (l), Juncaceae spp. (l), Lichens (w), Mosses (w), Myrica gale (l), and Vicia cracca (l). The contribution of these taxa (excluding E. arenarius and M. gale) to summer diet of white-tailed deer on Anticosti Island summed up to 75% (S1 Table). We collected only the leaves or needles of most taxa because they represent 80% of deer summer and fall diets on Anticosti Island . In littoral sites, we aimed to collect at least one tablespoon of thallus of the following seaweed taxa consumed by deer on Anticosti Island, as revealed by behavioural observations: Laminaria spp., Fucus spp., and Ulva spp. To collect enough C. canadensis to compare its isotope values between open and closed habitats, we also randomly sampled the following number of additional sites in each of the central and eastern sectors of the island: 6 sites in closed canopy, 3 sites in fens and canopy openings, as well as 3 sites in burned areas (eastern sector only). We preserved all samples in polyethylene airtight bags, froze them within 8 hours of collection, then processed and analyzed their δ13C and δ15N according to laboratory procedures described in S1 Methods.
We compared isotopic signatures of vegetation and seaweed samples between the 6 following habitat categories: closed, fens, canopy openings, clear-cuts, supralittoral, and littoral. We used univariate linear mixed models (LMMs) to quantify the influence of habitat on δ13C and δ15N on deer forage, respectively. We specified the site as a random variable. Because we did not find all consumed taxa in every site of the same habitat, we also specified taxa as a random variable to take into account that sample sizes differed between taxa within an habitat (Table 1). We used multiple comparisons to compare the differences in isotopic signatures between all open habitats (fens, canopy openings, and clear-cuts althogether) and closed, supralittoral, or littoral habitats, as well as between all terrestrial habitats (open and closed habitats) and supralittoral or littoral habitats. Because the assumption of homogeneity of residuals was not respected for δ15N, we modelled the variance structure to allow residual variance to vary between habitats .
To verify whether we could detect isotopic differences between open and closed habitats in a single preferred taxum (C. canadensis), we fitted linear models to quantify variations in δ13C and δ15N, respectively, between open and closed habitats. We retained C. canadensis because it was the only forage species sampled in sufficient number in both open and closed habitats. We also took into account the sector and its interaction with habitat in the set of models compared to verify whether the eventual difference between habitats was maintained across sectors.
For both objectives, we used Akaike’s Information Criterion corrected for small sample size (AICc; ) to select the best-fitting model. Unless otherwise mentionned, all analyses respected the assumptions of normality and homoscedasticity of residuals and were performed with R version 2.15.1 .
Objective 1: Isotopic differences between habitats
δ13C of forage taxa was 3.0‰ (95% CI = [1.3, 4.6]), 2.6‰ (95% CI = [1.0, 4.1]), and 16.2‰ (95% CI = [14.4, 18.1]) higher in open, supralittoral, and littoral habitats, respectively, than in the closed habitat (Fig 2). δ15N did not differ between closed (-0.9‰, 95% CI = [-2.1, 0.4]) and open (mean = -1.1‰, 95% CI = [-2.3, 0.1]) habitats, but was 2.0‰ (95% CI = [0.3, 3.7]) and 7.4‰ (95% CI = [4.5, 10.1]) higher in supralittoral and littoral habitats, respectively, than in terrestrial habitats (closed and open habitats; Fig 2).
We show the average signatures (± SD) of the four habitat types.
The isotopic difference in δ13C tended to be larger between the closed habitat and peatlands (4.4‰, 95% CI = [2.6, 6.1]), than between the closed habitat and the canopy openings (1.9‰, 95% CI = [0.3, 3.5]) or cuts (2.5‰, 95% CI = [0.8, 4.2]). δ15N did not differ between the closed habitat and any of the open habitats (peatlands: 0.4‰, 95% CI = [-1.3, 2.1], canopy openings: 0.2‰, 95% CI = [-1.3, 1.7], cuts: -1.4‰, 95% CI = [-2.9, 0.2]).
Objective 2: Isotopic differences in a single species between open and closed habitats
For Cornus canadensis, the best model explaining variation in isotopic signatures included only habitat as a predictor variable for δ13C and included no variables for δ15N (null model; see S2 Table). δ13C was 2.6 ‰ (95% CI = [1.7, 3.5] higher in open (mean = -27.7‰, 95% CI = [-28.3, -27.0], n = 15) than in closed (mean = -30.3‰, 95% CI = [-30.9, -29.6], n = 14) habitats. δ15N did not differ between open (mean = -1.8‰, 95% CI = [-2.9, -0.8]) and closed (mean = -2.3‰, 95% CI = [-3.4, -1.2]) habitats.
We showed an isotopic distinction between forage consumed by a large herbivore in open, closed, supralittoral, and littoral habitats. As predicted, δ13C of forage consumed by deer was lower in closed than in open, supralittoral, and littoral habitats, whereas δ15N was lower in terrestrial (open and closed) than in supralittoral and littoral habitats. The difference between δ13C of forage found in the closed habitat and in the different open habitats ranged from 1.9‰ to 4.4‰. When comparing isotopic signatures between open and closed habitats for a single preferred plant species, C. canadensis, we also observed higher δ13C in closed than in open habitats.
Isotopic differences between habitats
Detecting variations in the isotopic signatures of deer forage taxa between habitats was the first step in evaluating the potential for using SIA to quantify the contribution of resources from the mosaic of habitats found in coastal boreal forests to deer diet. The isotopic difference between open and closed habitats observed in our study (3.0‰ on average) was slightly larger than in other studies conducted in boreal forests (0–2.2‰: , reviewed in  and ), and this holds for cuts (2.5‰) and peatlands (4.4‰), but not for canopy openings (1.9‰). The isotopic differences between terrestrial vegetation (open and closed habitats) and seaweeds (littoral habitats) were similar to the ranges reported in the literature (our study: δ13C = 14.8‰ and δ15N = 5.9‰; other studies: δ13C = [9.3‰,19‰] and δ15N = [6.4‰,9‰], [26–28]). The difference in δ15N between inland vegetation and coastal plants was 2.6‰, which is smaller than what has been observed for inland and coastal terrestrial plants in Africa (5–10‰; ), but closer to what has been observed for inland and coastal terrestrial plants in northern California (~3‰;).
Our ability to explain the differences or similarities in the extent of isotopic contrasts compared to previous studies is limited by the fact that the mechanisms underlying isotopic differences between autotrophs are still largely hypothetical. To explain the isotopic differences between open and closed habitats, two potential mechanisms have been identified, but both remain to be tested. These mechanisms are: 1) the lesser extent of mixing of the ground level air layer (where 13C-depleted CO2 derived from litter decomposition is released) with the overlying air in less ventilated closed habitats , and 2) the influence of light level on the isotopic fractionation of carbon during photosynthesis . The isotopic contrast between δ13C in terrestrial and marine autotrophs may result from the use of bicarbonate, which has a higher δ13C than dissolved CO2, as a source of carbon by algae and/or from the slower diffusion of CO2 in water that can influence isotopic fractionation of carbon . In marine ecosystems, the presence of 15N-containing compounds that are enriched in 15N compared to sources of nitrogen in terrestrial ecosystems can likely explain the higher δ15N observed in marine autotrophs . Coastal plants may have access to 15N-enriched nitrate from sea-spray, potentially explaining their higher δ15N compared to terrestrial plants . Although our study was not designed to study the mechanisms underlying the isotopic contrast between open and closed habitats discussed above, we nonetheless showed that such a contrast could be observed in a single species sampled in both open and closed habitats (C. canadensis). The isotopic difference between C. canadensis sampled in open and closed habitats (2.6‰) was within the range of the differences observed between open and closed habitats when comparing multiple taxa consumed by deer (1.9‰ to 4.4‰). This result indicates that the isotopic difference between open and closed habitats is not simply due to different vegetation communities occupying the two habitat types (Table 1). Indeed, isotopic signatures can vary between vegetation taxa due to differences in anatomy, physiology, and biochemistry . For example, conifers are enriched in 13C compared to deciduous trees because of different photosynthesis/transpiration ratios . Our study showed that a canopy effect can occur without variations associated with habitat-specific vegetation assemblages.
Further studies should aim at clearly assessing whether isotopic differences among habitats is a consequence of intraspecific variation in isotopic ratios among habitats, differences in community composition or both. In our study system, this could be performed by working at the species level rather than at the genera or higher taxonomic rank. Performing the study at the species level would likely result in both a larger set of species unique to each habitat and a larger set of species that could be sampled in more than one habitat (e.g. C. canadensis in our study). This would allow to compare the isotopic differences among habitats obtained when analysing all species, species unique to certain habitats only, and species found in all habitats (or in pairs of habitats). In our study, working at the species level was limited by the resolution of micro-histological analyses of deer diet (S1 Table). In the future, this limit could be relaxed by conducting diet analyses using DNA barcoding.
Using habitat-specific isotopic signatures to reconstruct the diet of coastal herbivores
In our study, the isotopic values of forage from open, closed and supralittoral habitats were relatively similar compared to the isotopic values frequently used in diet reconstruction studies (e.g. the contrasting isotopic values of marine and terrestrial autotrophs; [25, 26, 28]). In addition, there was a high degree of overlap between isotopic values from closed, open and supralittoral habitats. Given the relative isotopic similarity and the overlap between forage from open, closed, and supralittoral habitats, the next step would be to determine whether using these resources in isotopic diet reconstruction may help understand the relationships between diet estimates and predictors describing environmental or individual variations. Recent advances in isotopic diet reconstruction have stressed the need to account for different sources of uncertainty in isotopic analyses, even though doing this would result in uncertainty around diet estimates [51, 52]. All other variables being equal, smaller isotopic contrasts induce greater uncertainty around isotopically derived diet estimates [52, 53]. Although diet estimates obtained with such models are more robust, both interpretation and statistical analyses should take this uncertainty into account [52, 54]. Hence, further studies should determine whether the estimated contributions of forage from open, closed, and supralittoral habitats to the diet of herbivores would be precise enough to either show the effects of predictors describing environmental or individual variations on those diet estimates or an effect of diet variation on individual performance.
Determining how the use of open, closed, and coastal habitats influences the performance of forest dwelling large herbivores would deepen our understanding of how they adjust their foraging behaviour to maximize survival, growth and reproduction when exposed to various trade-offs (e.g. forage quantity vs. quality or forage quantity vs. predation risk). SIA may represent an additional technique to address such questions, and may prove especially useful to determine the influence of consumption of terrestrial vs. marine autotrophs on the performance of herbivores. Yet, the potential to address such questions with respect to the relative use of forage from open, closed, and supralittoral habitats remains to be demonstrated.
S1 Data. Data set used in objective 1 to verify whether we could detect isotopic differences among habitats.
S2 Data. Data set used in objective 2 to verify whether we could detect isotopic differences between open and closed habitats in a single preferred taxum (C. canadensis).
S1 Methods. Additional methodological details of isotopic analyses.
S1 Table. Contribution of different vegetation taxa to the summer diet of white-tailed deer on Anticosti Island during summer, as assessed by micro-histological analyses of 15 pellet groups collected on Anticosti Island (Québec, Canada) in July 2004 (data from Massé and Côté).
Taxa in bold were those included in the comparison of isotopic signatures between habitats described in the main text.
S2 Table. Model selection for linear models fitted to determine if δ13C (A) and δ15N (B) of Cornus canadensis varied between habitat (open terrestrial vs. closed), sector (western, central, and eastern) and their interaction.
We report the number of parameters (k), Akaike’s information criterion for small sample sizes (AICc) relative to the model with the lowest AICc (ΔAICc), as well as the AICc weight (ωAICc). Models are ranked by their AICc values. The best model is shown in bold.
This study was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC)–Produits forestiers Anticosti Industrial Research Chair (to S.D.C), Université Laval. MAG received scholarships from NSERC (Alexander Graham Bell, Michael Smith, and EnviroNorth scholarships). We thank J.-F. Adam, J. Arsenault-Hétu, C. Bajzak, C. Hins, A. Hemery, F. Lebel, S. Ménard, G. Reuchet, M. Sicilia, and M. Veilleux-Nolin for their valuable assistance in the field. We are grateful to S. De Bellefeuille and C. Hins for logistical support, as well as to D. Gallant, E. Sotka, A. Stien, M.-A. Villard, and two anonymous reviewers for insightful comments on earlier versions of the manuscript. MAG is grateful to N.G. Yoccoz and R.A. Ims for hosting her during her research stay in Tromsø (Norway).
Conceived and designed the experiments: MAG SDC JPT. Performed the experiments: MAG EV. Analyzed the data: MAG EV. Contributed reagents/materials/analysis tools: SDC. Wrote the paper: MAG EV JPT SDC.
- 1. Mysterud A, Østbye E. Cover as a habitat element for temperate ungulates: effects on habitat selection and demography. Wildl Soc Bull. 1999;27(2):385–94.
- 2. Godvik IMR, Loe LE, Vik JO, Veiberg V, Langvatn R, Mysterud A. Temporal scales, trade-offs, and functional responses in red deer habitat selection. Ecology. 2009;90(3):699–710. pmid:ISI:000263776800011.
- 3. Johnson AS, Hale PE, Ford WM, Wentworth JM, French JR, Anderson OF, et al. White-tailed deer foraging in relation to successional stage, overstory type and management of Southern Appalachian forests. Am Midl Nat. 1995;133(1):18–35.
- 4. Massé A, Côté SD. Habitat selection of a large herbivore at high density and without predation: trade-off between forage and cover? J Mammal. 2009;90(4):961–70.
- 5. Dussault C, Ouellet J-P, Courtois R, Huot J, Breton L, Jolicoeur H. Linking moose habitat selection to limiting factors. Ecography. 2005;28(5):619–28. pmid:ISI:000231858800006.
- 6. Winnie J, Creel S. Sex-specific behavioral responses of elk to spatial and temporal variation in the threat of wolf predation. Anim Behav. 2007;73:215–25.
- 7. Verdolin JL. Meta-analysis of foraging and predation risk trade-offs in terrestrial systems. Behav Ecol Sociobiol. 2006;60(4):457–64.
- 8. Happe PJ, Jenkins KJ, Starkey EE, Sharrow SH. Nutritional quality and tannin astringency of browse in clear-cuts and old-growth forests. J Wildl Manage. 1990;54(4):557–66.
- 9. Ford WM, Sydney Johnson A, Hale PE. Nutritional quality of deer browse in southern Appalachian clearcuts and mature forests. For Ecol Manage. 1994;67(1–3):149–57. http://dx.doi.org/10.1016/0378-1127(94)90013-2.
- 10. Hebblewhite M, Merrill E, McDermid G. A multi-scale test of the forage maturation hypothesis in a partially migratory ungulate population. Ecol Monogr. 2008;78(2):141–66.
- 11. Spiller DA, Piovia-Scott J, Wright AN, Yang LH, Takimoto G, Schoener TW, et al. Marine subsidies have multiple effects on coastal food webs. Ecology. 2010;91(5):1424–34. pmid:20503874
- 12. Conradt L. Use of a seaweed habitat by red deer (Cervus elaphus L.). J Zool. 2000;250:541–9. pmid:ISI:000086948400010.
- 13. Hansen BB, Aanes R. Kelp and seaweed feeding by High-Arctic wild reindeer under extreme winter conditions. Polar Res. 2012;31:17258.
- 14. Carlton JT, Hodder J. Maritime mammals: terrestrial mammals as consumers in marine intertidal communities. Mar Ecol Prog Ser. 2003;256:271–86.
- 15. Croll DA, Maron JL, Estes JA, Danner EM, Byrd GV. Introduced predators transform subarctic islands from grassland to tundra. Science. 2005;307:1959–61. pmid:15790855
- 16. Lastra M, Rodil IF, Sánchez-Mata A, García-Gallego M, Mora J. Fate and processing of macroalgal wrack subsidies in beaches of Deception Island, Antarctic Peninsula. J Sea Res. 2014;88(0):1–10. http://dx.doi.org/10.1016/j.seares.2013.12.011.
- 17. Owen-Smith N, Fryxell JM, Merrill EH. Foraging theory upscaled: the behavioural ecology of herbivore movement. Philos Trans R Soc Lon B Biol Sci. 2010;365(1550):2267–78. pmid:20566503
- 18. McLoughlin PD, Gaillard J-M, Boyce MS, Bonenfant C, Messier F, Duncan P, et al. Lifetime reproductive success and composition of the home range in a large herbivore. Ecology. 2007;88(12):3192–202. pmid:18229853
- 19. Litvaitis JA. Investigating food habits of terrestrial vertebrates. In: Boitani L, Fuller TK, editors. Research techniques in animal ecology: controversies and consequences. New York: Columbia Universtiy Press; 2000.
- 20. Dalerum F, Angerbjörn A. Resolving temporal variation in vertebrate diets using naturally occurring stable isotopes. Oecologia. 2005;144(4):647–58. pmid:16041545
- 21. van der Merwe NJ, Medina E. Photosynthesis and 13C/12C ratios in Amazonian rain forests. Geochim Cosmochim Acta. 1989;53:1091–4.
- 22. Li S-G, Tsujimura M, Sugimoto A, Davaa G, Oyunbaatar D, Sugita M. Temporal variation of δ13C of larch leaves from a montane boreal forest in Mongolia. Trees. 2007;21(4):479–90.
- 23. France R. Carbon isotope ratios in logged and unlogged boreal forests: examination of the potential for determining wildlife habitat use. Environ Manage. 1996;20(2):249–55. pmid:ISI:A1996TZ62800009.
- 24. West AG, Midgley JJ, Bond WJ. The evaluation of δ13C isotopes of trees to determine past regeneration environments. For Ecol Manage. 2001;147(2–3):139–49. http://dx.doi.org/10.1016/S0378-1127(00)00474-6.
- 25. Balasse M, Tresset A, Ambrose SH. Stable isotope evidence (δ13C, δ18O) for winter feeding on seaweed by Neolithic sheep of Scotland. J Zool. 2006;270(1):170–6.
- 26. Bessa F, Baeta A, Marques JC. Niche segregation amongst sympatric species at exposed sandy shores with contrasting wrack availabilities illustrated by stable isotopic analysis. Ecol Indicators. 2014;36:694–702. http://dx.doi.org/10.1016/j.ecolind.2013.09.026.
- 27. Anderson WB, Polis GA. Marine subsidies of island communities in the Gulf of California: evidence from stable carbon and nitrogen isotopes. Oikos. 1998;81(1):75–80.
- 28. Ladin ZS, Williams CK, Castelli PM, Winiarski KJ, Osenkowski J, McWilliams SR. Regional and intraseasonal variation in diet of wintering and staging Atlantic brant. The Journal of Wildlife Management. 2014;78(7):1206–15.
- 29. Heaton THE. The 15N/14N ratios of plants in South Africa and Namibia: relationship to climate and coastal/saline environments. Oecologia. 1987;74(2):236–46.
- 30. Gannes L, Del Rio C, Koch P. Natural abundance variations in stable isotopes and their potential uses in animal physiological ecology. Comp Biochem Physiol A. 1998;119(3):725–37.
- 31. Tremblay J-P, Huot J, Potvin F. Divergent nonlinear responses of the boreal forest field layer along an experimental gradient of deer densities. Oecologia. 2006;150(1):78–88. pmid:ISI:000241552400009.
- 32. Hidding B, Tremblay J-P, Côté SD. A large herbivore triggers alternative successional trajectories in the boreal forest. Ecology. 2013;94(12):2852–60. pmid:24597230
- 33. Environment Canada. Climate normals and averages daily data reports of Port-Menier’s station from 1995 to 2005. Ottawa, Canada: Atmospheric Environment Service; 2006.
- 34. Environment Canada. Canadian climate normals, temperature and precipitation, 1951–1980, Québec. Ottawa, Canada: Atmospheric Environment Service; 1982.
- 35. Simard MA, Huot J, de Bellefeuille S, Côté SD. Influences of habitat composition, plant phenology, and population density on autumn indices of body condition in a northern white-tailed deer population. Wildl Monogr. 2014;187:1–28.
- 36. Saucier J-P, Grondin P, Robitaille A, Gosselin J, Morneau C, Richard PJH, et al. Écologie forestière. In: MultiMondes , editor. Manuel de foresterie. Québec: Ordre des ingénieurs forestiers du Québec; 2009. p. 164–316.
- 37. Tremblay J-P, Thibault I, Dussault C, Huot J, Côté SD. Long-term decline in white-tailed deer browse supply: can lichens and litterfall act as alternative food sources that preclude density-dependent feedbacks. Can J Zool. 2005;83(8):1087–96.
- 38. Pellerin S, Huot J, Côté SD. Long term effects of deer browsing and trampling on the vegetation of peatlands. Biol Conserv. 2006;128:316–26.
- 39. Huot J. Body condition and food resource of white-tailed deer on Anticosti Island, Québec. Fairbanks: University of Alaska; 1982.
- 40. Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. Mixed effects models and extensions in ecology with R. Gail M, Krickeberg K, Samet JM, Tsiatis A, Wong W, editors. New York: Springer; 2009.
- 41. Burnham KP, Anderson DR. Model selection and multimodel inference. Second ed. New-York: Springer-Verlag; 2002. 488 p.
- 42. R Development Core Team. R: A language and environment for statistical computing 2.15.0 ed: R Foundation for Statistical Computing, Vienna.; 2012.
- 43. Darling AF, Bayne EM. The potential of stable isotope (δ13C, δ15N) analyses for measuring foraging behaviour of animals in disturbed boreal forest. Ecoscience. 2010;17(1):73–82.
- 44. Medina E, Montes G, Guevas E, Rokzandic Z. Profiles of CO2 concentration and delta13C values in tropical rain forests of the upper Rio Negro basin, Venezuela. J Trop Ecol. 1986;2(3):207–17.
- 45. Farquhar G, O'Leary M, Berry J. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Aust J Plant Physiol. 1982;9(2):121–37. http://dx.doi.org/10.1071/PP9820121.
- 46. O'Leary MH. Carbon isotopes in photosynthesis. Bioscience. 1988;38(5):328–36.
- 47. Montoya JP. Natural abundance of 15N in marine planktonic ecosystems. Stable isotopes in ecology and environmental science: Blackwell Publishing Ltd; 2008. p. 176–201.
- 48. Virginia R, Delwiche CC. Natural 15N abundance of presumed N2-fixing and non-N2-fixing plants from selected ecosystems. Oecologia. 1982;54(3):317–25.
- 49. Smith BN, Epstein S. Two categories of 13C/12C ratios for higher plants. Plant Physiol. 1971;47(3):380–4. pmid:16657626
- 50. Leavitt SW. Seasonal 13C/12C changes in tree rings: species and site coherence, and a possible drought influence. Can J For Res. 1993;23(2):210–8.
- 51. Parnell AC, Jackson AL. siar: Stable Isotope Analysis in R 2011 [Available at http://cran.r-project.org/src/contrib/Archive/siar/ (accessed 12 September 2013).].
- 52. Phillips DL, Gregg JW. Source partitioning using stable isotopes: coping with too many sources. Oecologia. 2003;136(2):261–9. pmid:ISI:000184092500011.
- 53. Phillips DL, Gregg JW. Uncertainty in source partitioning using stable isotopes. Oecologia. 2001;127(2):171–9. pmid:24577646
- 54. Parnell AC, Inger R, Bearhop S, Jackson AL. Source partitioning using stable isotopes: coping with too much variation. PLoS ONE. 2010;5(3):e9672. pmid:20300637