Editor’s Choice - Identifying Habitat for Endangered Species: Making Habitat Selection Models More Realistic

June 2012 (Issue 49:3)

Moreau, G., Fortin, D., Couturier, S. & Duchesne, T. (2012) Multi-level functional responses for wildlife conservation: the case of threatened caribou in managed boreal forests.Journal of Applied Ecology, 49, 611-620.

One of the most persistent challenges faced by Applied Ecologists is the identification and mapping of key habitat components for managed, threatened or endangered species. Habitat conservation is often specifically targeted in conservation policy, for example, through formal critical habitat protection under the United States Endangered Species Act (ESA) or the Canadian Species at Risk Act (SARA).

Despite the central role of habitat, the definition, identification, and ecological drivers of habitat are often quite challenging to uncover (see recent reviews by McLoughlin et al. 2009; Gaillard et al. 2010). A common approach is to conduct a species-specific study through radiotagging in which the resources associated with the species presence are compared to ‘random’ available sites where the animal might have occurred, or within some defined study area. Much has been written on which habitat modelling approaches, such as Environmental Niche Modeling, MaxEnt, or Resource Selection Function (RSF) models, are more appropriate (Manly et al. 2002; Phillips & Dudik 2008). Basically, all these models statistically compare used (presence) or unused/available locations to understand what resources comprise ‘habitat’ for a species (McLoughlin et al. 2009). By virtue of their shared statistical framework, such approaches also make a number of very questionable assumptions that often limit or restrict their generality. In essence, many studies suffer from being one-off descriptions of patterns in one limited geographic area.

A critical weakness is the assumption that habitat selection by a species remains constant over a wide range of conditions, or availabilities, of a key resource. Imagine a desert bighorn sheep Ovis canadensis nelsoni studied under limiting water conditions. It would be expected to show very strong selection for water, but might be expected to relax its selection as water availability increases. This phenomenon is called a functional response in habitat selection whereby selection changes as a function of availability. First identified as a critical weakness of habitat selection studies in the 1990s by Mysterud & Ims (1998), it is now widely known that this assumption permeates many habitat selection models (Beyer et al. 2010). Functional responses are also common where there is a trade-off between two competing resource types. And yet, it has taken applied ecologists the last 15 years to finally integrate these concepts into conservation modelling. This is because the statistical framework to allow flexibility in modelling functional responses has just emerged through mixed-effects generalized linear models and their extensions (Gillies et al. 2006; Hebblewhite & Merrill 2008; Matthiopoulos et al. 2011).

It is exactly this issue of incorporating biologically meaningful and conservation relevant functional responses in studies of habitat selection that this Issue’s Editor’s Choice exemplifies. Photo supplied by author Daniel Fortin
Woodland caribou Rangifer tarandus caribou are declining rapidly across their North American range (Environment Canada), following declines of the species across the world (Vors & Boyce 2009) because of habitat destruction and fragmentation. One of the challenges in understanding the responses of caribou to human-caused habitat destruction has been reconciling apparently varying responses of woodland caribou to human activity across studies. One study might show strong avoidance of human-caused clearcuts, for example (DeCesare et al. 2010), whereas another shows ambiguous avoidance or even positive selection (reviewed in Moreau et al. 2012). Such discrepancies between studies may then be cited by industry or apathetic governments as evidence for uncertainty in the underlying mechanisms driving habitat selection, and be used to defer conservation planning.

Tackling this problem head on, Moreau et al. (2012) use an ingenious approach to explicitly ask how woodland caribou habitat selection for old growth forest (a key resource for winter lichen, their main forage) changes as a function of availability of human-caused clearcuts. Old growth forests are strongly selected by industrial forestry for their high commercial value. Obviously, as availability of old growth decreases with increased forestry, one might expect caribou to show stronger selection for this key resource. Moreau et al. (2012) explicitly modelled caribou habitat selection responses to a multivariate suite of realistic habitat covariates over a huge 60,674 km2 area. They then used a flexible extension of recent mixed-effects modelling approaches to allow functional responses to the availability of clearcuts to be tested at both the individual and population scales. Indeed, they found a nested-hierarchical response of caribou to clearcuts wherein caribou showed stronger selection for old growth forests in areas of their home ranges with heavy forestry activity. Next, this translated up to the home-range scale where caribou again selected old growth most strongly when the availability of clearcuts in a 5 km radius was the greatest. Combined, their work shows a strong selection by caribou for old growth, especially above a relatively low threshold of logging (>2.5% of their home range) that emphasizes the importance of protection of this key resource at large spatial scales.

 The consequences of increased forestry driving this trade-off for caribou are well known across Canada. Recent meta-analyses of over 26 caribou populations across Canada show that as the proportion of a caribou range within 250 m of a clearcut or road increases above 35%, caribou populations decline because of increased predation by predators artificially buoyed by early seral ungulates such as moose Alces alces through apparent competition (DeCesare et al. 2010). The most important impact of Moreau et al. (2012) is to provide a unifying conceptual and empirical framework to understand varying responses of individual caribou populations and previous studies by explicitly integrating changing availability into a habitat selection model. The advances made by Moreau et al. (2012) provide new ways to do this across threatened and endangered species that will be a great benefit to applied ecologists working on habitat conservation.

Mark Hebblewhite

mark.hebblewhite@cfc.umt.edu

References

Beyer, H. L., Haydon, D. T., Morales, J. M., Frair, J. L., Hebblewhite, M., Mitchell, M. & Matthiopoulos, J. (2010) The interpretation of habitat preference metrics under use–availability designs. Philosophical Transactions of the Royal Society B: Biological Sciences, 365, 2245-2254.
DeCesare, N., Hebblewhite, M., Robinson, H. & Musiani, M. (2010) Endangered, apparently: the role of apparent competition in endangered species conservation. Animal Conservation, 13, 353-362.
Gaillard, J.-M., Hebblewhite, M., Loison, A., Fuller, M., Powell, R., Basille, M. & Van Moorter, B. (2010) Habitat–performance relationships: finding the right metric at a given spatial scale. Philosophical Transactions of the Royal Society B: Biological Sciences, 365, 2255-2265.
Gillies, C., Hebblewhite, M., Nielsen, S. E., Krawchuk, M., Aldridge, C., Frair, J., Stevens, C., Saher, D. J. & Jerde, C. (2006) Application of random effects to the study of resource selection by animals. Journal of Animal Ecology, 75, 887-898.
Hebblewhite, M. & Merrill, E. (2008) Modelling wildlife-human relationships for social species with mixed-effects resource selection models. Journal of Applied Ecology, 45, 834-844.
Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L. & Erickson, W. P. (2002) Resource selection by animals: statistical analysis and design for field studies. Second Edition. Kluwer, Boston, USA.
Matthiopoulos, J., Hebblewhite, M., Aarts, G. & Fieberg, J. (2011) Generalised functional responses for species distributions. Ecology, 92, 583-589.
McLoughlin, P. D., Morris, D., Fortin, D., Van der Wal, F. & Contasti, A. (2010) Considering ecological dynamics in resource selection functions. Journal of Animal Ecology, 79, 4-12.
Moreau, G., Fortin, D., Couturier, S. & Duchesne, T. (2012) Multi-level functional responses for wildlife conservation: the case of threatened caribou in managed boreal forests. Journal of Applied Ecology, 49, 611-620.
Mysterud, A. & Ims, R. A. (1998) Functional Responses in Habitat Use: Availability Influences Relative Use in Trade-Off Situations. Ecology, 79, 1435-1441.
Phillips, S. J. & Dudik, M. (2008) Modeling of Species Distributions With Maxent: New Extensions and a Comprehensive Evaluation. Ecography, 31, 161-175.
Vors, L. S. & Boyce, M. S. (2009) Global declines of caribou and reindeer. Global Change Biology, 15, 2626-2633.

Search the Site

Search

Site Adverts

 

South America Virtual Issue
South America Virtual Issue
Practitioner's Perspective 
 

 

Site Adverts

 

South America Virtual Issue
South America Virtual Issue
Practitioner's Perspective