Modelling species distributions using regression quantiles

Abstract

Summary 1.Species distribution modelling is an important and well-established tool for conservation planning and resource management. Modelling techniques based on central estimates of species responses to environmental factors do not always provide ecologically meaningful estimates of species-environment relationships and are being increasingly questioned. 2. Regression quantiles (RQ) can be used to model the upper bounds of species-environment relationships and thus estimate how the environment is limiting the distribution of a species. The resulting models tend to describe potential rather than actual patterns of species distributions. 3. Model selection based on null hypothesis testing and backward elimination, followed by validation procedures, are proposed here as a general approach for constructing RQ limiting effect models for multiple species. 4. This approach was applied successfully to 16 of the most abundant marine fish and cephalopods in the eastern English Channel. Most models were validated successfully and null hypothesis testing for model selection proved effective for RQ modelling. 5. Synthesis and applications . Modelling the upper bounds of species-habitat relationships enables the detection of the effects of limiting factors on species' responses. Maps showing potential species distributions are also less likely to underestimate species responses' to the environment, and therefore have subsequent benefits for precautionary management.

Quantification and prediction of the impact of fishing on epifaunal communities

“…The 75th, 80th, 85th, 90th and 95th quantiles of epifaunal coverage and biomasses measurements of different functional and taxonomic groups were tested in relation to fishing frequency and to each environmental variable to determine which parameters were inducing a response of the epifauna and at which quantile levels they showed an effect. The choice of these 5 quantiles was adapted from an approach used to detect species responses (Vaz et al 2008). The use of different quantiles did not impair the interpretation of the results as the regressions were only used to detect potential limiting factors rather than to compare responses among variables.…”

“…Additionally, sponges are known to be able to adapt their body form in response to physical environmental conditions (Bell & Barnes 2000). If the sponge species observed in our study are able to adapt their body form to a less vulnerable encrusting morphology, they may benefit from the impact of fishing on the competing epifauna.Quantile and PLS regressions are rarely adopted in studies of human and environmental impacts (see Vaz et al 2008, Carrascal et al 2009 for further details), but, in the present study, were valuable tools to reveal relationships in a dataset with multiple interacting environmental factors. Quantile regressions can handle zero-inflated datasets (Vaz et al 2008), which was useful when studying variation in rare taxonomic groups.…”

Curvas de índices de local em povoamentos de eucalipto obtidas por regressão quantílica

“…Na pesquisa florestal, ela tem sido empregada em trabalhos sobre a predição da distribuição de diâmetros (Mehtätalo et al, 2008), do crescimento de árvores (Coomes & Allen, 2007a), da mortalidade (Coomes & Allen, 2007b), da sucessão ecológica (Dahlgren et al, 2006), do crescimento em altura após incêndio (Meyer et al, 2005), de interação genótipo x ambiente (Vaz et al, 2008) e para o ajuste de funções de afilamento (Cao & Wang, 2015).…”

Interaction between classification detail and prediction of community types: implications for predictive modelling of benthic biotopes

“…Richards et al 2007) and ecology (e.g. Berger & Hildenbrandt 2000, Moore et al 2009, Vaz et al 2008, Kearney & Porter 2009, Sundblad et al 2009, Chatfield et al 2010, Pleydell & Chretien 2010. The underpinning ecological and evolutionary theories and assumptions have been the major driving force in ecological studies (Guisan & Thuiller 2005).…”