Article

Predicting the Fate of Biodiversity Using Species’ Distribution Models: Enhancing Model Comparability and Repeatability

Ecology and Evolution, Stony Brook University, Stony Brook, New York, United States of America.
PLoS ONE (Impact Factor: 3.23). 09/2012; 7(9):e44402. DOI: 10.1371/journal.pone.0044402
Source: PubMed

ABSTRACT

Species distribution modeling (SDM) is an increasingly important tool to predict the geographic distribution of species. Even though many problems associated with this method have been highlighted and solutions have been proposed, little has been done to increase comparability among studies. We reviewed recent publications applying SDMs and found that seventy nine percent failed to report methods that ensure comparability among studies, such as disclosing the maximum probability range produced by the models and reporting on the number of species occurrences used. We modeled six species of Falco from northern Europe and demonstrate that model results are altered by (1) spatial bias in species' occurrence data, (2) differences in the geographic extent of the environmental data, and (3) the effects of transformation of model output to presence/absence data when applying thresholds. Depending on the modeling decisions, forecasts of the future geographic distribution of Falco ranged from range contraction in 80% of the species to no net loss in any species, with the best model predicting no net loss of habitat in Northern Europe. The fact that predictions of range changes in response to climate change in published studies may be influenced by decisions in the modeling process seriously hampers the possibility of making sound management recommendations. Thus, each of the decisions made in generating SDMs should be reported and evaluated to ensure conclusions and policies are based on the biology and ecology of the species being modeled.

Download full-text

Full-text

Available from: Roland Jansson
    • "Poor model performance stemming from the use of training datasets where the full bioclimatic range of the species is not included is thought to result from: (1) the inability of statistical SDMs to forecast species occurrence probabilities into nonanalogous climate space (e.g. Harrison et al., 2006; Williams & Jackson, 2007; Rodríguez-Castañeda et al., 2012) and/or (2) the instability of model algorithms, as seen in unstable response functions – where the true relationship between the probability of species occurrence and climate is not properly captured – and in arbitrary variable selection (Thuiller, 2004; Thuiller et al., 2004; Araújo et al., 2005b; Barbet-Massin et al., 2010). In the former case, maximizing the climate space used in the training data as much as possible would reduce the chance of SDM errors, whereas in the latter case the reasons for instability of the response function and arbitrary variable selection – and whether they are at all related to truncation of the species' climate space in the training data – would need to be established. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Aim Species distribution models (SDMs) are commonly used to determine threats to biodiversity and opportunities under climate change. Despite SDMs being based on the assumption of complete knowledge of the climate space of the modelled species, truncated occurrence datasets (and hence truncated climate spaces) such as national inventories are often employed. This may lead to prediction errors, which have been proposed to stem from: (1) the degree of climate space truncation and/or (2) instability of the modelling algorithms. Our aim was to explore the potential causes of prediction errors in SDMs using truncated training datasets. Location Europe 11° W–32° E, 34°–72° N. Methods SDMs employing commonly used bioclimatic variables were applied to seven forest tree species. We created two model training datasets covering: (1) Germany only (significantly truncated climate space) and (2) Europe (minimally truncated climate space). Differences between the climate space represented by Germany-only and European data were measured on two-dimensional climate spaces obtained through principal component analysis of the bioclimatic variables. Seven SDM algorithms were run, and the stability of the response function and variable selection for each species and model type were analysed. Results The degree of climate space truncation was less important for model performance than the instability of model algorithms and indiscriminate variable selection. The latter led to irrelevant relationships of species occurrence with bioclimatic variables. These instabilities caused pronounced prediction errors. Main conclusions Our results strongly suggest that erroneous model predictions stem from instability and ecological irrelevance of the statistical functions relating the probability of a species' occurrence to bioclimatic variables, compounded by a lack of consistency in variable selection. Models displaying these characteristics showed lower overall performance when trained with truncated datasets. Further, commonly used ensemble approaches do not compensate for the shortfalls of individual models. Detailed model-by-model and species-by-species analysis of response functions and variable importance is recommended.
    No preview · Article · Nov 2015
  • Source
    • "Although there are various well known methodological issues with species distribution models (e.g. Elith and Leathwick, 2009; Heikkinen et al., 2006; Rodríguez-Casta~ neda et al., 2012), this result is likely to be robust. H. salicifolia is widely planted in the winter-rainfall region of South Africa (Fig. 3B), but in its native range summer-rainfall dominates (Fig. 3A). "
    [Show abstract] [Hide abstract]
    ABSTRACT: The outcome of plant introductions is often considered in binary terms (invasive or non-invasive). However, most species experience a time lag before naturalization occurs, and many species become naturalized at some sites but not at others. It is therefore important to understand the site-specific mechanisms underlying naturalization. Proteaceae is an interesting case as some species are widespread invaders, while others, despite a long history of cultivation, show no signs of naturalization. At least 26 non-native Proteaceae species have been introduced to, and are cultivated in, South Africa. We mapped populations and examined differences between naturalized and non-naturalized populations (e.g. propagule pressure, land use and bioclimatic suitability). Of the 15 species surveyed, 6 were naturalized at one or more sites. Of these, Hakea salicifolia is most widely cultivated, but is only naturalizing in some areas (32 naturalized populations out of 62 populations that were surveyed). We found propagule pressure to be the most important determinant of naturalization for H. salicifolia. However, in suboptimal climatic conditions, naturalization only occurred if micro-site conditions were suitable, i.e. there was some disturbance and water available. For the other naturalized species there were few sites to compare, but we came to similar conclusions – Banksia integrifolia only naturalized at the site where it was planted the longest; Banksia serrata only naturalized at a site influenced by fire regimes; while Banksia formosa naturalized at sites with high propagule pressure, absence of fires and where there is no active clearing of the plants. Naturalization of Proteaceae in South Africa appears to be strongly mediated by site-specific anthropogenic activities (e.g. many planted individuals and water availability). More broadly, we argue that invasion biology needs to focus more closely on the mechanisms by which species and pathways interact to determine the likelihood and consequence of an invasion.
    Full-text · Article · Aug 2014 · Acta Oecologica
  • Source
    • "2011; Rodríguez-Castañeda et al. 2012). Several SDM methods have been developed and applied to investigate species' geographic ranges and possible shifts under global climate change. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Aspalathus linearis (Burm. f.) R. Dahlgren (rooibos) is endemic to the Fynbos Biome of South Africa, which is an internationally recognized biodiversity hot spot. Rooibos is both an invaluable wild resource and commercially cultivated crop in suitable areas. Climate change predictions for the region indicate a significant warming scenario coupled with a decline in winter rainfall. First estimates of possible consequences for biodiversity point to species extinctions of 23% in the long term in the Fynbos Biome. Bioclimatic modelling using the maximum entropy method was used to develop an estimate of the realized niche of wild rooibos and the current geographic distribution of areas suitable for commercially production. The distribution modelling provided a good match to the known distribution and production area of A. linearis. An ensemble of global climate models that assume the A2 emissions scenario of high energy requirements was applied to develop possible scenarios of range/suitability shift under future climate conditions. When these were extrapolated to a future climate (2041-2070) both wild and cultivated tea exhibited substantial range contraction with some range shifts southeastwards and upslope. Most of the areas where range expansion was indicated are located in existing conservation areas or include conservation worthy vegetation. These findings will be critical in directing conservation efforts as well as developing strategies for farmers to cope with and adapt to climate change.
    Full-text · Article · Apr 2014 · Ecology and Evolution
Show more