Description:
<jats:title>Summary</jats:title><jats:p>
<jats:list>
<jats:list-item><jats:p>Species distribution models (<jats:styled-content style="fixed-case">SDM</jats:styled-content>s) have become a standard tool in ecology and applied conservation biology. Modelling rare and threatened species is particularly important for conservation purposes. However, modelling rare species is difficult because the combination of few occurrences and many predictor variables easily leads to model overfitting. A new strategy using ensembles of small models was recently developed in an attempt to overcome this limitation of rare species modelling and has been tested successfully for only a single species so far. Here, we aim to test the approach more comprehensively on a large number of species including a transferability assessment.</jats:p></jats:list-item>
<jats:list-item><jats:p>For each species, numerous small (here bivariate) models were calibrated, evaluated and averaged to an ensemble weighted by <jats:styled-content style="fixed-case">AUC</jats:styled-content> scores. These ‘ensembles of small models’ (<jats:styled-content style="fixed-case">ESM</jats:styled-content>s) were compared to standard <jats:styled-content style="fixed-case">SDM</jats:styled-content>s using three commonly used modelling techniques (<jats:styled-content style="fixed-case">GLM</jats:styled-content>,<jats:styled-content style="fixed-case"> GBM</jats:styled-content> and Maxent) and their ensemble prediction. We tested 107 rare and under‐sampled plant species of conservation concern in Switzerland.</jats:p></jats:list-item>
<jats:list-item><jats:p>We show that <jats:styled-content style="fixed-case">ESM</jats:styled-content>s performed significantly better than standard <jats:styled-content style="fixed-case">SDM</jats:styled-content>s. The rarer the species, the more pronounced the effects were. <jats:styled-content style="fixed-case">ESM</jats:styled-content>s were also superior to standard <jats:styled-content style="fixed-case">SDM</jats:styled-content>s and their ensemble when they were evaluated using a transferability assessment.</jats:p></jats:list-item>
<jats:list-item><jats:p>By averaging simple small models to an ensemble, <jats:styled-content style="fixed-case">ESM</jats:styled-content>s avoid overfitting without losing explanatory power through reducing the number of predictor variables. They further improve the reliability of species distribution models, especially for rare species, and thus help to overcome limitations of modelling rare species.</jats:p></jats:list-item>
</jats:list>
</jats:p>