Are predators negative or positive predictors of farmland bird species community on a large geographical scale?. (March 2016)

• Record Type: Journal Article
• Authors: Kosicki, Jakub Z.; Zduniak, Piotr; Ostrowska, Marta; Hromada, Martin
• Is Part Of: Ecological indicators. Volume 62(2016)
• Page Start: 259
• Page End: 270
• Publication Date: 2016-03
• Subjects: Bird species richness -- Common Buzzard's density -- Common Raven's density -- Conservation -- Farmland -- Predictive mapping -- Surrogate -- Species distribution modeling; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2015.11.003
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: Common Buzzard and the Common Raven, as positive and negative predictors, respectively, of farmland bird species richness. Distributions of farmland species richness were best described by a combination of features measured by remote sensing and density of predators. Species' specific models can improve the predictive power of species distribution models. Abstract: Species distribution models (SDMs) are numerical tools that combine species occurrence/density or species richness with environmental data in order to predict particular species' distribution. In most cases only abiotic environmental parameters are used as predictors, while biotic interactions which control distribution of species and influence the goodness of fit of the SDM, such as predator–prey systems, have been broadly neglected. For this reason, we tested the usefulness of easy to detect predators, such as the Common Buzzard and the Common Raven, as positive and negative predictors, respectively, of farmland bird species richness. We analyzed factors affecting the density of both predators and farmland bird species using data from 958 1 × 1 km 2 study plots in Poland and a set of 22 environmental variables. Next, we also included these predators' densities as additional predictors of farmland bird species. Habitat and climatological predictors were aggregated using the Principal Components Analysis and then related to the Common Raven's and the Common Buzzard's densities as well as farmland bird species richness using General Additive Models. Finally, completed models were assessed according to information – theoretic criteria. Our results showed that all the analyzed groups occurred in open areas; the Common Buzzard and passerine bird species preferred traditional farmland, while the Common Raven reached its highest density in modern intensive farmland. Importantly, we documented a significant increase in the goodness of fit of SDMs for farmland bird species, having added the density of predators as negative (Common Raven) and positive (Common Buzzard) predictors. Consequently, our findings suggest that species' specific models can improve the predictive power of SDMs and can be used as an effective tool for predicting bird diversity with higher accuracy.