Browsing by Subject "MULTIMODEL INFERENCE"

Sort by: Order: Results:

Now showing items 1-3 of 3
  • dos Santos, J. W.; Correia, R. A.; Malhado, A. C. M.; Campos-Silva, J.; Teles, D.; Jepson, P.; Ladle, R. J. (2020)
    Scientific knowledge of species and the ecosystems they inhabit is the cornerstone of modern conservation. However, research effort is not spread evenly among taxa (taxonomic bias), which may constrain capacity to identify conservation risk and to implement effective responses. Addressing such biases requires an understanding of factors that promote or constrain the use of a particular species in research projects. To this end, we quantified conservation science knowledge of the world's extant non-marine mammal species (n = 4108) based on the number of published documents in journals indexed on Clarivate Analytics' Web of Science (TM). We use an innovative hurdle model approach to assess the relative importance of several ecological, biogeographical and cultural factors for explaining variation in research production between species. The most important variable explaining the presence/absence of conservation research was scientific capacity of countries within the range of the species, followed by body mass and years since the taxonomic description. Research volume (more than one document) was strongly associated with number of years since the data describing on that species, followed by scientific capacity within the range of species, high body mass and invasiveness. The threat status was weakly associated to explain the presence/absence and research volume in conservation research. These results can be interpreted as a consequence of the dynamic interplay between the perceived need for conservation research about a species and its appropriateness as a target of research. As anticipated, the scientific capacity of the countries where a species is found is a strong driver of conservation research bias, reflecting the high variation in conservation research funding and human resources between countries. Our study suggests that this bias could be most effectively reduced by a combination of investing in pioneering research, targeted funding and supporting research in countries with low scientific capacity and high biodiversity.
  • Sumasgutner, Petra; Terraube, Julien; Coulon, Aurélie; Villers, Alexandre; Chakarov, Nayden; Kruckenhauser, Luise; Korpimäki, Erkki (2019)
    Selecting high-quality habitat and the optimal time to reproduce can increase individual fitness and is a strong evolutionary factor shaping animal populations. However, few studies have investigated the interplay between land cover heterogeneity, limitation in food resources, individual quality and spatial variation in fitness parameters. Here, we explore how individuals of different quality respond to possible mismatches between a cue for prey availability (land cover heterogeneity) and the actual fluctuating prey abundance.
  • Kelm, Detlev H.; Toelch, Ulf; Jones, Mirkka M. (2021)
    Background: Mixed-species groups in animals have been shown to confer antipredator, foraging and other benefits to their members that may provide selective advantages. In most cases, however, it is unclear whether functional benefits are a principal driver of heterospecific groups, or whether groups simply result from simultaneous exploitation of common resources. Mixed-species groups that form independently of environmental conditions may, however, evidence direct benefits of species associations. Bats are among the most gregarious mammals, with sometimes thousands of individuals of various species roosting communally. Despite numerous potential functional benefits of such mixed-species roosting groups, interspecific attraction has never been shown. To explore alternative explanations for mixed-species roosting, we studied roost selection in a speciose neotropical understory bat community in lowland rainforest in Costa Rica. Long term roost data were recorded over 10 years in a total of 133 roosts comprising both natural roosts and structurally uniform artificial roosts. We modelled bat roost occupancy and abundance in each roost type and in forest and pasture habitats to quantify the effects of roost- and environmental variability. Results: We found that bat species presence in natural roosts is predictable from habitat and structural roost parameters, but that the presence and abundance of other bat species further modifies roost choice. One third of the 12 study species were found to actively associate with selected other bat species in roosts (e.g. Glossophaga commissarisi with Carollia sowelli). Other species did not engage in communal roosting, which in some cases indicates a role for negative interspecific interactions, such as roost competition. Conclusions: Mixed-species roosting may provide thermoregulatory benefits, reduce intraspecific competition and promote interspecific information transfer, and hence some heterospecific associations may be selected for in bats. Overall, our study contributes to an improved understanding of the array of factors that shape diverse tropical bat communities and drive the dynamics of heterospecific grouping in mammals more generally.