Abstract
Maintaining landscape connectivity through identifying movement corridors is the most recommended conservation strategy to reduce the negative impacts of habitat loss and isolation. The basis of most connectivity modelling approaches for modelling corridors is that species choose movement pathways based on the same criteria they used to choose habitats. However, species behave differently in using landscape elements for moving than for selecting habitat. In other words, suitability of a given landscape feature may differ between moving and habitat use stages. In this study, we evaluated how the availability of migration occurrence points for wild sheep (Ovis orientalis) could affect the outputs of distribution models and consequently the strength and extent of landscape connectivity for migratory movements of this species in central Iran. We employed concepts of Maximum entropy and circuit theory and developed 2 groups of habitat suitability and connectivity models with and without migration data. Comparing the results of the developed models showed that the main differences in the outputs of MaxEnt models were associated with suitability values predicted for the unprotected migration habitats. Without migration occurrence points, MaxEnt did not identify the traditionally used migration habitats. In contrast, Circuitscape represented a similar performance in predicting the main migration corridor of the species when using or not using the migration occurrence data. These differences could be associated with wild sheep's different behavior in the selection of habitat during movement and home range stages. Owning to this difference, we suggest using migration data in modelling landscape connectivity as these data may include different environmental conditions to those collected from home range habitats. For wild sheep, we recommend protecting the migration corridor at least during migration time. Maintaining such connectivity would also largely depend on managing the unprotected matrix through preventing expansion of human land uses in the vicinity of the corridor and buffering it to some distance.
Highlights
Herbivorous mammals demand large tracts of continuous marginal and/or seasonal habitats (Forman and Gordon, 1986) to meet the full suite of their daily and year-round requirements
By comparing results of the Species distribution models (SDMs), and the circuit models, we aimed to find out how well circuit model can predict landscape connectivity based on occurrence data related to home range habitat and compare how different preferences of wild sheep in selecting home range and migration habitats would affect the output of the circuit model
In the second MaxEnt model, unprotected highly suitable habitat patches were predicted in close proximity to the southern edge of the Haftad-Gholleh PA
Summary
Herbivorous mammals demand large tracts of continuous marginal and/or seasonal habitats (Forman and Gordon, 1986) to meet the full suite of their daily and year-round requirements. Landscape connectivity is defined as the ability of a landscape in facilitating or impeding individuals’ movements between suitable habitat patches (Taylor et al, 1993). This definition consists of 2 main components: structural connectivity which refers to the impact of. The second component is functional connectivity describing movements of individuals or genes across a landscape (Taylor et al, 1993) This kind of connectivity results from interactions between individuals’ ecological characteristics such as habitat preference and dispersal ability and structural characteristics of the landscape (Rudnick et al, 2012). An important benefit of this definition is that from this perspective, every feature across a landscape could have the potential to provide functional connectivity regardless of being structurally connected or not
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
