Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

Ecological Traps: Modeling habitat selection for mobile animals in heterogeneous landscapes
Jessica Hearns

Last modified: 2014-03-28


We used a research framework including sources, sinks, and ecological and perceptual traps to evaluate the role that habitat selection plays in native species adaptation to transformed landscapes. We built a simulation model in MATLAB to evaluate the correspondence between habitat selection and quality with forest generalist birds in tropical agroforestry plantations. We generated an algorithm that defines ecological neighborhoods, and evaluates the spatial and temporal patterns of habitat selection behaviors and habitat quality outcomes. We started with a continuous cover of forest that was converted to coffee in variable proportions by replacing individual cells of forest by coffee in different configurations. These included (1) random, (2) lateral, (3) radial, and (4) percolation replacement. Coffee and forest cells can have constant or variable canopy cover as identified with field sample in an actual plantation in Santa Marta, Colombia. We distinguished between three hypothesized possible roles for habitat selection (neutral, adaptive or maladaptive) and created a spatially-explicit population model to explore the consequences of the interplay between habitat selection, quality and availability. We use a novel and multi-scale methodological approach to link landscape structure, individual behavior, and population outcomes of mobile animals in heterogeneous landscapes. We defined bird attributes in both habitats based on data collected. We distinguished between juveniles, breeding adults and floaters whose survival and fecundity rates varied by habitat and stage. We allowed the birds to search for nesting territories within four different distances (defined by number of cells). We incorporated an algorithm that simulated field sampling to assess potential researcher biases. We tested theoretical predictions of habitat selection models using a study system to prompt generally applicable lessons for management and conservation. We found clear differences between scenarios of neutral, adaptive and maladaptive habitat selection consistent with current theoretical predictions and evaluated vital rates changes associated with these scenarios. The largest effects were associated with changes occurring during the transient demographics before all territories are occupied. Birds with neutral selection distributed homogenously in the landscape and were intermediate in population growth. Birds with adaptive selection settled in the favorable habitat and had the highest population growth. Birds with maladaptive selection concentrated in the most unfavorable habitat and had the lowest population growth. These differences were mostly due to lowest recruitment in the unfavorable habitat. Sampling affected the ability to identify these differences. We were able to evaluate how previous methodological limitations are hindering our ability to explore the ecological mechanisms allowing species persistence in anthropogenic and disturbed regions.