Landscape patterns and dispersal success: simulated population dynamics in the brown treecreeper

Abstract

We used a spatially explicit, individual-based simulation model to explain the possible role of patch isolation in causing observed declines of Brown Treecreepers (Climacteris picumnits) in northern New South Wales, Australia. Using aerial photographs and a geographic information system, we created a spatially realistic landscape in which territories were constrained to woodlands and the matrix composition mimicked the actual landscape in the study area. We compared observed population behavior to the outcome Of Simulations based on two dispersal movement rules and three dispersal mortality rules. Under the first movement rule, a dispersing bird's initial direction was selected at random (Random); under the second rule, a bird's initial direction was toward its nearest neighbor (Neighbor). The first mortality rule used a constant mortality rate for dispersing birds, whereas the second and third varied mortality rate dependent on habitat type traversed. In simulations using the Random movement rule, populations in contiguous habitat were relatively stable, whereas populations in fragmented habitat steadily declined due to low female recruitment, which is the same pattern observed in the real population. Populations in both contiguous and fragmented habitat increased under the Neighbor movement rule, suggesting that population dynamics may be sensitive to dispersal search patterns. Varying mortality rate with habitat type had a large effect on population behavior even though long-distance dispersal was infrequent. Thus, for this cooperative breeder, where territories become clumped due to a high rate of territorial budding, matrix habitats continue to influence overall population performance. In this system, conservation efforts must address the quality of the matrix habitats as well as the configuration of remaining habitat.