printbranchlizard

(dij) as the minimum edge-to-edge distance between patches. No lizards crossed water. Thus, we defined the distance between patches separated by aquatic habitats as the minimum edge-to-edge distance between the patches without crossing water (i.e., by measuring the minimum distance around the water). Therefore, to obtain an indirect measure of dispersal across habitat gaps, we measured the edge-to-edge distance between each scrub patch and its nearest occupied neighbor, and then calculated the ratio of occupied to unoccupied patches within distance categories (e.g., ≤100 m, 100 to 200 m, etc.). Second, to supplement this indirect assessment, we measured dispersal distances of juvenile lizards within a scrub patch using mark and recapture techniques. The maximum distance between a scrub patch and its closest occupied neighbor was 750 m. Thus, we used 750 m as an estimate of maximum dispersal distance for scrub lizards. However, the proportion of patches that were occupied by scrub lizards declined sharply when the gap between patches exceeded 200 m (Fig. 2.5). Of the 277 juvenile lizards marked to assess dispersal distances within a scrub patch, 71 lizards were recaptured. Most lizards moved less than 100 m from their release site (Fig. 2.6). The maximum distance was 330 m. No interpatch dispersals were observed. Patches were grouped into clusters of patches defined such that the greatest distance between neighboring patches within a cluster was <750 m (the estimated maximum dispersal distance for scrub lizards). Because scrub lizards rarely survive more than 2 years (Branch et al. 1996), we described local population demographics using a 3-stage Table 3.2. Primary assumptions associated with the numerical simulation model for Florida scrub lizards. 1 Local demographics are adequately described by a 3-stage matrix. 2 Small patches (<7 ha) have lower vital rates (survival and fecundity) and abundance than do large patches. 3 Stochastic effects are not correlated among patches. 4 Lizard abundance for each occupied patch starts out at carrying capacity and at a stable age distribution. 5 Ceiling density dependence operates in all patches (see Akçakaya 1994). 6 Only lizards in stage 1 migrate. 7 Only 10% of the stage-1 lizards from each local population (patch) migrate out of their natal patch. 8 Migrants are equally distributed among all patches within 750 m of the source patch. We assumed that females oviposit an average of 4 eggs per clutch and produce an average of 3 clutches per year (Jackson and Telford 1974). However, some eggs are probably not viable or fail to develop. We did not have estimates of egg viability for scrub lizards but assumed that egg mortality was similar to estimates (5 to 10%) for other Sceloporus species (e.g., Overall 1994). We used the incidence function model derived by Hanski (1994) and a numerical simulation approach to model regional population dynamics of scrub lizards. The incidence function approach uses presence/absence data to estimate metapopulation parameters that describe the relationship between colonization/extinction rates and landscape characteristics (e.g., patch size and patch configuration). In contrast, the numerical simulation model uses detailed demographic data to project population responses. With the proper input, the numerical simulation approach is capable of providing the most complete model of the regional dynamics of any particular cluster of habitat patches (e.g., Hanski and Thomas 1994, LaHaye et al. 1994).