9307 (2) - Laserfiche WebLink

Home Browse Search

- RELATIONSHIPS OF BIRDS, LIZARDS, AND NOCTURNAL RODENTS TO "I'1-iFIR HABITAT IN TUCSON, ARIZONA METHODS Study Design I sampled birds, lizards, and land cover across the entire range of residential development present in the study area. I used a randomized sampling design to minimize biases associated with misrepresentation of habitat availability (Johnson 1980). I used a Geographical Information System (GIS) to generate 33 random points along roads and trails throughout the study area to serve as starting points for 5-km or 5.5-km census routes. I placed a point every 0.5 km along each route as the center of census plots for birds, lizards, and land cover descriptors. This design resulted in 334 census plots within the study area. All routes followed existing roads or trails. I used the following criteria to avoid clustering census plots, to avoid censuring edges between primary land cover types, and to keep within-plot habitat uniform: 1. Starting points were placed >_ 1 km apart; 2. Census plots on adjacent transects had to be >_ 200 m apart; 3. Census plots were not placed along 4-lane roads; 4. Census plots were not placed within a 0.5-km buffer zone around Interstates 10 and 19; 5. All census plots were located below 975 m elevation; 6. No census plot was established adjacent to an urban setting with residential densities differing >_ 1 housing density class from that of the plot itself (Shaw et al. 1993); and, 7. Commercial and industrial areas were excluded (they were beyond the scope of this study). In addition, census plots adjacent to and including portions of vacant lots were moved entirely within the lots and as close as possible to the lot center. I rejected a plot if any of the above 7 criteria were violated. If a plot was rejected, I added 1 to the end of the route. If > 2 plots were rejected per route, the entire route was replaced. 91Vildlife Data Collection Breeding Birds. I collected abundance data for breeding birds from March 15 through June 24, 1994. This period coincided with or overlapped peak nesting activity for most species present (Davis and Russell 1990, Corman 1993). For several other species, the census period coincided with the establishment of territories and nest site selection. I did not consider migrant species in this study. I made no effort to eliminate unmated individuals (floaters) from population estimates, so my estimate of breeding habitat for individual species was likely broader than actually used. Since floaters often occur in marginal habitat, comparisons of bird abundances among habitat types may have artificially inflated probabilities, thus differences would be harder to detect. Since both females and males are likely under similar energetic constraints during the breeding season, and visually distinguishing between sexes of some species is difficult, both sexes were included in counts of abundance at each census plot. I conducted censuses from 0530 to 1030 Mountain Standard Time (MS'1~ using 50-m fixed- radius, circular plots (Fowler and McGinnes 1973, Verner 1988). This early morning period coincided with the peak in avian daily activity (Robbins 1980). I chose fixed-radius circular plots because accurate distance estimation to singing birds would have been too difficult with the background noises in urban settings. In addition, a stationary observer can detect slow moving birds, cryptic birds, and birds moving into and out of the census plot better than even a slow moving observer (Verner 1988). Moreover, I selected fixed plots over transects because a primary objective was to collect data from independent and homogeneous plots, and area is usually positively correlated with habitat heterogeneity. I completed 2 census routes each morning, conducting a 5-min. census at each plot. Each plot was visited 4 times. Visits were separated by 3-4 weeks. Iran each route twice in the early morning and twice in the late morning. Iran each route backward once during each time period. I conducted all censuses and identified all birds by sight or sound. Censuses were temporarily halted during interruptions due to jet, traffic, or human distractions. Censuring was not conducted in rainy (continual, heavy enough to keep pavement wet) or windy (> 3 on Beaufort scale) weather. Wintering Birds. I censused winter resident birds from January 2 through February 20, 1995, using the same methods as described for breeding ~TEPHEN $. GERMAINE 1995 ARIZONA GAME Fs FISH DEPARTMENT, TECH. REP. 20