-We studied the location and site selection of burrows of black iguana, Ctenosaura similis, at Palo Verde, Guanacaste, Costa Rica, in 1989 and 1990. Our study site included the environs of buildings as well as dry lowland deciduous forest. Iguana had burrows on earth banks, and in trees, logs and rocks with a higher frequency than occurred at randomly-selected points. Burrows differed from random points in having less cover over the entrance and within 1 m (both rock and vegetation) than did the random points. Temperatures at the burrow entrance were cooler than at the random points, although air temperatures did not differ significantly. We suggest that the use of burrows and selection of burrow sites by black iguanas may relate to predation pressures, thermal constraints and daily activity patterns. Descriptions of general habitat use are numerous for reptiles (e.g., Heatwole, 1977), although there are fewer descriptions of nest sites (see Van Devender and Howard, 1973; Burger and Zappalorti, 1986, for review) or of nonnesting burrows. Studies of site selection, whereby the authors have shown that the reptiles chose particular sites over all available sites, are even less common. Studies of selection require the location of a sufficient sample of nests or burrows for statistical analyses, and it is often difficult to find reptile nests or burrows. Nonetheless, nest site selection has been studied in alligators (Alligator mississippensis; Garrick and Lang, 1977), green iguanas (Iguana iguana; Rand, 1968), turtles (Burger and Montevecchi, 1975; 430 This content downloaded from 157.55.39.153 on Mon, 19 Sep 2016 04:52:46 UTC All use subject to http://about.jstor.org/terms BURROW SELECTION BY IGUANAS Ewert, 1976; Stoneburner and Richardson, 1981), and pine snakes (Pituophis melanoleucus; Burger and Zappalorti, 1986). Most reptiles do not regularly dig burrows for daily occupation (see Wright and Wright, 1957), although many reptiles use subterranean burrows dug by mammals. However some lizards (Rand and Dugan, 1983) and a snake (Carpenter, 1982; Burger et al., 1988) dig burrows for nests or for use throughout the year. Presumably the location of these burrows reflects evolutionary constraints such as predation pressures, conspecific competition, and thermoregulation. In this paper we examine burrow site selection in black iguana (Ctenosaura similis) at Palo Verde, Guanacaste, Costa Rica. We were interested in whether the same burrows were used from year to year, whether burrow site characteristics varied between years, and whether burrow sites differed from the available habitat. Black iguanas are large iguanid lizards ranging from southern Mexico to Panama. The flesh is edible, and they are often hunted for food (Fitch and Hackforth-Jones, 1983). They range in size from 20-50 cm snout-vent length (SVL), with overlap between the sexes, although males are generally larger (Fitch and Hackforth-Jones, 1983). METHODS AND STUDY AREA We studied black iguanas at the Palo Verde National Park in Guanacaste, northwestern Costa Rica, in March of 1989 and 1990. Palo Verde contains mainly tropical lowland deciduous forest that borders the extensive marshes of the Tempisque River. Further description of the study area can be found in Gill (1989) and Burger and Gochfeld (1991). In 1989 we surveyed the area immediately around the OTS biological station, the deciduous forest within 25 m of the road between the OTS station and the park headquarters (hereafter called the hacienda), and around the hacienda (1 km from the station). We searched by walking transects through each area in a manner that allowed us to see nearly all of the habitat. Burrows included all holes where iguanas lived, but did not include nests or spaces under buildings. We included only burrows actively being used by black iguanas. We used photographs and sketches from 1989 to locate the sites of all burrows for comparison with 1990. To compare overall habitat choice (bank, ground, flat rock, rock or trees, see Table 2) with available habitat, we used a table of random numbers to generate coordinates for a grid used to select points for the entire study site. General habitat types included bank (slope over 15?), ground (flat area), flat rock, rocks or boulders, and trees, logs or roots. We recorded specific site characteristics (see Table 3) from all burrows and from an equal number of random points in each of the three plac s (OTS station, deciduous forest, hacienda). Specific site characteristics recorded included: slope; substrate (ground, rock, bank); percent cover over the opening; percent vegetation cover within 1 m of the point (or burrow entrance estimated); percent rock within 1 m; percent overall cover (vegetation, trees, rock) within 1 m; distance to the closest burrow and closest tree; and temperature of the air and at the mouth of the burrow entrance or spot. We so recorded the height and width of the burrow entrance for all burrows whether they were dug or were in rock crevices. Random points were selected by using a table of random numbers to generate a compass direction and a distance (1 to 4 m) from each burrow. Thus, each random point was matched to a burrow site. We estimated iguana length by noting where an iguana started and ended on the habitat and then measuring this distance, and checked our estimates against iguanas of known length (N = 20 from marked individuals, r = 0.9, P < 0.001). We were usually only 1-3 m from the iguana, so the estimates were possible. In 1989 both general habitat type (bank, ground, flat rock, rock, logs) and specific site characteristics were recorded for random points and for burrow sites. In 1990 we recorded the general habitat types for burrows at the OTS station, hacienda, and forest in between, but we tripled the size of the areas searched in equal proportion to the three types. We also recorded some site characteristics, including location, height and width of burrow entrance, percent cover at the entrance, overall slope of area with burrow, slope at the burrow entrance, nearestneighbor distance, and the length of the burrow. Burrow lengths are minima because we had difficulty measuring beyond the bend if a burrow made a major turn. Indeed it was impossible to determine if the stick was at the end. We used Wilcox X2 tests to determine differences in the distribution of site characteristics between the burrows and random sites, and Contingency x2 tests to determine differences between the general habitat types of random sites and burrows in 1989 and 1990, and between the 1989 and 1990 burrow sites.