FIRST STEP IN DEVELOPMENT OF A QUADRAT-BASED PROTOCOL TO SAMPLE GOPHER TORTOISE (GOPHERUS POLYPHEMUS) BURROWS

FIRST STEP IN DEVELOPMENT OF A QUADRAT-BASED PROTOCOL TO SAMPLE GOPHER TORTOISE (GOPHERUS POLYPHEMUS) BURROWS

I developed a preliminary quadrat-based protocol to survey low density populations of burrows of the Gopher Tortoise (Gopherus polyphemus) and applied the protocol in four upland habitats at Avon Park Air Force Range, Florida in 1997–98. I randomly selected and surveyed 55 1-ha plots in oak scrub, 2...

Ausführliche Beschreibung

Bibliographische Detailangaben
Veröffentlicht in:Florida Scientist. - The Florida Academy of Sciences, Inc.. - 75(2012), 3, Seite 189-208
1. Verfasser: Kinlaw, Alton E. (VerfasserIn)
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2012
Zugriff auf das übergeordnete Werk:Florida Scientist
Schlagworte:Biological sciences Arts Physical sciences Mathematics
Beschreibung
Zusammenfassung:I developed a preliminary quadrat-based protocol to survey low density populations of burrows of the Gopher Tortoise (Gopherus polyphemus) and applied the protocol in four upland habitats at Avon Park Air Force Range, Florida in 1997–98. I randomly selected and surveyed 55 1-ha plots in oak scrub, 23 in sand pine scrub, 53 in pine flatwoods, and 17 in oak hammock. Detection probability was maximized by using 10 transects and multiple observers at most quadrats. Oak scrub had the highest density of burrows per ha (1.47) and sand pine scrub had the lowest density (0.17). Spatial distribution of oak scrub burrows fit the negative binomial distribution. Likelihood ratio tests showed that the best model for all four habitats had a dispersion parameter (k) of the negative binomial that did not vary but did have a variable mean (m). This protocol is appropriate to survey low density populations of any object in situations where distance sampling could be inefficient, and can be improved by conducting double-observer surveys at a subset of quadrats to calibrate detection probability. This approach is supported by two subsequent studies that showed that distance sampling is not efficient in cases where densities of burrows are low and that quadrat sampling may offer advantages over other tortoise survey techniques.
ISSN:00984590