Avian predation upon lizards and frogs in a neotropical forest understorey

Citation
B. Poulin et al., Avian predation upon lizards and frogs in a neotropical forest understorey, J TROP ECOL, 17, 2001, pp. 21-40
Citations number
46
Categorie Soggetti
Environment/Ecology
Journal title
JOURNAL OF TROPICAL ECOLOGY
ISSN journal
02664674 → ACNP
Volume
17
Year of publication
2001
Part
1
Pages
21 - 40
Database
ISI
SICI code
0266-4674(200101)17:<21:APULAF>2.0.ZU;2-S
Abstract
Although tropical forest birds are known to prey upon small lizards and fro gs, no study has documented the attributes of vertebrate-eating birds or wh ether birds prey opportunistically on the different elements of the herpeto fauna within tropical communities. This study is based on a 14-mo investiga tion on avian diet, supplemented with a 3-y census of frogs and a 1-y censu s of lizards in a humid forest of central Panama. From 91 bird species, 108 6 regurgitates were collected, in which were found 75 lizards and 53 frogs. Over 50% of the common, primarily insectivorous bird species preyed upon l izards or frogs, with a mean frequency of 0.26 prey/sample. These birds (22 species, nine families) foraged on various substrates from different strat a of the forest, fed on invertebrates averaging from 3.3 to 17.2 mm in leng th, weighed from 11 to 195 g, and had bill lengths that varied from 12.2 to 49.8 mm. Based on a logistic regression analysis, intensity of foraging at army-ant swarms was the variable that best explained the likelihood that a bird species preyed upon lizards, leading to a classification that was 91% correct. In contrast, bill length and body length classified correctly 88% of the frog-eating birds, which showed a fairly constant 1:7 bill length/b ody length ratio (as opposed to a mean but highly variable 1:10 ratio in ot her species). A multiple regression analysis showed that seasonal variation in intensity of lizard predation was positively related to arthropod abund ance except during the breeding season when lizard intake decreased, presum ably because nesting birds did not follow ant swarms. Intensity of frog pre dation correlated with frog abundance over time, the latter being inversely related to arthropod availability. Ninety-seven per cent of all lizards an d frogs identified in the diet samples (n = 105) were from two genera, Anol is and Eleutheradactylus, respectively. Prey size distribution in the regur gitates suggested an optimal prey size of 33.5 mm snout-vent length (SVL) f or lizards and 14.5 mm SVL for frogs. Birds preyed opportunistically on the different Anolis species, but almost exclusively upon juvenile individuals . Abundances of the different Eleutherodactylus species correlated with the ir predation rates, but these frogs represented only 10% of all the frogs o bserved during the censuses. The two most common local anurans, Colostethus flotator and Bufo typhonius, were not taken by any bird species.