Tadpoles involved in predator-prey interactions were studied in tropical wet forest in Costa Rica under laboratory and field conditions. Larvae of the frog Leptodactylus pentadactylas and naiads of the odonate Pantala flavescens are important predators on larvae of several species of frogs. The predators discriminate the prey on the basis of size and species, but not type of habitat in which predation occurs. A graphical model is proposed to illustrate the relationships between species diversity and habitat complexity as they affect the composition of tadpole communities. The model is used to evaluate the relative importance of abiotic and biotic factors in determining the use of specific kinds of aquatic habitats by frogs with larval stages. Predation by permanent aquatic predators (primarily fish) is considered to be the most important biotic factor influencing the temporal and spatial composition of tadpole communities. The development and maintenance of predatory feeding modes, including cannibalism, in certain tadpoles is examined in light of the model.MOST TADPOLES are generalized herbivores (Jenssen 1967) that either scrape plant material from a substrate or filter planktonic food directly from the water. Only a few species are known to be carnivorous although some instances of cannibalism among tadpoles have been reported (Bragg 1964). We were surprised, therefore, to find a well-developed predator-prey system in which tadpoles of one species were carnivores on other tadpoles. This finding was unusual not only because documented instances of tadpoles feeding on other tadpoles are relatively rare, but also because the carnivorous tadpoles were found in relatively small puddles.The predator-prey interactions studied occur in small rain-filled tropical puddles characteristically utilized by opportunistic breeders. Experiments were run to study the relative ability of predators to capture prey and their preference for certain prey species, to ascertain the influence of microhabitat on predator-prey interactions, and to evaluate the influence of prey size on their ability to escape predation.Results of these experiments led to a consideration of tadpole habitats and to those factors controlling the diversity of tadpole communities. A model is used to illustrate the relationship between pond complexity and predation as they affect the species composition of a tadpole community. The predatory feeding mode of tadpoles is considered in light of this model. METHODS AND MATERIALSObservations and experiments were made during two weeks in June 1973 (WRH and RWM) and two weeks in August 1973 (DLW and RWM) in the vicinity of the Osa Field Station, Rincon de Osa, Puntarenas Province, Costa Rica. The station is surrounded by the remnants of the lowland wet forests that once covered much of the Golfo Dulce region. This area receives more than 4000 mm of rain each year and is the only remaining large tract of wet forest on the Pacific coast of Middle America. Most of the work was done in the immediate vicinity of ...

No abstract

No abstract

Immediate and delayed (96 h) mortality, descaling, and injury rates of striped bass Morone sax‐atilis and rainbow trout Oncorhynchus mykiss that were inserted into the entrance (suction side) of a Hidrostal pump were compared to those of control fish that were inserted at the exit (pressure side). Comparisons were made in 158‐paired trials of striped bass and 86 trials with rainbow trout. The Hidrostal pump had no significant effect on immediate or delayed (96 h) mortality. Immediate mortality for striped bass and rainbow trout averaged 0.1 and 0%, respectively, and delayed mortality averaged 2.9 and 0.1%, respectively. Mean scale loss after passage for striped bass and rainbow trout was low (0.2 and 1.0%, respectively). Frequency of injury to the head, eyes, skin, and tins of pumped striped bass averaged 1.9, 2.8, 1.9, and 18.7%, respectively, and those of rainbow trout averaged 2.3, 0, 2.4, and 3.1%. No significant relationships were detected between fish mortality and pump speed, injected fish density, and debris load. Results suggest that large Hidrostal pumps have the capacity to transport live striped bass and rainbow trout at high density (1‐6 fish/L) with little mortality and body injury.