Many marine and freshwater fish move about their environment in a cohesive group called a school and the question, "why do fish school?", has often been raised (e.g., Cushing and Harden Jones, 1968; Weihs, 1973). In numerous species schooling is thought to have evolved primarily as an antipredator strategy but few experimental data are available to refute or support this hypothesis (Brock and Riffenburgh, 1960; Manteifel' and Radakov, 1961; Williams, 1964; Breder, 1967; Shaw, 1970). This report presents a new source of evidence favoring an antipredator role for schooling. Because most evolutionary changes occur imtially at the population or subpopulation level, studies of intraspecific (geographic) variation may yield valuable clues to the adaptive significance of schooling behavior. Natural populations of a small tropical freshwater fish, the guppy (Poecilia reticulata) , provide excellent material for studying behavioral responses to predation. In the Northern Range Mountains of Trinidad, West Indies, several geographically isolated (no interconnecting freshwater) and semi-isolated populations of guppies are exposed to different intensities of predation by other fish species (Haskins et al., 1961; Seghers, 1973; Liley and Seghers, 1974). In some rivers predation by large characid and cichlid piscivores, notably Hoplias malabaricus and Crenicichla alta, is severe. In other rivers these predators are absent and a smaller, carnivorous cyprinodontid fish, Rivulus hartii, is the major predator. In a few rivers Rivulus density is very low and guppies enjoy virtual freedom from fish predation. (Detailed measurements of other environmental variables are in Seghers, 1973, and Liley and Seghers, 1974.) I made observations on the schooling tendencies of 5 populations in the Northern Range area (Table 1). The tendency for fish to remain together was estimated semi-quantitatively. These observations were supplemented with observations on wild fish placed in aquaria and outdoor pools. Schooling is poorly developed in the Upper Aripo and Paria R. The most common social groups involve fish engaged in courtship activity (e.g., three males following and displaying to a female). These groups do not persist for more than a few
[1]
D. Cushing,et al.
Why do Fish School?
,
1968,
Nature.
[2]
C. Breder.
On the survival value of fish schools
,
1967,
Zoologica : scientific contributions of the New York Zoological Society..
[3]
I. Vine,et al.
Risk of visual detection and pursuitby a predator and the selective advantage of flocking behaviour.
,
1971,
Journal of theoretical biology.
[4]
E. Hobson.
Predatory behavior of some shore fishes in the Gulf of California
,
1968
.
[5]
D. Weihs.
Hydromechanics of Fish Schooling
,
1973,
Nature.
[6]
B. Seghers.
Analysis of geographic variation in the antipredator adaptations of the guppy : Poecilia reticulata
,
1973
.
[7]
I. Vine.
Detection of prey flocks by predators.
,
1973,
Journal of theoretical biology.
[8]
H. Pulliam,et al.
On the advantages of flocking.
,
1973,
Journal of theoretical biology.
[9]
Robert H. Riffenburgh,et al.
Fish Schooling: A Possible Factor in Reducing Predation
,
1960
.
[10]
Manteifel' Bp,et al.
The adaptive significance of schooling behaviour in fishes.
,
1961
.
[11]
W. Hamilton.
Geometry for the selfish herd.
,
1971,
Journal of theoretical biology.