Fm. Jaksic et al., A LONG-TERM STUDY ON THE DYNAMICS OF GUILD STRUCTURE AMONG PREDATORY VERTEBRATES AT A SEMIARID NEOTROPICAL SITE, Oikos, 67(1), 1993, pp. 87-96
For 48 mo we monitored diets and populations of 10 species of predator
y vertebrates (two foxes, four falconiforms, and four owls) in semi-de
sert scrub at Auco, Chile. Pellets (birds) and feces (foxes) collected
monthly contained identifiable remains of small mammals, birds, repti
les, amphibians, large insects, other arthropods, and (in some fox fec
es) fruits. For non-mammalian prey, seasonal variation in availability
appeared to be more pronounced than longer-term variation. In contras
t, availability of the eight species of small mammals varied relativel
y little between predators' breeding and non-breeding seasons but decl
ined by an order of magnitude through the first 3 yr of the study. Rat
her than responding to these shifts in food availability by diverging
in diet during ''lean'' times and converging opportunistically during
''fat'' times, predators at Auco responded in idiosyncratic ways. Pred
ators of all species tended to maintain species-specific means in size
of prey taken regardless of shifts in the relative availability of pr
ey size classes. Two owl species maintained strict mammal diets throug
hout the decline in mammal abundance, whereas two other owls plus the
two fox species maintained a tight feeding guild based on arthropods a
s well as mammals and other food groups. The four falconiforms failed
to maintain consistent diets from season to season, and consequently e
nded up temporarily in unique positions in ''diet hyperspace'' or else
joined temporarily one of the two consistent guilds. The degree to wh
ich falconiform diets were unique in a given season showed no clear re
lation to food abundance. Changes in use of habitat did, however, rela
te to changes in food abundance. None of the falconiforms resided cont
inuously in the study area throughout the period, and their occurrence
was particularly sporadic late in our study even for falconiform spec
ies that were not particularly mammalivorous. The lack of clearcut pat
terns among Auco predators in response to resource-''lean'' and resour
ce-''fat'' times may result from one or more of the following: (a) fro
m predators' viewpoints, food classes at Auco may be quite discrete, i
nhibiting gradual diet shifts by owls and foxes at least; (b) truly op
portunistic immigrants that might have increased overall diet overlap
during ''fat'' times, as sometimes occurs in other consumer assemblage
s, do not exist in the Auco landscape; and (c) despite the pronounced
decline in small mammal densities. the predator assemblage may have re
mained above the threshold of food limitation through part or all of t
he study.