Thomas W. Schoener and David A. Spiller
This passage is adapted from Thomas W. Schoener and
David A. Spiller, "Trophic Cascades on Islands." ©2010 by
Island Press
In the 1970s, one of us visited more than 500
Bahamian islands to survey distributions of
vertebrates, with special emphasis on lizards and
birds. A key objective was to determine the threshold
island area on which vertebrate populations could just
survive. We were astonished to find lizards,
particularly Anolis sagrei, on some tiny islands, a
discovery that multiplied by at least two orders of
magnitude the list of Bahamian islands surmised or
known to have resident populations of vertebrates.
We realized we had to check many quite small islands
to determine such thresholds, and in the course of
that endeavor we came upon a large number of
islands without lizards. This led to a second, even
more exciting discovery: Such islands sometimes had
extraordinarily high densities of spiders, the
omnipresent webbing giving them the appearance of
the proverbial grandmother's attic.
In 1981, we had to investigate this phenomenon
systematically for the many small islands in the
central Bahamas near the relatively large island of
Staniel Cay, a major stopover in our earlier survey.
Our first study found that spiders were about an
order of magnitude denser on no-lizard than lizard
islands (adjusted for the positive and negative
correlations with area and distance from large
landmasses, respectively). A second observational
study in 1982 examined numbers of spider species,
finding that no-lizard islands had 1.5-2 times the
number of species as had lizard islands (again
adjusted for area and distance, and for the maximum
height attained by the vegetation on the island, which
correlated positively with number of spider species).
This result was quite different from Paine's (1966)
famous one in the rocky intertidal, in which diversity
increased increasing predation, and it presaged other
such results for terrestrial arthropods in our system
and in others also.
Such comparative data pointed to a strong
negative effect of lizards on spiders, but as is true of
all comparative studies, the observations did not
suffice to eliminate alternative hypotheses about why
islands with and without lizards might differ. A more
reliable investigation would be experimental, and
toward that end we staked out nine approximately 83-
square-meter plots on Staniel Cay in 1985. Three of
the plots were unenclosed, and the others had wood-
framed fences made of hardware cloth topped with
smooth plastic to impede lizard locomotion in and
out. Three of the enclosed plots were randomly
chosen to maintain lizards at natural densities,
whereas the other three had lizards removed. Thus we
had three treatments: The two types of enclosed plots
tested the lizard effect, and the unenclosed plots were
a cage control, to be compared with the enclosed
lizard plots. The 18-month experiment showed that
lizard removal enclosures had spider densities three
times higher than those control enclosures and the
unenclosed (also having lizard) plots. Numbers of
spider species were higher without lizards as well, in
parallel to the comparative studies. Numbers and
biomasses of insects caught in the sticky traps were
also higher in lizard removal enclosures; therefore, an
increase in spiders did not completely compensate for
the absence of lizards. There was some effect of the
enclosures: Sticky traps in the enclosed plots caught
about 20 percent fewer arthropod individuals than
those in open plots.
What was the mechanism of the (now firmly
established) lizard effect on spiders? The obvious one
is predation. However, a second is competition for
food: Spiders consume prey large in relation to their
own size, so lizards and spiders might overlap in prey
size well beyond their relative body sizes alone.
تعليقات
إرسال تعليق