Some detailed biogeographical/macroecological patterns - Nonequilibrium Ecology - page 158

strong support to the view that saturation, overall, has not been reached;
(see Chapter 6 ).
Nonsaturation of habitats with species, i.e., the existence of vacant
niches, was discussed in detail on pp. 39-48. Here, I repeat the main
points. There is indeed ample evidence that many vacant niches exist.
This can be shown by comparing species numbers in similar habitats. For
example, examinations of many species of marine fish have shown
remarkable differences in numbers of parasite species (Figure 2.6 ) .
Rohde ( 1998a ) estimated that a maximum of only about 16% of all niches
available to metazoan ectoparasites of marine fishes are filled, if the
assumption is made that the maximum reached by some fish species is
indeed the maximum possible, for which - however - there is no
evidence. Walker and Valentine ( 1984 ) estimated that 12-54% of niches
for marine invertebrates are vacant, and Lawton and collaborators (e.g.,
Lawton 1984b ) demonstrated the availability of many empty niches for
insects of bracken (p. 130).
Several studies deal with the effects of temperature on generation times
and the effect of generation time on speed of evolution (for an early
discussion see Rensch 1954). Tyler et al.( 1994 ) have shown that butterflies
have shorter generation times in the tropics (for poikilotherms and homo-
iotherms in general see discussion in Rohde 1992 and references therein).
Although there is no evidence for shorter generation times of tropical birds
(Mayr 1976 ), speed of selection (independent of generation time) and
mutation rates may still be affected by temperature, and, furthermore, the
evolutionary speed of birds may be determined by diversification of taxa
with shorter generation times lower down in the ecological hierarchy.
Sibley and Ahlquist ( 1990 , further references therein) gave a detailed
discussion of the effects of generation time on the rate of DNA evolution,
and concluded that ''generation time does have an effect on the average rate
of genome evolution'' (see also Martin and Palumbi 1993 ,furtherrefer-
ences in Cardillo 1999 ). Bromham et al.( 1996 ) showed, using 61 mammal
species and controlling for phylogeny, that substitution rates at fourfold
degenerate sites in two of three protein sequences were negatively corre-
lated with generation time. They found no evidence for any effect due to
metabolic rates on sequence evolution (see also Mooers and Harvey 1994 ).
Barraclough et al.( 1996 ) found a positive correlation between the rate of
evolution of the rbcL chloroplast gene with the number of species per family
within families of angiosperms, unrelated to generation time. Rosenheim
and Tabashnik ( 1993 ) evaluated a large data base on the evolution of
pesticide resistance of insects and found that the number of generations