BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 38 PT. 1(2)


activity and a bimodal pattern centered on spring and autumn activity peaks. On
the Ordway Preserve, such patterns were observed, but the patterns were more
complex with less adherence to a particular modality. Failure to conform to
recognized patterns suggests that sampling may be biased, that 1989 may have
been an unusual year in terms of snake activity, that concepts of modality in
subtropical snake activity patterns need to be refined to incorporate the possibility
of both variation and complex patterns, or that all of these factors may be true. In
areas with year-round activity, snakes seem to keep the general patterns observed
in more northern latitude relatives but retain the plasticity to modify activity in
accordance with local environmental conditions (Dalrymple et al. 1991b).

 Sampling considerations.-Funnel traps have been used successfully to obtain
data on both individual snake species and communities (Fitch 1960 1987; Ford et
al. 1991). Clearly, the technique is effective at capturing certain upland species,
such as Coluber, Micrurus, and Sistrurus. Six or more snakes were captured in
only five traps throughout the study, providing little evidence of biased trapping.
Although more traps had zero capture success in xeric habitats in 1990 than in
1989, the sampling effort was only 16% of that in 1989. Likewise, many mesic
habitat traps had zero capture success during the two-year study, but mesic habitats
were sampled for only 28% of the sampling total. Snakes did not shy away from
any cluster of traps in any habitat type.
 On the other hand, why did most traps capture few snakes? Several
hypotheses are possible, including insufficient sampling effort, generally poor trap
placement, low snake density, or escape from traps before the observer checked
them. Snakes also may avoid traps in which other snakes had been caught, perhaps
due to chemoreceptive cues left by previous occupants (but see Weldon et al. 1990).
None of these hypotheses can be ruled out, and all may affect capture success.
 The 14 species of snakes trapped during the study represent 61 percent of the
snakes known from the Ordway Preserve (Franz this vol.). Other snake species
(Diadophis punctatus, Drymarchon corals, and Tantilla relicta) are known from
upland habitats on the Ordway Preserve but were not trapped. In addition, general
collecting, radio-telemetry studies, and subjective impressions suggest that
additional species, such as Crotalus adamanteus (Timmerman 1989), Elaphe
guttata, E. obsoleta, and Pituophis melanoleucus, were underrepresented in funnel
traps in relation to their probable abundance. Such discrepancies suggest that
funnel trapping alone is inadequate to sample all species of an upland snake
community.
 Snake size, habitat specificity, and foraging mode may play an important role
in the effectiveness of funnel traps to sample communities. Those upland species
either not trapped or underrepresented were generally small as adults (Diadophis,
Tantilla) or very large and robust as adults (Crotalus, Drymarchon, Pituophis). On
the other hand, small robust Sistrurus and large slender Masticophis were trapped.
Habitat specificity, such as fossorial (Tantilla) or arboreal (E. guttata, Opheodrys)