ecology of phasmids - KLUEDO - Universität Kaiserslautern

Community structure & host range 24
2.4 Discussion
This study is the first to demonstrate diversity and distributional patterns of the phasmid community of a
tropical rainforest. The phasmid community of BCI was poor in species and in density. Spatially
phasmids were largely restricted to forest edge habitats resulting in higher diversity and density
compared to the forest understory. While phasmid abundances varied throughout the study period, no
seasonal effect on phasmid abundance was detectable. The distribution pattern of phasmid species was
clearly linked to restrictions in their host range that seemed to reflect the successional status of host
plant species.
2.4.1 Phasmid diversity, density, and distribution
With a total of 12 species, the phasmid community of Barro Colorado Island was expectedly low in
species richness. Worldwide, the order Phasmatodea contains about 3000 described species that occur
mainly in tropical regions (Whiting et al. 2003) and is species poor compared to other insect orders that
are highly diverse in herbivores like Lepidopterans, Coleopterans, and - to a lower extent –
Orthopterans. For example, on BCI there are approximately 274 species of true butterfly (Lepidoptera,
Papilionoidea; DeVries 1994) and 250 species of katydids (Orthoptera, Tettigoniidae; A. Lang,
unpublished data). I found reports of 42 phasmid species occurring in Panamá (see Appendix 1 for
species and references) whereof BCI accommodates 26 %; plus one species not formerly reported from
Panamá (Otocrania sp.). Analysis of collection records (Windsor, Aiello, Berger, unpublished data)
indicates that Gamma-Diversity in Panamá is high with discrete regions giving shelter to different
phasmid species. Together with the fact that I recorded all but one species previously reported from BCI
(Trychopeplus laciniatus; Aiello, unpublished data) I am confident that I assessed the phasmid
community adequately.
I showed that forest edges and gaps represented important habitats to the majority of phasmid species on
BCI. In contrast, the forest understory seemed to provide a suitable habitat for M. diocles only. Higher
abundance of insect herbivores in gaps is often inferred from higher herbivore damage on gap-adapted
versus understory plant species (Coley 1980, 1982, 1983; Coley & Barone 1996). Smiley (1978) found
that Heliconius butterflies were more abundant in young successional than in understory habitats
(Table 2-3). My results support this pattern on the phasmid community and species level. In forest edges
phasmid density was 35 times, and M. diocles density three times higher than in the understory (cf.
Table 2-1). Low densities of phasmids in the BCI forest understory correspond with Elton (1973) who
found low total arthropod densities of 0.92 ind.*m -3 in the same habitat (18400 ind.*ha -1 , extrapolated
on two meter height; Table 2-4).