ecology of phasmids - KLUEDO - Universität Kaiserslautern

Community structure & host range 9
2 Community structure and host range of phasmids on BCI
2.1 Introduction
Biodiversity of tropical forests is markedly higher than in their temperate counterparts (e.g., Mac Arthur
1972; Erwin 1982) and herbivorous insects constitute a major fraction of species (Erwin 1982; Stork
1993; Ødegard 2000; Novotny et al. 2002a). The origin of tropical species diversity is partly
hypothesized to be based on the arms raise between plants and their consumers. Their impact on leaves
is thought to represent an important selective force for the evolution of plant defenses (Ehrlich & Raven
1964; Rhoades & Cates 1976; Coley et al. 1985). Also, herbivory on seeds and seedlings has been
attributed to the high diversity of tropical plant communities (Janzen 1970; Connell 1971). In return the
greater variety and higher regime of plant defenses in tropical forests are discussed to favor narrower
diets of tropical insect herbivores thereby leading to a higher arthropod diversity than in temperate
forests (Janzen 1973; Coley & Aide 1991).
Aside from plant defenses, natural enemies can select for narrower diets since they have different
abilities to locate a herbivore on variable plant species (Price et al. 1980, Bernays & Graham 1988).
In contrast, the scarcity of most plant species in tropical forests (Hubbell & Foster 1986) may favor the
generalist habit if locating them is difficult and costly for a specialist herbivore, both in terms of time
and exposure to predators (Jaenike 1990; Basset 1992; Coley & Barone 1996).
Indeed most tropical insect herbivores seem to be relatively specific. Even though the debate on the
degree of host specificity remains controversial (e.g. Erwin & Scott 1980; Erwin 1982; Basset 1999;
Basset et al. 1992, 1996; Barone 1998; Ødegard 2000; Novotny et al. 2002a, 2002b) there is evidence
that 90 percent of all phytophagous insects feed on plants in less than three different plant families
(Bernays & Graham 1988; Marquis & Braker 1994; Barone 1998).
The majority of insect herbivore species in tropical forests is rare (Basset 1996, 1999; Basset et al. 1992,
1996; Barone 1998; Novotny et al. 2002b) and abundances vary temporally and spatially. In seasonal
tropical forests, abundances of many insect species are positively correlated to rainfall (Wolda 1978,
1979, 1982, 1992) and leaf production (Murali & Sukumar 1993). Concerning the maintenance of low
overall population densities, the relative roles of lower or higher trophic levels are subject to ongoing
debate but evidence suggests that top-down effects are more important (Coley & Barone 1996; Pace et
al. 1999; Persson 1999; Polis 1999; Dyer & Coley 2001). On a spatial scale insect herbivore densities
are generally assumed to be higher in light gaps than in the forest understory, but this has mainly been
inferred from leaf damage (e.g., Coley, 1980, 1982, 1983; Coley & Barone 1996 and references therein).
Empirical studies on spatial variation in tropical insect herbivore abundances are scarce (e.g., Elton