A lack of functional wings is at the heart of two obstacles to understanding the evolutionary biology of some earth-bound cockroaches. First, aptery and brachyptery are associated with a developmental syndrome that reduces morphological complexity, making it difficult to distinguish among closely related taxonomic groups. Second, the loss of mobility associated with aptery can result in complex geographic substructuring of these morphologically ambiguous groups.
Wing reduction or loss is the best indicator of paedo-morphosis, defined as the retention of juvenile characters of ancestral forms in the adults of their descendents (Matsuda, 1987; Reilly, 1994). Not all short-winged insects retain juvenile characters, but in other cases, it is clear that many so-called adult characters are absent in short-winged or apterous morphs (Harrison, 1980). The diminishment or loss of structures such as ocelli, compound eyes, antennal and cercal segments, and some in-tegumental structures such as sensilla often accompanies aptery and brachyptery (Matsuda, 1987). These reductions are common in cockroaches (Nalepa and Bandi, 2000), and like other animals (Howarth, 1983; Juberthie, 2000b; Langecker, 2000) occur most often in species that inhabit relatively safe, stable environments, such as caves, burrows, logs, social insect nests, leaf litter, and other cryptic environments. Lefeuvre (1971) found that some cockroach species with reduced wings have fewer developmental stages than macropterous relatives, and that juvenile features can be retained in the tracheal system, peripheral nervous system, and integument. Warnecke and Hintze-Podufal (1990) concluded that the reduced wings of female Blaptica dubia are the result of larval characters that persist into maturity, rather than the growth inhibition of adult wings. Other examples include the retention of styles in wingless adult females of Noc. termitophila (female cockroaches normally lose their styles prior to the adult stage) (Matsuda, 1979), and the reduced sensory and glandular systems of the myrmecophile Att. fungicola (Brossut, 1976). Cryptocercus has reduced eyes and cercal segmentation, and exhibits marked paedomorphic traits in its genital morphology (Walker, 1919; Crampton, 1932; Klass, 1995). Females of the desert cockroach A. in-vestigata are "generally nymphlike," lack the wings and ocelli seen in the male, and have shorter antennae and cerci (Friauf and Edney, 1969). Because wing loss in cockroaches is female biased, it is most often females that exhibit correlated paedomorphic characters.
The systematics of paedomorphic organisms can be frustrating. Because many structures never develop or develop variably within a group, they cannot be used to delimit taxa, or to infer phylogenetic relationships. Independent losses of ancestral postmetamorphic features is an important source of homoplasy and can confound cladistic analysis (Wake, 1991; Brooks, 1996; Hufford, 1996). The morphological homogeneity of the Poly-phagidae has caused quite a few problems with attribution, not only to species but also to genera (Failla and Messina, 1987). Members of the genus Laxta "vary so much in color and size and have genitalia so similar as to make distinguishing taxa difficult" (Roth, 1992). Paedo-morphic characters and mosaic evolution in the woodfeeding cockroach Cryptocercus strongly contribute to problems in determining the phylogenetic relationships of this genus at all taxonomic levels (Klass, 1995, 1998a; Nalepa and Bandi, 1999, 2000; Nalepa et al., 2002). Cave cockroaches, like other cave dwellers (Howarth, 1983; Ju-berthie, 2000a; Langecker, 2000), are prone to taxonomic problems associated with paedomorphosis. Roth (1990b) noted that Para. stonei from different caves all had reduced hindwings but varied in body size, in the development of pulvilli, and in length of tegmina. The genitalia were so similar, however, that he assigned them to different races within the species. A morphometric study by Slaney and Weinstein (1997b) subsequently supported Roth's conclusions.
Molecular and chemical tools are increasingly required to provide characters to distinguish among these morphologically ambiguous cockroach taxa. Humphrey et al. (1998), for example, used protein electrophoresis to propose that morphologically similar populations of M. rhi noceros are comprised of three genetic species. Slaney and Blair (2000) used the ITS2 gene region of nuclear ribo-somal DNA in the Para. stonei group, and their results supported conclusions based on morphology. Molecular phylogenetic relationships, however, are not always completely congruent with relationships based on morphological characters. Basal relationships among species of the wood-feeding blaberid Salganea are poorly resolved by molecular analysis, probably because of rapid and potentially simultaneous radiation of the group (Maekawa et al., 1999a, 2001).
In flightless animals the pool of potential mating partners is limited to those that can be found within walking distance, resulting in restricted levels of gene flow. Populations may become subdivided and isolated to varying degrees, resulting in complex genetic substructuring and the formation of local species, subspecies, and races. This is common in caves, where subterranean spaces can be isolated or locally connected via mesocaverous spaces (Barr and Holsinger, 1985). It is also common on mountains, where endemic races and subspecies may be wholly restricted to single peaks (Mani, 1968). Cryptocerus pri-marius, for example, is found in an area of China with a dissected topography characterized by high mountain ridges sandwiched between deep river gorges, forming various partitioned habitats (Nalepa et al., 2001b). This genus of montane cockroaches is also dependent on rotting logs, which ties their distribution to that of mature forests. Any event that has an impact on the distribution of forests, including glaciation (Nalepa, 2001; Nalepa et al., 2002) and deforestation (Nalepa et al., 2001b) will affect the population structure of the cockroach. Consequently the geographic distribution of genetic populations and species groups in both Northeast Asia (Park et al., 2004; Lo et al., 2000b) and the eastern United States (Nalepa et al., 2002) can be unexpected. Cryptocercus found in southern Korea, for example, are more closely related to populations in Northeast China than they are to all other Korean members of the genus.
Was this article helpful?