THE COST OF MALE SECONDARY SEXUAL TRAITS - DEVELOPMENTAL CONSTRAINTS DURING ONTOGENY IN A SEXUALLY DIMORPHIC WATER STRIDER

Authors
Citation
G. Arnqvist, THE COST OF MALE SECONDARY SEXUAL TRAITS - DEVELOPMENTAL CONSTRAINTS DURING ONTOGENY IN A SEXUALLY DIMORPHIC WATER STRIDER, The American naturalist, 144(1), 1994, pp. 119-132
Citations number
37
Categorie Soggetti
Ecology
Journal title
ISSN journal
00030147
Volume
144
Issue
1
Year of publication
1994
Pages
119 - 132
Database
ISI
SICI code
0003-0147(1994)144:1<119:TCOMSS>2.0.ZU;2-S
Abstract
Natural selection is generally thought to constrain the effects of sex ual selection on secondary sexual traits. In hemimetabolous arthropods , the evolution of complex secondary sexual morphologies may be constr ained by interference during the molting process. This study assesses the effects of phenotypic variation in secondary sexual traits during ontogeny in the water strider Gerris odontogaster. In this species, ma les are provided with a grasping apparatus consisting of two abdominal , ventral processes. It is demonstrated that the length of the process es is positively related to the duration of the ultimate larval molt, presumably because the processes constitute a mechanical hindrance dur ing molting. It is further shown that larvae experience a high risk of mortality from cannibalism during their ultimate molt and that this r isk is density-dependent. On the basis of the assumption that cannibal istic events occur at random during molting, males with long processes will suffer higher risk of mortality during molting than will males w ith short processes. It is concluded that this form of developmental c onstraint can affect the evolution of male abdominal processes, and qu antitative cost curves for the length of male abdominal processes are presented. Simulations of natural selection due to larval mortality sh ow positive density dependence, in contrast to sexual selection in thi s species. It is suggested that the net effect of selection on seconda ry sexual traits in this species varies between negative and positive values and that net selection is zero only within a narrow range of en vironmental conditions. A comparison between natural populations showe d that mean trait value co-varies negatively with environmental variab les related to the cost of the trait in accordance with this suggestio n.