196 
Psyche 
[Vol. 91 
in amplexus upon our arrival in the field were still together several 
hours later when we departed. Pairs on flowers in the laboratory 
stayed together overnight. We found that these males were guarding 
females both prior to and following copulation (Dodson and 
Marshall 1984). 
Presumably time spent with one female, particularly after 
copulation has taken place, decreases a male’s opportunities for 
copulations with additional females. Why, then, might males be 
employing this guarding tactic? One possible answer is that the low 
population densities favor males that, upon finding a female, remain 
with her and actively repel other males. If such an attempt to 
monopolize a potential mate increased the chance of successfully 
copulating with her, precopulatory guarding might be a better tactic 
than spending this time searching for other females. This, however, 
seems an unlikely explanation for postcopulatory mate guarding. 
Parker (1970) argued persuasively that postcopulatory defense of a 
mate (preventing other males access to her) is an evolved strategy to 
counter sperm competition from potential, subsequent male mates. 
Although nothing is known regarding sperm precedence in ambush 
bugs specifically, most evidence for insects shows an insemination 
advantage for the last male to mate (Walker 1980). 
We feel there is another factor that may be important to our 
understanding of mate guarding in these ambush bugs (and perhaps 
in other insects). This factor is male preference for particular 
females. At one of our study sites, males were paired with females 
that were heavier than females without males (Dodson and 
Marshall 1984). These paired females also had more eggs than single 
females. The apparently superior fecundity of these females seemed 
to us to be a potential fitness gain for males. To further investigate 
the possible adaptiveness of selective pairing by males we looked at 
egg size within females. We had previously determined the wet 
weight (WW), pronotal width (PW) and femur length (FL) of these 
same females. All 3 measurements were significantly correlated with 
egg volume (BW: r = .60, p = .003; PW: r = .61, p = .003; FL: r = 
.62, p = .002). The 3 measurements are so highly correlated with 
each other, the residual variation examined by partial correlations 
was found to be negligible (r E v fl • ww pw = .04, p = .84; r E v pw • ww fl 
= .24, p = .3 1 ; r EV ww • pw fl — .27, p = .25). Therefore, all 3 variables 
are approximately equivalent predictors of egg volume. We believe. 
