Introduction
Parasitic life strategies dominate biological systems and can profoundly influence several aspects of host biology, ecology, and community structure (Windsor 1989; Jones et al. 1994; Thomas et al. 1999; Moore 2002; Thomas et al. 2005; Peay et al. 2008; Anderson et al. 2009; Pontoppidan et al. 2009). Several studies have demonstrated a propensity for certain fungal lineages to actively manipulate their hosts through complex changes in behavior and body plan, and these results have captured the attention and imagination of evolutionary ecologists in recent years (Moore 2002; Thomas et al. 2005; Anderson et al. 2009; Pontoppidan et al. 2009; Goodman and Johnson 2011; Hughes et al. 2011; Hughes et al. 2016). The majority of these studies have focused primarily on myrmecophilous Ophiocordyceps species, namelyO. unilateralis , for their particularly spectacular influence on their ant-hosts (Anderson et al. 2009; Hughes et al. 2011). Parasite induced changes in host physiology force ants infected with O. unilateralis to permanently secure themselves in locations specifically suited to fungal development and dispersal via irreversible “death-grip” behavior (Andersen et al. 2009; Hughes et al. 2011; Weinersmith and Faulkes 2014; Thomas et. al. 2005, 2010). Scores of ants killed in this way tend to aggregate into so-called “graveyards”, a phenomenon thought to represent an “extended phenotype” of parasite gene expression (Dawkins 1982; Anderson et al. 2009; Pontoppidan et al. 2009). In the context of parasitic manipulation of host behavior, an extended phenotype refers to the influence of parasite gene expression on host phenotype, particularly when this behavioral manipulation serves to increases parasite transmission and fitness (Dawkins 1982; Hughes et al. 2011; Bailey 2012). Such dramatic examples of extended phenotypes are not limited to Ophiocordyceps species, however, and empirical evidence suggests that the same phenomena may occur among certain specialized fungal parasites of spiders.