Dynamics of Insect Mutualisms
Context-Dependent Outcomes. Much of my early research evaluated questions concerning the dynamics of insect mutualisms. I was particularly interested in how the strength and sign of interactions, and the selection pressures which accompany them, changed in space and time. Although a number of studies had previously recognized the context-dependent nature of antagonistic interactions, my work provided one of the first attempts to experimentally examine the extent and consequences of such variation in mutualistic associations.
Focusing on two homopteran-ant associations in Arizona and Colorado, I showed that the occurrence and strength of benefits was highly variable. These interactions were 1) temporally-variable, as homopterans benefitted from the association only in years when ants significantly deterred spider predators (Cushman and Whitham 1989); 2) age-specific, as only nymphs benefitted from ant tending due to their extreme susceptibility to predators (Cushman and Whitham 1989); 3) density-dependent, as large homopteran aggregations benefitted proportionally more from ant tending than small aggregations (Cushman and Whitham 1989); and 4) mediated by competition, as homopterans on neighboring host plants competed with each other for the protective services of ant mutualists (Cushman and Addicott 1989; Cushman and Addicott 1991; Cushman and Whitham 1991). I also proposed and evaluated a fifth kind of variation in which host plant quality can mediate the outcome (both the sign and strength) of homopteran-ant and lepidopteran-ant interactions (Cushman 1991).
Mutualism Between Hosts and Visitors. Previous studies of putative mutualisms have concentrated primarily on the effects of animal visitors on the survival and reproduction of their plant or animal hosts. Considerably less attention has been directed toward potential host benefits related to development time, which is surprising given that development time can have substantial effects on reproductive success and population dynamics. In addition, although the definition of mutualism requires that both participants benefit from their association, most studies have neglected to consider the perspective of visitors, or have simply assumed that visitors benefit from their interactions with hosts (Cushman and Beattie 1991).
In southeastern Australia, my collaborators and I addressed these two issues in a study of the interactions between lycaenid butterfly larvae and ants (Cushman et al. 1994). Our work showed that lycaenid larvae reared with ants were 31-76% heavier, developed 37% faster, and commonly completed one or two fewer instars than larvae reared without ants. Ant tending also resulted in 20% heavier pupae and 69% shorter pupal stages. With respect to the visitor’s perspective, ant workers had 40% greater survivorship when butterfly larvae were present, presumably because ants consumed the lycaenid’s nectary gland secretions, which contained considerable amounts of glucose and amino acids. These results are important because they show that both hosts and visitors benefit from their association, that lycaenid larvae benefit from ants in ways other than, or in addition to, protection from natural enemies, and that lycaenids did not incur substantial metabolic costs from associating with ants.
Context-Dependent Outcomes. Much of my early research evaluated questions concerning the dynamics of insect mutualisms. I was particularly interested in how the strength and sign of interactions, and the selection pressures which accompany them, changed in space and time. Although a number of studies had previously recognized the context-dependent nature of antagonistic interactions, my work provided one of the first attempts to experimentally examine the extent and consequences of such variation in mutualistic associations.
Focusing on two homopteran-ant associations in Arizona and Colorado, I showed that the occurrence and strength of benefits was highly variable. These interactions were 1) temporally-variable, as homopterans benefitted from the association only in years when ants significantly deterred spider predators (Cushman and Whitham 1989); 2) age-specific, as only nymphs benefitted from ant tending due to their extreme susceptibility to predators (Cushman and Whitham 1989); 3) density-dependent, as large homopteran aggregations benefitted proportionally more from ant tending than small aggregations (Cushman and Whitham 1989); and 4) mediated by competition, as homopterans on neighboring host plants competed with each other for the protective services of ant mutualists (Cushman and Addicott 1989; Cushman and Addicott 1991; Cushman and Whitham 1991). I also proposed and evaluated a fifth kind of variation in which host plant quality can mediate the outcome (both the sign and strength) of homopteran-ant and lepidopteran-ant interactions (Cushman 1991).
Mutualism Between Hosts and Visitors. Previous studies of putative mutualisms have concentrated primarily on the effects of animal visitors on the survival and reproduction of their plant or animal hosts. Considerably less attention has been directed toward potential host benefits related to development time, which is surprising given that development time can have substantial effects on reproductive success and population dynamics. In addition, although the definition of mutualism requires that both participants benefit from their association, most studies have neglected to consider the perspective of visitors, or have simply assumed that visitors benefit from their interactions with hosts (Cushman and Beattie 1991).
In southeastern Australia, my collaborators and I addressed these two issues in a study of the interactions between lycaenid butterfly larvae and ants (Cushman et al. 1994). Our work showed that lycaenid larvae reared with ants were 31-76% heavier, developed 37% faster, and commonly completed one or two fewer instars than larvae reared without ants. Ant tending also resulted in 20% heavier pupae and 69% shorter pupal stages. With respect to the visitor’s perspective, ant workers had 40% greater survivorship when butterfly larvae were present, presumably because ants consumed the lycaenid’s nectary gland secretions, which contained considerable amounts of glucose and amino acids. These results are important because they show that both hosts and visitors benefit from their association, that lycaenid larvae benefit from ants in ways other than, or in addition to, protection from natural enemies, and that lycaenids did not incur substantial metabolic costs from associating with ants.