Pollinating seed eaters: why is active pollination so rare?

Ecology, Sept, 1997 by Olle Pellmyr

INTRODUCTION

Many seed-eating insects have life histories that straddle the interface between pollination and herbivory. Those that oviposit during the flowering period face the risk that a flower receiving eggs will not develop into a fruit, and their progeny accordingly face starvation (Rattray 1913, Brantjes 1976, Stirton 1976, Pellmyr 1989, Straw 1989, Petterson 1991, 1992, Brody 1992, Donaldson 1992). For this reason, one might expect that assuring pollination should be an adaptive character in many such insect lineages, with active pollination being a possible outcome. I use "active pollination" here in contrast to "passive pollination," to refer to cases where specific morphological structures and behavior components exist in the pollinator for the purpose of picking up and transporting pollen, and depositing it on stigmas.

Contrary to this expectation, there are only two known origins of active pollination in the hundreds of lineages of seed-eating insects, namely in yucca moths and in fig wasps (Bronstein 1992, Pellmyr and Thompson 1992, Powell 1992, Brown et al. 1994, Herre et al. 1996). Here I ask what factors may explain this scarcity. First, I ask whether the costs of being an active pollinator are so high as to make it less likely to evolve. Concluding that the cost is modest, I then discuss ecological factors that affect the probability of selection for active pollination. In a final section, I review similarities between active pollination and analogous systems of brood substrate creation involving insects that transport fungal spores (rather than pollen), and suggest issues where comparative analyses across these subdisciplines could prove profitable.

THE COSTS OF BEING A POLLINATOR

There are many estimates of the plant's cost of mutualism in pollination systems based on seed eaters (Riley 1892, Janzen 1979, Keeley et al. 1984, Addicott 1986, Pellmyr 1989, Bronstein 1992, Pellmyr and Thompson 1992, Thompson and Pellmyr 1992, West and Herre 1994, Richter and Weis 1995), but none of the animal's cost of being a pollinator. This asymmetry may stem in part from the ease of measuring the plant's cost, which is the magnitude of seed destruction. Costs can be measured, but their fitness consequences are not so readily quantified in the pollinators.

For this purpose, I studied yucca moths, in which females use unique mouthparts to pick up host pollen and then use them to pollinate flowers immediately after ovipositing into the ovary (Riley 1892). I assume that pollinating yucca moths incur three types of costs: (1) the time spent collecting and depositing pollen, (2) the biomass allocated to the unique maxillary tentacles used for pollen manipulation, and (3) the weight of transporting a pollen load. I measured these three parameters for the moth that pollinates Yucca filamentosa in Ohio, a member of the taxonomically unresolved Tegeticula yuccasella complex. Behavior data were collected throughout the flowering period, during 927 June 1992 and 13-30 June 1993, in a long-established population at the Cincinnati Zoo, Cincinnati, Ohio. Moths became active at [approximately equal to]2100 (LDT), and their activity was recorded until it ceased, which typically happened between 2300 and 0100. Each observer followed one female moth at a time, using a handheld data logger (Psion Organizer II; Psion, Watertown, Connecticut) to record behavior types and durations. Behaviors were readily classified into either of seven categories: search, pollen collection, oviposition, pollination, mating, harassment by males, and rest. Search behaviors included search within and among flowers, and flight to new inflorescences. Each observation series was terminated after a female flew off and was lost, or after she had remained at complete rest on a flower for 5 min; in our experience, such females rarely became active again in the course of the night. A total of 175 395 sec (48.7 h) of moth activity was recorded [ILLUSTRATION FOR FIGURE 1 OMITTED]. Pollen collection was rare and constituted 0.7% of the total active time, whereas pollination occupied 3.4% of the active time of the females. Total time allocated to the behaviors that are specific to pollinators was thus 4.1% of their active time.

The second cost, biomass of unique morphological structures of the pollinators, was estimated by determining the proportion of the body mass allocated to the maxillary tentacles. Forty-nine females were collected throughout Cincinnati during the flowering periods in 1993 and 1994. Judging from wing wear and abdominal distention, they were a mixture of young and old adults. The moths were stored at -20 [degrees] C until they could be weighed. The pollen load of each female was removed, the entire moth weighed to nearest 0.1 mg, and the tentacles then cut off with iris scissors and weighed to the nearest 0.1 [[micro]gram]. Repeat measures were performed for ten moths to assure that water loss did not affect the measurements. Moth body masses ranged between 5.4 and 29.5 mg (mean [ or -] SD = 13.9 [ or -] 6.2 mg). Tentacle masses were 31.5-86.3 [[micro]gram] [ILLUSTRATION FOR FIGURE 2A OMITTED], constituting an average of 0.42% of the total body mass of the moths [ILLUSTRATION FOR FIGURE 2B OMITTED]. Only in one moth did the mass exceed 1% of the body mass. The average is an overestimate, because the sample included many females that had laid eggs and thus had lost some of their original body mass.