Applied pollination ecology

TREE vol. 6, no. I, January 1991 Applied Pollination Ecology Sarah A, Corbet AN AC ID TEST of our understanding of an ecological system is our ability to manipulate it. In manipulating the plant-soil-weather system called a crop, we are helped by a long history of agricultural experience. To include another interaction - insect pollination - increases the challenge, the academic interest and, potentially, the commercial reward. The aim is to achieve an interaction of mutual benefit to the three partners - plant, insect and human. The Sixth International Symposium on Pollination, held at Tilburg, Netherlands, in August 1990, brought together an international murmuration of ecologists and evolutionary biologists, apiculturalists and horticulturalists to consider a broad range of aspects of pollination. Its encouraging conclusion is that the effective management of pollination systems is a goal within reach. At least in theory, growers can already choose crops from a world list, to suit the local climate, soil and market. To achieve adequate pollination, they usually trust to luck. This symposium has shown that luck is not always enough; pollinator management can often improve yield, or quality, or both, even in self-fertile crops. Pollinator management may be necessary to rescue alien crops introduced without regard to pollination requirements, or established crops stranded by a decline of native pollinators (Peter Kevan, University of Guelph, Canada). Until recently, managing pollinators has meant introducing honeybee colonies. The growers’ choice of pollinator is still seriously limited by availability and cost. But soon, perhaps, they will be able to choose pollinators, as well as crops, from a world list. Already, several bee species are managed profitably to increase yield (Philip Torchio, Utah State University, Logan, USA). No doubt as experience accumulates the list of species will grow ever faster. In particular, dramatic success with the opportunist bumble bee Bombus terrestris for pollination of glasshouse tomatoes gives promise that further bumblebee species will soon be brought into harness. Systematic choice based on biological evaluation is a hope for the future. Not all plant-pollinator relationships depend on mutual benefit. Some flowers offer no reward, but gain insect visits by deceit; some insects harvest the floral reward with- Sarah Corbet is at the Zoology Dept, Cambridge University, Downing St, Cambridge CB2 3EJ, UK. out transferring pollen. Typically, nectar thieves are insects with a wideranging diet, a feature often associated with social species that must harvest abundant resources from a limited area to sustain a growing colony over a long season. Not surprisingly, these versatile opportunists include the bees that have been easiest to domesticate: the honey bee Apis mellifera, some stingless bees in the tropics and the bumble bee B. terrestris. Hence another example of Murphy’s Law: floral larceny is particularly prevalent among the bees most commonly managed for pollination, and may reduce their effectiveness. In a stable natural community, insects visit flowers from which they can make a profit, and flowers are expected to evolve so that pollen is transferred by the insects that visit them. Butwhen a managed pollinator is introduced into an alien crop, it may profit from the floral reward without performing a good pollination service, or it may gain inadequate reward from the flowers. A misalliance that disadvantages the insect must increase management costs, because it is necessary to force the pollinator to visit the crop, perhaps by flooding the crop with bees or by supplementary feeding, as in kiwifruit (Actinidia deliciosa) orchards, to compensate for the lack of nectar (Mark Goodwin, Ruakura Agricultural Centre, Hamilton, New Zealand). The evolutionary ecologists’ preoccupation with the assessment of benefit is not without economic relevance here. In a well-matched pollination system, both plant and insect should benefit. The principle of mutual benefit is recognized by those who work with alfalfa leafcutter bees (MegaChile rotundata) in North America. For some growers, profits from bee production are comparable with those from the seed yield of the crop. If this bee is to be recommended for a new crop, the crop must not only benefit from its pollination, but must also support good reproduction by the bee. The best alliances yield more bees, as well as more seed (Daphne Fairey and L.P. Lefkovitch, Agriculture Canada, Beaver Lodge, Alberta). Wider application of this principle might reduce the cost to growers of pollinator management. Misalliancesthat give dispcoportionate advantage to the insect are equally to be avoided. The opportunists A. mellifera and B. terrestris are experts at harvesting reward from flowers with which they are not coevolved. On flowers with long corollas, such as the field bean Vicia faba, they often remove nectar without pollinating; only in their pollencollecting visits do they routinely touch the stigma. Growers sometimes compensate for inadequate pollen transfer by flooding a crop with more honey bees. The resulting depletion of floral resources may be counterproductive for the plant if it reduces the crop’s attractiveness for small native populations of more suitable bees (H. Pechhacker and E. Huttinger, Bundesanstalt fur Bienenkunde, Lunz am See, Austria; A. Boonithee and N. Juntawong, Kasetsat University, Bangkok). Of a range of options in pollinator management (Klaas Brantjes, Barendrecht, Netherlands), probably the simplest is habitat management to encourage selected native pollinators. Existing native pollinators are promising candidates for costeffective pollination systems because of the relatively low cost of management by habitat enhancement (Maria Maciel Correia, Universidade de Porto, Portugal), and this option is expected to receive more attention as techniques for exploring the role of different pollinator species become more sophisticated and habitat degradation increasingly threatens native pollinator populations. As well as modified pesticide usage, it may involve provision of supplementary forage or artificial nest sites, as for Osmia in Yugoslavian apple orchards (Miloje Krunic and Miloje Brajkovic, University of Belgrade, Yugoslavia), or for a carpenter bee that pollinates passionfruit (Passiflora) in Malaysia (Makhdzir Mardan, University Pertanian Malaysia, Selangor). A more invasive option, necessary if wild populations are sparse or inappropriate for the crop, is to introduce pollinating species. Here the biological control of pests by natural enemies offers a model. Inoculative release establishes a self-sustaining population of an introduced biological control agent; this continues to operate without further intensive management, although habitat management (and especially restraint in pesticide use) may be necessary. Such a situation is rarely achieved with pollinators, but the weevil Elaeidobius kamerunicus, introduced from Cameroon to Malaysia to pollinate oil palms (Elaeis guineensis), provides a spectacular example (M. Hussein, N. Lajis and J. Ali, University Pertanian Malaysia, Selangor). 3 TREE vol. 6, no. I, January lnundative release, involving repeated mass introduction, is a mode of biological control that is more costly, and much more open to commercial exploitation. It is paralleled by many managed pollination systems. Honeybee colonies do not thrive on kiwifruit, which provides no nectar, and the annual production of colonies for kiwifruit pollination is a major commercial activity for New Zealand beekeepers (Andrew Matheson, MAFF, Tauranga, New Zealand). Management of leafcutter bees comes somewhere between the inoculative and the inundative options. Unlike most biological control agents, these bees return faithfully to artificial nest sites, and can therefore be managed in winter to counter parasites and disease and to improve winter survival, before being released, perhaps in a different site, the following season. This level of management is commercially viable and biologically effective, and will surely be emulated in the developing management of other solitary bee species. A hallmark of a pollinator species whose management is becoming commercially viable is the flowering of research on its parasites and diseases (Klaas Brantjes; Richard Rust, University of Nevada, USA, and Philip Torchio; Gerald and Frances Rank, University of Saskatchewan, Canada). To make an informed choice of pollinator, it is necessary to balance the cost of management against the relative effect on yield of different pollination systems. The cost of management is sometimes easily determined. The value of the resulting increase in crop yield or quality is harder to estimate, but the effort is worthwhile because even supposedly self-fertile crops may benefit from insect visits (Borje Svensson, Bikonsult HB, Sala, Sweden). Individual plant species are usually tested by comparing yield in plots caged with bees (a no-choice situation for the bees), plots caged without bees, and open pollination. A study on leguminous forage crops in Canada is one of the few to involve more than one pollinator species, permitting comparison (Ken Richards, Agriculture Canada, Lethbridge, Alberta). One drawback of cage studies is the lack of choice for the bees. A bee that visits a crop when confined with no alternative may fail to visit it when more choice is available in the field. Thus, in Thailand A. mellifera sneaks off to visit opium poppies, leaving the task of fruit pollination to native bee species (H. Pechhacker, E. Huttinger, A. Boonithee and N. Juntawong). To promote visitation of non-preferred crops by manipulating competition flooding alternative forage with large numbers of bees - is an expensive solution to this problem. Another drawback of cage studies is the possibility that the modified climate inside the cage will affect reward, bee activity or even seed yield (J. Woyke, Agricultural University, Warsaw, Poland). Pollinator value in unconfined conditions is sometimes estimated by protecting flowers, unbagging them * Bridging the gap between ecology and genetics: the case of freshwater zooplankton, MA. Mart * The evolution of developmental strategy in marine invertebrates, G.A. Way and R.A. Raf? * Cytoplasmic incompatibility in insects: why sterilize females2 F. Rousset and M. Raymond * Planktonic microbes: tiny cells at the base of the ocean’s food webs, E.B. Sherr and B.F. Shw * Trophic levels and trophic dynamics: a consensus emerging? L. Oksanen * Implications of ‘supply-side’ ecology for environmental assessment and management, P. G. Fainweather 4 7991 briefly to allow known numbers of visits from selected pollinators under observation, and then rebagging (Barry Donovan, DSIR Lincoln, New Zealand). This method is timeconsuming, but it is one of the few ways to disentangle the potential of individual species in a mixed community of native pollinators. Research on pollinator manageis gathering momentum, ment helped by the increasing availability of managed colonies of a range of bee species. A weak link remains in the transfer of information from research workers to the growers who make the decisions (Borje Svensson; Robert Couston, ex Scottish College of Agriculture, Perth, UK). Until growers have that information, they will be at the mercy of fashion and a commercial market sometimes led by beekeeping or pollinator-producing interests. On a vast hectarage of glasshouse tomatoes in the Netherlands, the costly use of bumble bees is one step ahead of the research that might demonstrate their superiority to honey bees on this crop (Hestern Banda and Robert Paxton, University College Bee Research Unit, Cardiff, UK; Gerard Welles, Glasshouse Crops Research Station, Naaldwijk, Netherlands). Again, alfalfa leafcutter bees in Canada set an example worth emulating. The growers are well informed about the management and biology of leafcutter bees, and themselves participate in, and profit from, bee production as well as seed production. When other bee species achieve this status, and a range of potential pollinators can be evaluated on each crop, so that growers have a real choice, we can expect a corresponding improvement in yield, fruit or seed quality, and profit. Further research will be stimulated by the recognition of economic benefit, and by the ready availability of captive colonies of various species of bee. Their intricate interactions with flowers will be accessible to pollination ecologists, whose findings will help in the development of mutually beneficial pollination systems. Perhaps this symposium, and the avalanche of experience in bumblebee management that it has witnessed, heralds a transition from the honeybees-or-nothing phase to a phase of informed choice of pollinator system (Klaas Brantjes). Misalliances will be a thing of the past. We can look forward to a day when balanced mutual benefit for plant, insect and human is a feature of many aged pollination systems, and yield is limited by the resources available in the plant, not by pollinators.