2
Status of Pollinators

A definitive assessment of the status of pollinator populations in North America will hinge on the quality and availability of data from a variety of well-corroborated sources, and such information is not available for every taxon. Because of their economic importance, actively managed pollinators are more likely than are wild pollinators to be closely and systematically monitored. But even when standardized data are available, interpretation of patterns of population change can be difficult. Ascertaining a pattern of decline in wild pollinator species involves consideration of a broader range of sources of information, including historical accounts, natural history collections, recently published observations, and comparative analyses. For some species, population data that are sufficient to inform an assessment of pollinator status simply do not exist.

POLLINATORS AND THE CONCEPT OF DECLINE

Identifying population declines, particularly for insects, is problematic primarily because, for many species, there are no historical data on absolute abundance. Historical accounts (for example, Jones and Kimball, 1943) often described abundance not quantitatively but qualitatively—a species might be called “common,” “uncommon,” or “rare”—so the information is difficult to interpret or compare. There are, however, numerous reports of declines of pollinating insects that have been documented according the strict criteria of federal or state law or regulations or by nongovernmental organizations. A case in point is the Massachusetts Endangered Species Act (MESA; 321 CMR 8:00), which requires demonstration of habitat



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 34
Status of Pollinators in North America 2 Status of Pollinators A definitive assessment of the status of pollinator populations in North America will hinge on the quality and availability of data from a variety of well-corroborated sources, and such information is not available for every taxon. Because of their economic importance, actively managed pollinators are more likely than are wild pollinators to be closely and systematically monitored. But even when standardized data are available, interpretation of patterns of population change can be difficult. Ascertaining a pattern of decline in wild pollinator species involves consideration of a broader range of sources of information, including historical accounts, natural history collections, recently published observations, and comparative analyses. For some species, population data that are sufficient to inform an assessment of pollinator status simply do not exist. POLLINATORS AND THE CONCEPT OF DECLINE Identifying population declines, particularly for insects, is problematic primarily because, for many species, there are no historical data on absolute abundance. Historical accounts (for example, Jones and Kimball, 1943) often described abundance not quantitatively but qualitatively—a species might be called “common,” “uncommon,” or “rare”—so the information is difficult to interpret or compare. There are, however, numerous reports of declines of pollinating insects that have been documented according the strict criteria of federal or state law or regulations or by nongovernmental organizations. A case in point is the Massachusetts Endangered Species Act (MESA; 321 CMR 8:00), which requires demonstration of habitat

OCR for page 34
Status of Pollinators in North America threat and population decline before an animal or plant can be listed as endangered, threatened, or of special concern—terms that themselves are suggestive of particular patterns of population change. However, different jurisdictions can define terms differently, and that causes difficulty for comparative studies of decline or endangerment. Some species also have inherently small populations and restricted ranges, and their relative rarity might not be the result of declining population. In determining whether pollinator populations are declining, it is important to acknowledge the distinction between a “decline” and a “shortage.” An economically driven shortage of pollinators that occurs as a result of increased demand could be entirely independent of the condition of pollinator populations. In this report, the term “decline” is applied to populations for which the number of individuals is decreasing over time; “shortage” means that the supply of pollinators or their services is insufficient to meet demand. The status of pollinator populations and assemblages can be assessed in many ways, both direct and indirect (see Appendix G for examples of methods for analyzing pollinator status). POPULATION TRENDS Insect Pollinators Although more than 750,000 insect species have been described (Grimaldi and Engel, 2005), possibly as many as 30 million more await discovery and formal description (Erwin, 1982; Stork, 1988, 1996; see also May, 1999, and Erwin, 2004). Insects comprise the most diverse assemblage of terrestrial animals, including within their ranks some of the most economically important pollinators and the dominant pollinators in a variety of natural systems. In some communities, insects pollinate as many as 93 percent of the flowering plants (Bawa, 1974, 1990; Kato, 2000). Unfortunately, the available taxonomic expertise does not exist to document fully the Earth’s insect biodiversity (Box 2-1); it is a virtual certainty that many insect pollinators have yet to be discovered and identified. Notwithstanding the existence of taxonomic impediments, a substantial body of information is available on pollinator population trends. The quality of this information, however, varies with taxon as, accordingly, do conclusions about the status of pollinators in these groups. Ants, Bees, and Wasps (Order Hymenoptera) The order Hymenoptera is a diverse and economically important group of approximately 125,000 described species comprising plant-feeding sawflies, parasitic and nonparasitic wasps, ants, and bees (Zayed and Packer,

OCR for page 34
Status of Pollinators in North America BOX 2-1 Diversity and the “Taxonomic Impediment” Insects account for more than half of the estimated 1,586,800 species that have been formally described by scientists (Grimaldi and Engel, 2005). The most current estimates of species undescribed or unknown to science range from 10 million to 30 million (Grimaldi and Engel, 2005; Stork, 1988, 1996); and many of the most species-rich groups are among the least thoroughly characterized. Because of a lack of available expertise, it is often impossible to identify (or “determine”) specimens. Taxonomy and its applied interface, identification, are fundamental to continuing the study and conservation of organisms. As knowledge of living systems grows more comprehensive, the scientific community demands more from taxonomy than simply identifying which species to avoid and which are edible or otherwise useful. That the rate at which species are becoming extinct appears to exceed the rate at which new species are described (Hambler and Speight, 1996) poses not merely an academic problem but a daunting challenge to understand biodiversity with economic potential before it disappears. The problem applies to the study of plant-pollinator interactions in North America as some pollinating insects, particularly beetles and flies, are yet to be discovered and described. The Global Taxonomic Initiative is attempting to reduce the bottleneck in taxonomic research resources in the face of what has been called the greatest extinction crisis in roughly 60 million years (J.A. Thomas et al., 2004). Under 2005). The order includes within its ranks the principal managed pollinators of the world, bees in the genera Apis, Bombus, Megachile, Osmia, and Melipona, as well as numerous unmanaged species of bees (Box 2-2) and wasps that represent a variety of families. Honey Bees (Apis mellifera) Nearly 17,000 species of bees have been formally described, and as many as 30,000 are estimated worldwide (Michener, 2000; T. Griswold, U.S. Department of Agriculture [USDA] Bee Biology and Systematics Laboratory, presentation to the committee, October 18, 2005). Although other species are often more efficient pollinators than are honey bees on a flower-by-flower basis, honey bees are, for many reasons, the pollinator of choice for most North American crops. A. mellifera is highly suitable as a commercial pollinator because of its biology (Hoopingarner and Waller, 1992;

OCR for page 34
Status of Pollinators in North America the leadership and authority of the Convention on Biological Diversity, the initiative has two aims: (1) to increase the efficiency of biological systematics, and (2) to bolster the number of practicing, professional systematists. Critical to the development of greater understanding is a supply of professional taxonomists, usually university-trained scientists with doctorates in their disciplines. The taxonomic impediment is far from an insoluble problem. The Consortium for the Barcode of Life is an international initiative to develop DNA barcoding as a global standard for identifying specimens. DNA barcoding uses a short gene sequence from a specific region of a genome as an identifying marker for a species (http://barcoding.si.edu). DNA barcoding promises to provide a rapid and inexpensive means of identifying specimens by matching barcode sequences with those of taxonomically validated vouchers. In the United States, steps to ameliorate the shortage of professionals include the highly successful program of the National Science Foundation (NSF) Partnerships for Enhancing Expertise in Taxonomy, which supports taxonomic research and training (Rodman and Cody, 2003). Assembling the Tree of Life—another NSF effort—involves advanced molecular and optical technology, readily disseminated Web-based initiatives, and increasingly advanced analytical software. Whether the federal govenment will continue to support and expand such programs is an open question. Hence, the first challenges to solving the taxonomic impediment in North America and globally are to assess available resources and identify the support and resources needed to reduce or eliminate taxonomic impediments. Winston, 1987). In contrast to most other species of bees that have annual nests founded by individual, overwintered females each spring, honey bee colonies are perennial. Honey bee populations range between 10,000 and 30,000 individual worker bees, even at their nadir in late winter and early spring. Thus, honey bee colonies are able to muster large numbers of pollinators when they are needed for late winter and early spring blooms, as well as throughout the rest of the growing season. As a generalist, the honey bee can pollinate many agricultural crops, including almond and blueberry. Because it forages over long distances (up to 14 km from its nest), it is useful in expansive monocultures where wild bees of other species with more limited foraging ranges are restricted to field margins. In addition, honey bees exhibit sophisticated communication, which increases foraging efficiency, and floral constancy; individuals repeatedly visit a single plant species during each foraging trip and can recruit nestmates to flowers of that species (von Frisch, 1967). Thus, honey bees’ behavior increases the

OCR for page 34
Status of Pollinators in North America BOX 2-2 Sociality and Bee Pollination Of the nearly 17,000 described species of bees (Michener 2000), the vast majority are solitary. Each female makes her own nest and cares for her own offspring. Among the species of pollinators that are actively managed, Megachile rotundata, Nomia melanderi, Osmia cornifrons, and Osmia lignaria all exhibit this solitary lifestyle. The other species of bees that are actively managed for pollination in North America, Apis mellifera and various species of Bombus, are “eusocial.” Eusociality is defined by three traits: (1) cooperative care of young by members of the same colony; (2) reproductive division of labor, with more or less sterile individuals (“workers”) working on behalf of fecund colony members (“queens”); and (3) an overlap of at least two generations of adults in the same colony (Michener, 1969; Wilson, 1971). Eusociality is the most extreme form of social organization in the animal kingdom (Wilson, 1971). It is relatively rare, limited to termites (order Isoptera), several groups of Hymenoptera (all ant species and several lineages of bees and wasps), and a few species of aphids, thrips, beetles, shrimp, and mammals (Crespi and Yanega, 1995; Sherman et al., 1995). Eusociality plays a prominent role in pollinator behavior, especially in the case of the honey bee. Division of labor for reproduction lies at the heart of eusociality. Hymenoptera display the haplodiploid mode of sex determination; fertilized diploid eggs develop into females and unfertilized haploid eggs develop into males. Females can develop into either queens or workers. Queens specialize in reproduction, laying up to several thousand worker eggs per day. Workers engage in little if any personal reproduction, and perform all tasks related to colony maintenance and growth, including foraging. Eusocial species are divided into two groups: primitively eusocial and advanced eusocial. In most primitively eusocial species, colonies have annual life cycles and populations are relatively small, typically a few dozen to a few hundred individuals. There are no morphological differences between queens and workers, but there can be differences in physiology and size. Division of labor for reproduction is achieved by a dominance hierarchy that is established and maintained by direct behavioral mechanisms, including pushing, biting, and physical prevention of egg laying. Aggression is a common occurrence in a primitively eusocial colony. Bumble bee species exhibit a primitively eusocial lifestyle. In advanced eusocial species, colonies are typically perennial and populations number in the thousands to even millions of individuals. Queens and workers are distinguished by striking morphological differences. In advanced eusocial species, queen inhibition of worker reproduction is achieved by chemical communication—queen pheromones—rather than by direct physical aggression. In advanced eusocial species, the fate of an individual—queen

OCR for page 34
Status of Pollinators in North America or worker—is determined before adulthood, and there is far less dominance-related aggression among individuals than in other animal societies. This sets the stage for natural selection, acting on the phenotypes of colonies, to fashion systems of division of labor among groups of highly specialized workers and intricate forms of communication to integrate their activities. Honey bees exhibit an advanced eusocial lifestyle. Several aspects of eusociality contribute to the value of the honey bee as a commercial pollinator: (1) Perennial colonies result in large forces of foraging worker bees, especially early in the growing season, when pollination is required for many crops. Noneusocial species, with annual population cycles, have far smaller populations early in the growing season. (2) Foraging in honey bee colonies is based on division of labor. There is an age-related division of labor among worker honey bees, which is based on a process of behavioral maturation (Robinson, 1992). After working in the hive for 2 to 3 weeks, worker honey bees specialize in foraging for the remainder of their 4- to 7-week adult life. They take about 10 foraging trips per day and log up to 800 km over the course of their foraging career (Winston, 1987). Workers become more efficient at foraging with experience (Dukas and Visscher, 1994), which likely increases their efficacy as pollinators. (3) Foraging in honey bee colonies also is enhanced by communication. Foragers communicate the location of particularly rewarding food sources by means of the famous “dance language,” elucidated by Nobel laureate Karl von Frisch (1967), the only nonprimate symbolic language. Honey bees are thus able to rapidly and effectively direct their foraging force toward a particular field or orchard in bloom. This can enhance pollination by mobilizing a large group of foragers during what is sometimes a relatively short window of opportunity. Pollination often is constrained temporally by floral phenology or adverse weather conditions that limit bee flight (Delaplane and Mayer, 2000). Other traits that enhance the value of A. mellifera as a pollinator are described in the section entitled “Honey Bees (Apis mellifera)” in this chapter. Other species of bees display levels of sociality that are intermediate between solitary and eusocial. “Communal” species nest in aggregations but do not display any of the three defining traits of eusociality. Megachile rotundata nests in aggregation, which facilitates their use as an actively managed pollinator. “Quasisocial” species display cooperative brood care, but no reproductive castes or generational overlap. “Semisocial” species display cooperative brood care and reproductive castes, but no generational overlap. Some species exhibit different levels of sociality during different phases of the colony lifecycle. A bumble bee colony, for example, is established by a single individual, acting in a solitary manner. When the first brood emerges and assumes responsibility for all colony activities except egg laying, the colony then becomes primitively eusocial. Bumble bee colonies are most valuable for pollination during the eusocial phase, when they have an active group of worker foragers.

OCR for page 34
Status of Pollinators in North America efficiency of pollination by ensuring that compatible pollen is transferred among conspecific flowers when needed. Perhaps of greatest significance to the economic importance of A. mellifera is that apiculture—the management of honey bees—is a highly developed discipline that has made bees and beekeeping equipment widely available. Honey bees have been used in North Amercia to provide pollination services for crops in bloom in extensive areas. Typically, one-quarter to one-third of workers in a colony during flight season are foragers. Honey bees can be concentrated in very high densities, which are required for effective pollination in large monocultures with extremely high floral densities, and they can be transported by truck to any location at any time crops are in bloom. Finally, because honey bees can be cared for and maintained by humans, they are buffered to some extent from declines in environmental quality. Honey bee populations have followed different trends in the three North American nations. In the United States, data from the USDA National Agricultural Statistics Service (NASS) reveal declines in the number of honey bee colonies producing honey during 1947–1972 and 1989–1996 (Figure 2-1) (USDA-NASS, 1995, 1999, 2004a, 2005, 2006a). Overall, the number of managed colonies dropped from 5.9 million in 1947 to 2.6 million in 1996–2004. That number fell again in 2005 to 2.4 million. The decline from 1985 to 1996 is likely linked to the occurence of the tracheal mite, FIGURE 2-1 U.S. honey bee colonies, 1945–2005. Data compiled from USDA-NASS (1995, 1999, 2004a, 2005, 2006a).

OCR for page 34
Status of Pollinators in North America Acarapis woodi (first detected in 1984) and to the Varroa destructor mite (first detected in 1987) (Chapter 3). The precipitous drop between 1985 and 1986 also is attributable to a change in NASS survey methods that could represent a statistical artifact. After not collecting data from 1982 to 1985 (official data were based on estimates only), NASS changed its surveys to exclude beekeepers who had fewer than five colonies. Assuming that estimates of honey bee pollination activity in agricultural crops would be improved by more accurate information on total commercial honey bee colony numbers, the U.S. data have four limitations, most of them linked to the NASS focus on honey production. First, the surveys count only honey bee colonies from which commercial honey is harvested; those that exclusively provide pollination services are not counted. Second, the same hives can be counted in several states if commercial honey is harvested in more than one state. Third, annual data are no longer collected on the number of colonies held by beekeepers who own fewer than five hives. Finally, no data are collected on colony health, a factor that has become more important since the parasitic mite invasions of the 1980s (Chapter 3). NASS also surveys beekeeping operations every 5 years for its census of agriculture (USDA-NASS, 2004a). The 2002 census included all honey bee colonies and reported them “in the county where the owner of the colonies’ largest value of agricultural products was raised or produced” (USDA-NASS, 2004a), thus addressing the first three limitations above. However, the agricultural census data are taken less frequently and the variable definitions are incompatible with the annual honey survey data. In contrast to the declines in the United States, Canada had important periods of growth in honey bee colony numbers between 1955 and 1986 and from 1996 to 2005 (Figure 2-2) (Statistics Canada, raw data and 2006). As in the United States, there was a decline after the period of mite invasions in the late 1980s, and the Canada-U.S. border was closed to the importation of live bees in 1987 to prevent the spread of mites from the United States to Canada (Saskatchewan Agriculture and Food, 2004). Statistics Canada collects data on honey bees kept for pollination and on those that produce honey (Statistics Canada, 2006), but there are some inconsistencies in data collection practices across provinces. Honey bee colony data from Mexico, available only for 1990–2003, show a decline in the total from 2.1 million colonies to 1.7 million colonies between 1990 and 1997 (SIAP, 2005). With minor fluctuations, colony numbers remained stable at 1.7 million during 1997–2003. Mexican honey production data are available for a longer period, 1980–2002, but those data do not show any clear trend. Honey production in the principal states of Yucatán, Campeche, Veracruz, Jalisco, Guerrero, and Quintana Roo has fluctuated from 42,000 to 75,000 metric tons, leveling off at 57,000– 59,000 metric tons in 2000–2004 (Ortega-Rivas and Ochoa-Bautista,

OCR for page 34
Status of Pollinators in North America FIGURE 2-2 Canadian honey bee colonies, 1945–2005. Data compiled from StatisticsYear Canada (raw data and 2006). 2004; SAGARPA, 2005). Details on data collection procedures were not available. In contrast to honey bees reared for commercial pollination, feral honey bees are not well studied (Buchmann and Nabhan, 1996; Hoopingarner, 1991). Because honey bees are not native to North America, feral honey bee populations (like those that are actively managed) represent races introduced to the United States from eastern and western Europe and from Africa since the 1620s (Schiff et al., 1994). Schiff et al. (1994) studied the genetic diversity of feral honey bee populations in the southern United States and found that 61 percent of the 692 colonies assessed were maternal descendants of the European races most commonly used for commercial pollination. Few studies have examined the population status of feral honey bees over time. The USDA Carl Hayden Bee Research Center has data on the survival of feral honey bee colonies in southern Arizona that span 19 years (Loper et al., 2006). Loper found that feral honey bee colonies in this area were decimated by tracheal mites in 1990 and varroa mites in 1995 (Loper, 1995, 1996, 1997). In 2002, Seeley (2003) repeated a survey of feral honey bee colonies of Arnot Forest, New York, that he conducted with Visscher in 1978 (Visscher and Seeley, 1982). He found that the number of honey bee colonies was about the same in 1978 and 2002. Kraus and Page (1995) studied the spread of varroa mites within California’s population of feral honey bees. In 1990, a sample of bees from 208 colonies located in feral hives revealed no varroa mites. In 1993, a survey of 124 of the same

OCR for page 34
Status of Pollinators in North America feral hives revealed 75 percent of these colonies no longer existed, and all surviving colonies were infested with varroa. The proximity of these feral colonies relative to commercial apiaries led Kraus and Page (1995) to suggest that the mites moved from commercial to feral colonies. Other than these studies, the committee is not aware of other surveys of feral honey bee colonies in North America. Bumble Bees (Bombus spp.) Approximately 239 bumble bee species are known worldwide (Williams, 1998), 49 of them in the United States. The 41 species in Canada are also all found in the United States. Twenty species are known in Mexico, nine of them also present in the contiguous United States (R.W. Thorp, University of California, Davis, presentation to the committee, January 14, 2006; personal communication, March 2006). Some species of bumble bees (Bombus impatiens and B. occidentalis) have been managed primarily for pollinating greenhouse tomatoes (Dogterom et al., 1998). In contrast with managed honey bee colonies (Apis mellifera) in the United States and Canada—for which agricultural monitoring agencies often have long-term records of honey bee colonies (Figure 2-1)—data on managed bumble bee colonies are not collected in the United States, Canada, or Mexico. Native bumble bees pollinate wild flowers and serve as alternative or complementary pollinators for some crops, such as watermelon and cucumber (Stanghellini et al., 1996a,b). Although many native bumble bee species in the United States were once common, entomologists and naturalists have been noting declines and regional absences of some species within the past decade. The Xerces Society for Invertebrate Conservation has placed four bumble bee species (Appendix H) on its Red List of at-risk pollinator insects of North America (Shepherd et al., 2005). Bombus (Bombus) franklini (Frison, 1921), the Franklin bumble bee, is (or was) an endemic species with the most restricted geographic range of any bumble bee in North America and possibly the world (Williams, 1998). Its range, known at one time to span from southwest Oregon to northwest California, encompasses a distance of 241 km north to south and 112 km east to west. Within that area, B. franklini could be found at elevations from 162 m in the north to above 2,340 m in the southern portion of its historic range. B. frankilini is thought to have become extinct recently in its native range of the U.S. Pacific Northwest (Buchmann and Ascher, 2005; Shepherd et al., 2003, 2005). Thorp (2003, 2005) first began to notice and document a precipitous decline in B. franklini at numerous localities in 1988. Extensive searching by R.W. Thorp and his colleagues over the last 4 years has failed to re-locate B. franklini populations or individuals across that region (Thorp, 2003). B. franklini is now treated as a “species of concern” or a “special

OCR for page 34
Status of Pollinators in North America status species” by the U.S. Fish and Wildlife Service, the California Natural History Data Base, and the Oregon Natural Heritage Information Center. It also appears on the Xerces Society Red List (Appendix H; Shepherd et al., 2005). B. occidentalis, at one time commonly observed in central California, began in the late 1990s to disappear from most of its known geographic range. Thorp (2003, 2005) conducted extensive searches for B. occidentalis and reported that it is now extremely rare in habitats where it was formerly common. The species is still present in some parts of its range, such as the Colorado Rocky Mountains; it is still relatively common near the Rocky Mountain Biological Laboratory as of 2006 (D. Inouye, University of Maryland, personal observation, 2006). B. affinis apparently disappeared from northern New York state about 1998 and from southern New York before 2004 (Day, in preparation; J. Ascher, American Museum of Natural History, personal communication, March 2006). John Ascher of the American Museum of Natural History reports that B. affinis was common on the Cornell University campus between 1996 and 1998, but that he and other entomology students and faculty have not observed it since 2001. Despite collecting more than 1,200 bumble bees in the Black Rock Forest of New York during 2003, Giles and Ascher (2006) failed to find any specimens of B. affinis. Because there is no long-term monitoring or corresponding baseline data for bumble bees or other species of wild non-Apis bees in the United States, Canada, or Mexico, the population status of bumble bees cannot be determined definitively in North America. The United Kingdom, in contrast, has a long and well-established tradition of monitoring by scientists and naturalists. Extensive standardized monitoring protocols are followed across a grid system covering the entire United Kingdom. The Bees, Wasps and Ants Recording Society was established in 1978 expressly to allow “anyone of any age or experience with an interest in aculeates” (ants, bees, and stinging wasps) to contribute to a recording scheme designed to obtain “proper, well coordinated data on the distribution and habitats of many species in order to support conservation programmes, ecological research, and to promote effective conservation strategies on a national basis” (http://www.searchnbn.net/organisation/organisation.jsp?orgKey=222). The ALARM project (Assessing Large Scale Risks for Biodiversity with Tested Methods) was established in 2004 with the objective of assessing changes in the richness, abundance, and distribution of pollinators across Europe (Box 2-3). This project and several other studies show that decline in species richness, frequency, and distribution of bees is evident (Box-2-4; Goulson et al., 2005; Westrich 1989, 1996) if these parameters are carefully monitored or observed. Records of species richness, frequency, and distribution of bees in North America are few in number.

OCR for page 34
Status of Pollinators in North America Thrips (Order Thysanoptera) The thrips (Thysanoptera) are slender, small insects (generally no more than one millimeter long), arranged into nine families of living species distributed worldwide, largely in the tropics and temperate regions, with a few species in Arctic regions (Lewis, 1997; Mound, 1997). Checklists of adult thrips have been produced by Stannard (1957, 1968) for North America. Thrips feed on a variety of plant tissues, including pollen, fungal mycelia, and spores, and they also are predatory (Grimaldi and Engel, 2005; Kirk, 1993, 1997). When they feed on pollen, thrips puncture the coat and drain the grains (Kirk, 1984, 1985, 1997). Grimaldi and Engel (2005) note that pollen feeding evolved several times in thrips; they are so numerous on flowers that they can be effective pollinators of a wide variety of plants in nature and agriculture (Ananthakrishnan, 1993; Endress, 1994; Kirk, 1988; Lewis, 1973, 1997; Terry, 1997). Generally, however, they are regarded as minor or secondary pollinators (Kirk, 1997; Lewis, 1973, 1997; Terry, 2001). As minor pollinators, thrips also pollinate such agricultural plants as beets, beans, onions, and cacao (Kirk, 1997; Lewis, 1973). Although thrips can pollinate plants in the absence of other pollinators, their importance in open-pollinated crops depends on whether other insects pollinate the flowers first (Kirk, 1997). Thrips can enter unopened buds (Mackie and Smith, 1935), but the peak number of thrips can occur after peak visits by other insects (Kirk, 1984). The grooming behavior of thrips contributes to both self- and cross-pollination in plants (Kirk, 1997). As thrips arrange the fringe hairs before and after flight, pollen grains are shed from their bodies (Kirk, 1997). The stigma is prominent in many flowers and because it is used by thrips for take-off and landing, the pollinator thus places pollen directly on the stigma (Kirk, 1997). Populations of thrips on crops grown in greenhouses and shade houses depend on breeding within the crop (Kirk, 1997). For example, young chrysanthemum plants are rooted from older plants, and when adult female Frankliniella occidentalis (western flower thrips) oviposit in apical leaves, growers can inadvertently raise their own pest populations and transport them to other sites in the cuttings (Kirk, 1997). The flower trade is responsible for the worldwide distribution of that thrips species, as well as others (Table 6.2 of Kirk, 1997). In part because of their size and their more frequent role as herbivorous plant pests and disease vectors (Ullman et al. 1997), North American thrips have not generally been the focus of concern about population decline; no thrips species is currently protected under the provisions of the Endangered Species Act (ESA). Because of restrictions on ESA, it is unlikely that any species that has had an adverse economic impact on a crop species would

OCR for page 34
Status of Pollinators in North America be eligible for listing, even if it could be shown that thrips provide essential pollination services (Chapter 6). Mammalian Pollinators Bats Estimates of the number of bat-pollinated plants species in the Americas range from 600 (Neuweiler, 2000) to 1,000 species (Winter and von Helversen, 2001). Most bat-pollinated flowers have intense scents that are different from those pollinated by other animals (von Helversen and Winter, 2003). Sulphur-based compounds are more common in bat-pollinated species than they are in other pollination systems (von Helversen et al., 2000). Given that the scents are produced in many phylogenetically unrelated species of plants, they are likely the result of long coevolutionary associations (Knudsen et al., 1993). The color of bat-pollinated flowers is normally inconspicuous, from whitish to green or brown. The color reflectance of the flower itself probably would not be a strong attractant for bats, because most bats are considered color-blind (Jacobs, 1992), although some species might see some color (von Helversen and Winter, 2003). Ultraviolet clues in several bat-pollinated flowers (notably on columnar cacti) prompted studies of bats’ ability to detect ultraviolet radiation. Von Helversen and Winter (2003) reported that Glossophaga soricina is highly sensitive to ultraviolet light. Other characteristics of bat-pollinated flowers include an outward-facing position at the edge or away from the plant’s foliage, thereby facilitating the bats’ access. Most bat-pollinated flowers are large, with sturdy petals and exposed stamens and pistil. They generally open at night, and many open only for a single night. The protein content in the nectar of bat-pollinated flowers is greater than it is in flowers pollinated by insects, and those plants generally have more nectar (Neuweiler, 2000). Plant families recognized for their many bat-pollinated species include Agavaceae, Bignoniaceae, Bombacaceae, Cactaceae, Caesalpiniaceae, Chrysobalanaceae, Convolvulaceae, Cucurbitaceae, Fabaceae, Malvaceae, Marcgraviaceae, Mimosaceae, Musaceae, Pandanaceae, and Tiliaceae (Neuweiler, 2000), although as many as 27 plant families in the New World have bat-pollinated species (Vogel, 1969). Most of those groups are tropical, and the number of bat-pollinated species increases as latitude decreases (Heithaus, 1982; von Helversen and Winter, 2003). There are at least 12 species of pollinating bats in North America, including southern Mexico (Baker et al., 2003; Ceballos et al., 1997; Medellín et al., 1997). The most prominently recognized—by virtue of their conservation status—are the three long-distance migratory species: the

OCR for page 34
Status of Pollinators in North America lesser long-nosed bat (Leptonycteris curasoae), the Mexican long-nosed bat (L. nivalis), and the hog-nosed bat (Choeronycteris mexicana). Few population data are available for pollinating bats in North America, given that populations are difficult to survey, few people are qualified to survey them, and few people study them. Most population data take the form of isolated reports that indicate local trends in abundance rather than precise estimates of population size. Information on both species of Leptonycteris (Ceballos et al., 1997; Fleming, 2004; Fleming and Nassar, 2002; Fleming et al., 2001a; Galindo et al., 2004; Moreno-Valdez et al., 2004; Stoner et al., 2003) covers regions of western, northwestern, and central Mexico and the researchers have presented data on population dynamics that cover only relatively short periods. An effort to monitor and assess the status of L. curasoae is being centralized by the Arizona Game and Fish Department (http://www.azgfd.gov/w_c/edits/documents/Leptcuye.fi.pdf), and the Program for Conservation of Mexican Bats has been compiling information on this and other species for several years (Medellín, 2003; Medellín et al., 2004). Pollinating bats can be divided into two behaviorally functional groups: species restricted to the tropical regions of southern Mexico (Glossophaga, Hylonycteris, Choeroniscus, Anoura, Lichonycteris, Musonycteris) and those that migrate, moving from central and southern Mexico to northern Mexico and the southern United States (Ceballos and Oliva, 2005; Reid, 1997). All pollinating bats provide important services for many species of North American plants. Many species of columnar cacti, species of Agave, trees in the family Bombacaceae, and many other plants rely heavily on bats to carry out sexual reproduction. Agaves are economically important throughout Mexico, but particularly in western regions where they are used for the production of tequila. Tequila production does not, however, depend on pollination by bats. By virtue of their common asexual mode of reproduction, agaves are planted from shoots associated with adult plants, and their flowering is prevented by premature harvest. Nevertheless, bats could be necessary to promote genetic diversity for the long-term viability of commercially grown agaves (Arizaga et al., 2002; Dalton, 2005; Rocha et al., 2005; Valenzuela-Zapata and Nabhan, 2004). Some other economically important plants linked to bat pollination are the balsa tree (Bombacaceae: Ochroma pyramidale), ceiba (Ceiba pentandra), and many columnar cacti whose fruits are used and commercialized fresh or dried, or processed into jams, jellies, and candies (Neobuxbaumia spp., Pachycereus spp., and others) (Bawa, 1990; Watson and Dallwitz, 1992). Ecologically important plants, such as the cardon (Pachycereus pringlei), saguaro (Carnegiea gigantea), and other columnar cacti, vary in their reliance on bats (Fleming et al., 2001b; Grant and Grant, 1979; Valiente-Banuet et al., 1996). Some populations of these cacti are frequently pollinated by white-winged doves (Zenaida asiatica), several species

OCR for page 34
Status of Pollinators in North America of hummingbirds, sphingid moths, bees, or beetles (Fleming et al., 2001b; Grant and Grant, 1979; Valiente-Banuet et al., 1996) on the day after their nocturnal anther dehiscence. The genus Glossophaga has four North American species, all present in tropical Mexico. Three of them have wider distribution, extending into Central and South America, and one, G. morenoi, is endemic to the dry tropical forest of western Mexico. Two species, G. commissarissi and G. soricina, are widespread and common and do not seem to face any threat; the other two tend to be locally rare (Ceballos and Oliva, 2005; Reid, 1997). No population estimates or trends have been obtained, but neither is considered to be facing conservation threats by the Mexican government and neither appears on the International Union for the Conservation of Nature and Natural Resources/The World Conservation Union Red List of threatened species (http://www.iucnredlist.org/). Leptonycteris curasoae and L. nivalis are migratory species considered threatened in Mexico and endangered in the United States. Their listing in the United States was prompted by surveys in some known roosts that indicated severe declines (U.S. Fish and Wildlife Service, 1988, 1994; Wilson et al., 1985), although subsequent studies suggested that the declines might not have been as severe as originally thought (Cockrum and Petryszyn, 1991). In Mexico, the species was listed as threatened when some winter maternity roosts (the species has a summer and a winter reproductive pulse) were found severely depleted; much of its original habitat in western and central Mexico has been destroyed for tourism and agricultural development (SEMARNAT, 2002). The Program for Conservation of Mexican Bats has monitored between 7 and 20 roosts per year, documenting population stability or growth, and noting temporary declines in some years (Medellín, 2003; Medellín et al., 2004). Another migratory species listed as threatened in Mexico, but not in the United States, is the hog-nosed bat, Choeronycteris mexicana (SEMARNAT, 2002). Although there are no reliable population estimates, since 1906, fewer than 1,500 individuals have been documented throughout the species’ range (Cryan and Bogan, 2003). In contrast with the long-nosed bats, this species roosts in small numbers, typically about 12 bats per roost. Because roosts tend to be scattered widely over landscapes, surveys are difficult (Arroyo-Cabrales et al., 1987; Cryan and Bogan, 2003; Tuttle, 2000). Two of the four North American Glossophaga species are very abundant (Ceballos and Oliva, 2005; Reid, 1997), and there is no evidence of decline. Although the one species endemic to Mexico appears to be less common, there are no population estimates that permit a firm assessment of status. The banana bat (Musonycteris harrisoni), a rare species that appears to be highly specialized, as evidenced by its extremely long snout and tongue,

OCR for page 34
Status of Pollinators in North America is endemic to western Mexico, from Jalisco to Oaxaca. Fewer than 70 individuals have ever been observed, and only 3 roosts with 3 individuals or fewer have been reported (Téllez and Ortega, 1999). The species is considered threatened by the Mexican federal government (SEMARNAT, 2002). The tailless bat (Anoura geoffroyi) is widespread in the southern half of Mexico and occurs frequently at medium elevations—1,000–2,500 m above sea level—and also, rarely, at lower elevations. Although this species is not very common, it is not thought to be facing any threats (Ceballos and Oliva, 2005; Reid, 1997). No population estimates are available. Three well-recognized pollinator species, Hylonycteris underwoodi, Choeroniscus godmani, and Lichonycteris obscura, tend to be rare. The former two have a wide distribution over the southern half of Mexico; the third is known only from the state of Chiapas south to South America. No population estimates or trends are detectable through peer-reviewed literature. The three species are associated with primary tropical forests, both dry and wet (Ceballos and Oliva, 2005; Reid, 1997), and none is on the Mexican list of species at risk of extinction. Three migratory bat species are considered threatened or endangered by the Mexican (SEMARNAT, 2002) and U.S. (U.S. Fish and Wildlife Service, 1988) federal governments. Conservation and recovery programs have been initiated and the bat populations are being monitored; surveys in recent years in several colonies suggest that the populations of at least two (L. nivalis and L. curasoae) of these three species in the genus Leptonycteris are stable. However, taking those species off the threatened or endangered species list may be premature. More local evidence is required before a firm conclusion can be drawn (Medellín, 2003; Medellín et al., 2004). Other pollinating bats might not be in decline, but those associated with primary habitats have long been considered rare, and their biology and importance are virtually unknown. Nonflying Mammalian Pollinators Among nonflying mammals, at least two species of opossum (Caluromys derbianus and Didelphis marsupialis; Tschapka and von Helversen, 1999) visit the flowers of Marcgravia in Central America. Coatis (Nasua nasua; Mora et al., 1999) and kinkajous (Potos flavus; Kays, 1999) have been documented as consistent flower visitors and potential or realized pollinators of various trees, including Ochroma, Pseudobombax, Tetrathylacium, and others. Janson et al. (1981) suggested that several primates (including the spider monkey, Ateles), opossums, and procyonids could be pollinators of several tree species in the rainforests, although at least the spider monkey has been shown to damage virtually all flowers it visits, apparently substantially decreasing fruit set (Riba-Hernandez and Stoner, 2005).

OCR for page 34
Status of Pollinators in North America These mammals’ habitats span southern Mexico to central Mexico and from the northern coatis range to southwestern United States. Opossums can also be found in southern Canada. The woolly opossum (Caluromys derbianus) is considered under special protection, and two of its subspecies are considered endangered. An endemic species of coati is threatened, and the kinkajou is under special protection (SEMARNAT, 2002). Pollination by nonflying mammals is reported more often for other continents (Carthew and Goldingay 1997; Goldingay et al. 1991). Avian Pollinators Pollination by birds is well known and recognized in North America, largely because of hummingbirds (Faegri and van der Pijl, 1966). Population data for avian pollinators are available from a variety of sources, including the North American Breeding Bird Survey (BBS) (see http://www.mbr-pwrc.usgs.gov/bbs/genintro.html for a history of this effort), which is now coordinated through the U.S. Geological Survey (USGS) Patuxent Wildlife Research Center. Bird banding data collected by individuals or at bird banding stations are compiled by the North American Bird Banding Laboratory, (http://www.pwrc.usgs.gov/bbl/) which also is part of the Patuxent Wildlife Research Center. Some summary BBS statistics are available from a USGS website (Sauer et al., 2005); some results are presented here. Cautious interpretation is necessary, however, because at least for hummingbirds, the BBS methodology is less than ideal. Hummingbirds Eighteen hummingbird species are known in the United States, 9 are known from Canada (although some of these are rare visitors from Mexico; Sibley, 2000), and 63 are known from Mexico. Most hummingbirds do not sing (even though they vocalize in aggressive interactions), so they can be more difficult to detect. Males and females can occupy different habitats, and males are polygynous—that is, they mate with more than one female in a breeding season. Although a few species overwinter in the United States, most migrate southward, depending on migration corridors or nectar corridors (Nabhan et al., 2004). It can be difficult to assess hummingbird populations because some surveys (such as the Christmas Bird Counts) are conducted when individuals might have left for wintering grounds. There is evidence that a high percentage of rufous hummingbirds (Selasphorus rufus) lose body weight during migration, requiring longer stopover times if floral resources are scarce (Russell et al., 1994). The relationship between hummingbirds and the flowers they visit

OCR for page 34
Status of Pollinators in North America is well studied (Arizmendi and Ornelas, 1990; Calder, 2004; Grant and Grant, 1967), and hummingbirds are important pollinators in much of North America. Although hummingbirds visit a wide variety of flowers, many hummingbird-pollinated flowers share some general characteristics. The flowers tend to be tubular, brightly colored (red, orange, bright yellow), and relatively odorless, and their nectar is often more diluted than that of bee-pollinated flowers (Baker, 1975; Pyke and Waser, 1981) but could contain higher levels of sucrose (Baker et al., 1998). Hummingbirds display variation in bill shape and length (Stiles and Skutch, 1989), and some studies indicate that all hummingbirds can extract nectar from long-tubed, wide-opening flowers, but that only long-billed hummingbirds do so from long-tubed, narrow-opening flowers (Temeles et al., 2002). The list of species of plants that are visited and pollinated by hummingbirds is extensive (Bertin, 1982; Grant and Grant, 1968), and it encompasses plants in many families: Acanthaceae, Asteraceae, Bromeliaceae, Campanulaceae, Ericaceae, Fabaceae, Gentianaceae, Heliconiaceae, Loranthaceae, Malvaceae, Onagraceae, Polemoniaceae, Rubiaceae, Zingiberaceae, and many others (Knudsen et al., 2004; McDade and Weeks, 2004). Although hummingbirds might be minor as pollinators of agricultural crops (cacti; Griffith, 2004), many species of wildflowers have coevolved with hummingbirds and exhibit morphological, phenological, or other traits that facilitate interaction (Fenster et al., 2004). Data from BBS with a high credibility index (at least 14 samples in the long term, of moderate precision, and of moderate abundance on routes) are available for 8 hummingbird species. Data cited below come from the BBS website (http://www.mbr-pwrc.usgs.gov/bbs/bbs.html). In the states where the credibility index is high (such as North Carolina, Oklahoma, and West Virginia), the trend (percentage change per year) shown for the ruby-throated hummingbird (Archilochus colubris) from 1966 to 2005 is positive. Overall, trends in the United States (2.5 percent per year) and Canada (2.5 percent per year) are positive (http://www.mbr-pwrc.usgs.gov/cgi-bin/atlasa99.pl?04280&1&05). For the black-chinned hummingbird (Archilochus alexandri), for one site (Edward’s Plateau) for which the credibility index is high, the trend from 1966 to 2005 is positive (1.2 percent per year). Overall, the trend for A. alexandri is positive in the United States (1.6 percent per year), and negative in Canada (−3.2 percent per year). For Anna’s hummingbird (Calypte anna), in the states where the credibility index is high, the trend from 1966 to 2005 is positive; the data set includes California, a state with a few regions—California, Southern California grasslands, foothills, and Fish and Wildlife Service Region 1. Overall, the trend is positive (1.2 percent per year) in the United States. For the broad-tailed hummingbird (Selasphorus platycercus), in the states where the credibility index is high, the trend from 1966 to 2005 is mixed;

OCR for page 34
Status of Pollinators in North America it is slightly negative in Colorado and slightly positive in New Mexico. Overall, the trend is slightly negative (–0.2 percent per year) in the United States. Calder et al. (1983) reported that, over a 10-year period at the Rocky Mountain Biological Laboratory, the population of S. platycercus appeared to be declining, although nest counts of breeding females remained fairly constant. In the years since that study, the population has remained variable, but with no discernable long-term trend (D. Inouye, University of Maryland, and Rocky Mountain Biological Laboratory, Colorado, personal observation). For the rufous hummingbird (Selasphorus rufus), in Oregon and Washington, the two states where the credibility index is high, the trend from 1966 to 2005 is negative. Overall, trends in the United States (–2.0 percent per year) and Canada (–2.1 percent per year) are negative. An iteresting development over the past decade is that rufous hummingbirds are commonly found in the eastern United States, where they previously were thought to be absent. The status of several species, because of a lack of information, is more difficult to determine. For the Costa hummingbird (Calypte costae), there are no states where the credibility index is high, but the trend from 1966 to 2005 is positive for one state with a few regions—the Great Basin deserts and Mexican highlands. Overall, the trend is positive (0.5 percent per year) in the United States. Similarly, for the calliope hummingbird (Stellula calliope), there are no states where the credibility index is high, but the trend from 1966 to 2005 is positive in Idaho and Wyoming and negative in California, Montana, Oregon, and Washington. Overall, the trend is negative (–0.9 percent per year) for the United States and slightly positive (0.8 percent per year) for Canada. There is no statistically significant detectable trend. Finally, for the Allen hummingbird (Selasphorus sasin) there are no states with a high credibility index, but in the southern Pacific rainforest region—the only region with a high index—the trend from 1966 to 2005 is negative (–1.2 percent per year) but edging upward over the past several years. Overall, the trend for S. sasin in the United States is negative (–2.0 percent per year). This species is on the Audubon Society Watch List because of its very restricted range in the United States. There are two subspecies, identified primarily through their distribution (mainland or the Channel Islands off the coast of Southern California), and although the subspecies with the wider range (coastal Mexico to Oregon) appears to be in decline, the other appears to be spreading. Unfortunately, there do not appear to be any long-term data about population trends of species of hummingbirds that are distributed in Mexico exclusively. Although these species could be under pressure from habitat alteration and fragmentation, there is no equivalent of the BBS data and no systematic banding efforts that the committee could discover.

OCR for page 34
Status of Pollinators in North America Nonhummingbird Avian Pollinators Although not typically thought of as pollinators, at least one dove species, the white-winged dove (Zenaida asiatica), is an important pollinator of saguaro and possibly other cacti (Fleming, 2000). The population trend for that species is slightly negative from 1966 to 2005 in Arizona, the Sonoran Desert, and the Mexican highlands. Overall, in the United States and in the survey area, the trend is positive (1.6 percent per year). A few other nonhummingbird avian species also pollinate North American plants, and there is evidence that they also affect flower morphology. The genus Erythrina (Fabaceae), a pantropical leguminous small tree, shows two distinct flower and nectar types: hummingbird-pollinated species have upright inflorescences with tubular, radially arranged flowers and nectar that has relatively high concentrations of sugar. The species pollinated by passerine birds (including swallows) have horizontal inflorescences held upright, with the flowers arranged radially along the axis, the narrow standard petal folded to form a pseudotube, and relatively dilute nectar (Bruneau, 1997). Several species have been identified: verdins (pollinating ocotillo, Fouquieria splendens; Waser, 1979), oriole (Icterus spp.: Etcheverry and Aleman, 2005; Toledo and Hernandez, 1979), parrot (Aratinga), woodpecker (Centurus), tityra (Tityra), warbler (Dendroica), wren (Campylorhyncus), jay (Psilorhinus), vireo (Vireo), blackbird (Dives), grackle (Cassidix), oropensola (Psarocolius), honeycreeper (Cyanerpes), tanager (Thraupis, Piranga), euphonia (Euphonia), mockingbird (Mimus), thrasher (Toxostoma), and finch (Carpodacus) visit flowers of a diverse array of species including Bernoullia, Ceiba, Tabebuia, Spathodea, and Agave (Ornelas et al., 2002; Toledo, 1977). Some bird species of these genera are on the Mexican federal list of endangered species: six vireo species are under special protection, two species are threatened, and one is endangered. Three orioles and two oropendolas are under special protection; two parrots of the genus Aratinga are under special protection, two more are threatened, and one subspecies of one threatened species is considered endangered. Two species of warbler are threatened, and one wren species is under special protection, one is threatened, and one is endangered. Two euphonias are under special protection, one thrasher is endangered, and two subspecies of finch are endangered (SEMARNAT, 2002). Information Needs As the number of individual hummingbird banders in the United States has grown, so has the amount of information about the birds’ abundance and migration paths. Additional data might have been collected at some

OCR for page 34
Status of Pollinators in North America long-term bird banding stations or by long-term individual banders that could add to the incomplete data for hummingbirds. Long-term monitoring of bird-pollinated plants also could provide useful information on the stability of the ecosystem services they provide, particularly if plants pollinated solely by the birds are chosen. CONCLUSIONS Despite the paucity of long-term data, collectively there is reliable evidence that some North American pollinator species have become extinct or locally extirpated, or have exhibited decreases in population size (Table 2-6). At least two bumble bee species could face imminent extinction, and several other pollinators have declined significantly (honey bees and U.S. and Mexi- TABLE 2-6 Illustrative Examples of Pollinators in North America for Which Evidence of Decline Is Available Common Name Species Name Location Species for Which Quantitative Data Are Available   Hymenoptera   Honey bee Apis mellifera United States Honey bee A. mellifera Mexico Franklin’s bumble bee Bombus franklini Pacific Northwest of the United States Western bumble bee B. occidentalis Central California Bumble bee B. affinis New York   Lepidoptera   Bay checkerspot butterfly Euphydryas editha bayensis Palo Alto, California and other localities   Chiroptera   Long-nosed bat Leptonycteris curasoae United States and Mexico Long-nosed bat L. nivalis United States and Mexico   Apodiformes   Rufous hummingbird Selasphorus rufus United States and Canada Allen’s hummingbird S. sasin United States Species for Which Quantitative Data Are Not Available   Hymenoptera   Stingless bees Melipona spp. Southern Mexico   Trigona spp.   Pollen wasps Pseudomasaris micheneri Inyo County, California Pollen wasps P. macswaini     Chiroptera   Hog-nosed bat Choeronycteris mexicana Mexico Banana bat Musonycteris harrisoni Mexico

OCR for page 34
Status of Pollinators in North America can pollinating bats), although many have populations that are stable or perhaps even increasing, as are a few of the hummingbird species. It should be noted that there is no evidence of population decline for some species merely because their populations have not been monitored over time. Overall, whether there is a “pollinator crisis” is difficult to ascertain inasmuch as there is no definition of “crisis” that is universally accepted; however, if “decline” is defined as a systematic decrease in population size over time, then there is evidence that some pollinators in North America representing a diversity of taxa are, in fact, in decline. It is accordingly important to ascertain the causes and the consequences of those declines as a step toward informed decision making about action to be taken and what would most likely ensure successful reversal.