Conservation Status of the Endemic Bees of Hawai‘i, Hylaeus (Nesoprosopis) (Hymenoptera: Colletidae)1

Karl N. Magnacca2,3

Abstract: The 60 species of native Hylaeus bees in the Hawaiian Islands are im- portant pollinators in native ecosystems, but they have been almost completely ignored in conservation studies. The only previous assessment of the conserva- tion status of the individual species was not based on recent collections. Here I report on conservation status of all known species, based on collections made from 1999 to 2002. Species are arranged into six categories according to degree of threat, and species considered to be threatened are discussed individually. Five species have not been collected recently from one or more islands from which they are historically known, seven are restricted to endangered habitat, 10 are considered to be very rare and potentially endangered, and 10 have not been collected recently and could be extinct. With such a high proportion of rare species and the importance of Hylaeus species as pollinators, further work on their ecology is needed.

The Hawaiian Islands are famous for more undescribed). The remainder includes their numerous evolutionary radiations. A a second small group of eumenines (the Ody- large number of species derived from a rela- nerus nigripennis group, probably belonging tively small number of introductions is the in Euodynerus, three species), another group rule for both plants and insects, due to ex- of bethylids (Sclerodermus, 17 species), and treme isolation from continents and other two Crabronidae: one group of crabronines high islands (Zimmerman 1948, 1970, Wag- (Ectemnius, 20 species) and two sister genera ner et al. 1990). The result is a fauna in which of pemphredonines, Deinomimesa and Neso- many higher taxa are absent, including some mimesa (five and six species, respectively, de- that dominate continental areas. The aculeate rived from a single introduction [Nishida (stinging) Hymenoptera are a striking exam- 2002, Daly and Magnacca 2003]). Entirely ple. The native fauna contains 404 described lacking are ants, social vespids, and the enor- endemic species, nearly 90% of which are de- mous array of other solitary hunting wasps rived from only three introductions: Hylaeus and bees. The groups that are native to Ha- (Nesoprosopis) bees (Colletidae, 60 species), wai‘i make up a small fraction of the fauna in the ‘‘Nesodynerus’’ group of eumenine wasps continental areas, in terms of both numbers (Vespidae, 112 species), and Sierola bethylid of species and numbers of individuals. For wasps (Bethylidae, 180 species with many example, of the more than 3,000 species of bees in North America, there are only 48 Hy- laeus species (fewer than occur in Hawai‘i), distributed among seven subgenera. More- 1 Manuscript accepted 5 May 2006. over, the radiation in Hawai‘i is extraordinary 2 Department of Entomology, Cornell University, for the genus; with the inclusion of the eight Ithaca, New York 14850. 3 Current address: Department of Environmen- known Japanese species, Nesoprosopis is the tal Science, Policy, and Management, University of second largest subgenus of Hylaeus (Mich- California–Berkeley, Berkeley, California 94720 (e-mail: ener 2000) (the Australian subgenus Prosopis- [email protected]). teron is also larger when undescribed species are included, but it is highly polyphyletic [un- Pacific Science (2007), vol. 61, no. 2:173–190 publ. data]). : 2007 by University of Hawai‘i Press The 60 Hawaiian Hylaeus species arose All rights reserved from a single introduction, probably from Ja-

173 174 PACIFIC SCIENCE . April 2007

Figure 1. The main islands of the Hawaiian chain, with 300-m contours. Under the name of each island is its approx- imate age in millions of years (Moore and Clague 1992, Carson and Clague 1995). Numbers for species occurring on ‘‘all islands’’ of Maui Nui do not include Kaho‘olawe, which has a much more depauperate fauna. pan or East Asia where other species of the Very little has been done on conservation subgenus live (Hirashima 1977). Being strong assessment for these bees. Information is so fliers, they have a lower rate of endemism scattered that a paper recently appeared in a than most Hawaiian groups; only 40 of the major journal stating that almost the entire species are single-island endemics, and many group was extinct (Cox and Elmqvist 2000). species found on Maui Nui are shared among Thirty-four currently recognized species were two or more islands (Figure 1). Nevertheless, listed en masse as Category 2 candidate en- they form one of the larger Hawaiian insect dangered species in 1984 (see Table 1) (U.S. radiations (Liebherr 2001). They were con- Fish and Wildlife Service 1984), meaning that sidered by Perkins (1913:lxxix) to be ‘‘almost they are likely to be endangered but that not the most ubiquitous of any Hawaiian in- enough information was available to justify sects.’’ Unlike nearly all other native groups, listing as endangered. This candidate listing they can be found across virtually the entire was based on a very preliminary assessment range of rainfall and elevation in the Islands of conservation needs (Gagne´ 1982), and and in all vegetation types (Figures 1 and 2). further determinations of listing status were Yet despite the widely recognized importance made with little solid information. Eight of bees as pollinators (Proctor et al. 1996), names now considered synonyms were also their role in maintaining the health of Hawai- included in the Category 2 list, and 17 species ian forests and other ecosystems has not been were listed as Category 3A (probably extinct). investigated. Only recently has the basic tax- Five of the latter group are now considered onomy been revised for the first time in over synonyms of species known to be extant and 100 yr (Daly and Magnacca 2003). eight have been recollected recently, but four Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 175

Figure 2. Approximate extent of native habitat suitable for bees (adapted from Juvik and Juvik [1998]; see Results). For coastal habitat, see Figure 3. Closed circles denote recent collection sites; open circles are historic collection sites now dominated by nonnative vegetation.

(H. finitimus, H. mauiensis, H. melanothrix, of threats to Hylaeus species to be based on and H. perspicuus) have still not been recov- recent field data. ered (Daly and Magnacca 2003). Since that original listing, virtually no follow-up work materials and methods has been done to ascertain the true status of the Hawaiian bees. The impetus to do so has Bees were collected from all main islands dur- declined even further since the Category 2 ing the summers of 1999–2002. Collecting candidate list was eliminated and its members was done primarily with a net at flowers or reclassified as ‘‘species of concern,’’ with no over ground; at some localities yellow pan official federal status (U.S. Fish and Wildlife traps were used. An attempt was made to col- Service 1996). lect from all ecosystems where Hylaeus species In this paper, I present a summary of are likely to occur. Due to a number of re- the distribution and abundance of Hawaiian strictions, collections could not be made at Hylaeus species, based on recent and historic some sites. These restrictions included physi- collections. Rare species are considered indi- cal access, land ownership, weather, and time. vidually, followed by discussion of factors in- Bees are also capable of maintaining popu- fluencing their distribution and abundance. It lations in small areas, and sites that are is intended as a first step toward assessment largely dominated by exotic vegetation may of the state of Hawaiian bee populations and still harbor Hylaeus species in pockets of na- cannot be regarded as comprehensive while tive habitat. No collections have been made several large areas of the Islands remain un- on Ni‘ihau (except the islet of Lehua), off- surveyed. Nevertheless, it is the first report shore islets, or the Northwestern Hawaiian 176 PACIFIC SCIENCE . April 2007

Islands for several decades; among the lat- specimens date from 1964 (Beardsley 1966). ter, Hylaeus species are recorded only from The island’s vegetation appears to be largely Nı¯hoa. Additional records from other col- intact, including as major components several lectors were also examined; details of collec- plant species that are rare on the main islands tion records used in this study can be found (Conant 1985), and H. perkinsianus is proba- in Daly and Magnacca (2003). Unless other- bly safe. No bees were collected during a re- wise cited, factual references here are from cent trip to Nı¯hoa to assess the impact of an that reference or from personal experience. outbreak of Schistocerca locusts (S. Montgom- For the purpose of conservation status ery, pers. comm.), but bees in other localities analysis, the species were sorted into six cate- have been observed to go through boom-and- gories. Placing species into discrete categories bust cycles. Although the locusts had a severe as done here is somewhat subjective but is impact on the island’s vegetation, a later trip necessary as a first step in assigning priorities. found a low population of them and the veg- Because there is a gap of approximately 70 yr etation recovering. in regular collections of Hylaeus species in Figure 2 shows recent collection sites and Hawai‘i and much is still unknown about approximate extent of habitat; empty areas their distribution and abundance, it is difficult of native habitat indicate regions that have to use more objective categories such as those been searched inadequately or not at all. Na- used by the IUCN (2001). The categories tive habitat is generally patchier in reality are (in order of increasing threat): (1) un- than depicted on the map. This is particularly common to abundant, widespread, habitat the case for regions on the island of Hawai‘i relatively safe; (2) generally found in low such as South Kona where ranch land alter- numbers but widespread, with a large poten- nates with state forest reserves in mauka- tial habitat; (3) abundant in some areas, but makai strips, and upper areas of Mauna Loa some populations threatened; (4) abundant that are largely bare rock but contain numer- where found but habitat restricted and/or ous kı¯pukas, ‘‘islands’’ of older substrate threatened; (5) very rare, potentially endan- with more-developed vegetation than the sur- gered; (6) not recently collected, probably en- rounding, newer lava flows. On older islands, dangered or extinct. Species in category 1 are high ridges with native vegetation alternate considered to be safe, barring a dramatic with valleys that are often dominated by ex- increase in habitat destruction, and are not otic plants. discussed. Those in categories 2 and 3 are probably safe, at least as a species, and are Rare but Widespread covered briefly. Category 4 species are dis- cussed in the context of their habitats, which The species in this category have been also include some category 5 species. The re- collected only in low numbers and could po- mainder of the category 5 and 6 species are tentially be threatened but due to the rela- considered individually. tively large available habitat probably occur in much larger populations than collections results indicate. Hylaeus hirsutulus from Kaua‘i; H. angustu- The distributions and conservation status of lus from Maui Nui; H. kukui from Maui Hylaeus species are listed in Tables 1 and 2. and Hawai‘i; H. crabronoides, H. filicum, H. In addition to those shown, H. flavifrons is muranus (¼ insignis), and H. rugulosus from known from Ni‘ihau (including the islet of Hawai‘i; and H. specularis from Hawai‘i, Mo- Lehua), and H. anthracinus and H. assimulans loka‘i, O‘ahu, and Kaua‘i (and probably are known from Kaho‘olawe (all recent col- Maui) have all been collected only sporadi- lections). The only species recorded from cally. All are found in wet forest (several also the Northwestern Hawaiian Islands, H. per- occur in mesic areas), and all except H. mura- kinsianus, is endemic to Nı¯hoa. Little collect- nus are known from widely separated locales. ing has been done there, and the most recent These species are found in habitat that is TABLE 1 Island Distribution, Habitat Preference, Currently Assessed Conservation Status (See Text and Table 2 for Details), and U.S. Fish and Wildlife Service (1984) Candidate Listings for All Hawaiian Hylaeus (Nesoprosopis) Species

Habitata Islandb FWS Species (1984) Status c d m w s K O Mo L Ma H akoko Magnacca & Daly, 2003 nc 5 andrenoides (Perkins, 1899) 2 1 angustulus (Perkins, 1899) 3A 2 anomalus (Perkins, 1899) 2 6 anthracinus (F. Smith, 1853) 2 4 assimulans (Perkins, 1899) 2 5 chlorostictus (Perkins, 1899) 2 1 coniceps (Blackburn, 1886) 2 1 connectens (Perkins, 1899) 3Ad 3 crabronoides (Perkins, 1899) 2 2 difficilis (Perkins, 1899) 2 1 dimidiatus (Perkins, 1899) 2 4 dumetorum (Perkins, 1899) 1 facilis (F. Smith, 1879) 2 5 filicum (Perkins, 1911) 2 2 finitimus (Perkins, 1899) 3A 6 flavifrons (Kirby, 1880) 2 4 flavipes (F. Smith, 1853) 2 3 fuscipennis (F. Smith, 1879) 2 3 gliddenae Magnacca & Daly, 2003 n 6 haleakalae (Perkins, 1899) 2 1 hilaris (F. Smith, 1879) 3A 5 hirsutulus (Perkins, 1899) 2 2 hostilis (Perkins, 1899) 2 1 hula (Perkins, 1911) 2 5 inquilina (Perkins, 1899) 1 kauaiensis (Perkins, 1899) 2 1 kokeensis Magnacca & Daly, 2003 n 2 kona (Blackburn, 1886) 2 4 kuakea Magnacca & Daly, 2003 n 5 kukui Magnacca & Daly, 2003 n 2 laetus (Perkins, 1899) 2 3 longiceps (Perkins, 1899) 2 4 mana Magnacca & Daly, 2003 n 5 mauiensis (Perkins, 1899) 3A 6 melanothrix (Perkins, 1899) 3A 6 mimicus Magnacca & Daly, 2003 n 1 muranus (Warncke, 1970) 2 2 mutatus (Perkins, 1899) 3A 1 nalo Magnacca & Daly, 2003 n 6 niloticus (Warncke, 1970) 2 6 nivicola Meade-Waldo, 1923 3A 1 ombrias (Perkins, 1910) 2 4 paradoxicus (Perkins, 1899) 5 pele (Perkins, 1911) 3A 1 perkinsianus (Timberlake, 1926) 2 1 perspicuus (Perkins, 1899) 3A 6 psammobius (Perkins, 1911) 3A 4 pubescens (Perkins, 1899) 2 1 rugulosus (Perkins, 1899) 3A 2 satelles (Blackburn, 1886) 2 6 setosifrons (Perkins, 1899) 1 simplex (Perkins, 1899) 2 6 178 PACIFIC SCIENCE . April 2007

TABLE 1 (continued)

Habitata Islandb FWS Species (1984) Status c d m w s K O Mo L Ma H solaris Magnacca & Daly, 2003 n 4 specularis (Perkins, 1899) 2 1 sphecodoides (Perkins, 1899) 2 1 takumiae Magnacca & Daly, 2003 n 1 unicus (Perkins, 1899) 2 1 volatilis (F. Smith, 1879) 2 3 volcanicus (Perkins, 1899) 1 Collected 12 9 10 8 18 25 Missing 2 6 5 7 5 4 Proportion collected 0.86 0.60 0.67 0.53 0.78 0.86

a c, coastal; d, dry forest and shrubland; m, mesic forest; w, wet forest; s, subalpine shrubland. Open circles indicate only historical collections. b K, Kaua‘i; O, O‘ahu; Mo, Moloka‘i; L, La¯na‘i; Ma, Maui; H, Hawai‘i. c n, new species described in Daly and Magnacca (2003). d Although H. connectens was listed as 3A, two junior synonyms (H. koae and H. vicina) were listed as category 2. difficult to access as well as difficult to collect species typically forage only during sunny bees in; wet forest makes up the bulk of the weather, the chances of successfully collect- areas that have not been surveyed (aside ing them in these almost perpetually cloudy from vast areas of montane shrubland on the regions are low, and it is a testament to their island of Hawai‘i [Figure 2]). Because Hylaeus adaptability that they are able to persist in

TABLE 2 Hylaeus Species Listed by Conservation Category

123456

Uncommon Rare but Missing from Restricted to Very rare, No recent to abundant, widespread, one or more endangered probably collections, habitat safe habitat safe islands habitat endangered endangered andrenoides angustulus connectens anthracinus akoko anomalus chlorostictus crabronoides flavipes dimidiatus assimulans finitimus coniceps filicum fuscipennis flavifrons facilis gliddenae difficilis hirsutulus laetus longiceps hilaris mauiensis dumetorum kokeensis volatilis ombrias hula melanothrix haleakalae kukui psammobius kona nalo hostilis muranus solaris kuakea niloticus inquilina rugulosus mana perspicuus kauaiensis specularis paradoxicus satelles mimicus simplex mutatus nivicola pele perkinsianus pubescens setosifrons sphecodoides takumiae unicus volcanicus Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 179 such places. Thus, although they have been larly on Hawai‘i. At least in part this is be- collected only rarely, the fact that their habi- cause it can also be found in some upland tat is largely unsurveyed and is relatively pris- areas; the largest known population is at the tine over large expanses (Cuddihy and Stone coast at South Point, but it is relatively abun- 1990) suggests that they may be much more dant at Kı¯puka Ne¯ne¯ in Hawai‘i Volcanoes abundant than their collection records imply. National Park, at approximately 900 m, and Although H. angustulus has not been found on the western slope of Mauna Kea at recently on La¯na‘i, that is probably due to 1,700–2,800 m. The Maui and La¯na‘i popula- poor weather during collecting trips and it tions were originally described as H. black- likely persists on the island. burni and can usually be distinguished from Hylaeus kokeensis, a newly described species the Hawai‘i populations by their smaller size from Kaua‘i, appears to be restricted to mesic and more extensive coloration. The initial and dry forest. Both of these area types are Maui site was the Wailuku sandhills (see dis- generally under greater threat from develop- cussion of coastal areas in the next section); it ment, agriculture, and invasive plants than has not been collected on Maui since 1912. It are wet areas. However, the species occurs in was recently rediscovered on La¯na‘i at an ele- a protected area (Ko¯ke‘e State Park/Na¯pali- vation of about 450 m, but the size and stabil- Kona Forest Reserve), and the habitat is in- ity of the population there is not known. The tact over relatively large areas. Although it La¯na‘i site is in a heavily eroded, largely bar- has been collected only twice, it was found in ren area on the windward side of the island, high numbers. Invasive exotic plants are in a small patch of Myoporum trees. Although probably the greatest threat to this species. other such patches presumably occur, the bees were not found on other Myoporum nearby. The coast of Lana‘i has not been adequately Some Populations Threatened ¯ explored for bees, but H. flavipes was not Hylaeus connectens, though usually not greatly found in a small area of native vegetation at abundant, is the most widespread species in the coast where H. longiceps was present. the Islands. It occurs on all the main islands Hylaeus fuscipennis, like its close relative H. and in all habitats except subalpine shrubland, pubescens of Hawai‘i, is one of the most com- though it seems to prefer mesic forest. Nev- mon wet-forest bees of Maui Nui. It can be ertheless, it has not been collected recently found there in mesic forest and even in rela- from Moloka‘i. The habitat it is typically tively marginal habitat at the edge of native found in on Maui—open ‘o¯hi‘a-uluhe wet forest but has not been collected recently forest, with abundant Scaevola gaudichaudiana on O‘ahu. Why this should be, when several shrubs—is relatively common on Moloka‘i. other wet-forest species (H. connectens, H. On other islands (Hawai‘i and Kaua‘i, in par- mimicus, and H. unicus) have been collected ticular) it tends to be collected as scattered in- in good habitat there, is not clear. dividuals, so it is not entirely surprising that it Hylaeus laetus is a dryland species and is has not been found on Moloka‘i. Only one also very abundant on some islands but rare collection is known from the island, so it may on others. The only species besides H. con- never have been abundant there. It is unusual, nectens to occur on all the main islands, it is however, that it is quite abundant on O‘ahu, common on Hawai‘i, La¯na‘i, and Kaua‘i, where many other bee populations are de- even in relatively poor habitat (i.e., low plant pleted. Unknown species-specific factors are diversity, high cover of invasive exotics, few also likely at work: H. fuscipennis is abundant other Hylaeus species). On Maui it has been on Moloka‘i but is now absent from O‘ahu collected recently only from a few lowland (see later in this section), although H. unicus areas of West Maui, and on O‘ahu from a is common on both islands, and all three oc- single locality in the northern Wai‘anae cur sympatrically on Maui. Range; no recent collections have come from Hylaeus flavipes is the only coastal species Moloka‘i. Unlike its close relatives H. difficilis with large, widespread populations, particu- and H. hirsutulus, it is almost never found in 180 PACIFIC SCIENCE . April 2007

Figure 3. Rare habitats (primarily coastal), described in Results. wet or even mesic forest, and this may be a sons: it is highly valued for development, primary reason for its rarity on Maui, Molo- popular for recreation, typically dry and ka‘i, and O‘ahu, where intact native dry forest therefore vulnerable to fire, susceptible to in- and shrubland are scarce. However, all three vasion by exotic plants, and it covers a small islands have unsurveyed areas that may sup- area by definition. On most of the Islands, port larger populations of H. laetus. only one coastal site with diverse native vege- Hylaeus volatilis is one of the five clepto- tation is protected, making the bees that in- parasitic species of Hawaiian Hylaeus, the habit them vulnerable to single catastrophes. only known cleptoparasitic colletids (Mich- As a result, they make up the bulk of the areas ener 2000). It is fairly common in the sub- of concern for species in this category. alpine shrubland in and around Haleakala¯ At Polihale, a state park on the western National Park, where its presumed hosts, H. coast of Kaua‘i, the vegetation is almost all difficilis and H. nivicola, form extremely dense native along the first few banks of dunes be- populations. Although it has historically been fore changing to mostly exotic trees about recorded from various parts of all the islands 100 m back from the water. The bee fauna of Maui Nui and from O‘ahu (Perkins 1899), consists of H. chlorostictus, H. connectens, H. the only other recent collection is from West hostilis, H. flavifrons, and H. solaris. The first Maui, above Lahaina at about 600 m. As three are also found in montane areas, but noted earlier, other dry areas are highly de- the last two are apparently restricted to the graded, although some of those that remain coast. The site is open to the public for camp- have not been surveyed. ing and other recreational activities, making it vulnerable to physical destruction of plants by people and vehicles. However, the Pacific Restricted to Rare Habitats Missile Range Facility at Barking Sands is Coastal strand habitat (Figure 3) is the most directly to the south and probably also in- endangered in Hawai‘i for a number of rea- cludes intact habitat. There are also areas Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 181 on the southeastern coast where native vege- In addition, this area may be developed in the tation persists, but these are threatened by near future. Hylaeus difficilis, but apparently development. not H. anthracinus, can be found in the pro- Ka‘ena Point, on the western tip of O‘ahu, tected Kaloko-Honoko¯hau National Historic is a state Natural Area Reserve. The bee Park just to the south. The difference in dis- fauna includes H. anthracinus and H. longiceps; tribution between the two species may be re- this is the only known locality for these spe- lated to greater ant tolerance by H. difficilis. cies on O‘ahu. However, H. assimulans and Mo‘omomi, on the northwestern coast of H. volatilis, both previously recorded from Moloka‘i, is a Nature Conservancy of Hawai‘i the leeward coast of O‘ahu, are notable in preserve. Like Polihale, it consists of a native their absence from such prime habitat. Illegal beach flora backed by mostly exotic trees. off-road driving has long been a problem The bee fauna includes H. anthracinus and H. there, but the site is now blocked off, and longiceps, as well as the extremely rare clepto- the diverse vegetation is recovering and being parasite H. hilaris. The latter has been col- restored (B. Gagne´, pers. comm.). lected only twice in recent years (with only South Point (Ka Lae) is the best coastal a single specimen each time), and its appar- site for bees on Hawai‘i. Most of the species ent extirpation from Maui and La¯na‘i leaves found there—H. anthracinus, H. difficilis, H. this as the only known site for it. Based flavipes, H. ombrias, and H. sphecodoides—can on the few visits I have made, it is difficult also be found at other sites, either farther to estimate numbers of the other two species, east along the coast or at higher elevations, its presumed hosts. However, I have never but South Point is the only place where all caught more than 10 bees in an hour, a much the coastal species on the island (with the ex- lower rate than at most other sites. The pop- ception of H. psammobius) are found together. ulation of H. hilaris must therefore be ex- Much of the area is under the jurisdiction of tremely small, and among species that have the Department of Hawaiian Home Lands, been collected recently it is probably the but access is largely unregulated. Although most endangered. It is the most colorful of driving may injure vegetation, the native the Hawaiian bees: the abdomen is largely plants extend over a wide enough area that red, with unusual white hair bands, and the the bee populations there are probably not front of the male’s head is almost entirely yel- threatened. However, H. anthracinus appears low. As one of only five species of parasitic to be restricted to regions of more-recent Colletidae in the world (the others are also lava flows east of the point, on flowers of Hawaiian but are not threatened), it would Scaevola and the introduced Tournefortia. It is be a particularly tragic loss if it were to be- not clear why this should be when the other come extinct. Even with a well-protected site, species are abundant on the carpet of Sida the apparent lack of a second population (it is (‘ilima) and Jacquemontia (pa¯‘u¯ o Hi‘iaka) unknown how far H. hilaris and its hosts ex- growing on the older substrates and H. an- tend along the coast) makes the known popu- thracinus is found at these plants on other is- lation highly vulnerable to fluctuations in lands. However, the high degree of genetic host numbers, which can drop very low dur- divergence between island populations of H. ing periods of drought or other poor flower- anthracinus (Magnacca and Danforth 2006) ing periods. may point to behavioral differences between The Kalaupapa peninsula on the northern them. coast of Moloka‘i has an extensive area of na- A second population of H. anthracinus was tive coastal vegetation on its east side. It is recently discovered at Pu¯hili Point on the similar to the site at South Point, with pros- Kona coast of Hawai‘i, also on Tournefortia. trate, windswept plants including Chamaesyce Although the extent has not been fully inves- degeneri (‘akoko), Heliotropium spp. (hina- tigated, it appears to live in a narrow area be- hina), Jacquemontia, and Sida, as well as Tour- tween a population of Anoplolepis gracilipes nefortia in places. Hylaeus anthracinus and H. ants to the south and barren lava to the north. facilis have been collected there on the latter; 182 PACIFIC SCIENCE . April 2007

H. longiceps and possibly H. hilaris are proba- Hylaeus species, very different from that bly also present on the native species during of other coastal species. Phylogenetic data flowering season but were absent when I vis- (Magnacca and Danforth 2006) show that it ited during a dry period. is sister to H. anthracinus þ H. flavifrons. There are few coastal sites of the quality Very few historic collections are known, and of those just described on the island of Maui. it has been found recently at only one site, at Areas of native vegetation exist on the north- ‘Eleilei Point on the northern coast of East ern coast of West Maui (F. Warshauer, pers. Maui (similar sites exist elsewhere along the comm.) but have not been investigated. Per- coast but have not been surveyed). Although kins’ primary site for coastal bees on Maui it is reasonably abundant there, the area is so was the Wailuku sandhills, which once sup- small and close to the shoreline—the plants ported a diverse bee fauna: H. anthracinus, H. grow on rocks barely above the spray zone, assimulans, H. facilis, H. flavipes, H. hilaris, H. in an area @10 m wide between the ocean laetus, H. longiceps, and H. volatilis (Perkins and cliffs—that a single large storm could 1899). These dunes are now built upon or seriously damage or destroy the site as habitat covered with the exotic tree Prosopis pallida for the bees. The species has not been col- (kiawe); native plants are scarce, and bees are lected on Hawai‘i since 1908, but the wind- absent. Hylaeus flavipes and H. hilaris have not ward coast there should be checked for intact been collected on Maui since 1912; their ab- habitat of this type. sence from areas where other coastal or low- Coastal habitat is even scarcer on La¯na‘i. land species have been collected suggests that Two of the typical coastal species, H. flavipes they may have been extirpated. and H. longiceps, are primarily found at middle Two of the species listed above, H. facilis elevations (around 300–450 m); H. anthraci- and H. volatilis, are (or were) commonly nus has not been collected there recently. found at higher elevations. The other more A number of other potential coastal strictly coastal species, H. anthracinus, H. as- sites that have not been investigated should simulans, and H. longiceps, have been found be checked for Hylaeus populations. These in- recently at a few scattered sites around the is- clude Keoneloa on Kaua‘i, Kahuku on O‘ahu, land. The first has been collected at Kanaio at and the northeastern coasts of Maui and 600 m on the lower southern slopes of Halea- Hawai‘i. Bees have also been collected on kala¯, an unusual location for this otherwise Kaho‘olawe (H. assimulans in 1997 and H. exclusively coastal species (it was also found anthracinus in 2002), but the island has not at the coast nearby, at Manawainui). Hylaeus been searched widely. assimulans was found in dry forest at 600 m The most notable threatened habitat on the western side of the island, together in montane regions is Chamaesyce olowaluana with H. laetus and H. volatilis. A small popula- (‘akoko) trees in mesic or dry forest in west- tion of H. longiceps occurs at Waiehu, on a ern Hawai‘i. At both Po¯hakuloa Training very small (<1 ha) patch of relictual native Area and Pu‘u Wa‘awa‘a Wildlife Sanctuary, dune vegetation near a golf course. A few ad- isolated trees of this species appear to support ditional specimens have been collected from large populations of a diverse group of bees elsewhere on West Maui, but that area is rarely found elsewhere. These include H. largely dominated by exotic plants and has akoko, H. dimidiatus, H. filicum, H. hula, H. not been adequately searched for patches of kona, and H. paradoxicus, along with more native vegetation. common species such as H. coniceps, H. diffici- Hylaeus psammobius, known historically lis, H. laetus, and H. pele. The species of the from Maui and Hawai‘i, is the only native first group were not found at nearby, more bee to inhabit windward coastal regions. It common flowering plants, such as Bidens, My- lives in Sesuvium (‘a¯kulikuli) herbland (Gagne´ oporum, and Styphelia, where those of the sec- and Cuddihy 1990), where it collects pollen ond group were collected. Moreover, of the of Sesuvium, Bacopa, and Lycium (unpubl. first group, only H. dimidiatus and H. paradox- data). This is a unique habitat for a Hawaiian icus were found at both sites, and none was Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 183 found in the Mauna Loa–Mauna Kea saddle strange that it does not seem to occur at or Mauna Kea north slope regions, where C. Ka‘ena Point, the best coastal site on these is- olowaluana is more abundant. This suggests lands. Recent collections have been sporadic that conservation of C. olowaluana in the and few, from West Maui, La¯na‘i, and Ka- Pu‘u Wa‘awa‘a and Po¯hakuloa areas is im- ho‘olawe. Like its sister species, H. ombrias portant for the survival of the bees, even if of Hawai‘i, it appears to be more closely asso- other populations of the tree remain. Despite ciated with Sida than other species are, visit- abundant seed set and germination, repro- ing it for both nectar and pollen. Thus, it duction of C. olowaluana is impeded by sheep, may be more common in areas with large which devour the seedlings before they have a amounts of Sida, for example, the Prosopis chance to grow. (kiawe) forests of southern La¯na‘i. Unfortu- In addition to the presence of C. olowa- nately, I visited this area during a prolonged luana, Po¯hakuloa and the adjoining western drought during which nearly all of the Sida slope of Mauna Kea are unique upland sites had died. I collected only a few H. assimulans, for H. flavipes and H. ombrias. Most of the but it may be more common under favorable other sites for the former are below 400 m, conditions. and all (except the Halepo¯haku record) are The closely related H. facilis and H. simplex below 1,000 m; the only other recent collec- present a striking case. According to Perkins tions of the latter are from South Point, at (1899), they were among the most common the coast. Why both species should occur and widespread species on O‘ahu/Maui Nui there, at over 1,700 m and on an entirely dif- and Hawai‘i, respectively. Dozens of speci- ferent set of host plants, but not in other up- mens in the Bishop Museum collection attest land areas is unknown. It may be that the to this. But between the 1930s, when collect- ‘‘coastal’’ species also formerly inhabited all ing of Hylaeus species tapered off, and recent areas of upland dry forest but not the shrub- years, when it picked up again, the species land that makes up most of the dry habitat virtually disappeared. Only three specimens that remains today. This would also explain of H. facilis have been taken in the last 30 yr: the presence of H. anthracinus in dry forest one from O‘ahu in 1975, one from Maui in at over 600 m at Kanaio on Maui, when it is 1993, and one from Moloka‘i in 2005. I was otherwise strictly coastal. A single H. anthra- unable to recover either species during ex- cinus was recently collected at Po¯hakuloa, but tensive searches in areas where they had pre- the existence of a viable population there still viously been recorded. It is a mystery why needs to be verified. these two formerly abundant species should have decreased to the point of near extinction while substantial habitat remains and other Rare Species similar, closely related species, such as H. Hylaeus akoko, H. hula, H. kona, and H. para- chlorostictus and H. difficilis, are still very doxicus were discussed earlier. All four might abundant. be found farther south in the uplands of the The answer may be complex. It may be Kona and Ka‘u¯ Districts, but much of that that H. simplex was never as common as it area has been heavily grazed and has not been was thought to be. The expansion of the searched for bees since Perkins at the turn of male gonocoxae is the primary character that the last century. Aside from Pu‘u Wa‘awa‘a, unites H. simplex and H. facilis, and it is the H. hula is known from a small area of mesic only reliable means of distinguishing them forest at Kı¯lauea but is rare there as well; the from H. difficilis (females of the three species other species have been collected recently cannot be separated at all). The genital cap- only on C. olowaluana at Po¯hakuloa or Pu‘u sule is usually retracted within the abdomen, Wa‘awa‘a. and the gonocoxae are not visible in most of Hylaeus assimulans is a large coastal and the museum specimens labeled as H. simplex lowland species, known historically from that I have examined. This makes their deter- O‘ahu, Maui, Kaho‘olawe, and La¯na‘i. It is minations questionable, especially in light 184 PACIFIC SCIENCE . April 2007 of the current great abundance of H. difficilis. Uncollected Species It is still strange that it should have disap- peared while its habitat seems to remain but Ten species have not been collected recently, perhaps not as much so as if it had been very including the previously mentioned H. sim- abundant. plex. Most are known from very few speci- The gonocoxae of H. facilis are usually vis- mens. ible externally even without being extracted, Hylaeus anomalus occurred on southern and there is no question that it was formerly leeward Ko‘olau ridges on O‘ahu. Although quite common. However, all dry- and mesic- apparently never highly abundant, it was not forest species are now rare on Maui Nui and uncommon during the early collecting period O‘ahu. Even H. laetus, one of the most com- through the 1930s. Like H. fuscipennis, which mon species in drier areas of Kaua‘i and lived in the same area, it has not been col- Hawai‘i, is difficult to find on Maui and lected in decades. In light of the apparent O‘ahu, and has not been recently collected disappearance of these two species from the on Moloka‘i. Native dry and mesic habitats area, it is particularly strange that the newly have been decimated on the latter three described H. mimicus had not been collected islands (Mueller-Dombois 1973, Medeiros before. Even on the ridges directly above Ho- et al. 1986); much of what remains is in nolulu, I found the latter species in sufficient very small patches that may be incapable of numbers that a collector could hardly fail supporting a population of bees. It may be to recover it; yet in the Bishop Museum col- that if H. facilis still survives on Maui and lection (which contains dozens of H. facilis O‘ahu, it is in small remnant patches of mesic from this area) only a single specimen was or dry forest that have yet to be searched for found, and that was from the northern Ko‘o- bees. lau Range where little collecting was done. Two recently described O‘ahu species, H. Clearly a shift in species composition has oc- kuakea and H. mana, are also inhabitants of curred in the area despite the apparently good mesic forests. Each is known from a single quality of the vegetation. collection record, with two and four speci- Hylaeus finitimus is known from two speci- mens, respectively. The former is from a Na- mens collected by Perkins (1899) on the coast ture Conservancy preserve in the Wai‘anae at Makaweli (probably in the vicinity of Range (Honouliuli); this region contains a Ho‘a¯nuanu Bay, east of Waimea) on Kaua‘i. relatively large amount of potential habitat, Although there are patches of Scaevola at the and current protection and restoration efforts coast in that area, there do not appear to be may help maintain it there. It is not known any intact coastal communities there. Perkins what its specific habitat requirements are. Hy- (1899) stated that he searched there and in laeus mana was collected in mesic forest on other places for more specimens and did not Santalum at a relatively low elevation (430 find any, and the species was not found at m) in the leeward Ko‘olau Range. Although Polihale. Whether it is phylogenetically there is a considerable amount of native wet closer to H. longiceps or H. flavipes is uncertain forest above 500 m, at lower elevations and because the male specimen cannot be located; rainfall levels the vegetation is largely alien Perkins’ description (1899) and later key on most of the ridges with trails. That no (1910) are in conflict regarding the characters other species were found with either H. kua- that separate H. finitimus from the other two. kea or H. mana, and both were collected more The female is indistinguishable from H. long- or less by accident, suggests that more inten- iceps, and it may simply be a Kaua‘i population sive and systematic searching will turn up of that species. more specimens and possibly even more spe- Hylaeus gliddenae is known from a single cies. Nevertheless, the fact that it has taken so male collected at Kı¯lauea in 1934. Its habitat long for them to be discovered, coupled with is unknown, but the label states that it came the general scarcity of their habitat, means from a nest burrow in Myrsine, a tree that is they are likely to remain rare. widespread but most common and largest in Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 185 mesic forest. Its obvious close relationship to collect in, so it is not surprising that they H. paradoxicus, a mesic- or dry-forest species, have not been collected recently. Large areas supports this conclusion. In fact, it is conceiv- of intact wet-forest habitat remain on the is- able that at least some of the smaller, rare land, so more-intensive collecting will proba- females described by Perkins (1899) as H. ery- bly turn up these species. throdemas (a name later synonymized with H. Hylaeus niloticus (¼ obscuratus) is known paradoxicus) were actually females of H. glid- from a few specimens collected on the coasts denae. Kı¯lauea straddles a steep rainfall gradi- and lowlands of Hawai‘i, La¯na‘i, and Molo- ent, so wet, mesic, and dry habitats can be ka‘i. Little or no native habitat remains in found within a few kilometers (Figure 2). It the areas where it was collected previously has also been protected as part of Hawai‘i (Kona coast of Hawai‘i, dry lowlands of Mo- Volcanoes National Park since 1916. As a re- loka‘i), and it has not been found in similar sult, it is the most diverse locality for bees: 22 areas where other coastal Hylaeus species have of the 28 Hawai‘i Island species have been been collected (e.g., South Point, Po¯hakuloa, collected there, all except four coastal species Mo‘omomi). A few other areas of native dry and two that are confined to Kona. Eighteen habitat remain, such as at Kamiloloa in cen- of the 22 have been collected at Kı¯lauea tral Moloka‘i, but have not been searched. recently, but several mesic- to dry-forest Hylaeus perspicuus is known from a few species, including H. paradoxicus, have been specimens collected by Perkins in mesic to found exclusively or more abundantly at lo- wet forest at Makaweli on Kaua‘i. This area calities in northwestern Hawai‘i such as Pu‘u is privately owned and has not been surveyed Wa‘awa‘a and Po¯hakuloa. It therefore seems for bees recently, and the species has not more likely that H. gliddenae will be found at turned up in similar habitat around Waimea these sites than at its type locality. Canyon and the Alaka‘i region. Still, the wet- Hylaeus nalo is also known only from a sin- ter areas east of the canyon where it is most gle male, collected on O‘ahu in 1914. The likely to be found have not been searched in- specific locality is not recorded on the label, tensively, and it is also possible that it is lim- and it does not show any clear affinities to ited to the Makaweli area. other species. The most likely sister species is probably H. rugulosus, although there are several other possibilities. The lack of infor- discussion mation about it obviously makes it hard to Vegetation and Habitat search for, which is especially frustrating in light of the difficulty in interpreting its rela- The strong dependence of all species of Hy- tionships and its unique morphology. It is a laeus on native plants to the near-complete large species, but its impunctate metasoma exclusion of exotics (with the sole exception and broad scape separate it from the other of Tournefortia among the latter) means that groups of large species (the pubescens group, to conserve them native forests and shrub- and H. assimulans and H. ombrias, respec- lands must be preserved. Conservation of the tively). It also possesses truncate gonocoxae, plants and bees is a reciprocal situation, given which are unique in the subgenus, and pecu- the likely status of Hylaeus species as primary liar, poorly defined facial marks. pollinators of many important plants. Based Hylaeus mauiensis, H. melanothrix, and H. on current distributional information, it is satelles are all recorded from wet forests of clear that for bees to be present, at least some Maui (the last has also been collected on Mo- level of vegetation diversity is required. This loka‘i and La¯na‘i). The first is known from is probably due to a combination of temporal only a single male (the female was described (i.e., year-round availability of floral re- by Perkins [1899], but the specimen could sources) and nutritional factors. Visitation not be located); modest series of the other records (Daly and Magnacca 2003) and iden- two exist. All three were collected from very tification of pollen loads (unpubl. data) indi- wet areas that are difficult to access and to cate that Hylaeus species are frequent visitors 186 PACIFIC SCIENCE . April 2007

TABLE 3 Favored and Avoided Plant Species (See Table 1 for Explanation of Habitat Types)

Habitat

Scientific Name Family Common Name c d m w s

Favored flowers Chamaesyce spp. Euphorbiaceae ‘akoko Cheirodendron trigynum Araliaceae ‘o¯lapa Dodonaea viscosa Sapindaceae ‘a‘ali‘i Metrosideros polymorpha Myrtaceae ‘o¯hi‘a Myoporum sandwicense Myoporaceae naio Scaevola sericea Goodeniaceae naupaka kahakai Sida fallax Malvaceae ‘ilima Sophora chrysophylla Fabaceae ma¯mane Styphelia tameiameiae Epacridaceae pu¯kiawe Tournefortia argenteaa Boraginaceae tree heliotrope Rare plants Argyroxiphium sandwicenseb Asteraceae ‘a¯hinahina (silversword) Chamaesyce olowaluana Euphorbiaceae ‘akoko Sesbania tomentosab Fabaceae ‘ohai Dominant dry-forest trees (rarely visited) Acacia koa Fabaceae koa Diospyros sandwicensis Ebenaceae lama Erythrina sandwicensis Fabaceae wiliwili Nestegis sandwicensis Oleaceae olopua Psydrax odoratum Rubiaceae alahe‘e

a Nonnative species. b Listed endangered.

to many of the community-dominant plants Sophora, which is abundant in montane dry of Hawaiian ecosystems. All of the top 10 forest on Maui and Hawai‘i but relatively pollen types collected by Hylaeus species are rare on the other islands, and Myoporum, the dominant or codominant plants in wet which has a similarly patchy distribution, are forest or dry or coastal shrubland (Table 3, commonly visited dryland trees. Although it Favored flowers). In addition, several rare is a favored plant in all habitats, Metrosideros species are regularly visited where they are appears to be incapable of supporting bees found (Table 3, Rare plants). Although polli- on its own; they are rarely found where Met- nation rates have not been investigated, for rosideros is the only native flower available. many of these plants Hylaeus species are Instead, Hylaeus species primarily visit under- almost the only regular floral visitors and are story shrubs such as Dodonaea and Styphelia. undoubtedly important for native plant re- As a result, composition of the flora, not production and hence ecosystem health. merely diversity, is probably a major con- Dry and mesic forests seem to be an excep- straint on bee populations in this habitat. Al- tion to the dominant-plant rule. Aside from though there are still some large, native dry Metrosideros, most of the dominant trees char- and mesic areas on Kaua‘i and Hawai‘i, much acteristic of dry and mesic forest (Table 3, of the latter is pioneer vegetation on barren Dominant dry-forest trees) are not attractive lava flows, consisting almost exclusively of to bees; many of those that potentially are scattered Metrosideros ( Jacobi and Scott 1985). attractive, such as Reynoldsia sandwicensis Even in many remaining forests with a native (Araliaceae, ‘ohe makai), are now rare. Only canopy, the native understory shrubs that Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 187 form a large part of the diet of Hylaeus spe- in particular use many different plants as food cies are often absent due to the presence of sources, not only seasonally but at any given invasive grasses and shrubs (Cuddihy and time. In contrast, most native wet forests are Stone 1990, Gagne´ and Cuddihy 1990). On capable of supporting bees, at least partially O‘ahu and Maui Nui, dry and mesic forest because the pollen diet of wet-forest species and shrubland is restricted to small patches is overwhelmingly made up of only two (Figure 2) (Mueller-Dombois 1973, Medeiros species, Metrosideros polymorpha and Cheiro- et al. 1986). At least partly as a result, the dendron trigynum (unpubl. data). Both of middle islands have a much higher propor- these are usually present at relatively high tion of unrecovered species than do Hawai‘i frequency even in moderately disturbed for- and Kaua‘i (Table 1). est. Despite its reputation for being gravely Chamaesyce olowaluana represents a striking threatened worldwide, wet forest is the most example of the impact that only a few individ- intact ecosystem in Hawai‘i (Figure 2) (Cud- uals of a preferred host can have. It is one of dihy and Stone 1990). Large expanses of the most favored host plants despite its rarity forest are protected, to a greater or lesser de- and peculiar flowers, which are minute and gree, in montane areas of central Kaua‘i; effectively petal-less. Nine species of bees windward O‘ahu, Moloka‘i, Maui, and Ha- were collected on an isolated tree in remnant wai‘i; and Kona and Ka‘u¯ Districts on Ha- mesic forest at Pu‘u Wa‘awa‘a, but there were wai‘i. almost none on other flowering trees nearby. Similar areas a few hundred meters away, Nesting Biology even at sites with greater overall diversity and density of native plants, also had very low The nesting habits of most Hawaiian Hylaeus bee numbers and diversity, consisting only of species are unknown. However, many bees are common, widespread species. Another single more limited by availability of nest sites than tree in dry forest at Po¯hakuloa had similarly host plants (Westrich 1989), and this may be high diversity relative to nearby areas. This the case for Hawaiian species as well. Al- suggests that dry and mesic forests may once though most Hylaeus species worldwide nest have supported many more bee species that in dead pithy stems, examples of species that are now extirpated because certain favored nest in a variety of other substrates are known plants are gone. It also indicates that the (Michener 2000 and references therein). remaining dry forests, especially those with Among the Hawaiian species, it appears that high diversity and an intact native understory, those inhabiting wet areas nest in stems or should be searched intensively for bees, rather wood, whereas dryland species nest in the than passed over because a quick search finds ground. Hylaeus anomalus, H. dumetorum, H. none on the most common trees. Given the gliddenae, H. pubescens, H. setosifrons, and H. widespread destruction of dry forests and unicus are known as examples of the former, drastic reduction in diversity in much that re- and H. difficilis, H. flavipes, H. laetus, H. nivi- mains, it is probable that some species of Hy- cola, and H. rugulosus the latter. At least some laeus may have become extinct due to habitat ground-nesting species are also able to nest in loss before being discovered. crevices under rocks (Cole et al. 1992) or in Coastal species face a similar situation. rock walls. The cleptoparasitic species appear Although the indigenous Scaevola sericea is to attack only ground or crevice nesters. common and widespread, Hylaeus species Hylaeus species lack modifications for dig- are apparently not capable of surviving on it ging, such as strong mandibles and a pygidial alone. Although their exact requirements are plate that would allow them to excavate solid not known, coastal bees are almost exclusively wood and hard-packed soil. Almost certainly found in areas dominated by a variety of na- in the case of wood nesters, and possibly for tive shrubs and herbs, including Chamaesyce, ground nesters as well, the nest holes are usu- Jacquemontia, and Sida. Analysis of pollen ally not initiated by the bees but are the result loads (unpubl. data) shows that coastal species of burrows made by other insects, such as 188 PACIFIC SCIENCE . April 2007

Plagithmysus beetles (Cerambycidae) or Ec- Scaevola, and Sophora, honeybees are rarely temnius wasps (Crabronidae). If this is the found on some of the other important pollen case, they are dependent on the abundance plants for the native bees, such as Cheiroden- of other species for nest sites. dron, Dodonaea, and Styphelia. Although Hy- laeus species do not occur in native habitat where there are large numbers of honeybees, Alien Invertebrates the impact of moderate populations of Apis Competition with introduced bees is a poten- species is not known. The other common tially important factor that has not been in- species, Ceratina spp., H. albonitens, and L. im- vestigated in Hawai‘i. A large number of alien pavidum, are more similar in size and flower species have been introduced to Hawai‘i over visitation to native Hylaeus species and may the past 200 yr, including some hymenopter- have a greater impact through pollen collec- ans (such as ants and yellowjacket wasps) tion. All five are almost completely unstudied that have wreaked havoc on native arthropods in Hawai‘i, and only recently has their pres- (Perkins 1913, Zimmerman 1948, Gambino ence and/or extent of distribution been docu- 1992, Gillespie and Reimer 1993). Fifteen mented (Snelling 2003). species of exotic bees have been introduced Ants are considered the greatest threat to into Hawai‘i (Snelling 2003); 14 are adven- native arthropods (Perkins 1913), but their tive, nearly as many as in all of North effect on Hylaeus species is not fully known. America ( J. Ascher, pers. comm.). Some in- Although Cole et al. (1992) found that Argen- troduced species (Hylaeus leptocephala, Lithur- tine ants (Linepithema humile) on Haleakala¯ gus scabrosus, Megachile spp., and Xylocopa eliminated bees from ant-infested areas, they sonorina) occur in relatively low numbers or only compared the abundance of bee nests almost exclusively in areas dominated by under rocks in ant and non-ant areas. It may introduced plants and are generally not con- be that bees in small burrows in sand or soil sidered a threat to native species. Others, in- (the typical nesting medium for dryland Hy- cluding Apis mellifera (honeybee), Ceratina laeus species in Hawai‘i) are less vulnerable arizonensis, C. nr. dentipes, C. smaragdula, Hy- than those in unexcavated nests under loose laeus albonitens, and Lasioglossum impavidum, rocks. Reduction or extirpation of their Hy- occur abundantly in company with native Hy- laeus pollinators by ants and/or alien bees laeus species. Apis species can be found in ex- may have been a factor in elimination of na- tremely high numbers in some areas. The tive plants from much of the coast. parasites that have decimated feral honeybee populations in North America, the Asian bee conclusions mite (Varroa jacobsoni) and tracheal mite (Acarapis woodi) (Hoopingarner and Waller Nearly half of the Hawaiian Hylaeus species 1992), are not present in Hawai‘i. Apis species face immediate threats, primarily due to hab- readily visit a wide variety of both native and itat loss or alteration, and 10 are already pos- exotic plants (Arita et al. 1989) and forage sibly extinct. As a potential keystone group in throughout the day and in cloudy or even all Hawaiian ecosystems, they deserve greater rainy weather. In some areas, it seems that study of their behavior and ecology, as well as by sheer numbers they could hardly fail to their broader influence on native habitats have an impact on Hylaeus species, especially through pollination. The degree of reliance with regard to the nectar supply. The pollina- of both common and rare plants on pollina- tion relationship between Hylaeus species and tion by Hylaeus species, tolerance of bees for the endangered plant Sesbania tomentosa at alien bees and ants, and nesting biology are Ka‘ena Point on O‘ahu was studied by Hop- all fertile subjects for future research that per (2002), who found that Hylaeus species would greatly enhance our understanding of are the primary pollinators, with Apis species the role of Hylaeus species in Hawaiian eco- mainly acting as nectar robbers. Still, al- systems and the steps necessary for their though abundant on Metrosideros, Myoporum, conservation. Conservation Status of Endemic Hylaeus (Nesoprosopis) of Hawai‘i . Magnacca 189

acknowledgments Hawaii’s endemic arthropods, with an em- phasis on the moths or Lepidoptera. Pages Thanks to Bryan Danforth, Jim Liebherr, 63–72 in Proceedings of the 4th Hawaii and Pamela Scheffler for reviewing the manu- Volcanoes National Park Natural Science script. Funding for collecting trips came from Conference. University of Hawai‘i Coop- a grant to Howell Daly (University of erative Park Studies Unit, Honolulu. California–Berkeley) from the U.S. Fish and Gagne´, W. C., and L. W. Cuddihy. 1990. Wildlife Service, Pacific Islands Ecoregion, Vegetation. Pages 45–114 in W. L. Wag- Ecological Services, Honolulu. ner, D. R. Herbst, and S. H. Somer, eds. Manual of the flowering plants of Hawai‘i. Literature Cited University of Hawai‘i Press, Bishop Mu- seum Press, Honolulu. Arita, L. H., W. S. Sakai, and J. J. Moniz. Gambino, P. 1992. Yellowjacket (Vespula pen- 1989. Pollen types collected by honey bees sylvanica) predation at Hawaii Volcanoes at three localities on the island of Hawaii. and Haleakala National Parks: Identity of Part 1: Floral source identification. Proc. prey items. Proc. Hawaii. Entomol. Soc. Hawaii. Entomol. Soc. 29:125–131. 31:157–164. Beardsley, J. W. 1966. Insects and other Gillespie, R. G., and N. Reimer. 1993. The terrestrial arthropods from the leeward effect of alien predatory ants (Hymenop- Hawaiian Islands. Proc. Hawaii. Entomol. tera: Formicidae) on Hawaiian endemic Soc. 19:157–185. spiders (Araneae: Tetragnathidae). Pac. Carson, H. L., and D. A. Clague. 1995. Geol- Sci. 47:21–33. ogy and biogeography of the Hawaiian Is- Hirashima, Y. 1977. Revision of the Japanese lands. Pages 14–29 in W. L. Wagner and species of Nesoprosopis, with descriptions of V. A. Funk, eds. Hawaiian biogeography. two new species (Hymenoptera, Colleti- Smithsonian Institution Press, Washing- dae, Hylaeus). Esakia 10:21–43. ton, D.C. Hoopingarner, R. A., and G. D. Waller. Cole, F. R., A. C. Medeiros, L. L. Loope, and 1992. Crop pollination. Pages 1044–1082 W. W. Zuehlke. 1992. Effects of the in J. M. Graham, ed. The hive and the Argentine ant on arthropod fauna of Ha- honey bee. Dadant and Sons, Hamilton, waiian high-elevation shrubland. Ecology Illinois. 73:1313–1322. Hopper, D. 2002. The reproductive biology Conant, S. 1985. Recent observations on and conservation of the endangered Ha- the plants of Nihoa Island, Northwestern waiian legume Sesbania tomentosa, with em- Hawaiian Islands, USA. Pac. Sci. 39:135– phasis on its pollination system. Ph.D. diss., 149. University of Hawai‘i at Ma¯noa, Hono- Cox, P. A., and T. Elmqvist. 2000. Pollinator lulu. extinction in the Pacific Islands. Conserv. IUCN. 2001. IUCN Red List categories Biol. 14:1237–1239. and criteria: Version 3.1. IUCN Species Cuddihy, L. W., and C. P. Stone. 1990. Survival Commission. IUCN, Gland, Alteration of native Hawaiian vegetation: Switzerland, and Cambridge, United Effects of humans, their activities and in- Kingdom. troductions. University of Hawai‘i Coop- Jacobi, J. D., and J. M. Scott. 1985. An assess- erative Park Studies Unit, Honolulu. ment of the current status of native upland Daly, H. V., and K. N. Magnacca. 2003. In- habitats and associated endangered species sects of Hawaii. Vol. 17. Hawaiian Hylaeus on the islands of Hawaii. Pages 3–22 in C. (Nesoprosopis) bees (Hymenoptera: Apoi- P. Stone and J. M. Scott, eds. Hawaii’s dea). University of Hawai‘i Press, Hono- terrestrial ecosystems: Preservation and lulu. management. University of Hawai‘i Press, Gagne´, W. C. 1982. Working toward an Honolulu. assessment of the conservation status of Juvik, S. P., and J. O. Juvik, eds. 1998. Atlas 190 PACIFIC SCIENCE . April 2007

of Hawai‘i, 3rd ed. University of Hawai‘i waiiensis. Vol. 2. Cambridge University Press, Honolulu. Press, London. Liebherr, J. K. 2001. Preface to the reissue of ———. 1913. Introduction. Pages i–ccxxvii volume 1. Pages v–xv in E. C. Zimmer- in D. Sharp, ed. Fauna Hawaiiensis. Vol. man. Insects of Hawaii. Vol. 1. University 1. Cambridge University Press, London. of Hawai‘i Press, Honolulu. Proctor, M., P. Yeo, and A. Lack. 1996. Magnacca, K. N., and B. N. Danforth. 2006. The natural history of pollination. Timber Evolution and biogeography of native Ha- Press, Portland, Oregon. waiian Hylaeus bees (Hymenoptera: Colle- Snelling, R. R. 2003. Bees of the Hawaiian tidae). Cladistics 22:393–411. Islands, exclusive of Hylaeus (Nesoprosopis) Medeiros, A. C., L. L. Loope, and R. A. Holt. (Hymenoptera: Apoidea). J. Kans. Ento- 1986. Status of native flowering plant mol. Soc. 76:342–356. species on the south slope of Haleakala, U.S. Fish and Wildlife Service. 1984. Endan- East Maui, Hawaii. Technical Report 59. gered and threatened wildlife and plants: University of Hawaii Cooperative Park Review of invertebrate wildlife for listing Studies Unit, Honolulu. as endangered or threatened species. Fed. Michener, C. D. 2000. The bees of the Regist. 49:21664–21673. world. The Johns Hopkins University ———. 1996. Endangered and threatened Press, Baltimore. wildlife and plants: Review of plant and Moore, J. G., and D. A. Clague. 1992. animal taxa that are candidates for listing Volcano growth and evolution of the is- as endangered or threatened species. Fed. land of Hawaii. Geol. Soc. Am. Bull. Regist. 61:7596–7613. 104:1471–1484. Wagner, W. L., D. R. Herbst, and S. H. Mueller-Dombois, D. 1973. A non-adapted Somer. 1990. Manual of the flower- vegetation interferes with water removal ing plants of Hawai‘i. University of Ha- in a tropical rain forest area in Hawaii. wai‘i Press, Bishop Museum Press, Trop. Ecol. 14:1–18. Honolulu. Nishida, G. N. 2002. Hawaiian terrestrial Westrich, P. 1989. Die Bienen Baden– arthropod checklist, 4th ed. B. P. Bishop Wu¨rttembergs, Volume 1. Ulmer Verlag, Museum, Honolulu. Stuttgart. 431 pp. Perkins, R. C. L. 1899. Hymenoptera Acu- Zimmerman, E. C. 1948. Insects of Hawaii. leata. Pages 75–113 in D. Sharp, ed. Fauna Vol. 1. Introduction. University of Hawai‘i Hawaiiensis. Vol. 1. Cambridge University Press, Honolulu. Press, London. ———. 1970. Adaptive radiation in Hawaii ———. 1910. Supplement to Hymenoptera. with special reference to insects. Biotrop- Pages 600–612 in D. Sharp, ed. Fauna Ha- ica 2:32–38.