Madrono, Vol. 54, No. 4, pp. 286-292, 2007 POLLINATORS OF AN ENDEMIC AND ENDANGERED SPECIES, MAMMILLARIA GAUMERI (CACTACEAE), IN ITS NATURAL HABITAT (COASTAL DUNE) AND IN A BOTANICAL GARDEN Manuela Giovanetti^ J. Carlos Cervera^, and Jose Luis Andrade^-^ ^Universita degli Studi di Milano, Dip. Biologia, Via Celoria 26 - 20133 Milano, Italy [email protected] ^Unidad de Recursos Naturales, Centro de Investigacion Cientifica de Yucatan, A. C, Calle 43 No.130 Colonia Chuburna de Hidalgo, C. P. 97200, Merida, Yucatan, Mexico Abstract Mammillaria gaumeri (Britton & Ross) Orcutt (Cactaceae), an endemic plant of the Yucatan Peninsula, is included bytheMexican government inthelistofspeciesthat require special protection. Its natural habitat is now restricted to fragmented areas and protection programs involve botanical gardens in growing individuals rescued from disturbed areas. Little information is available on the reproductivecharacteristicsofthisspeciesandnothingisknownofitspollinators. Weinvestigatedthe visitors ofM. gaumeriflowers, collecting and observing bee species in its natural habitat (i.e., coastal dune) and in a botanical garden, where coastal dune vegetation had been created. Observationswere made on plants whose density was artificially increased by grouping flowering individuals. At each site, we: 1) collected insects visiting the flowers; 2) recorded number ofvisits; and 3) video-recorded bee movements on the flowers. As expected, the number ofbee species and visitation frequency were higher at the botanical garden than at the coastal dune. After landing on a flower, bees either inspected the anthers or dived among them. These behaviors, carried out by all observed species, seemed relatedtothestateoftheanthers(fulloremptyofpollen)andstigmalobes(openedorclosed). Specifically, visits lasted longer when anthers were full ofpollen and stigma lobes were opened. The samebeespeciesrecorded on thedunewere also recorded at thebotanical garden, suggestingthat the artificial dune at the botanical garden offered suitable conditions for the natural pollinators ofthis endangered cactus. Resumen Mammillariagaumeri(Britton & Rose) Orcutt (Cactaceae)esendemicade la Peninsula deYucatan y esta incluida en la lista de especies que requieren proteccion especial por decreto del gobierno de Mexico. Presenta un habitat restringido en areas fragmentadas y los programas de proteccion consideran el rescate de individuos de areas perturbadas para crecerlos enjardines botanicos. De la poca informacion disponible sobrelascaracteristicas reproductivasdeestaespecie, no seconocenada desuspolinizadores. Enesteestudio,nosotrosinvestigamoslosvisitadoresdelasfloresde M. gaumeri y observamos y colectamos las especies de abejas en su habitat natural (duna costera) y en unjardin botanico, donde el habitat de la duna costera ha sido imitado. Las observaciones fueron hechas en plantascuyadensidadfueaumentadaartificialmenteconvariosindividuosenfloracion. Encadasitio: 1) colectamos los insectos que visitaron las flores; 2) registramos el mimero de visitas; y 3) filmamos losmovimientosdelasabejassobrelasflores. Comoesperabamos,elnumerodeespeciesdeabejasyla frecuencia de visitas fueron mayores en eljardin botanico que en la duna costera. Al aterrizar sobre una flor, las abejas se comportaron de dos diferentes maneras: ya sea que ellas inspeccionaban las anteras o se zambullian entre ellas. Estos comportamientos, realizados por todas las especies visitadoras, parecen estar relacionados con el estado de las anteras (llenas o vacias de polen) y a la posicion de los lobulos del estigma (abiertos o cerrados). Especificamente, las visitas duraron mas tiempo cuando las anterasestaban llenas de polen y los lobulosde los estigmas estaban abiertos. Las mismasespeciesdeabejassecolectaronenambossitios, loque sugierequeladunaartificial deljardin botanico tambien ofrece condiciones adecuadas para los polinizadores naturales de este cactus amenazado. Key Words: botanical gardens, Cactaceae, endangered cactus, endemiccactus, Mammillariagaumeri, pollinators. When dealing with the conservation ofendan- reserve design (Deltoro et al. 2004), demography gered plant species, research usually has focused for short-term population management (Fox and on habitat preservation (Hall and Gillespie 2004), Gurevitch 2000), and the possibihty ofreproduc- tion in botanical gardens for future reintroduc- tion in their natural range ofdistribution (Perini ''Corresponding author, e-mail: [email protected] and Tornadore 2004). But, it is also frequently! I 2007] GIOVANETTI ET AL.: POLLINATORS OF MAMMILLARIA GAUMERI 287 We reported that endangered plant species experience quency. described the bees' behavior to infer reproductive failure that may be strongly related each species' interest in the resources offered by to the loss of pollinators (Aronne and Wilcock the flower and possible elements crucial in flower 2004). Many species may vegetatively reproduce, selection by pollinators. especially when pollinators are rare or absent and flower fertilization is low. However, pollination Methods (i.e., cross-pollination) of flowering plants pro- vides the enormous benefit of gene flow in the We conducted field observations on a natural population and helps to maintain the species' population ofMammillariagaumeriin thecoastal ability to respond to a changing environment dune scrubland of San Benito (2rT9'10"N, (Proctor et al. 1996; Price 2002). 89°30'40"W) and in an artificial coastal dune To formulate effective strategies for manage- recreated at the Centro de Investigacion Cienti- ment and conservation, it is essential to un- fica de Yucatan (CICY) Botanical Gardens in the derstand the ecological and evolutionary nature city of Merida (2r02'38"N, 89^38'22"W). The ofplant-pollinator interactions. Notwithstanding subject ofthe study is a small globularcactus that that pollinators ofmany plant species are not yet grows on exposed areas as well as under cover of identified (Kearns et. al. 1998), there is already vegetation. evidence that a deficit of pollinators in fragmen- At the two sites, the micro-habitats were very ted habitats threatens plant populations and similar, but differed in the respective surround- plant diversity (Lennartsson 2002). ings. On the coastal dune, the vegetation was Regarding plant conservation and reintroduc- mainly characterized by small bushes and succu- tion, one potential problem is that some native lents that can reach heights from 3 to 5 m, and pollinators may have disappeared. Nonetheless, common species were Coccoloha uvifera (L.) L. most flowering plants are pollinated by moderate (Polygonaceae), Cordia sebestena L. (Boragina- to high numbers ofinsect species, and most floral ceae), Bravaisia herlandieriana (Nees) T.F. Daniel visitors usually visit many different hosts (Corbet (Acanthaceae), Agave angustifolia Haw. (Agava- 1997; Kearns et al. 1998). Thanks to the multiple ceae), Opuntia dillenii (Ker Gawl.) Haw. (Cacta- pollinators observed for the majority of plant ceae) Acanthocereus tetragomis (L.) Hummelinck species, "new" pollinators might adopt a plant if (Cactaceae) and Myrmecophila tihieinis var.chris- its original pollinators are absent. The adoption tinae Carnevali & Gomez-Juarez (Orchidaceae); of endangered species by alternative pollinators the area surrounding the dunes included urban may also be crucial when trying to propagate areas and natural vegetation. The CICY Botan- them under artificial conditions, as in botanical ical Garden recreates many habitats on an area of gardens. Thus, priorities for research programs 27,000 m-, and maintains 107 families and 592 should include the identification of native polli- plant species. Around it, land was urbanized. nators as well as other possible pollinators ofthe We established a 2-m X 2-m observation area threatened plant species. at each site. At San Benito, where we located Mammillaria gaumeri (Britton & Rose) Orcutt about 4 flowering individuals of M. gaumeri on (Cactaceae, tribe Cacteae) is a globular cactus the dunes in open areas, 7 plants growing in pots endemic to the northern coast of Yucatan. This were added. Plant density at this study site was species is under special protection in the Pro- 0.86 individuals per m-, so density was increased tection of Natural Resources Document by the to 2.75 m ^ At CICY Botanical Gardens, 9 Mexican government (NOM-059-ECOL-2001). It plants were growing in pots and located in open grows only in some restricted areas, due to areas (plant density = 2.25 m -). Flower density fragmentation of its natural habitat by human was then increased compared to natural condi- disturbances such as urbanization, cattle-range tions, but kept constant at both sites (about 15 management, and agriculture (Duran et al. 1998). opened flowers during observations). Some studies deal with its distribution and We collected specimens and made observations abundance, seed germination, and seedling sur- during the early dry season (January 27 to vival (Leirana-Alcocer and Parra-Tabla 1999; February 3, 2005), which lasts from November Lopez-Jimenez 2001; Cervera et al. 2006), but to February and is marked by an alternation of accurate information on its reproductive biology full sun and cloudy days, with infrequent pre- is still lacking. Our aim was to identify the native cipitation and strong winds. Because Hymenop- pollinators of this rare species, through the tera are often inactive on cloudy days, we used collection of pollinators in one of its natural data from sunny days only. Status of flowers habitats, the coastal dune scrubland. We were (open/closed) was recorded every hour for nine also interested in the flower visitation rate in flowers on four plants. a botanical garden, where plants rescued from Considering weather conditions (around mid- disturbed areas have been preserved on an day clouds appeared) and flower opening, we artificial dune. Between these two sites we based comparisons on visitation frequency be- compared pollinator species and visitation fre- tween 10:30 and 12:00. The number of flower- 4 MADRONO [Vol. 54 Fig. 1. FlowersofMammillariagaumeri. Theanthersarecurvedtowardsthestigma;thelobesarestillcloseinthe flower on the left and open in the flower on the right. ( I visiting bees was recorded during one day at each However, we observed bees entering flowers with site; attention was directed to bees landing petals only shghtly opened. | independently on each flower by only one observer at each site. Temperature and relative Number of Bee Species humidity were also recorded during observations, using a digital thermo-hygrometer (Deltha Ohm, Table 1 reports all species collected at the two HD201-1; Italy). sites. We collected two species at San Benito: an witThheaviddiegoi-traelcorvdiidnego-ocfamseormae v(iSsiotnsy,(n D=CR31-) AHpyildaaeeu,s CqeuraadtriantaifLeartraei(lCloecskpr.el1l,).andAlasoC,ollweetidoabe-, TRV80E; Japan) equipped with magnifying lens served an ant, which apparently inspected a provided data on: landing surface elected by the flower, and we saw a winged wasp (possibly bee; flower characteristics at the moment of the a Sphecidae or a Mutillidae) visiting a flower. visit (empty/full anthers, open/closed stigma We collected five bee species at CICY Botan- lobes); bee behaviors on the flower; and visit ical Gardens. Among them, there was one ofthe duration. Video-recording was done on January bee species collected at San Benito (Colletidae, 27, 29 and February 3, 2006 only when bees were Hylaeus quadratifera), another Apidae belonging approaching flowers. Total duration ofthe video to the genus Ceratina Latreille sp. 2, and other i record was about 30 min, obtained during9 hr of three species belonging to two families (Mega- | observations. chilidae, Megachile Latreille sp.; Halictidae, Dia- | Due to difficulties of identification related to their small size, bees were divided into three recognizable classes: small black bees, megachihd ClosedPetalsDOpenPetals bees, green bees. Additionally, some scattered observations at both sites and specimencollection were done, on sunny days, from January 15 to February 10, 2005. Statistical analysis was performed using Statis- tica (StatSoft Inc., Tulsa, Oklahoma, USA) and following Zar (1974). Results Characteristics of Mammillaria gaumeri Flowers The flowers of this species are of a pale color, ooooooooo with whitish-yellow petals and light yellow coc*5c^cococococr)co anthers. Anthers are very numerous and curved ooi-i-----^-^-^-^ towards the stigma (Fig. 1). The stigma has four lobes that changed in color from yellow to Timeoftheday reddish-pink and from a closed to an open stance. Flowers were fully open (Fig. 2) during Fig. 2. Opening time of the nine flowers under the central hours of the day (12:30-15:30). observation on a sunny day. |j j 2007] GIOVANETTI ET AL.: POLLINATORS OF MAMMILLARIA GAUMERI 289 Table L Bee Species Collected during Observations on Flowers of Mammillaria gaumeri. Classification following Michener (2000). SAN CICY Botanical raQltvnli1i1\y/ ^11 m11v RFNITO Apidae Xylocopinae Ceratina Latreille, 1802 Unknown 1 Present Unknown 2 Present CoUetidae Hylaeinae Hyleaus Fabricius, 1793 quadratifera Present Present (Cockrell) Halictidae Halictinae Dialictus Robertson, 1902 Unknown Present Augoclora Smith, 1853 Unknown Present Megachilidae Megachilinae Mt'gflc/z/Ve' Latreille, 1802 Unknown Present lictus Robertson sp. and Augochlora Sinith sp.). Description of Visits During observations, no ant or wasp visited the flowers, although some were present at the site. The following descriptions regard the behavior Observed species correspond to the bee classes of Ceratina sp. 1, Ceratina sp. 2, Hylaeus as follows: Small black bees = Ceratina sp. 1, quadratifera, and Megachile sp. Ceratina sp. 2, Hylaeus quadratifera; Dialictus Bees landed on the flower rapidly: in 69.3% of sp.; Megachilid bees = Megachile sp.; Green bees the cases (n = 26), they landed mainly on the = Augochlora sp. petals and walked to the anthers; in 19.2% ofthe cases, they landed on the stigma lobes, then move to the anthers, whereas in 11.5% of cases, they Visitation Frequency landed directly on the mass of anthers. Petals obsAetrvabtoitohns swieterse, ocpltiimmaatlicforcobnedeitaicotnivsityd,urwiintgh w1e5.r4e6,thPe <mor0.e00f1r).equAefntterlarnedaicnhgingsurtfhaecean(tAh^ers=, tfuelmlpesruantuarnedshiwgehretemlpoewreartur(ersa.ngeAt27S.a0n-2B9e.n1i°tCo), (po4l2l%inaotforcsasae)s;innsp=ect3e1d),thoremb)widtohvetheiimrmaendtieantnealey than at CICY Botanical Gardens (32.0-34.7°C); among them (42%), and in a few cases, we relative humidity was higher at San Benito (64.3- recorded c) both behaviors (16%). No statistical 67.9%) than at CICY Botanical Gardens (36.9- differences among the frequency of the three 45.1%)., the number of visits recorded at CICY behaviors emerged (JV^ = 4.13, n. s.). Botanical Garden was 2 times higher than that Mean and SE ofvisit duration of each case is recorded at San Benito (A^ = 5.32, P < 0.05). reported in Table 2. The first strategy (i.e., While in San Benito small black bees were inspecting the anthers) consisted of very short responsible for all the visits (n = 16), at CICY visits: bees landed on the flower as described Botanical Garden they made 27 out of 32 visits above and quite immediately flew away. The (84%). The other visits were distributed among second strategy (i.e., diving among anthers) megachilid bees (3 visits) and green bees (2 visits). included the bee disappearing under the anthers, (These observations were very limited, however a lapse of time during which movements were [from 10:30 to 12:00, one day per site].) perceived from oscillation ofanthers and stigma. Table 2. Descriptive Statistics of Visit Duration. Observations based on: bee behavior, flower characteristics (anthers and stigma lobes), and bee group. Data obtained from analysis of video-recorded images. Statistical analysis between groups is reported in the text. Average (seconds) SE (seconds) Bee behavior inspecting 2.18 0.50 11 diving 11.08 L39 13 Anthers characteristics full ofpollen 3.08 0.80 13 empty ofpollen 9.87 L43 16 Stigma lobes characteristics opened 9.35 1.44 17 closed 3.25 0.84 12 Bee group megachilid bees 8.60 3.44 5 small black bees 6.46 1.10 24 MADRONO 290 [Vol. 54 Beesbehavior behaviors was statistically significant (Mann- Whitney z adjusted = -4.186, n 1 = 11, n 2 = 13, P < 0.001; Fig. 3, upper graph). There also was a significant difference between visits to flowers with full or empty anthers (Mann- Whitney z adjusted = -3.454, n 1 = 13, n 2 = 16, P < 0.001; Fig. 3, middle graph) and with opened or closed stigma lobes (Mann-Whitney z adjusted = 2.996, nl = 17, n2 = 12, P< 0.005; Fig. 3, lower graph). When lobes were opened, flowers received 58.6% of visits (n = 29); when stigma lobes were closed, flowers received the 41.4% ofvisits. In 16 out of25 visits, the body of a small black bee touched the stigma lobes, while a)inspect b)dive it happened in 4 out of 5 visits for megachilid bees. Flowercharacteristics: anthers 14 I Discussion This preliminary study identified some of the pollinators ofan endangered and endemic cactus of the Yucatan Peninsula, both in its natural habitat and in a botanical garden, where its habitat conditions were recreated. Mammillaria gaumeri can flower many times during a year, a trait shared with other, more common species ofMamniillaria (Bowers 2002). Flowering several times within a year may increase the probability ^ qI . . . of pollination by enabling flowers to encounter emptyantiiers pollinators under different environmental condi- fullantiiers tions. Flowercharacterisctics: stigmalobe—s Ceratina species, as well as Hylaeus quadrati- 14I • fera and Dialictus sp., accounted for most visits to Maimnillariagaumeriflowers at both sites. The tribe Ceratinini (Apidae, Xylocopinae)consists of small and slender species, similar to the world- wide subfamilyHylaeinae(Colletidae), a group of minute to moderate-sized, mostly slender bees with a limited yellow or white mark on the face (Michener 2000). Ceratina usually nest in dead plant material, such as hollow or pithy stems or burrowsmade in rotten woodyvines or stems. As do many Allodapinae in the Old World Tropics and Hylaeus, they tolerate exposure and highly q\ : . . , 1 variable nesting temperatures (Roubik 1989). openlobes closedlobes Michener (1970) found more nests of Ceratina in the introduced plant species ofgardens than in Fig. 3. Average duration of visits depending on bee normal vegetation, which may account for the behavior and floral characteristics (anthers and stigma lobes). Data were obtained from analysis of video- ability of this genus to invade non-natural areas recordedimages. Uppergraph: visitdurationdepending and consequently provide the pollination service on behavior; middle graph: visit duration depending on needed. anthers (full or empty); lower graph: visit duration The two species {Ceratina sp. and Hylaeus dependingon stigma lobes (opened orclosed). Data are quadratifera) collected on the dune scrubland means(opensquares) ± SE(graysquares) ± 1.96 X SE. were the only flower visitors during observations. On the dune, the vegetation provides plenty of and partial visualization of the bee among the nesting substrate (dead plant material). We can stamens. After that, they emerged and cleaned/ imagine that in such an environment a strong moved pollen grains to the legs, possibly resting relationship between a pollinator and this plant for a few seconds, while lying on the anthers or species may exist (flowers are small, pale, located on the petals, and then they flew away. The at surface level, often hidden by other plants). difference in visit duration between the two Both bee species are characterised by a small size 2007] GIOVANETTI ET AL.: POLLINATORS OF MAMMILLARIA GAUMERI 291 (four to six millimetres long) and by the fact that Although there were some differences in pollen is carried in the crop moistened with foraging and nesting characteristics, all species nectar. This feature may justify the observed performed different behaviors depending on the behaviour on the flowers: bees diving among state ofthe anthers, as well as that ofthe stigma anthers may be searching for the nectaries. lobes. These differences highlight the importance Nectar is even the main resource found in of recognition by individual bees of the flower Hylaeinae nests, so accurate measurement of status and possibly also ofthe reward to be found nectar crop in M. gaumeri population may in the flower. Flower attributes need to be further indicate its importance as a nectar resource for studied in order to inferthe information obtained the corresponding population of bees. The by the approaching bees and for a better un- percentage of individuals observed touching the derstanding of bee reactions when landing on stigma (64%; n = 31), as well as the ones landing a flower. This was only a preliminary work, but on it, accounts for a potentially good pollination its results have raised many questions related to service. the recovery of the natural populations of M. The simulated coastal dune habitat at CICY gaumeri and its pollinators, and its cultivation in Botanical Gardens (about 30 km from San botanical gardens. Benito) has favored the establishment ofthe bees found on the dune. At CICY Botanical Gardens Acknowledgments many different habitat have been created, such as We thank Luis Sima for help in the field, Martha tropical dry deciduous forest, dry deciduous Mendezand Gabriel Dzib forlending us plants, and we forest with columnar cacti, and coastal dune are especially indebted to Dr. R. Villanueva for the scrubland; the high amount of flowering species identification of bee specimens. M. Giovanetti was may also attract more bee species. This is actually supported by an Italian grant, FIRB (Fondo Italiano confirmed by the bee collection on M. gaumeri per la Ricerca di Base), n. RBAU019H94-001 C. . made at this location and by the significant Cervera was supported by a fellowship from Fondo higher visitation rate recorded compared to the Mixto CONACYT-Gobierno del Estado de Yucatan inantfuorramlatcoiaosntalondunMe.ingaSaunmeBreinitroe.soAugraciens mmoarye 4(s8Yu3pU4pC4o-/r22t04e05d838-.Cb0yW2e-0F4to2hn)ad.nokTthShieecstCoIrreiCsaYelarBScohtEaPnwi-acsaClOpGNaarArtCidaeYlnlTsy explain such a high rate of visit, considering the for the logistic support. diverse floweringplants present at the site and the possible competition among them. Literature Cited Megachilidae are famous for collecting pollen on the underside of the abdomen. They do not Aronne, G. and C. C. Wilcock. 2004. Reproductive characteristics and breeding systems of shrubs of line their cell, but use building materials as lead the Mediterranean region. Functional Ecology or petal pieces, chewed leaf pulp, hairs, nectar, 8:69-76. resin, pebbles, mud and combinations of these Bowers, J. E. 2002. Flowering patterns and reproduc- (Michener 2000). Provisions are not shaped in tive ecology of Mammillaria grahamii (Cactaceae), any form, but pollen is loose; some species are a common, small cactus in the Sonoran Desert. pollen specialist (Roubik 1989). They are bigger Madroiio 49:201-206. than Ceratina and Hylaeus and this may result Cervera, J. C, J. L. Andrade, J. L. SimA, and E. A. a disadvantage for the flower because bee Graham. 2006. Microhabitats, germination, and movement among the anthers may damage the establishment for Mammillaria gaumeri (Cacta- tissues. Also, considering that in their cells the ceae), a rare species from Yucatan. International voenrliyfyprwohviestihoenristphoelsleenb,eietswowuelrde bfeeeidnitnegresotningthteo CorbJeotu,rnaSl. oAf.Pl1a9n9t7.SciReonlceeso1f67p:o3l1l1in-a3t1o9r.s in species nectar or collecting pollen when diving among preservation, conservation, ecosystem stability and genetic diversity. Acta Horticulturae 437:219-229. the anthers. Halictinae includes some of the Deltoro, v., J. Perez-Botella, L. Serra, P. Perez- commonest bees; they mainly nest in the soil, in Rovira, a. Olivares, S. Fos, G. Ballester, banks or flat soil, or rarely in rotting wood (but AND E. Laguna. 2004. Plant Microreserves: Augochlora females do). Provisions are usually frequently asked questions. Planta Europa IV firm, sub-spherical and lie on the ventral surface Proceedings - 4th European Conference on the of the cell. The majority ofspecies are polylectic Conservation of Wild Plants - Valencia 17-20 (thMeiscehetnweorf2a0m0i0l)i.esFheawveibnedeivnidrueaclosrdbeedl,onbguitngthetyo DurAOnct,obRe.r, 2J0.04C. Trejo-Torres, and G. Ibarra. are probably more widespread than supposed. It 1998. Endemic phytotaxa of the Peninsula of would be especially interesting to compare the Fox,YuGc.ataAn.. HAaNrDvaJr.d GPuapReErVsiTinCHB.ota2n00y0.3:2P6o3pu2l7a3t.ion collection habits of these bees with those species numbers count: tools for near-term demographic observed on the dune, to estimate their potenti- analysis. The American Naturalist 156:242-256. ality as pollinators of M. gaumeri. In fact, the Hall, J. M. and T. W. Gillespie. 2004. Rarity and intensity or quality ofpollination may be affected conservation of Florida scrub plants. Florida ifpollinator species changed (Corbet 1997). Scientist 67:9-17. MADRONO 292 [Vol. 54 Kearns, C. a., D. W. Inouye, and N. M. Waser. NOM-059-ECOL-2001. Norma Oficial Mexicana que ! 1998. Endangered mutualism: the conservation of determina las especies y subespecies de flora y I plant-pollinator interactions. Annual Review of fauna silvestres terrestres y acuatica en peligro de Ecology and Systematic 29:83-112. extincion, amenazadas, raras,ylassujetasaprotec- i LEIRANA-ALCOCER, J. AND V. Parra-Tabla. 1999. cion especial y que establece especificaciones para | Factors affecting the distribution, abundance su proteccion. Diario Oficial de la Federacion de ! and seedling survival of Mammillaria gaumeri, Mexico, D.F., Mexico. j' an endemic cactus of coastal Yucatan, Perini, C. and N. Tornadore. 2004. The Italian Mexico. Journal of Arid Environments 41:421- Botanical Society and Botanic Gardens in service 428. to raise awareness ofplant diversity conservation. Lennartsson, T. 2002. Extinction thresholds and Planta Europa lY Proceedings - 4th European disrupted plant-pollinator interactions in frag- Conference on the Conservation of Wild Plants - mented plant populations. Ecology 83:3060- Valencia 17-20 October 2004. [http:www.nerium. 3072. net/plantaeuropa/main.htm] Lopez-Jimenez, L. 2001. Distribucion, abundancia y Price, P. W. 2002. Speciesinteractionsandevolutionof estructura poblacional de Mammillaria gaumeri biodiversity.Pp. 3-25inM.HerreraandO.Pellmyr, Orcutt, especie rara y endemica de Yucatan. BSc (eds.). Plant-animal interactions. An evolutionary Thesis. Instituto Tecnologico Agropecuario No. 2. approach. Blackwell Science Ltd. TJ International , Conkal, Yucatan, Mexico. Ltd, Padstow, Cornwall, Great Britain. MiCHENER, C. D. 1970. Nest sites of stem and twig Proctor, M., P. Yeo, and A. Lack. 1996. The inhabiting African bees. Journal of the Entomo- natural history of pollination. Harper & Collins logical Society ofSouth Africa 33:1-22. Publishers. The Bath Press, Great Britain. . 1974. The social behavior ofbees. A compar- ROUBIK, D. W. 1989. Ecology and natural history of ative study. Belknap Press, Harvard University tropical bees. Cambridge University Press, New Press, Cambridge, MA. York, NY. . 2000. The bees of the world. The Johns Zar, J. H. 1974. Biostatistical Analysis. Prentic Hall ' Hopkins University Press, Baltimore, MD. International Edition, Englewood Cliffs, NJ. s 1