REVISTA DE BIOLOGIA TROPICAL 1022 TABLE 1 Percent of Life Zones wherespecies offive bromeliad genera were expected (according to the Holdridge Life Zone System) but were not found. Genus Mean Tillandsia Vriesea Aechmea Catopsis Guzmania 47 51 51 47 47 * Sample size in number of species. Standard deviation 17 15 9 13 17 1988). The number of specimens per Life Zone Maximum Minimum Sample size* 85 75 67 64 82 17 14 36 29 12 31 24 12 6 22 DISCUSSION were tabulated by species and distribution maps were copied to smaller and simplified maps for Biogeographic studies based on herbarium visual examination. To assess agreement records are subject to collection bias. In this between the Holdridge Life Zone System and case, however, the extensive Costa Rican road bromeliad occurrence, the percent of zones system has made all of the Life where each species was expected but has not accessible to coHectors and an examinatíon of been collected was calculated. A species is Fig. 34 suggests that we have a reliable view of expected to occur in life zone "B" if it occurs in bromeliad occurrence across the country. Zones zones "A" and "C" that are immediate to it in The greater number of records for montane the Holdridge classification, because zone "B" habitats, previously noted by Gómez (1986) can is expected to have intermediate characteristics. be explained by the need for moisture al.1d light Here only the genus means (and related that is common in epiphytic bromeliads (Picado summary statistics) of those percents are 19 1 1, 19 13, Benzing 1994). Relatively low presented (detailed species tables are available diversity in NW Costa Rica was reported earlier from the authors). by Burt-Utley and Utley (1975), and attributed to the prolonged dry season there. Moisture RESULTS supply strongly influences the distribution of Bromeliaceae in broad (regional) and finer A total of 7 15 valid records spanning almost a century were found (Fígs. 1-34). Most records of Costa Rícan epiphytic bromeliads are from the mid-altitudes along the mountainous (local) scales (Gómez 1986, Brown 1990, Gómez and Winkler 1991, Benzing 1994). Sorne Vriesea are known to require high humidity, low líght levels and thin phorophyte backbone of the country (Fig.3 4). When branches (Brown 1990, Fontoura 1995) which numbers of records are pooled for genera agrees with the over-representation of the genus (because not enough data are available for in moist Costa Rican highlands (the same may individual species) Vríesea and Guzmania are apply to Guzmania). more frequent in moist highlands (Fig. 33), Tillandsia and Catopsi s peak at varíous The need for high humidity characterizes many species of Ti l l a ndsia, but this altitudes (Fig. 33) and Aechmea most often ecophysiologically diverse genus has species inhabits moist lowland habitats (Hg. 33). with very contrasting habitat requirements The percents of Life Zones where each (García-Franco and Peters 1987, Brown 1990) species was expected (see Materials and which may explain why abundance varied Methods) but has not been recorded had a mean greatly with Life Zone in this study. Finally, the of 50 % (Table 1), that is, there are no records apparent exclusion of most Aechmea for most species in about half the Life Zones open habítats (Brown 1990) is also consistent from where they should occur according to the with our results, except for specíes such as Holdridge classification. Aechmea mexicana. Rivas el al: Distribution of Costa Rican epiphytic bromeliads 1023 2 � TROPICAL • T1LLANDS/A A COSTA E TROPICAL-PREMONTANE • T/LLANDS/A A DPRESSA PREMONTANE MONTANE e TlLLA NDS/A A NCEPS O PARAMO TlLLANDS/A BALB/SIANA • T/LLA N DS/A BUTZII • T/LLANDSIA BRACHYCAULOS • e T/LLANDSIA EXCE LSA e TlLLANDS/A F/LlFOLlA O 6 T/LLANDS/A /NS/GNlS • TlLLANDS/A BULBOSA • • TlLLANDS/A FASC/CULATA TlLLANDS/A CAPUT-MEDUSAE e TlLLANDS/A COMPLANATA o TlLLANDS/A CONTORTA TlLLANDS/A FES TUCO/DES TlLLANDS/A FLEXUOSA • TlLLANDS/A JUNCEA • TlLLANDS/A LE/BOLDIANA e TlLLANDS/AMONADELPHA o TlLLANDS/A MULTlCAULlS Figs_ 1-6. Costa Rica: GeneralLife Zones (1) and localities where several species ofTillandsia havebeen collected (2-6).ln this and following figures asterisks indicate data from the literature (see main text). REVISTA DE BIOLOGÍA TROPICAL 1024 8 • TlLLANDSIA PAUClFOUA • TlLLANDSIA PRUINOSA IJ TlLLANDSIA PUNCTULATA O TlLLANDSIA ROTHSCHUHIANA • • IJ O TILLANDSIA SCHIEDIANA TILLANDSIA SINGULARIS TILLANDSIA SPICULOSA TlLI:ANDSIA SUBULlFERA 10 • TlLLANDSIA USNEOIDES • TlLLANDSIA UTRlCULATA IJ TlLLANDSIA VENUSTA O TlLLANDSIA COMPRESSA • AECHMEA MARIAE-REGINAE • AECHMEA MEXICANA e AECHMEA NUDlCAUUS o AECHMEA PlmERI • AECHMEAANGUSTlFOUA • AECHMEA BRACTEATA IJ AECHMEA DACTYLlNA O AECHMEA MAGDALENAE • AECHMEA PUBESCENS • AECHMEA TlLLANDSIOIDES IJ AECHMEA TONDUZlI O AECHMEA VElTCHII Figs. 7-12. LocaIities where several species oC Tillandsia and Aechmea have been collected. - REVISTA DE BIOLOGÍA TROPICAL 1026 20 111 GUZMANlA MUSA/CA • GUZMANlA NlCARAGUENS/S iJ GUZMAN/A OBTUS/LOBA O GUZMANlA PATULA III GUZMANlA PLlCATlFOLlA • GUZMANlA SANGUINEA ¡;¡ GUZMANlA SCI-IERZERIANA O GUZMANlA SPECTABIUS 21 11 GUZMANIA STENOSTACHYA • l!II VRIESEA ACUMINATA GUZMANIA SUBCORYMBOSA • ¡;¡ GUZMANIA ZAI-INIJ e o 11 • VRIESEA COMATA VRIESEA CI-IONTALENS/S e VRIESEA DIFFUSA o VRIESEA GLAD/OLlFLORA VRlESEA APICULATA VR/ESEA ATTENUATA VRlESEA BICOLOR iII VRIESEA GRAMINlFOLlA • VRIESEA GREENBERGI/ !J VRIESEA I-IEUCONlOIDES e VI�/ESEA HYGROMETRICA Figs.l9-24. Localities where severa! species of Guzmania and Vriesea have been collected. - IUvas et al: Distribution of Costa Rican epiphytic bromeliads 1027 26 ID VRIESEA INCURVA iD 111 VRIESEA LEPTOPODA VRlESEA KATHYAE e VRIESEA KUPPERIANA o VRIESEA LAnSSIMA iD VRIESEA LEUCOPHYLLA e VRIESEA LYMAN-SMITHI/ o VRIESEA MARNIER-LAPOSTOLLEI 18 VRIESEA PEDlCELLATA 11 VRIESEA MONSTRUM • VRIESEA NEPHROLEPIS e VRlESEA NOTATA o VRIESEA PlmERI e VRIESEA SANGUINOLENTA VRIESEA SUBSECUNDA VRIESEA ORORIENSIS 30 III! VRIESEA TONDUZlANA iD e o VRIESEA TRIFLORA lIi • e VRIESEA UMBROSA VRIESEA VIRIDIFLORA o VRiESEA VIRIDlS VRIESEA VITTATA VRIESEA WERCKLEANA VRIESEA WlLLlAMSII Figs. 25-30. Localities where severa! species of Vriesea have been collected. 1028 REVISTA DE BIOLOGÍA TROPICAL Figs. 31-32. Additionallocalities where several species of Vriesea have been collected (al1 from the literature). Aechmea 1 castelnavii, 2 penduliflora; Catopsis 3 mutans, 4 werkleana. 5 Guzmania milis; Tillandsia 6 abdita. 7 biflora, 8 cal/lif/ora. 9 lampropoda. 10 longifolia, 11 makoyana, 12 oerstediana, 13 tricolor; Vriesea 14ampla, 15 bala/lophora, 16 brunei, 17 burgeri, 18 picla, 19 ringens, 20 stenophyla. - REVISTA DE BIOLOGíA TROPICAL 1030 ,. . ' " .. , . .. . .. " · " • . " l. . . .-: .. r :� .. • . . : . :·AI 0• .. :t .. . .. .� ... l.� · ,. • .... . ::t; :. ". .. . ... . 0. ,.,:;.. ·-" �V. ·.... I . . ' . � :.\... ., - .. . . . . \ .' '.. . :.'. ,- . . Fig. 34. Sununary of Costa Rican bromeliad locaIity records considered in tbis study. Rivas et al: Distribution of Costa Riean epiphytic bromeliads Individual species were absent in about half of the expected Life Zones: light, humidity and temperature, i. e. the factors used by the Holdridge life System:, faíl to fully predict bromeliad occurrence. Assuming that sampling is consistent across Life Zones, this study rejects the hypothesis proposed by Burt-Utley and Utley (1975), even for Vriesea, which in their opinion followed the Life Zone ' classification System closely (Burt-Utley and Utley 1975). G6mez (1989) stated that Holdridge's Life Zone Ecology map of Costa Rica was not intended to be a vegetation map but this fact does not justify the System's failure because bromeliads are known to depend on environmental factors considered b y the system (Burt-Utley and Utley 1975). Future workers may test the b r omeliad hypothesis distribution that epiphytic reflects factor interactions rather than humidity, light or temperature individually. This project was financed by Centro de Investigaci6n General, UNED. We thank Mariela Bermúdez, Luis D. G6mez and B. for authorizing study cuales se les esperaba según la clasificación de Holdridge. Se presenta la hipótesis de que el sistema falló porque la distribución de las brolilelias epifitas depende de las interacciones entre factores, más que de humedad, luz o temperatura individualmente. REFERENCES Benzing, D.H. 1994. How much is known Bromeliaceae in 19947 Selbyana 15: 1-7. about BrowIi, AD. 1990. El epifitismo en las selvas montanas del Parque Nacional "El Rey", Argentina: Composición florística y patrón de distribución. Rev. Biol. Trop. 38: 155-166. Burt-Utley, K. & J.F. Utley. (1975). Supplementary notes: Phytogeography, physiological ecology and the Costa Rican genera of Bromeliaceae. Historia Natural Costa Rica 1: 9-29. Fontoura, T. 1995. Distribution pattems of five Bromeliaceae genera in Atlantic rainforest, Rio de Janeiro State, Brazil. Selbyana 16: 79-93. García-Franco, J.G. & C.M. Peters. 1987. Patrón espacial y abundancia de Tillandsia spp. a través de un gradiente altitudinal en los altos de Chiapas, México. Brenesia 27: 35-45. Gómez, M.A. & S. Winkler. 1991. Bromelias en manglares del Pacifico de Guatemala. Rev. Biol. Trop. 39: 207214. Gómez P., L.D. 1986. Vegetación de Costa Rica. Apuntes para una biogeografía costarricense, p. 1-328. In L.D. Gómez P. (ed.). Vegetación y clima de Costa Rica, vol. 1. Universidad Estatal a Distancia, San José, Costa Rica. ACKNOWLEDGMENTS Hammel 1031 of their institutional collections, Luis Fournier O., Luis D. G6mez P. and two anonymous reviewers for valuable suggestions to improve an earlier draft, and Martin Law Romero for his assistance. RESUMEN Se preparo mapas detallados de distribución para los principales géneros de bromelias epifitas nativos de Costa Rica, con base en las colecciones de los tres principales herbarios del país y de infonnes en la literatura. La mayoría de los registros corresponde a los hábitats montaiiosos, lo cual probablemente refleja la necesidad de humedad que tienen estas plantas. Vriesea y Guzmania han sido recolectadas principalmente en regiones altas y húmedas; Tillandsia y Catopsis muestran picos de abundancia en varias altitudes y A e c hmea está más restringido a los hábitats ht1medos de bajura. En 1975, Burt-Utley y Utley habían sugerido la hipótesis de que el Sistema de Zonas de Vida de Holdridge debería calzar con la distribución de las bromelias epifitas porque el suelo (un factor no considerado por el sistema) no es importante en su presencia. Sin embargo, en promedio, las especies estuvieron ausentes en aproximadamente la mitad de las zonas de vidas en las Gómez P., L.D. 1989. Costa Rica, p. 305-308. In D.G. Campbell & D. Hammond (eds.). Floristic inventory of tropical countries. NY Botanical Gardens, New York. Herrera, S.W. & L.D. GÓmez. 1993. Mapa de unidades bióticas de Costa Rica. Incafo, San José, Costa Rica. 1:685000. Holdridge, RL. 1967. Life zone ecology. Tropical Science Center, San José, Costa Rica, 206 p. Holmgren, P.K., W. Keuken &E. K. Schofield. 1981. Index herbariorum. I. W. Junk, The Hague. 417 p. Luther, H.E. 1995. An annotated check1ist of Bromeliaceae of Costa Rica. Selbyana 16: 230-234. Picado, C.. 1911. Les broméliacées épiphytes comme milieu biologique. Compt. Rend. Acad. Sci. 153: 960963. Picado, C .. 1913. Les broméliacées épiphytes considérées comme milieu biologique. Bull. Sci. France Belgique 47:1-398. Tosi, JA 1967. Mapa ecológico: República de Costa Rica según la clasificación de zonas de vida del mundo de L.R. Holdridge. Centro Científico Tropical, San José, Costa Rica (map). Tosi, J.A. 1988. Mapa de las zonas de vida de Costa Rica. Centro Científico Tropical, San José, Costa Rica (map, 1:200 000). Utley, J.F. 1994. Bromeliaceae, p: 89-156. In G. Davidse (ed.). Flora mesoamericana, vol. 6. Universidad Nacional Autónoma de México, México, D.F.