The effects of host species and substratum factors on the abundance and growth of epiphytic orchids Anna Keller Department of Environmental Analysis, Pomona College _________________________________________________________________ ABSTRACT It has been observed that epiphytic orchids are less abundant on trees of the genus Quercus Fagaceae than o n other Neotropical hosts. This study investigates this claim and attempts to link that pattern to substrate inhospitability. The results of orchid abundance counts on Quercus and non Quercus hosts yielded statistically non significant results, although the data appear to support this apparent distribution. Specklinia aristata were placed in six experimental trea tments in order to examine the effects of pH, fissure depth, and water holding capacity of the substrate on orchid growth. The results of these experiments also yielded no statistically significant differences between orchid growth on conditions mimicking oak and non oak bark. However, average new leaf data implicates pH as the most inhibitory factor of orchid growth. RESUMEN Ha sido observado que hay menos orquÃdeas epiffititas que estÃ¡n en los Ã¡rboles del gÃ© nero Quercus Fagaceae que en otros tipos de Ã¡rboles neotropicales. Este estudio investiga esta declaraciÃ³n y trata de conectar ese modelo a la inhospitabilidad de la cortez a. La cantidad de orquÃdeas en Ã¡ rboles del gÃ©nero Quercus y los de otros tipos mostraron resultados estadÃsticos no significativ os, pero los datos parecen como si firmaran esta distribuciÃ³n. Specklinia aristata estuvieron sometidos en seis condiciones experimentales diferentes para examinar los afectos del pH, la profundidad de las hendiduras y la capacidad de retener agua en la c orteza con el crecimiento de las orquÃdeas. Los resultados de estas investigaciones tampoco mostraron una diferencia significativa entre el crecimiento de las orquÃdeas y las condiciones que representaban los elementos de la corteza de Quercus y otros tipo s de Ã¡ rboles tropicales. Sin embargo, los datos del nÃºmero de hojas nuevas sugiere que el pH puede ser un factor que inhibe el crecimiento de las orquÃdeas. INTRODUCTION Epiphytic orchid habitat preferences provide important information regarding orchid abundance and distribution. Neotropical cloud forests have been reported as a "hot spot" of orchid diversity, containing approximately 60% of all species in 2% of their natural territory Hagsater and Arenas 1997. In Costa Rica in general, Dressl er 1993 reports nearly 1200 species of orchids, and Atwood 2000 reports ten endemic species in the cloud forests of Monteverde a lone. This study investigates how host species affects epiphytic orchid abundance and distribution, and the effects of substrate char acteristics on orchid growth in
laboratory conditions. The limited abundance of vascular epiphytes on certain tree species demonstrates their uneven natural distribution Withner 1959 in Frei et al. 1972. Epiphytic orchids appear to be less abundant on Ne otropical trees of the genus Quercus oaks than on other canopy trees of similar geometry. Thorton 1998 reported zero orchids found on an individual Quercus brenesii Fagaceae in her survey of 15 trees in Monteverde, Costa Rica. The distribution, growt h rate, and abundance of epiphytic orchids on host trees may be affected by numerous factors such as substrate texture, water holding capacity, pH, nutrient content, toxins, and bryophyte cover Went 1940, Frei 1973a b, Johansson 1974 , Dressl er 1981, whic h could explain the apparent unsuitability of Quercus hosts. Frei and Dodson 1972 suggested that the unequal distribution of orchids on Mexican Quercus species is related to inhibitory substances and toxins found in the bark. Bark humidity and water hold ing capacity influence orchid abundance directly or indirectly through lichen and bryophyt e abundance. According to Dressl er 1981, lichen presence is often necessary for orchid germination. Correlations between bryophyte and orchid abundance can pot entia lly be explained by the bu ffer bryophytes may provide against bark toxins ter Steege and Cornelissen 1989. However, orchids that have been observed to typically occur with abundant bryophyte populations have a wide vertical distribution on humid trees su ch as Eperua grandiflora Fabaceae/Caesalpinoidea, and are exclusive t o the lower canopy on more xeric dry trees [ E. falcata Fabaceae/Caesalpinoidea], indicating the importance of substrate humidity in orchid success and potentially providing an alternate explanation for the correlations observed between bryophytes and certain orchid species ter Steege and Cornelissen 1989. Similarly, bark fissure depth positively affects bryophyte growth, as a significantly higher abundance of bryophytes is fou nd on host trees with deep fissures ter Steege and Cornelissen 1989. In addition, a study performed in Monteverde by Broadbent 1999 revealed increasing epiphytic Ericaceae species abundance with increasing bark fissure depth. Fissures have been reporte d to increase colonization by increased anchorage potential Pittendrigh 1948 in Broadbent 1999, and have been demonstrated to contain high levels of humus accumulation due to increased water and nutrient retention Ingram 1989 in Broadbent 1999. These s tudies suggest that water holding capacity and bark fissure depth are important physical determinants of orchid substrate preference. In addition, chemical factors, such as pH, may greatly affect orchid abundance and distribution. This study compares orchi d abundance and substrate properties of Quercus and other genera of Neotropical canopy trees, and experimentally investigates the effects of pH, water holding capacity, and bark fissure depth on the growth of one orchid species, Specklinia aristata Pleuro thallidinae. METHODS Study Sites This study was conducted at 1540m in Monteverde, Costa Rica on the properties of the EstaciÃ³n BiolÃ³ gica de Monteverde, the Galleria Extasis, and Richard LaVal. Ten pairs of trees, one Quercus and one non oak were chosen for orchid abundance counts Table 1. They were chosen on basis of similarity in height, diameter at brest height
DBH, and geometry. Initial measurements were conducted in order to create experimental conditions based on natu ral parameters. Height was estimated to the nearest five meters, and DBH was calculated by measuring the circumference of the tree and subsequently calcula ting diameter [circumference = Â€ diameter]. The tree pairs were compared with respect to their geome try of Zone 4 Figure 1, which was the area studied for abundance counts. Zone 4 was the focus of this study because it has been shown to demonstrate the highest species richness and abundance of epiphytes Thornton 1998, especially epiphytes with a wide height range ter Steege and Corn elissen 1989. For six pairs of oaks and non oaks, bark was collected from Zone 4 and its pH, bark fissure depth and water holding capacity was measured. From these measurements natural control conditions were created base d on the parameters found from the non oak values, and experimental conditions used in the manipulative experiment were based on the results of the oak trees. For initial measurements, pH was measured using ground up substrate from each of the 12 trees and a water pollution analysis kit. Bark fissure depth was determined by taking the average fissure depth from four measurements on each of the 12 hosts on Zone 4 branches, and measured using a caliper and a pin. Water holding capacity was calculated by takin g the difference in wet weight and dry weight of three samples of each substrate, yielding ml of water held. Each substrate was submerged in approximately two liters of water for three different time periods in order to correct for saturation rate: 15 minu tes, one hour, and two hours, and then massed using a laboratory balance. For the statistical analysis the value from the two hour condition was used to assure saturation of the bark samples. Orchid abundance was estimated by a simple count of individuals belonging to the family Orchidaceae. Abundance was sampled using binoculars and assessing orchids on approximately three meters of one Zone 4 branch of each tree, starting one meter out from the trunk. Each of the three laboratory experiments pH, fissure depth, and water holding capacity contained a natural control non oak and a experimental oak treatment based on initial field measurements. The sample size was ten Specklinia aristata Pleuothallid inae in each of the six conditions, 60 individuals total . In the pH treatment, the experimental group of orchids grew on Q. insignis pH=5.5 , and the control group were grown on A cnistus arborescens Solanaceae substrate pH=6.5. The two groups of Pleu rothallids in fissure treatments were grown on A. arborescens with varying mean fissure depth, 1.7 mm experimental and 2.85 mm deepened manually. The water holding capacity treatments were two differing thicknesses of A. arborescens, 1.3 mm experiment al and 3.4 mm. Experiments were set up in the laboratory following initial determinations of treatments, and watered 35 squirts of a spray bottle and monitored daily for 17 days, and then examined for new growth, determined by number of new leaves. In o rder to quantitatively compare the results of this study, a Mann Whitney U test was implemented for the initial pH, water holding capacity, and fissure measurements between the six pairs of oak and non oak hosts, and used to compare new growth for each of the three variables experimentally analyzed oak pH vs. non oak pH; oak fissure depth vs. non oak fissure depth, and oak water holding capacity vs. non oak water holding capacity. A Wilcoxon sign rank test was used to compare the difference in abundance counts between the ten pairs of oak and non oak hosts.
RESULTS From the initial measurements, the pH of the six Quercus individuals was significantly more acidic than that of the non oak trees, and the average pH of oak trees was 5.917, and 6.5 for non oaks Figure 2. The mean fissure depth was slightly shallower on the oaks, with a value of 2.921 mm compared to 3.304 mm on non oaks Figure 3. The water holding capacity of oak substrate was approximately half of that of non oak substrat e, with a mean value of .774 ml on oaks, and 1.277 ml for non oak substrate Figure 4. The statistical results of the Mann Whitney U tests yielded no significant p values <0. 05 except for the initial pH measurement of six Quercus vs. non oak pairs p v alue = 0.022. The results of orchid abundance counts yielded higher counts on non oak hosts for five out of the ten pairs observed Figure 8, although the difference in average orchid abundance on oak vs. non oak hosts was minimal average on oaks = 2.4; non oaks = 4.4. The results of the Wilcoxon sign rank test yielded an insignificant tied p value, even when the three pairs of hosts that contained zero orchids were omitted p value = 0.4982 in both cases. The effect of fissure depth on orchid growth was found to be inconsequential, and the mean number of new leaves was equivalent in both the shallow and deep fissure treatments Figure 5. The results from the water holding capacity treatments displayed a sl ightly higher mean of new leaves in the thin low water holding capacity condition. The average growth on thin substrate was 1.7 leaves, and on thick substrate it was 1.2 leaves Figure 6. The most striking result was from the pH experiments, which demon strated a mean of 0.8 leaves on non oak substrate vs. 0.3 on oak bark Figure 7. However, the statistical results of the Mann Whitney U tests yielded no significant p values <0. 05. DISCUSSION Orchid abundance in Monteverde suggests that Quercus trees are less suitable hosts than similar non Quercus Neotropical individuals. On five of the ten pairs of hosts observed for orchid abundance, non oaks consistently held more orchids than their oak counterparts Figure 8. A difference in physical characterist ics may provide explanatory evidence for this observed phenomenon. In natural conditions, oaks have significantly more acidic pH than non oak hosts, and generally contain shallower fissures and a lower water holding capacity . In the initial measurements it was established that Quercus substrate had a significantly more acidic pH than non oak substrate. The greatest experimental difference in Pleurothallid growth was between the oak pH mean new leaves = 0 . 3 and non oak pH mean new leaves = 0. 8. Although this is not a statistically significant r esult, it implicates pH as a possible inhibitory factor of orchid growth on Quercus individuals. The combined laboratory and field results suggest that despite differences in other physical parameters such as water holding capacity and fissure depth, that pH is really the only inhibitory factor in orchid growth. Water holding capacity results demonstrated a difference in orchid growth, although it was in favor of oak conditions thin bark. The difference in mean gr owth was minimal mean new leaves on thin bark = 1.7, thick bark =1.2, and may suggest that other physical substrate factors have a greater effect. Both conditions in this experiment produced the highest number of new leaves overall 29 leaves total, sug gesting both were favorable treatments and that other factors are more influential, or
perhaps the difference indicates that the thick bark retained too much water. According to Dressl er 1981,. orchids are unable to tolerate extremely wet substrates for a long period of time as this causes the roots to rot, but instead soft, spongy bark with a rough surface is the most appropriate for adequate water retention and is able to provide places for seeds to lodge and germinate. Fissure depth had no observable effect on Pleurothallid growth in this study. However, fissures have been implicated in orchid growth due to increased colonization by increased anchorage potential Pittendrigh 1948 in Broadbent 1999 and by providing favorable germination conditions. In this experiment, the orchids were physically attached to the substrate and neither anchorage potential nor germination conditions were taken into consideration. The orchid abundance results may not be statistically significant due in part to pair #3 Tabl e 1, on which there were 14 orchids on the oak host and only two on the corresponding E. paniculata Figure 8. Orchid abundance was only higher on Quercus in this and one other pair, in which the Q. insignis host held six and the O. vestitus held only a single orchid. These results were atypical and could possibly be explained by a favorable location of the host or physical characteristics of the non oak host substrates. There were also three pairs in which neither oak nor non oak held any orchids. In two of these cases the trees were located near a road from which they received many pollutants and were more susceptible to high winds and other edge effects. In the third case both trees had a DBH less than 20 cm, most likely indicating a young age. I n this situation low orchid abundance could be explained by the fact that orchids had not had time to establish on younger trees. Based on previou s research on p hysical and chemical composition of substrates, it was expected that the S . aristata would be affected by the pH, water holding capacity, and fissure depth of the bark, and would demonstrate overall lower growth in the oak experimental treatments. Since these hypotheses were not statistically supported, there may be other physical substr ate factors influencing epiphytic orchid growth on Quercus hosts. These results may signify that the apparent uneven distribution of orchids on oak trees in the Monteverde area is purely by chance, and that there are no obvious inhibitory qualities that oc cur on members of the genus Quercus and not on other canopy trees of similar geometry. However, the raw data suggest support for previous observations that orchids are less abundant on oak hosts. In addition, there were several confounding variables that c ould explain these statistical results and provide a basis for future experimentation. Overall the non significant statistical results found in this study can be attributed to small sample size and the short duration of experimental treatments due to time constraints. Future experiments conducted on this topic would provide insight into the reasons behind the observed uneven distribution of orchids on oak hosts, and should include a larger sample size and increased exposure to experimental treatments in ord er to determine whether this observation is actually a naturally occurring phenomenon. In addition, future research on other physical substrate factors and their effects on orchid growth may provide a more detailed account of orchid habitat preferences. ACKNOWLEDGEMENTS I would like to thank Karen Masters for her guidance and her attraction to orchids that provided an irreplaceable inspiration for this study and an everlasting interest in the beautiful members of the family Orchidaceae. I would also like to thank the teaching assistants, Will Wieder for his plant identification assistance, shared interests and inspiration; and Andrew Rodstrom for his oak identifying abilities, without which this study would not have been possible. I would also like to than k Jeffrey Sanders and Christopher
Frohlich for their physical branch reaching abilities that facilitated the collection of fissure measurements. Quisiera dar gracias a Mauricio GarcÃa tambiÃ©n para su ayuda con mi espaÃ±ol en las Ãºltimas horas de escribir. F inally, I would like to thank the EstaciÃ³ n BiolÃ³ gica de Monteverde and all of the people involved with CIEE for allowing and facilitating this study. LITERATURE CITED Atwood, J. T. 2000. Orchids. In Nadkarni, N. M. and N. T. Wheelwright Ed. Monteverde: ecology and conservation of a tropical cloud forest; P.74 75. Oxford University Press, New York. Broadbent, E. 1999. Community structure of ericaceous epiphytes in secondary forest Patches of lower montane wet forest habitat. CIEE Spring 1999. Dressl er, R. L. 1981. The Orchids: Natural History and Classification. Harvard University Press, Cambridge. Dressl er, R. L. 1993 Field Guide to the Orchids of Costa Rica and Panama. Comstock Publishing Associates, Ithaca. Frei, S. J. K.and C. H. Dodson. 1972. T he chemical eff ect of certain hark substrates o n the germination and early growth of epiphytic orchids. Bulletin of the Torrey Botanical Club 99 6: 301 307 Frei, S.J.K. 1973a. Orchid ecology in a cloud forest in the mountains of Oaxaca, Mexico. Amer. Orchid Soc. Bull. 42: 307 314 Frei, S. J.K 1973b. Effect of hark substrate on germination of Encycl tampens is seeds. Amer. Orchid Soc. Bull. 42:701 708 Hagsater, E. and M. A S. Arenas. 1997. Orchid Conservation in Mexico. Orchid Conservation Proceedings, Marie Selby Botanical Gardens, Sarasota, Florida. Ingram, W. I. 1989. The abundance, vegetative composition and distribution of epiphytes in a Costa R ican lower montane r ain forest. Masters thesis, University of California, Santa Barbara, California. Johansson, D. 1974. Ecology of vascular epiphytes in West African rain forest. Acta Phytogeogr. Suec. 59: 1 136 Pittendrigh, C. S. 1948. The bromeliad anopheles malaria complex in Trinidad, I. The bromeliad flora. Evolution. 2:58 89 Richards, P. W. 1964. The ecology of tropical bryophytes. R. M. Schuster New Manual of Bryology 2: 1233 1270 Ruinen, J. 1953. Epiphytosis. A second view on epiphytism. Ann. Boqariensis 1:101 157 t er Steege, H. and J.H.C. Cornelissen. 1989. distribution and ecology of vascular epiphytes in lo w land rain forest of Guyana. Biotropica 21 4: 331 339 Thornton, H. 1998. Effects of host tree size and habitat heterogeneity on epiphytic orchid Diversity . CIEE Spring 1998. Valdavia, P. E. 1977 . Estudio botÃ¡nico y ecolÃ³gico de la regiÃ³n del Rio Uxpanapa, Veracruz 4: Las epifitas. Biotica 2: 55 81. Went, F.W. 1940. Soziologie eines tropischen Urwaldes. Ann. Jard. Bot. Buitenz. 50: 1 98 Withner, C.L. 1959. The orchids, a scientific survey. The Ronald Press Company, New York.
TABLE 1 : Host species studied organized by pair. Pairs 1, 2, 5, 6, 7, and 9 were used in initial measurements . Pair Oak Host Non oak Host 1 Quercus insignis Fagaceae Oreopanax vestitus Araliaceae 2 Quercus brenesii Fagaceae Cinnamomum tonduzii Lauraceae 3 Quercus brenesii Fagaceae Exothea paniculata Sapindaceae 4 Quercus brenesii Fagaceae Viburnum costaricanum Caprifoliaceae 5 Quercus brenesii Fagaceae Oreopanax xalapensis Araliaceae 6 Quercus brenesii Fagaceae Daphnopsis Americana Thymelaeceae 7 Quercus brenesii Fagaceae Oreopanax xalapensis Araliaceae 8 Quercus brenesii Fagaceae Ocotea whitei Lauraceae 9 Quercus brenesii Fagaceae Citharexylum costaricens is Verbenaceae 10 Quercus brenesii Fagaceae Citharexylum cos taricensis Verbenaceae
FIGURE 1. The vertical zonation of trees as described by Joha nsson 1974. Zone 4 was the fo cus of orchid abundance counts in this study. FIGURE 2 . Mean pH [and Standard Deviation bars SD] of oak and non oak hosts.
FIGURE 3 . Average fissure depth of oak and non oak hosts and SD bars. Fissure depths taken from four measurements on a single Zone 4 branch of each individual. Mann Whitney U test yielded a p value of 0.4233. FIGURE 4 . The average water holding capacity of oak and non oak hosts and SD bars. The water holding capacity was measured after being submerged in water for two hours. Mann Whitney U test yielded a p value of 0.87 28.
FIGURE 5 . Mean orchid growth and SD bars on A. arborescens substrate with different fissure depths shallow = 1.7 mm; deep = 2.85mm. Mann Whitney U test yielded a p value of > 0.9999. FIGURE 6 . Mean new orchid growth and SD bars on differing thicknesses of A. arborescens substrate thin = 1.3 mm. Thickness was used to mimic differing substrate water holding capacities. Mann Whitney U test yielded a p value of 0.5488.
FIGURE 7 . Mean new orc hid growth and SD bars on substrates of Q. insignis oak pH = 5.5 and A. arborescens non oak pH = 6.5. Mann Whitney U test yielded a p value of 0.3039. FIGURE 8 . Orchid abundance on oak and non oak hosts, organized by pair. The mean orchid abundanc e was 2.4 on oaks and 4.4 on non oaks + 1 SD. N = 20 individuals 10 oak and 10 non oak hosts.