|USFLDC Home | Search all Groups||| RSS|
This item is only available as the following downloads:
Liana size class diversity across three forest habitats Janelle Burke Department of Behavioral Biology, John Hopkins University __________________________________________________________________ ABSTRACT The purpose of this study was to see how forest edge, primary and secondary habitats affect average size and size class diversity of lianas in a cloud forest in Monteverde, Costa Rica. Three 625m 2 sites were studied in primary and secondary forests, and along the edge of a forest. Exhaustive sampling of lianas was taken of four continuous plots within each habitat: all individuals were measured for PBH (perimeter at breast height). The primary forest was the most diverse according to the Shannon Weiner index of ing by mean perimeter (3.87 0.34cm). In terms of significant differences in diversity indices, the secondary habitat differed significantly from the edge and primary habitats, while the primary and edge did not. The mean age of the primary varied signifi cantly from the other two habitats, while the secondary and edge did not show significant variation. Furthermore plot 1 varied significantly from plot 3 and 4 within the primary habitat. These results are probably a consequence of the ecotone qualities of the edge site and the small scale heterogeneity of disturbance history within the primary site. RESUMEN Examin diferencias entre tamao y edad de las lianas en un bosque nuboso en Monteverde, Costa Rica. El propsito de este estudio era para ver como el crecimiento del borde, primario, y secundario se afecta diversidad de la clase del tamao. Tres sitios de 625m 2 fueron estudiados en los bosques primarios y secundarios, y a lo largo del borde de un bosque. El muestreo exhaustivo de lianas fue tomado de cuatro diagramas continuos dentro de cada sitio por permetro en la altura del pecho (PBH). El bosque primario er a el ms diverso segn el ndice de Shannon permetro del medio (3, 87 0,34 cm). No todos los sitios variaron en una manera significativa a cada uno en trminos de la clase de edad de la diversi dad y del tamao. Esta es probablemente una consecuencia de las calidades como un ecotone del sitio el borde y la heterogeneidad en escala pequea de la edad dentro del sitio primario. INTRODUCTION Lianas are a plant growth form that peak in species dive rsity in the tropics. They are a large part of the composition of tropical forests, contributing to the cloud forest integrity. It is not uncommon for lianas to produce 40% of the leaves and 25% of woody stem density (Haber 2000; Laurance et al. 2001; Schi ntzer & Bongers 2002). High transpiration rates also make them an important tropical nutrient cycler (Gerwing & Vidal 2002). These woody vines are adapted to ascend trees before sprawling across the
canopy, where they damage trees by weighing them down, bl ock sunlight and absorb valuable nutrients from their hosts. Once a liana establishes itself on a tree, it provides an easier pathway for other climbers to wrap around (Forsyth & Miyata 1984), thus affecting forest structure by escalating even more tree da mage. More studies are focusing on liana growth at forest edges due to increased fragmentation created by deforestation in tropical forests (Laurance et al. 1998). The study of edges and their effects is essential to conservation because much of the remai ning forest has a high edge to interior ratio. Abiotic changes on the edge include increased exposure to wind and sunlight. These conditions cause changes in species composition, population dynamics, and a rise in invasive species. While both changes in po pulations and abiotic factors lead to increased tree mortality and damage (Pimm 1998). Lianas may compound edge effects by increasing tree mortality and damage already caused by increased exposure to wind and sun. Lianas are similar to pioneer species in their response to disturbances (Schintzer et al. 2000). Laurance et al. (1998) states that they are "light loving, and respond well to forest disturbances." It i s now well documented that liana abundance increases near forest edges, proliferating not only in numbers, but also in species richness (Haber 2000; Schnitzer & Bongers 2002). Liana abundance correlates negatively with tree biomass, showing they are likely to impede the regeneration of trees at forest edges (Laurance et al. 2002; Schintzer et al. 2000). Not only the abundance, but also the sizes of lianas vary across habitats. A tangle of small lianas is an indicator of a forest edge, while lianas with larg e diameter are indicative of an older growth forest (Laurance et al. 2001). Though many studies have focused on species diversities within different forest types, there are not many studies documenting liana size class diversity (Gerwing & Vidal 1998). Fo rest fragments are known to lose species richness, and may also lose liana size class diversity (Pimm 1998). Classifying lianas by size may be another way to quantify diversity, age, and overall health of a forest. I hypothesize that the diversity in size classes will vary between primary, secondary and edge habitats. I predict that class richness and evenness will be highest in primary forest, due to old age of growth and limited disturbance effects. I predict that secondary forest and edge affected fragments will exhibit lower diversity as well as greater stem abundance due to the younger growth and increased disturbance conditions. The age differences, as determined by liana size, between the forest types will be distinct. Thus liana site diversity may be a way to quantify the state and age of a forest. METHODS This month long study, from April May 2003, was conduct ed on the property of La Estacin Biol gica Monteverde, Puntarenas, Costa Rica. This area is describ ed as Lower Wet Forest with an altitude of 1500 m (sensu Holdridge). Data were collected from April May 2003. HABITAT SELECTION Three habitat sites, each 625 m 2 were chosen based upon forest growth type and accessibility. The primary forest site was selected 700 m behind the station.
The secondary forest sample was 400 m behind the station (to avoid edge effects which can penetrate 300 m into the interior). The edge habitat was chosen alongside a pasture near the station. Each habitat was divided into four smaller continuous plots. The secondary plots were 12.5 x 12.5 m. Edge and primary habitats were rectangular and had plots arranged in linear 10.5 x 15 m segments. LIANA SAMPLING An exhaustive census of all lianas I each plot and habitat was taken. Liana perimeter was measured in centimeters at breast height and marked to prevent redundancy. Vines and hemiepiphytes were disregarded. Then size classes were assigned to each liana to all ow am ample amount of classes. Lianas were classified by size class starting from <1 cm, and going up by increments of 1 cm until >20 cm. Lianas over 20 cm were considered a class of their own. STATISTICAL ANALYSES A Shannon Weiner index calculated H' (di versity) and E (evenness) between sites (Zar 1984). A pairwise comparison statistically compared H' values for differences. A one way ANOVA compared the average sizes of each habitat and the plots within the primary forest habitat. RESULTS A total of 400 lianas were measured: 120 in the edge, 142 in the primary and 138 in the edge site (Table 1). Pairwise comparisons of H' using a modified t test showed a significant variation of liana size class diversity between primary and secondary forest and significa nt difference between secondary forest and edge (Figure 1). A one way ANO VA showed a significant difference in mean perimeter of lianas between the primary forest and the edge and the primary and secondary forest (Figure 2). This measure of mean age variance was also showed a significant difference between the plots 1: 3 and 1 :4 of the primary forest (Figure 3; Fisher's PLSD p = 0.0141; p = 0.0143). Within the primary site, the H' values also varied between plots: plot 1 = 1.07,2 = 0.790, 3 = 0.62 6, and 4 = 0.784. The Edge and secondary plots did not show significant variation between plots (One way ANOVA, secondary p = 0.4258, edge p = 0.8197). DISCUSSION This study investigated the differences in liana size class diversity and mean liana age ac ross three habitats. The size class diversity of the secondary forest was significantly lower than the edge and primary as expected, however edge and primary forest diversity did not differ significantly. The primary habitat had a higher mean size than did the other two habitats, and it also varied between plots within the site due to small scale heterogeneity. I expected there to be a significant difference in the diversity of size classes between primary and edge habitats because of the differences in abiotic factors of lianas varie s across habitats (Ibarra Manr quez & Martinez Ramos 2002). Instead, I found that there was no significant difference between the two (Figure 1). This lack of difference may be due to the
history and conditions of the habit at sites chosen. The edge disturbance created by fragmentation allows ideal conditions for liana proliferation due to increased sunlight because this disturbance occurs on a short temporal scale; the new proliferation of lianas will be the same age. This is true for the clear cutting regrowth of the secondary forest as well. I thought this would create a lower evenness in the edge due an abundance of lianas within the same size class. The edge hab itat exhibited larger perimeters of lianas within more size classes than expected (Figure 2). The presence of large lianas are usually only found in old growth forests due to their slow annual growth in diameter (Gerwing & Vidal 2002). The interior of the forest experienced another type of disturbance as well: selective logging. Ninety percent of lianas survive tree falls, indicating they could survive selective logging (Schnitzer & Bongers 2002). A few large trees remained standing, and the lianas along wi th them. Therefore the interior of the forest had more of a habitat of a closed, older growth forest while the edge is more open with abiotic characteristics of a pasture. The influence of forest types within the edge habitat makes it an ecotone. The edge ecotone will show the limited size classes like the disturbed secondary forest, and larger lianas like the primary forest. The mean liana perimeter in each habitat can indicate forest age. The results show the mean for the secondary was lowest, followed by edge and primary (Figure 2). Recent disturbance will lower the overall age of the habitat. The edge and secondary means were not statistically distinct, presumably due to the high levels of disturbance at both sites. The secondary habitat, which had been clear cut in the past, showed the lowest age, while the edge had the second highest mean perimeter, probably due to the ecotone qualities of the habitat. Though there was a recent disturbance, the interior of the forest had older sectors. As expected, the primary forest reported the highest mean perimeter since it is the oldest. I assumed that age within habitats would be uniform; however between plot age comparisons revealed inter site heterogeneity in the primary site (Figure 3). Plot one showed the highe st mean perimeter, indicating the oldest forest. This plot showed significant variation from plots 3 and 4. This shows subtle differences in age within the primary forest mosaic. This is an important contrast to the secondary plot, which all plots had been going through succession on the same temporal scale. In the past, the primary plots probably had limited impact disturbances on a smaller scale, which created this heterogeneity in age. The data did now show a high abundance of lianas on the edge This may be explained by the observed presence of other climbers. The edge also shows a proliferation of vines and secondary growth (Laurance et al. 1998). Smilax, aroids and other vines were very common and created a tangled, dense understory. I observed an in creased percentage of trees with some type of climber. A lower instance of lianas than expected could be due to the abundance of vines. The vines fill the same niche by responding the same way to disturbances and using the same growth form. This may impede the growth of lianas if their growing substrates are already filled with vines. Furthermore, level of forest disturbance and tree biomass are other predictors of liana abundance, not just distance from edge (Laurance et al. 2001). The disturbance within t he forest interior or the density of trees could also play a role in the abundance of lianas in each habitat.
In conclusion, variations in liana size class diversity, age and abundance between habitats have wider implications for the composition of the fo rest. The proliferation of lianas of the same size may cause increased tree damage with succession. Over time, the older secondary forest will probably have a greater instance of larger perimeter lianas. This effect may be worse in the edge disturbance due to the thicker liana sizes in the interior and smaller ones caused by the disturbance. These liana distributions will limit light resources for the regeneration of trees. They may increase the frequency of tree fall disturbances with their added weight. A more heterogeneous distribution of liana sizes, as exhibited by the primary habitat, is probably beneficial for tree productivity because the trees will not be laden with lianas within the same size class. Overall, the disturbances created by edge effect s are beneficial to liana species diversity, but they may inhibit species diversity and regeneration of other growth forms. ACKNOWLEDGEMENTS I would like to thank my advisor Karen Masters for giving me guidance throughout this project. I would also like to thank Rick Fr ancis Xavier Smith V and Andrew Rodstrom for much random support. Eleonore Wesserle provided me with data entry in my time of need. La Estacin Biolgica for the use of their land. And lastly I would like to thank my parents, because withou t them I would never be able to come on this program. ___________________________________________________________________________ LITERATURE CITED Forsyth, A. and K. Miyata. 1984. Tropical Nature. Simon and Schuster, New York, pp. 48 50. Gerwing, J.J. and E. Vidal. 2002. Changes in liana abundance and species diversity eight years after liana cutting and logging in an eastern Amazonian forest. Conservation Biology 16 : 544 548. Haber, W. 2000. Plants and Vegetation. In N.M. Nadkarni and N.T. Wheelwright (Eds.) Monteverde: Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, New York, pp. 40,54. Ibarra Manrquez, G., and M. Martnez Ramos. 2002. Landscape variation of liana communities in a Neotropical rain forest. Plant Ecology 160 : 91 112. Laurance, W.F., L.V. Ferreira, J.M. Rankin de Merona, and S.G. Laurance. 1998. Rain forest fragmentation and the dynamics of Amazonian tree communities. Ecology 79 : 2032 2040. Laurance, W.F., T.E. Lovejoy, H.L. Vasconcelos, E.M. Bruna, R.K. Di dham, P.C. Stouffer, C. Gascon, R.O. Bierregaard, S.G. Laurance, and E. Sampaio. 2002. Ecosystem decay of Amazonian forest fragments: a 22 year investigation. Conservation Biology 16 : 605 618. Laurance, W.F., D. Prez Salicrup, P. Delamnica, P.M. Fearnside, S. Pohl, and T.E. Lovejoy. 2001. Rain forest fragmentation and the structure of Amazonian liana communities. Ecology 82 : 105 116. Pimm, S.L. 1998. The forest fragment classic. Nature 393 : 23 24. Schnitzer, S.A., an d F. Bongers. 2002. The ecology of lianas and their role in forests. Trends in Ecology and Evolution 17: 223 229.
Schnitzer, S.A., J.W. Dalling, and W.P. Carson. 2000. The impact of lianas on tree regeneration in tropical forest canopy gaps: evidence for a n alternative pathway of gap phase regeneration. Journal of Ecology 88 : 655 666. Zarr, J.H. 1984. Biostatistical Analysis, Second Edition. Prentice Hall, Inc., Englewood Cliffs, N.J. pp. 144 147.