Site location, administration, and scientific infrastructure: 0° 41' S, 76° 24' W. The study site is located in mature forest on terra firme in Yasuní National Park and Biosphere Reserve. The park and the adjacent Huaorani territory comprise 1,600,000 ha, representing the largest protected area of mature forest in the Amazon region of Ecuador. The 50-ha Yasuní Forest Dynamics Plot is located in the northwestern corner of the park, on a ridge above the Tiputini River, a tributary of the Napo.

Topography and soil: The 50-ha Forest Dynamics Plot lies along two ridges that are dominated by red clays. Bottom lands, characterized by brown or gray alluvium, separate the ridges. The bottom land includes several small permanent streams and a small swamp (frequently flooded) in the eastern half of the plot. The ridge is composed mostly by gently sloping hills, but steep slopes are usually associated with erosion gulleys. The difference between the lowest and the highest point is 33.5 m.  Detailed soil analyses have not been carried out. However, Two layers of about 10 cm of volcanic ash have been found about 5 m below the surface in a wetland located 4 km away from the study plot (Athens 1997), providing evidence for some volcanic influence in the soil formation. A major influence might have had the weathering of the geologic formation..

Forest type and characteristics: The forest is evergreen lowland wet forest, with a canopy mostly 15-30 m tall, with some emergent trees reaching 40 and rarely 50 m. The largest stem diameters are usually 2 m, frequently Ceiba pentandra [Bombacaceae]. The forest around the plot appears to be maturing, undisturbed for several centuries and possibly much longer.

Species composition changes slightly between the ridges and the bottom land. In a comparison between a hectare on a ridge and an adjacent hectare on bottom land, there were about twice as many mid-canopy species (with 10 individuals) growing exclusively on the ridge than on the bottom land (19 vs. 37). Among the dominant species that preferred the ridge forest were Brownea loretensis, Macrolobium sp. nov., Tachigali sp. [Leguminosae], Guarea kunthiana [Meliaceae], Talauma ovata [Magnoliaceae], Protium aracouchini [Burseraceae], and Ocotea javitensis [Lauraceae], whereas, Guapira sp. [Nyctaginaceae], Bauhinia brachycalyx [Leguminosae], Coccoloba densifrons [Polygonaceae], Guarea grandifolia, Guarea pubescens [Meliaceae], Maquira calophylla [Moraceae], and Astrocaryum murumuru [Arecaceae] were among the dominants in the bottom land. The most common species, such as Iriartea deltoidea, [Arecaceae] Matisia oblongifolia, Matisia malacocalyx [Bombacaceae], and Marmaroxylon basijugum [Leguminosae], grow abundantly in either habitat.

The swampy area in the eastern half of the plot is most notably different, with the palm Mauritia flexuosa, Sapium sp., several species of Piper only found there. This small swamp is topographycally a depression where the water remains even in the less rainy periods of the year.  Families and genera listed in the tables, as well as census information.

Natural disturbances: Most canopy disturbances are from small treefall gaps created when one or a few trees fall. The importance of large-scale windstorms is unknown, but there is no evidence for extensive blowdowns. Nor is there any indication that El Niño events have any impact in the region.

Human disturbance: In 1995, an archeological survey was carried out on a hilltop near the northwest corner of the study plot. There were ceramic shards just 50 cm below the forest floor, estimated to be roughly 500 to 1,000 years old (Netherly 1997.) The artifacts may belong to the nomadic Huaorani, who formerly opened small clearings or used natural gaps for plantations of manioc and temporary homesites. Evidence of prehistoric burnings, presumably for agriculture or subsistence, were found in soil cores taken near the study site and preliminarily dated to 7,700 years B.P. (Athens 1997.) There is evidence of Native American settlements in the area, but the existence of extensive clearings is unknown.

The most conspicuous present day human disturbances are consequence of oil exploitation and new settlements of indigenous groups around a new road. The large oil reserves found in the national park were conceded to various oil companies for prospecting and exploitation in the 90's. From 1992 to 1995 the oil company Maxus opened a road of 150 km that crosses the northwestern part of the park and established facilities for oil exploration, including an underground pipeline. Although the new road is mostly used for oil exploitation, and the oil company in charge of the camps control the access to the road preventing colonization, in the last four years Huorani and Quichua settlements have appeared within about 20 km of the station as well as in kilometers 68 and 99 of the main road. All Huaorani communities have given up their nomadic lifestyle for permanent houses. Because of these changes, hunting intensity has increased along the road. The Huaorani even hunt periodically in the 50-ha Forest Dynamics Plot, which is only 100 m from an oil road.  A relatively small sign of disturbance is also found inside the study plot: an area of about one-hectare near the southwestern corner of the plot --presently dominated by Ceropia species--, was apparently a heliport used for oil exploration before 1994.

Funding Sources: The Yasuní Forest Dynamics Plot has been funded by the Mellon Foundation (U. S.), the U.S. National Science Foundation, the Diva Project (Ecuador-Denmark), the Tupper Family Foundation, and the Smithsonian Tropical Research Institute.

• Romoleroux, K., R. Foster, R. Valencia, R. Condit, H. Balslev, and E. Losos. 1997. Arboles y arbustos (dap >1 cm) encontrados en dos hectáreas de un bosque de la Amazonía ecuatoriana. In Estudios sobre diversidad y ecología de plantas, ed. R. Valencia and H. Balslev, 189-215. Quito, Ecuador: Pontificia Universidad Católica del Ecuador.

Information about this research project will be provided in the near future.
 

•  R. Valencia, H. Balslev, W. Palacios, D. Neill, C. Josse, M. Tirado & F. Skov. 1997 In Press. Diversity and family composition of tress in different regions of Ecuador: a sample of 18 one-hectare plots. Pp xx-xx in F. Dallmeier (ed.). Forest Biodiversity in North, Central and South America and the Caribbean: Research and Monitoring. Parthenon Press. 
• Valencia, R., Balslev, H. & G. Paz y Miño. 1997. Tamaño y distribución vertical de los árboles en una hectárea de un bosque muy diverso de la Amazonía ecuatoriana. pp. 173-187 in R. Valencia & H. Balslev (eds.). Estudios sobre diversidad y ecología de plantas.  Memorias del II Congreso Ecuatoriano de Botánica realizado en la Pontificia Universidad Católica del Ecuador. Univ. Catolica, Quito. 
• Bernal, R. & H. Balslev. 1997. Strangulation of the palm Phytelephas seemannii by the pioneer tree Cecropia obtusifolia: The cost of efficient litter trapping. Ecotropia: in press. 

Dissertation abstract
The thesis consists of five papers, one accepted for publication and four pubished in scientific journals. The main topic is how small-scale environmental variability affects the population and community ecology of neotropical rain forest palms (Arecaceae). The results presented in this thesis are based on field-work in the lowland tropical rain forests of Amazonian Ecuador.  Forests exhibit a high degree of heterogeneity in numerous factors capable of affecting plants: canopy conditions and light, conspecifics, other plants, litter layer, soil factors, topography, and animal mutualists and pests. Any single plant population is subject to a high degree of small-scale environmental heterogeneity, but while such heterogeneity is present in all vegetation, some gradients in tropical rain forests are wider than in other vegetation types. Among the models proposed to explain how plant species richness in tropical rain forests is maintained, some of the most important are based on this small-scale environmental heterogeneity, notably
the niche-differentiation hypothesis. Environmental heterogeneity is not only important in the ecology, but most likely also in plant evolution and diversification, and the role of ecology in speciation has recently been re-emphasized. 

In the first paper, the introduction to this thesis, I review and discuss the importance of small-scale environmental heterogeneity in neotropical rain for-ests in the ecology and diversification of the rich palm flora inhabiting these forests. I show that environmental heterogeneity at small, 0.1-10² m, scales affects the individual performance and the small-scale distribution of palms in numerous ways and often affects different species differently. I also show that this heterogeneity promotes the local coexistence of palm species by niche differences among the species and probably also by mass effects and negative density dependence. Based on the observation that in species-rich palm genera and species complexes, sympatric species or morphs often differ in edaphic- topographic preferences or in characteristics conferring differing light requirements and in traits favouring reproductive isolation, I hypothesise that small-scale environmental heterogeneity is an important diversity-generating factor in the neotropical palm flora through the process of parapatric speciation.

The following four papers present my own (sometimes with Henrik Balslev) empirical results. The second paper documents population instability in a local population of the palm Iriartea deltoidea in Cuyabeno, Amazonian Ecuador. The juvenile classes are represented by very few individuals and a population ma-trix model based on the static population structure suggests that the population is declining. The third paper provides documentation that topographic heterogeneity is a strong structuring force in a species-rich Amazonian palm community: Yasuní, Ecuador. Mantel and logistic regression analyses of the distribution of palm species in a 50 ha plot in unflooded upland forest show that species composition is strongly influenced by topography, while drainage and canopy height (at 400 m 2 scale) are less important ecological factors. Antagonistic patterns of microhabitat preferences are found between several species pairs of small and medium-sized palms, indicating some importance of
topographic niche differences for species coexistence. The distribution patterns also suggest that mass effects occur frequently, both among upland microhabitats and between floodplains and swamps and the upland habitat, and may thus also contribute to small-scale species richness. The fourth paper evaluates three hypotheses regarding tall palm species: (1) Light requirements increase with increasing size; (2) canopy palms in general depend on large (  0.10 ha) tree fall gaps for recruitment; or alternatively (3) stilt-rooted palms do not. The first and third hypotheses are accepted (at least for the two common species
Oenocarpus bataua and Iriartea deltoidea), while the second hypothesis is rejected.  The conclusions are based on analyses of the size, crown position and forest phase of the individuals belonging to tall palm species at the Yasuní site. The final paper shows how the presence of a brook affects the recruitment and distribution of Iriartea deltoidea (same population as in paper 1) in a bottomland in Cuyabeno. Mantel analyses show that Iriartea deltoidea recruits preferentially on the brook banks and that this recruitment pattern causes the adults to be concentrated around the brook. The increased recruitment on the brook banks could reflect the increased light levels or better drainage there. Overall, my thesis research has shown that environmental heterogeneity, notably in topographic and canopy conditions, is important in the ecology of neotropical rain forest palms, determining small-scale distribution patterns and probably contributing to the maintenance of local species richness.

• Svenning, J-C. 1999. Microhabitat specialization in a species-rich palm community in Amazonian Ecuador. Journal of Ecology 87: 55-65.
• Svenning, J-C. 1999 Recruitment of tall arborescent palms in the YasunÌ National Park, Amazonian Ecuador: Are large treefall gaps important? Journal of Tropical Ecology 15: 355-66.

Svenning, J-C. and H. Balslev. (1999) Microhabitat-dependent recruitment of  Iriartea deltiodea (Arecaceae) in Amazonian Ecuador.  Ecotropica 5: 
• Svenning J.-C.. (2000) Small canopy gaps influence plant distributions in the rain forest understory. Biotropica. 32: 252-261. 
• Svenning Jens-Christian.(2000)  Growth strategies of clonal palms (Arecaceae) in a neotropical rainforest, Yasuni, Ecuador.  Australian Journal of Botany.  48: 167-178. 

 

Proposal summary
For the past 6 years, my interests in modern forests have focused on a group of plants that are not phylogenetically allied: the woody climbers, or lianas. Lianas can be found in about 140 families of plants and have probably existed on earth almost as long as there were trees up which to climb. Lianas are often excluded from large tree plot censuses (BCI, Yasuní, etc.) because of time and funding limitations involved in these studies. They are often included in smaller area plots but at the diameter limit of >10cm, which is a pretty large stem for a liana. So we know relatively little about tropical liana communities, compared to tropical tree communities. Lianas contribute roughly 10-35% of the species diversity to tropical and temperate forests (if we count just the woody species), and usually less than 10% of the biomass, based on litter fall or stem diameter estimates.

I am currently working in two areas: Yasuní National Park in eastern Ecuador and Manu National Park in eastern Peru. Both areas are reasonably well-protected parks, with adequate access and biological stations out of which I can work. 

In Yasuní, the objectives of my project are many: 
 [1] provide species lists and identification guides for the lianas of the park,
 [2] provide census information on species that are rarely encountered,
 [3] determine the pattern of ecological and geographic distribution of liana species in the park: who are the kings, who are the oligarchs and who are the proletariat?
 [4] how are species added and subtracted from communities: what matters?
 [5] monitor changes over time in forest communities within Yasuní to assess the ecological impact of the current land use practices: 
 oil extraction by multinational companies, expansion of indigenous populations (both Quichua and Huaorani), scientific and ecotourism study of forested communities.

I have used the following sampling scheme in Yasuní National Park and Manu National Park. I invite your discussion and comments on this methodology.
 [1] Plots are selected based on prior establishment of one hectare tree inventory plots (dbh > 10cm) censused and maintained by Nigel Pitman (Yasuní) or John Terborgh (Manu). Not all of my liana plots have a corresponding tree plot, but most do. Not all of their tree plots have a corresponding liana plot.
 [2] Each one hectare plot is sampled using five 4 x 100m transects. Total sample area = 0.2 ha. If the hectare is set up as 100 x 100, then the transect are located every 16 meters (parallel to one another) within the plot. A few plots are hectares established as 10 x 1000 m plots, and in these cases my five transects are lain end to end for the first 500 meters of the hectare.
 [3] All lianas > 1cm diameter are counted, identified, collected if necessary, and located within the 10m segment of the transect. Stems are measured at the widest part of the stem, exclusive of abnormal bulges and nodes. Each stem is measured only once, so a great deal of following individual lianas through the forest to determine their point of rooting and connection to other individuals is involved.
 [4] Lianas are counted if they enter the transect anywhere from 0 to 2 m in height, such that some stems can be rooted outside the 4 x 100m transect. Lianas are not counted if they are not climbing or dependent on another plant in any way.  This eliminates the possibility of preferential recognition of some distinctive species over others. It rarely does exclude some reasonably large (up to 2 cm) stems.
 [5] In addition to the counting of all liana stems > 1cm diameter, I keep a tally of all other climbing species less than 1 cm in diameter that I find in the transects. These tallies are kept on a "every 10 m" basis, such that I start a new list every 10 meters of a transect. Although these data cannot be considered quantitative because I do not keep a count of how many stems of an individual species I find every 10 meters, it adds a good sense of the density of the smallest stems and gives me a complete species list per plot.
 [6] All data are compiled into Excel worksheets on a per hectare basis, recording stem diameter, names, notes on stem characteristics, and collection status. In recent revisits of plots in Ecuador, I found that I could relocate almost every stem I had originally censused (one year previously), in spite of the fact that I do not mark stems with aluminum tags.

Dissertation project
Visit data presentation for more information on methods and results
I studied the ecology of lianas in Yasuní National Park, Amazonian Ecuador in the period 1996-2000. The research was focused on the diversity and distribution of lianas in two 0.2 ha plots near the Estación Científica Yasuní  and on the population ecology of the most abundant liana species, Machaerium cuspidatum Kuhlm. & Hoehne (Fabaceae).  Lianas (woody vines, or lianes), are woody climbing plants that start their life as terrestrial seedlings. They are a characteristic feature of the tropical rain forests. In the South American rain forests they often contribute more than 20 percent of the total woody plant diversity. They can sometimes dominate the vegetation, especially in disturbed areas and along rivers. 

1.   Liana diversity study in Yasuní, Amazonian Ecuador
The rain forest of the upper Amazon Basin is probably the place on earth with highest plant species richness. In some areas it is possible to find nearly 800 species of woody plants in just one hectare of forest. The lianas often constitute about 20% of these species.  One of the most puzzling questions in tropical ecology is how so many species of plants are able to coexist. One hypothesis is that different species, or different groups of species (guilds), use resources differently, which reduces interspecific competition. I had two main reasons for doing a liana community study in the Yasuní National Park in eastern Ecuador. The first reason was to make an inventory of the liana flora of the area. The second was to investigate if the liana communities were structured by the arrangement of potential host trees (trellises) or other environmental factors. Only if lianas use ressources differently can their diversity be maintained by niche diversification.

      Results & discussion
In the liana community study I recorded 606 liana individuals belonging to 138 species in 43 different plant families. The liana species richness in Yasuní was higher than in any comparable study. The number of liana species in the 10 x 10-m subplots depended mainly on the number of liana individuals in the subplots, which in turn was strongly positively correlated with the density of trees with a diameter of 1?10 cm. The remaining variation in liana diversity was partly explained by the density of smaller trees, with more liana species in subplots with a higher density of trees >=1 cm diameter.

Different species of lianas use different climbing mechanisms, and the size of the host trees they can use for support depends on how they climb. I showed that the lianas that climbed by attaching themselves to tree trunks with their roots were growing on host trees with a relatively large diameter, whereas the ones that climbed with hooks or thorns were growing on relatively thin trees. Lianas climbing with tendrils and lianas that twined around their hosts grew on host trees of similar diameters.

The lianas had a clumped distribution in the plots, and presence of one liana on a tree increased its risk of infestation by additional lianas. Although most lianas are rarely observed to climb large-diameter host trees, I found that the probability of trees being colonized from the ground by lianas increased with tree diameter. The rare event that a liana climbs a large tree from the ground may be important for the ecology of many liana species, as large trees provide direct access to the favourable high-light environment of the canopy.

2.   Ecology of the liana Machaerium cuspidatum
The liana Machaerium cuspidatum Kuhlm. & Hoehne (Fabaceae) is the most abundant liana species in large parts of Amazonian Ecuador and Colombia. Lianas are known to affect tree growth, and due to its abundance M. cuspidatum is likely to be the species that has the largest impact on forest dynamics. The high prevalence of M. cuspidatum also makes it ideal for studying aspects of liana dynamics and distribution that cannot be studied in less common species.

Machaerium cuspidatum is able to reproduce sexually as well as clonally. The seedlings are mostly found immidiately below very large (>20 m long) sun-exposed individuals. Clonal offspring, on the other hand, mostly occurs in areas where large lianas fall to the ground together with their host trees. When the liana stem starts touching the ground, new roots are formed. Later, the newly rooted portions of the stem break off from the mother individual to produce new independent plants (ramets).

    Results and conclusions
The population structure of M. cuspidatum varied with forest structure and with topography. The probability of finding seedling-sized plants (length <30 cm) and saplings (diameter <1 cm) was highest in steep upland areas, whereas the probability of finding large individuals (diameter >=1 cm) was highest in subplots with a low canopy and dense understorey.   The large plants were very strongly associated with low, dense forest patches, as seen from the ability of a logistic regression to predict the presence of large plants in subplots 60 percent of the times. The plants in the two smallest size classes were randomly distributed with respect to canopy openness, whereas most of the large plants had grown up to reach the higher light intensities of the canopy.  The total population density was generally not higher in subplots with elevated light intensities. The population density of M. cuspidatum was nearly the same in all habitat types, but the mode of reproduction changed with the drainage conditions. The number of seedling-sized individuals of clonal origin was highest in the floodplains, whereas the number of true seedlings (of sexual origin) was highest in well-drained upland areas. 

If seedlings were produced in the floodplains at all, they apparently had a higher mortality than the clonal plants. The lower abundance of clonal plants in the steep upland areas may be due to a lower disturbance rate, which cause large plants to re-root less often. The results show that a species may expand its realised niche to habitats with a low potential for seedling establishment by shifting to clonal reproduction.

3.   Dynamics of the liana Machaerium cuspidatum
In the analysis of the dynamics of M. cuspidatum I focused on how well variations in the population growth rate could be explained by variations in light and trellis availability. The study is the first large-scale quantitative study of the demography of a liana.  Since the plants in disturbed areas received as much light as the ones in the tall forest, I concluded that the species was directly affected by trellis availability. The population growth rate decreased when large plants growing in high-light areas were excluded. Since the remaining large plants consisted mostly of non-climbing individuals, some of which had fallen down from their hosts, the decrease in growth rate was linked to the reduced health of these plants. The study therefore provides the first direct evidence that a liana population can be more dependent on trellis availability than on light.

Objectives and preliminary results
Over the last ten years we have established a network of tree plots across eastern Ecuador in order to answer simple questions about the structure and diversity of tree communities there.  The inventories follow a conventional sampling methodology for tropical forests, encompassing all trees >10 cm dbh in a hectare, so that results are comparable to those of inventories carried out elsewhere in the tropics.  The network now includes 27 ha of upland, floodplain, and swamp forest containing nearly 17,000 individual trees, and we continue to establish new plots in poorly sampled areas.

Intensive analysis of the data set has begun only recently, but the picture emerging is of a very diverse forest dominated everywhere by the same suite of species.  For example, while the upland plots contain >1,000 species, 83 of them account for half of all the trees there.  This means it is difficult to find a patch of terra firme forest in the Oriente that does not contain healthy populations of Iriartea deltoidea (Palmae), Matisia ochrocalyx sensu latu (Bombacaceae), Brownea grandiceps (Fabaceae), along with several dozen other frequently encountered taxa.  And this homogeneity reigns over a wide range of topographic, edaphic, and other environmental conditions.  Even floodplain forests are mostly dominated by the same species that dominate upland forests, though the two forest types become progressively more different as the duration and frequency of flooding increases.

These patterns of large-scale homogeneity have been slow to emerge in Amazonian forests, in part because their famously high alpha-diversity makes small samples of identical communities appear different due to sampling effects.  We have tried to get around this problem by taking an explicitly predictive approach to understanding tree species distributions.  We build very simple models that make concrete predictions about which species are likely to occur where on the landscape and at what abundance, test them with field data, and then tinker with them to improve their performance.  The largest surprise to date is that the simplest of models -- ones that assume that tree communities are identical across the landscape -- make extremely accurate predictions.

At this early stage, our inventories have probably raised more questions than they have answered.  The greatest mystery to date  concerns the hundreds of rare tree species recorded there.  Most species in our data set are known from only a handful of individual trees, and we can infer almost nothing about their actual landscape densities, habitat preferences, or conservation status.  Much of our current research is aimed at improving our understanding of the basic biology of these species.

The tree plot network was established and is maintained in collaboration with the National Herbarium of Ecuador, the Missouri Botanical Garden, the Catholic University of Ecuador, the Central University of Ecuador, and the University of San Francisco de Quito.  It has been funded by the Duke University Department of Botany, the Andrew W. Mellon Foundation, the National Security Education Program, the National Science Foundation, and the Garden Club of America.  The Ecuadorean network is currently being administered conjointly with a larger, sister-network of plots in the vicinity of Cocha Cashu Biological Station  in southeastern Peru.  Additional results from the Yasuní network are described in Natural History.

• Pitman, N.C.A., J. Terborgh, M.R. Silman, and P. Núñez V. 1999. Tree species distributions in an upper Amazonian forest. Ecology 80(8): 2651-61.
• Pitman, N.C.A. 2000. A large-scale inventory of two Amazonian tree communities. Ph.D. diss., Duke University.
• Pitman, N.C.A., Mogollón, H.F., Terborgh, J.W., Núñez V., P., Silman, M.R., Neill, D.A., Cerón, C.E., Palacios, W.A., and M. Aulestia. (manuscript, in review) Large-scale homogeneity of western Amazonian palm communities.
• Pitman, N.C.A., Silman, M.R., Terborgh, J.W., Núñez V., P., Neill, D.A., Cerón, C.E., Palacios, W.A., and M. Aulestia. (manuscript, in review) Commonness and rarity in upper Amazonian tree communities.

Purpose and objectives
This project will initiate a collaborative study of the reproductive biology of the Forest Dynamic Plots at Barro Colorado Island, Panama and Yasuní.  The main objective is to collect and analyze data on reproductive phenology, dispersal, germination and seedling establishment characteristics of tree species on the 50ha Forest Dynamics Plot (FDP), Yasuní National Park.  We have set up 200 seed traps in the FDP which are emptied every two weeks, collecting all reproductive parts and identifying them to species level.  Using this information, I hope to gather data on the seasonality of seed dispersal and determine what factors may play a part in the fruiting phenologies of the species on the plot.  In addition, we have 20 seedling transects set around trails around the station and around the boundaries of the FDP.  These are each 100m long, and all emerging seedlings within a meter boundary of the transect are collected every 2 weeks.  These are counted and identified with the help of growing seedlings from identified fruits in a purpose built greenhouse at the Estación Científica Yasuní (E.C.Y.).

I will be collecting data over a period of two years and am setting up a fruit and seedling collection which will be kept here at the E.C.Y  Photographs are taken of all fruits and seeds which will be used to produce a CD ROM of The Fruits of Yasuní.  This project is supervised by Nancy Garwood (Natural History Museum of London) as well as David Burslem and Mike Swaine (University of Aberdeen).  Furthermore, S. Joseph Wright (Smithsoniam Tropical Research Institute, Panama) will be collaborating.  I will be assisted in the field by Zornitza Aguilar, Milton Cambrano, Paola Barriga, Matthew Priest.

Purpose and objectives
In this study, distribution patterns of species across habitats are examined in order to assess how habitat diversity influences the species richness on local scale within tropical lowland rain forests.  Data are evaluated under the assumption that species are individually distributed along environmental gradients as proposed by Gleason 1926. 

In order to maximize the chance of finding any pattern in the distribution of species in the present study, slope sites, regarded as an intermediate or transitional forest between flooded and unflooded (upland) forest, were selected for study.  Furthermore, flooded forest is studied and compared with upland forest on level terrain and slopes to deduce how tolerant species are to different stress factors, such as flooding and water logging of the soil. These three 'forest types' are in the following categorized as habitats, as the forest with its characteristic height and structure make up a habitat for especially herbs, but also for many understorey palms.

To be able to analyze whether most species are ecologically specialized or have a wide ecological tolerance, a broad range of environmental variables apart from topography are recorded, such as soil chemical and physical properties, forest structure and light. 

As quantitative studies of the distribution patterns of palm and herb species in relation to variation in environmental conditions in the Amazon rain forest are rare (Poulsen & Balslev 1991, Kahn & Castro 1985, Balslev et al. 1987, Kahn & de Granville 1992, Scariot et al. 1989, Svenning 1999a, 1999b), these plant groups were chosen for further study. Moreover, the study of herbs is chosen because of ease of collection and the possibility of morphospecies recognition in the field for a great part of the common
species. In addition palms are included in the present study to describe the habitat herbs grow in. Palms growing in the study area were quite well known and one of the most dominant tree families in terms of density and basal area at least in upland or so-called terra firme forest (Romoleroux et al. 1997, Balslev et al. 1987). Finally, palms are included in attempt to compare the distribution of a taxonomic group of trees with the distribution of herbs.

To my knowledge, no study has previously compared the species composition of herbs in flooded and adjacent unflooded forest in Upper Amazonia and only a few studies have compared the composition of all palm species in these habitats at adjacent locations (Balslev et al. 1987, Kahn & de Granville 1992). 

To be able to assess whether differences in species composition and diversity among the three habitats are related to differences in environmental conditions among these, the differences in species composition and diversity among habitats should be greater than the differences within habitats. Therefore, the differences in species composition and diversity are examined both among and within habitats.

Specific objectives: 
1. To describe the flooded forest, the forest on slope sites and the forest on upland sites on level terrain in terms of soil properties, forest structure and forest floor conditions and to examine the differences among and within habitats.
2. To describe the species richness, diversity and composition of the herb and palm communities in the three habitats and to examine the differences among and within habitats.
3. To examine how differences (if any) in species richness, species composition and abundance of herbs and palms among and within the three habitats are related to differences in environmental conditions. Emphasis is put on the influence of flooding and topography. 
4. To examine how the species richness and abundance of different size classes of palms and different groups of herbs are related to environmental conditions. 
5. To examine if a gradient in species composition of herbs and palms can be characterized from flooded forest to slope forest to forest on upland plateau.

The influence of flooding is here referring to the differences between flooded and unflooded forest sites (on slopes and level terrain) 
The influence of topography is here referring to the influence of inclination and variation in terrain height between forest sites on slopes and forest sites on level terrain.

Information about this research project will be provided in the near future.

• Erwin, T.L. (1994) Arboreal beetles of tropical forest: The Xystosomi group, subtribe Xystosomina (Coleoptera: Carabidae: Bembidiini): Part I. Character analysis, taxonomy, and distribution. Canadian Entomologist 126:549-666. 
• Kress, W.J., W.R. Heyer, P. Acevedo, J. Coddington, D. Cole, T.L. Erwin, B.J. Meggers, M. Pogue, R.W. Thorington, R.P.  Vari, M.J. Weitzman, and S.H.Weitzman. (1998) Amazonian biodiversity: assessing conservation priorities with taxonomic data. Biodiversity & Conservation 7:1577-1587. 
• Southard, D. & T.L. Erwin (1999). Tropical trees across an equatorial landscape: the entomological plot at Onkone Gare Biological Station, Huaorani Territory, Ecuador.  Biotropica
• Erwin, T.L. (2000) Arboreal beetles of neotropical forests: Taxonomic supplement for the Agra virgata and ohausi groups with a new species and additional distribution records (Coleoptera: Carabidae). Coleopterists Bulletin 54:251-262. 
• Erwin, T.L. (2000) Arboreal beetles of neotropical forests: Agra Fabricius, the Novaurora complex (Coleoptera: Carabidae: Lebiini: Agrina). Smithsonian Contributions to Zoology (608) i-iii:1-33. 

Dissertation abstract
I studied the ecology of mixed-species understory bird flocks in a 100-ha plot of pristine rainforest in Amazonian Ecuador during 12 months in 1994 and 1995.  On average, these flocks contained pair or family group of 20 species that foraged together each day. One-third of the species present in a flock changed each hour.  Flocks were assembled each morning and led on long-distance movements by a pair of Cinereous Antshrikes, Thamnomanes caesius.  This species was also the primary flock sentinel.  Flocks maintained stable territories averaging six hectares that they defended in infrequent border
disputes.  They rarely approached adjacent borders simultaneously, although they frequently entered adjacent territories and foraged within them.  Flock movements through a territory were predictable on small temporal and spatial scales, but less so on larger scales.

The study site contained 284 species, 221 of which were present in measurable densities.  Comparisons with two other sites in Amazonia showed an underlying structure to these bird communities.  Roughly 30 percent of all species and 50 percent of all individuals present at two sites foraged with understory flocks.

Mixed-species vine-tangle flocks were discovered as part of this study and represent a new type of avian social structure.  These flocks provided an interesting insight into current hypotheses on the adaptive significance of flocking.  They appear to have a benefit of flock membership that is independent of predation avoidance, likely interspecific learning about food resources.

Understory flocks were internally organized into many smaller subgroups of two to three species each.  Species composition of these subgroups changed in response to microhabitat changes as a flock moved about its territory.  Flock member species associated with sentinel species in open forest and with aerial dead-leaf and vine tangle specialist species in dense forest.

Significant observer bias was present in data collected using a haphazard protocol equivalent to those used to collect virtually all avian behavioral data.  This bias obscured results found in analyses of data collected using a randomized protocol.  Using only a haphazard protocol to collect behavioral data in this study would have made detection of internal flock organization unlikely.

Information about this research project will be provided in the near future.

Information about this research project will be provided in the near future.

Proposal abstract
I will examine the small-scale successional dynamics of an eastern Ecuadorian rainforest in the Yasuní National Park using passive (LANDSAT TM and SPOT HRV) and active (ERS SAR) remote sensing technology combined with successional and economic models. Spectral mixture analysis will be used to determine successional stage coverage of individual image pixels and change detection techniques will be employed to generate transition matrices for the different successional stages. Fieldwork will verify the images and also serve to quantify relevant forest parameters including seed bank, soil composition and standing biomass. Economic data and land use patterns of the Huaorani people who occupy the Yasuní National Park will be determined from historical data and informal interviews. Data from these studies will be integrated into a GIS and used to parameterize predictive models. A model that combines spatially-explicit succession dynamics with economic pressures will generate a platform on which to investigate the effects of different land use practices on future forest states. If history does truly repeat itself, the forest will be the site of massive deforestation as the Huaorani enter the local market economy. The ultimate goal of this project is to generate ecologically and economically viable alternatives of land use while the forest is still in an early stage of human expansion .

Image shows area of Yasuní that includes Proyecto Primates (by short road in the mid-left of photo), the roads and camp (center of photo) of Repsol-YPF, and the Estación Científica Yasuní (on the right, where spur road interscts with river).

Proposal abstract
Historically, indigenous peoples have been portrayed as sustainable managers of their natural environment.  Thus, many indigenous groups were viewed as having a "balanced existence with nature" (Alvard 1993).  Alternatively, Hardin (1968) proposed that any system in which a group of people share common resources is bound to lead to the ruin of all, as each individual pursues only their own best interest.  This idea is labeled "the tragedy of the commons" (Hardin 1968).  The apparent ability of hunter-gatherer groups to maintain a sustainable use of resources in the context of common resource use led researchers to question the mechanisms of this conservation.  The suggestion that this conservation actually results from non-conservation-oriented foraging behavior was proposed and has been termed "epiphenomenal conservation" (Hunn 1982).  The question addressed by many researchers concerning foraging decisions has been: Do indigenous hunter-gatherers inherently understand and implement some level of conservation in their foraging practices, or has their low population density and nomadic existence saved them from a "tragedy of the commons" (see Vickers 1980, Redford 1991, Alvard 1993, 1995, Alvard et al. 1996, and Bodmer et al. 1996)? 

In this study I examine the effects of market access and wage employment on the traditional hunting practices of the Huaorani, an indigenous hunter-gatherer-horticulturalist group in the Ecuadorian Amazon.  The current situation of the Huaorani is representative of many indigenous groups throughout the world that have been recently exposed to a market economy.  Traditional subsistence practices are often compromised under these circumstances as new opportunities lead to the introduction of cash into communities.  Of great concern is the on-going trend towards commercialization of indigenous hunting which has been documented in several African (Wilkie et al. 1992, Lahm 1993, Fa et al. 1995, Wilkie et al. 1998) and Amazonian (Robinson and Redford 1991, Stearman and Redford 1992, Vickers 1993) communities.  The effects of this commercialization on indigenous hunting patterns are largely unknown.  However, such a change may lead to the overexploitation of game species with high market value that are heavily prone to extinction (Redford 1992, Robinson and Redford 1991, Bodmer et al. 1994).

This research is being conducted within an area of the Yasuní National Park and Biosphere Reserve in Ecuador that extends south of the Napo River.  The field site is in an area where the Huaorani Ethnic Territory and Biosphere Reserve overlap.  The Huaorani have secured the legal right to continue living in their traditional style within this region.

In 1992, Maxus Oil Company began construction of a road into this section of the Yasuní Reserve in order to access oil wells (Holmes 1996).  Prior to construction of the road there was little human presence in the area resulting in abundant wildlife and virtually pristine rainforest.  Since completion, many Huaorani families have migrated from settlements within the western region of the Huaorani Ethnic Territory (an area referred to as the "Protectorate") into settlements along this stretch of road. 

At the time of contact (in 1958), the Huaorani totaled 500 individuals spread across their traditional homeland of 20,000 square kilometers in four main groups termed "neighborhood clusters" (Yost 1991).  The total Huaorani population is recently estimated to be 1300 individuals (Lu 1999).  Current population estimates for the two Maxus Road settlements (located at kilometer 32 and kilometer 99) are 50 individuals and 105 individuals, respectively.  In addition there are several families scattered at various locations along the road separate from these distinct communities.

Cash is becoming more important in the lives of the Huaorani in these communities.  Opportunities now exist for the commercial sale of wildlife at the Saturday market in Pompeya and for temporary wage employment.  The Huaorani use cash for items such as clothing, ammunition, aluminum pots, food items, travel and medical expenses, and for sending children to school in the Huaorani settlement of Tonampari, located within the Protectorate. 

Investigation into the current situation faced by the Huaorani requires a move beyond questions of conservation versus energy-maximization as the mechanisms motivating resource use. The significance of this research is in its use of both evolutionary ecology and microeconomic theory in the investigation of indigenous resource use in the context of a mixed economy.  It involves expanding optimal foraging theory to include a cash currency as well as testing predictions concerning the different situations in which a time-minimizing (subsistence) versus a quantity-maximizing (commercial) foraging strategy can be expected. 
Specifically, my research objectives are:

• To test predictions formulated from an evolutionary ecology framework concerning the foraging behavior (i.e., commercial vs. subsistence hunting) of individuals under two distinct circumstances, wage labor and unemployment.
• To test predictions generated from microeconomic theory concerning market participation in the context of a mixed economy.

Research summary and conclusions
The present research is an ecological and ethnobotanical approach to NTFP in three broad forest types -- terra firme, floodplain, and swamp -- in an Amazonian forest.  We asked whether the most species rich habitat was also the most used by the local indigenous communities, and which of the woody life forms (large trees, small trees and lianas) was the most useful  in each of the three forest types studied.

This study was carried out in the Yasuní National Park and the Huaorani Ethnic Reserve, in Amazonian Ecuador, in collaboration with the Huaorani indigenous group.  The Huaorani ethnic group was contacted less than 50 years ago. They are self-sufficient in the basic needs, deriving their subsistence from hunting, fruit harvesting, and cultivation of cassava and banana. The Huaorani possess a great traditional knowledge of forest ecology; presently, however, this information is disappearing as a result of their rapid acculturation.

We established 25 plots of 0.1 ha (50 x 20 m) in three broadly defined forest types in the vicinity of two Huaorani communities located 67 km apart: 
     TF: Terra Firme (10 plots), well drained upland forest.
     FP: Flood Plain forest (8 plots), well drained flood plain forest.
     SW: Swamp forests (7 plots), dominated by Mauritia flexuosa.
Woody stems with dbh * 2.5 cm were measured and identified, and relevant ethnobotanic information obtained from old experienced informants from the two Huaorani communities.  We defined three life forms: large trees (dbh * 10 cm), small trees (dbh < 10 cm), and lianas (dbh * 2.5 cm).  The usefulness of each forest types values was evaluated via two indices: the SPECIES USE VALUE INDEX (Phillips & Gentry, 1993) and the PERCENTAGE OF USEFUL SPECIES (Prance et al., 1987).

A total of 6953 individuals included in 1094 species, 370 genera, and 84 families were sampled. The great majority of the species (87.4%), and almost all the individuals (96.5%), were useful.  The three forest types -- terra firme, floodplain and swamp -- are more or less well defined according to a DCA ordination using species abundance data of the 25 plots.  Swamp plots are well separated from terra firme plots, and floodplain plots show high variation in both floristic composition and community structure.  The terra firme forest had the highest alpha diversity among habitats, followed by floodplain and swamp forest.  According to the SPECIES USE VALUE INDEX, the swamp forest was the most important forest type, followed by floodplain and terra firme forests.   On the other hand, according to the PERCENTAGE OF USEFUL SPECIES, floodplain was the most useful habitat, followed by terra firme and swamp forests.  From the life forms studied, large trees showed both the highest SPECIES USE VALUE INDEX and the highest PERCENTAGE OF USEFUL SPECIES, followed by small trees and lianas.  Lianas varied greatly in their SPECIES USE VALUE INDEX and PERCENTAGE OF USEFUL SPECIES in the three forest types.  The values of both indices varied more in swamp forest than in other habitat types.

To conclude,  this study found no correlation between the most diverse forest type, and terra firme was the most useful or most species rich habitat, but swamp forest showed the highest SPECIES USE VALUE INDEX, while floodplain showed the highest PERCENTAGE OF USEFUL SPECIES.  Second, the SPECIES USE VALUE INDEX of terra firme forest is less variable than floodplain, and swamp forest showed the highest variability.  Trees > 2.5 cm were more useful than lianas, and large trees were the most useful life form in the three habitats compared.  Finally, large trees showed less variability from SPECIES USE VALUE INDEX and PERCENTAGE OF USEFUL SPECIES than small trees and lianas.

We thank the Huaorani communities of Tiputini and Dicaro in Yasuní National Park and the Huaorani Ethnic Reserve, and the staff of the Herbarium QCA at Pontificia Universidad Católica del Ecuador.  This project is collaborative effort among three Latin American universities and three European universities: Universidad de los Andes (Colombia), Universidad Nacional de la Amazonia Peruana (Peru), and Pontificia Universidad Católica del Ecuador (Ecuador), University of Amsterdam (The Netherlands), University of Turku (Finland), and University of Aarhus (Denmark).  Similar studies are being developed in Colombia and Peru. This study was supported by European Comission, INCO-DC, IC18CT-960038.

• In press: Non-timber forest plant resource assessment in NW Amazonia. University of Amsterdam.

• de Vries, T. and M. Cristian. (2000) First nesting record of the nest of a Slaty-backed Forest-Falcon (Micrastur mirandollei) in Yasuni National Park, Ecuadorian Amazon. Journal of Raptor Research. 34:148-150. 
• Roubik, D.W. (1998) Grave-robbing by male Eulaema (Hymenoptera, Apidae): Implications for euglossine biology.  Journal of the Kansas Entomological Society 71:188-191. 
• Freiberg, M. and E. Freiberg  (2000) Epiphyte diversity and biomass in the canopy of lowland and montane forests in Ecuador. Journal of Tropical Ecology 16:673-688. 
• Ron, S. and J.B. Pramuk (1999) A new species of Osteocephalus (Anura: Hylidae) from Amazonian Ecuador and Peru. Herpetologica 55: 433-446. 
• Jungfer, K., S. Ron, R. Seipp and A. Almendariz. (2000) Two new species of hylid frogs, genus Osteocephalus, from Amazonian Ecuador. Amphibia-Reptilia.  21: 327-340. 
• Salvador-Van Eysenrode, D., J.  Bogaert, P. Van Hecke, and I. Impens. (2000) Forest canopy perforation in time and space in Amazonian Ecuador. Acta Oecologica. 21:285-291. 
• K.Galacatos and R.Barriga-Salazar  (2001)  Aquatic insect diversity and seasonality of the lowe Yasuní National Reserve of the Ecuadorian Amazon.