Authors:

Huw Lloyd & Arturo Palomino Marín

Return to Home Page

Introduction

South-eastern Peru and adjacent parts of Bolivia contain what is probably the largest and least disturbed area remaining of Upper Amazonian and Lower Andean ecosystems (Foster et al. 1994). The Department of Madre de Dios contains extraordinary levels of bird species richness. Over 570 bird species have been recorded at two single site locations; Cocha Cashu Biological Research Station, in Manu Biosphere Reserve (MBR), and at the Explorers’ Inn (EI), in the Tambopata Candamo Reserved Zone (TCRZ)(see Foster et al. 1994 for a review of bird species located at the latter). Furthermore, over 440 species have been recorded at the Tambopata Research Centre, located in the Upper regions of the Tambopata River within the TCRZ.

It is evident that the number of lodges and tourist visitors has increased significantly over the last decade in the area under study (see General Introduction). The process of looking for tourism related impacts on bird populations is a very problematic exercise. Firstly one must chose a ‘unit’ of ecotourism to measure e.g. number of tourists walking the trails, the presence of the trail itself, etc. How does one quantify this unit of tourism? How can one relate this unit of ecotourism to the abundance of birds in the habitats surrounding the lodges? The hardest obstacle, however, is how to distinguish any apparent impact from other forms of disturbance as it is evident from many other studies that many bird species are habitat specialists and can be very susceptible to slight habitat changes brought about by natural and anthropomorphic phenomena not related to tourism.

The bird communities around the lodges in the TCRZ have been, as a rule, subjected to a number of different forms of disturbance over the years. Poor trail management and selective logging, which has been a practice at many lodges, are often characterised by changes in forest biomass and tree size parameters, different tree architectures and tree species compositions which are the important habitat parameters which can affect bird communities (Bibby et al. 1998). There has been exploitation over the years of the Brazil-nut tree and many lodges are situated relatively close to human settlements, and both variables tend to increase local hunting pressures. Some species may also be captured, noticeably members of the Psittacidae and Ramphastidae families, for the pet trade. The distance from the lodges to the nearest settlement may be correlated with the degree to which it has been disturbed or used by the local people for other purposes other than tourism. Furthermore, two of the main lodges have been operating for over twenty years (CAPH and EI). EI has been a major centre for bird-orientated tourism and research, the latter of which has exposed the bird communities to a relatively high level of "playback recordings", which can impact particularly territorial species. A severe form of natural disturbance in Tambopata which has been little studied although potentially very significant is the mass seeding events of Guadua bamboo thickets.

Even if one could find a relationship between anyone of these forms of disturbance and the bird communities in the Tambopata region, one is still confronted by the problem that there is no quantified baseline population data prior to the lodges becoming established, to compare it with. The only quantitative research of an undisturbed lowland rainforest bird community in Peru comes from the Cocha Cashu Biological Station in the Manu Biosphere Reserve (Terborgh 1986, Terborgh et al. 1990, Robinson & Terborgh 1990). We are unable to make direct comparisons between the avifauna in Tambopata and Manu due to a lack of comparative habitat information between the two areas.

Previous research on bird species in the Tambopata province over the last twenty years has largely focused on simple observations (Parker, 1982), or on building species lists for five localities in the region (Davies et al. 1991, Foster et al. 1994). Three of these are ecotourist lodges we investigated here, however the other two sites are the Pampas del Heath (see Figure 2.1 - now incorporated into the Bahuaja-Sonene National Park) and the Cerros de Tavara, in the Andean foothill region (Foster et al. 1994). Kratter (1995b, 1997) was the first biologist to provide population data and habitat associations on a ‘subset’ of the lowland bird community in the Tambopata region, with his research on the habitat-restricted bamboo specialists at the TRC. Dyrcz (1990) and Lloyd (unpubl.) have also studied smaller ‘subsets’ of the whole bird community at the EI (understorey bird species and terrestrial bird species, respectively).

Given the aforementioned problems with ecotourism impact studies on birds, we decided for this research, not to focus or examine a cause and effect relationship between tourism and the bird communities. Our aim was simple: to provide the first quantitative and descriptive population data on the lowland rainforest bird communities and to relate the findings to features of the lowland forest habitat at each site. Determination of avian habitat use is important for understanding both the ecology and conservation needs of a species or community (Rappole et al. 1998). The value of such an approach may come from predicting distribution and numbers in unsurveyed areas; providing an understanding of the nature of the relationship between a bird species and its habitat; predicting possible consequences of future changes of land use (Bibby et al. 1998); and in this case providing much needed background information on bird populations for the benefit of the participating tourism operations.

Objectives

• To examine the habitat structure of the predominant forest types at each lodge and to see whether this habitat structure differs significantly between lodges.

• To describe the bird community in each forest type and to investigate the status of each by examining the composition of indicator species, following the criteria of Stotz et al. (1996).

• To produce the first species lists for SACHA and ECO.

• To examine the abundance of indicator species and total number of species surveyed and their relationship with features of the habitat in each forest type.

• To calculate absolute population densities of indicator species per unit area of forest habitat, with the aim of estimating absolute densities for the whole Tambopata region for the more habitat-restricted species, and thus those species most at risk from habitat alteration. [Total population estimates and subsequent recommendations for the conservation status for these habitat-restricted species are addressed in the section ‘Species Accounts’]

• To examine the abundance of large frugivorous birds, mainly parrots and toucans, and relate this to the timing of the fruiting season in Tambopata. [Due to their colouration, behaviour and familiarity to tourists from the pet trade and television documentaries, these species are frequently the species that tourist most want to see in the rainforest (Munn, 1992).]

• To make recommendations to lodge administrators regarding the management of forest habitats with respect to birds.

EcoAmazonía Lodge (ECO)

The primary habitat around the point count stations at ECO was Seasonally Flooded Swamp forest (Type B). Type B forest is characterised by an abundance of tree palms e.g. Scheelea butyracea, Euterpe precatoria and Socratea exorrhiza. Broad-leaved tree species include Symphonia globulifera. The canopy is normally low (<20 m) and broken, with a dense understorey of shrubs and small palms, especially Geonoma and Bactris species. Vines and scandent herbs are moderately abundant but large lianas are rare. Ground vegetation is restricted to areas of drier ground atop low hummocks, dominated by the fern Adiantum latifolium.

Aguajals are numerous within an area of 10,000 ha. around the lodge. There are three small aguajals running from North to South, whilst a fourth larger aguajal is also present (see satellite image). Following the main tourist trail, at 5 km there is a fairly extensive stand of riverine scrub. At 6.1 km two canopy platforms are located at the juxtaposition of three distinct habitat types; the scrub, type B forest and a large stand of Permanently Flooded Swamp forest (type A forest). This latter forest occurs in former ox-bow lakes and other permanently flooded or waterlogged situations. It’s characterised by a predominance of the palm Mauritia flexuosa. Due to the extended periods of flooding, ground and understorey vegetation is scarce or absent. Aquatics such as Apalanthe granatensis and Lemna aequinoctialis are sometimes present in the standing water. A considerable amount of organic detritus is usually present, accumulating normally around the base of the trees so that these sometimes form small raised islets within the swamp and may be colonised by terrestrial trees and other vegetation. Water levels in the swamp forest may drop during the dry season by as much as one metre during the dry season.

Located in the river, opposite the lodge is Monkey Island (Isla de Monos), a 30 ha. island dominated by Type C Lower Floodplain forest. This consists of pioneer trees and herbaceous vegetation, and is found on low recently developed or developing floodplain beaches, but is more characteristic of river islands in both the Rio’s Madre de Dios and Tambopata. Cecropia membranacea, Ficus insipida and Sapium ixamasense dominate this open successional forest. Few trees reach more than 20-25 m in height. The cane Gynerium sagitatum is the dominant understorey species. Small liana’s such as Paullinia alata are common and form locally dense tangles. Ground cover is sparse, mainly consisting of dense clonal patches of Calathea and Heliconia species. Tessaria integrifolia and Salix humboldtiana are found growing on a narrow belt along the river edge, upon levee deposits.

Cusco Amazónico Pueblo Hotel (CAPH)

Predominant forest type at CAPH is a mosaic of type B forest with Old Floodplain Forest (type F). Extensive areas of pure type F forest are the second most common type within the 10,000 ha. of land surrounding the lodge. The forest canopy in areas is tall (>30 m) and is more continuous than type B forest. The palm species Iriartea deltoidea is particularly abundant. In total, palm species make up approximately 30% of the tree stems in this forest type. Large emergents such as Dipteryx species, Ceiba pentandra and strangling figs are common. The dense shade caused by the more closed canopy inhibits dense understorey growth, with the result that the understorey is relatively open. The ground fern Adiantum latifolium dominates the ground vegetation. In pure type F forest flooding no longer occurs but probably has happened in recent years. However, due to the presence of type B forest in this mosaic, this forest floods more frequently than normal.

North of the trail system there are larger areas of type B forest exist, whereas south-east of the lodge, is situated a 45 ha island, also named Isla del Mono. Larger in size than the island at ECO, it is also covered with type C forest.

Explorers’ Inn (EI)

The vegetation types of the reserve have previously been described by Phillips (1993), Nicholson and Edwards (1994), and Nicholson and Phillips (unpubl.). Nine distinct forest types are recognised. Type A forest is found along the northern region of Laguna Cocococha, a large oxbow lake in the north-eastern part of the reserve. Areas of type C forest are found along the banks of the Rio La Torre to the south of the lodge clearing, while in the southern area of the reserve, there are extensive areas of type E and type F forest (Nicholson and Edwards, 1994). A mosaic of type B and type G forest covers an extensive area to the east of the lodge clearing, whilst type H forest occupies the large areas of higher ground, to the south and east of the Laguna Cocococha. The light texture and free draining nature of the soils distinguishes this forest type from type G forest. Floristically it is also quite distinct from type G forest; palms such as Iriartea deltoidea and Socratea are almost completely absent, while emergents such as Bertholletia excelsa, Cedrelinga cateniformis, and Dipteryx species are common. Subcanopy species include Bixa arborea, Pourouma species, and Capirona species amongst others. The canopy is tall (>30 m), and closed, while the understorey is relatively open and composed of small palms such as Geonoma deversa plus a variety of broad-leaved shrubs.

Sachavacayoc (SACHA)

SACHA contains a similar variety of habitat types to that found at the nearby EI. We recorded around five different forest types in total. The initial 2.7 km along the main tourist trail is dominated by Old Floodplain forest (type F). Extensive areas of bamboo (Guadua weberbaueri) are common in the understorey vegetation here. In one area located between the nearby stream on the west of the lodge and the lodge itself, there is a large thicket of dense bamboo.

Around 2.5km along the main tourist trail and approximately 400 m east of it, there exist smaller areas of type G Terra Firme Clay Forest. This latter forest type has a very tall canopy layer (>30 m) but is much more open than type F forest, thus resulting in a more dense understorey and ground layer. This forest type is usually found on higher raised platforms than the corresponding floodplain forests. Prominent tree species include the palms Iriartea deltoidea, and Euterpe precatoria, and broad-leaves such as Pourouma and Pseudolmedia species. Large emergents present include Dipteryx sp., and Parkia nitidia. Characteristics of the understorey are shrubs such as Cephaelis tomentosa and smaller palm species such as Geonoma. More extensive however are patches of type B, Seasonally Flooded Swamp forest, of which some encroach to within 250 m on the main tourist trail.

The area beyond the 2.7 km mark along the main trail and extending for a further 3 km consists of an extensive area of Terra Firme Sandy Clay forest (type H). This forest continues up to a 45 ha. oxbow lake, Lago Sachavacayoc. On the east and west ends of the lake exist smaller areas of type A, Permanently Flooded Swamp forest. The type H forest type continues on the south side beyond the lake for an unknown but vastly extensive area until it reaches the foothill region.

Selective logging has occurred within the forests at SACHA. In 1985 small scale extraction of Cedrela odorata (Tropical Cedar) and Cedrelinga cateniformis began, while south of the lake in the type H forest, Sweitenia sp. were removed. Currently there are incidents of hunting near the lodge, along with small scale harvesting of Geonoma palms (Palmiche).

Tambopata Research Centre (TRC)

The different habitats at the TRC have already been described by numerous authors (Foster et al. 1994; Kratter, 1995a, b, 1997) and will only be briefly mentioned here. The main area of forest encompassing the majority of tourist trails is a mosaic of type E/F, Upper Floodplain/Old Floodplain forest. The Upper Floodplain forest component of the mosaic has relatively open canopy cover and a mean canopy height of 25-30 m. Palm genera such as Iriartea, Astrocaryum, and Socratea are frequent, as are emergents broad-leaved trees such as Ficus, Chorisia and Ceiba. Large lianas are more prominent than in type F forest. The understorey contains species such as Geonoma sp. and broad-leaved shrubs. The fern Adiantum latifolium is the dominant ground level species. This forest type is subjected to more frequent flooding than Old Floodplain forest, probably once per decade. In other areas there exist blocks of type B forest, and type F forest inundated with bamboo (see Kratter, 1995a, 1997). Bamboo also dominates the 2.5 km long and 50 m tall bluff which runs in a NW-SE direction and located just SW of the lodge, it also dominates the vegetation atop the macaw clay-lick situated on the western shore of the Tambopata River, 200 m from the lodge (see Foster et al. 1994; Kratter, 1995a).

Methods

Return to Home Page