BIRD COMMUNITIES OF THE ECUADORIAN CHOCO: A CASE STUDY IN CONSERVATION

Between 1995 and 1998 I studied bird communities through mist netting (MN) and the 'Multi Time-Window Transect-Mapping' (MTW) technique for audiovisual surveys of tropical bird communities. The research was carried out in Playa de Oro (0 degrees 52'35 '' N, 78 degrees 47'40 '' W; 45-590 m a.s.l.), Esmeraldas, northwestern Ecuador. A total of 344 bird species are known to occur, though I presume that >400 species might be present in the study area, or migrate through it, on a regular basis. I sampled 18 transects (25.14 km), using the MTW method for rapid assessments. In order to perform a comparative study of the effectiveness of mist-netting and transect-mapping protocols, I chose two independent transects, MNT1 (625 m) and MNT2 (550 m), which were studied with both methods. They were located in a similarly structured habitat mosaic of cultivated land and selectively logged forest. The greater variety of human-managed habitats was the principal cause for higher species richness and diversity at MNT1 (n(MN) = 117; n(MTW) = 162) in comparison with MNT2 (n(MN) = 96; n(MTW) = 144). Nevertheless, according to the MTW data, the number of forest-dependent species (MNT1 = 78, MNT2 = 90) and threatened taxa (world: MNT1 = 4, MNT2 = 7; Ecuador: MNT1 = 37, MNT2 = 48) was higher at MNT2, likely due to the lower degree of fragmentation, higher foliage complexity, and lower hunting pressure. Biomass density was almost identical, with an estimated 193 kg/km(2) for MNT1 and 198 kg/km(2) for MNT2. The number of breeding 'pairs' was roughly 2000 /km(2) and 1800 /km(2), and the average biomass per individual 36 g and 41 g, respectively. In conclusion, the conservation value of study sites should be ranked on the basis of a detailed comparison of habitat association as well as threat status of the bird species present, whereas species richness, diversity indices, and biomass density should not he used in this context. For example, the 'bird community index' (BCI) sensu Canterbury et al. (2000) clearly reflected the higher value of MNT2 in comparison with MNT1 for the conservation of forest-dependent birds (BCI: MNT1 = -0.049 versus MNT2 = 0.541). Regarding the comparative effectiveness and efficiency, the results indicate that, in only 8-9% (21-23 h) of the study time, 39-50% (45 and 48) more bird species were recorded by transect mapping than by mist netting. High costs and low time efficiency, as well as inaccurate representation of entire bird groups, make the use of mist netting prohibitive for most survey purposes, unless the status and habitat selection of altitudinal and latitudinal migrants are concerned. Transect mapping is highly effective at detecting avian taxa that regularly vocalize. However, dispersing birds, floaters, long-distance migrants on passage, and very silent breeding residents were overlooked or underestimated. Thus, I developed optimized MTW protocols to further increase bird species detectability. For instance, implementing the specifically designed MTW monitoring protocol, a well-trained observer might record c. 70% of the total species richness and 80-90% of all breeding residents present in the transect area within as little as six days. For transects of up to 1200 m, sufficient data for rough estimations of absolute population and biomass densities can be obtained in the same period.