Human-Altered Tree-Based Habitats and Their Value in Conserving Bird and Bat Diversity in Northeast Luzon, The Philippines - Smallholder Tree Growing for Rural Development and Environmental Services - page 355

of globally threatened species (Peh et al. 2005). Mixed-rural habitats in central
Sulawesi contain 76 percent of total forest bird species richness recorded in a study
of a variety of habitats including primary and secondary forest. The majority of
these forest bird species is endemic to Sulawesi (Sodhi et al. 2005). The same study
shows that within mixed agroforestry plantations 32 percent of all forest bird spe-
cies is present, a much higher representation of regional forest bird species com-
pared to the proportion of lowland forest birds of northeast Luzon found in our
study habitats (13 percent). Why does the human-altered landscape in northeast
Luzon have so few forest bird species?
A first explanation may be serious under-sampling of forest birds which are usu-
ally harder to detect than open area species. But although forest bird species accu-
mulation curves for all three studied habitats are not fully saturated, they show
flattening towards the asymptote (Fig. 16.2B). A further “check” on observed forest
bird species richness using the average of nine non-parametric species richness
estimators (Table 16.4) indicates that in Gmelina forest, homegardens and shrub-
land respectively 72 percent, 61 percent and 51 percent of forest bird species has
been detected. When compensating for “missed” species, the estimated forest bird
species richness is 13 to 15 species for Gmelina forest, 8 to 9 species for homegar-
dens and 11 to 13 species for shrub-land. Using these estimated forest bird species
richness figures, Gmelina forest and homegardens still have very little representa-
tion of forest birds occurring in nearby natural forest compared to studies of tree-
based human-altered habitats in tropical landscapes elsewhere.
A second explanation may be related to the characteristics of the habitats studied.
At local-scale level, the vertical complexity of vegetation, the presence of tall trees
and a relatively high canopy cover all enhance the suitability of human-altered habi-
tats for forest birds (Hughes et al. 2002; Sodhi et al. 2005; Peh et al. 2005; Sekercioglu
et al. 2007). In our study area, Gmelina forest has the highest canopy cover, the tallest
trees and the highest tree densities. The highest forest bird species richness and abun-
dance are also found in these plantations. Correlations are however not significant.
Although homegardens have a higher average canopy cover than shrub-land, other
habitat characteristics such as the tallest tree and tree densities are comparable.
Homegardens do not have more forest bird species than surrounding shrub-land. Both
homegardens and Gmelina forest have low mean canopy covers (28 percent and 46
percent respectively) and relatively low means for the tallest tree (16.5 and 19.9 m;
Table 16.5) compared to agroforestry habitats in other studies. Thiollay (1995) sur-
veyed birds in agroforestry systems with 35 to 45 m high near-closed canopy, i.e., a
height and cover comparable to that of natural forest in Sumatra. In the Subic Bay
area in Luzon forest birds hardly occur in human-altered habitats when canopy cover
is below 30 percent, with most forest birds requiring at least 60 percent cover (Posa
and Sodhi 2006). The Gmelina forest and especially the homegardens in the Cagayan
Valley seem to lack the vertical complexity and the minimum canopy cover needed
to be important habitats for forest birds.
At landscape level, the proportion and quality of forest or tree-based habitats in
a predominantly human-altered landscape influence forest bird persistence (Marsden
et al. 2006). Likewise, the quality of the human-altered landscape surrounding