Conservation of biodiversity in agricultural landscapes in Uganda: using birds as indicators.

Abstract

The rapid global loss of biodiversity and habitats is a central crisis now well understood. Conservationists are now putting most of their research efforts into finding ways of combating the resulting effects to conserve what is left of the world’s ecosystems. In this study, twenty-two small-scale mixed agricultural sites and four large-scale monoculture sites were surveyed for a period of one year in 2006 and 2007. Bird surveys were conducted using Timed Species Counts and Point Counts using Distance sampling at each of ten points in all 26 sites in five data-collection rounds. Data on tree density and farmland habitat characteristics were collected for each point. I investigated how the local and landscape habitat characteristics relate with species richness of sites and with the density of 35 common bird species on farmlands of central Uganda. In addition, I looked at the comparison of the two survey methods, Point Counts and Timed Species Counts in assessing the Species richness and density of birds in farmlands of Uganda. And lastly, I looked at the seasonal patterns of bird species richness and density in agricultural landscapes in relation to rainfall.

A total of 218 bird species was recorded, 31 of which are classified as species of conservation concern, globally and/or regionally. Species recorded included 40 forest dependents (FF and F), 75 forest visitors (f), 46 wetland species and 21 grassland species. Bird species richness decreased with increasing human population density and the proportion of cultivated land at the sites, but increased with increasing crop diversity. Species richness for species that are more sensitive to forest habitat loss; forest dependents, frugivores and nectarivores, increased with increasing numbers of trees at the sites, especially large trees and with increasing proportion of fallow land at the sites.

The Common Bulbul Pycnonotus barbatus and Eastern Grey Plantain-eater Crinifer zonurus were the most abundant species recorded during this study. The Red-Chested Cuckoo Cuculus solitarius, Eastern Grey Plantain-eater Crinifer zonurus and Red-eyed Dove Streptopelia semitorquata were found to be the most sensitive of the more common species as they reacted to most of the habitat variables considered. The tree density (especially of indigenous trees) and the proportion of crop land were the habitat variables highly correlated with the density of bird communities in the sites.

The results also indicate that small-scale mixed farms have a higher bird species richness and bird densities than large-scale monoculture farms. Small-sized birds like mannikins and waxbills react more towards local-scale habitat variables while larger birds like turacos and hornbills react more towards landscape habitat variables.

Further results indicate that the sites nearer to Lake Victoria have fewer fluctuations in the amount of rainfall received as compared to those further away (these fluctuations increase as you move away from the lake). The highest number and density of species were recorded in the wetter periods while the lowest species richness and diversity were in the drier periods of the year. Palearctic migrants plus the chats, finches, turacos, parrots, warblers, tinkerbirds and weavers reacted positively towards rainfall while the cuckoos, flycatchers, shrikes, and sunbirds reacted negatively. Wetland species and insectivores increased in the wetter periods, while forest visitors and predators increased at the end of the drier periods. Nectarivores on the other hand increased towards the beginning of the drier periods. This study therefore suggests that bird species trends seem to follow rainfall seasonality but this needs further research as other factors like behaviour of birds in different times of the year may be in play.

In addition, there was a highly significant relationship between the frequency of occurrence of bird species estimated from Ten-Minute Counts and that estimated from Point Counts (P<0.0001). Ten-Minute Counts accumulated bird species faster in more sites than Point Counts but this difference was not significant. The number of bird species recorded using Point Counts were highly correlated with Exotic tree density in the sites while a weak correlation existed with the number of birds recorded using Tenminute counts. Crop diversity was positively correlated with the number of bird species recorded by both methods, suggesting that either method can work well in small-holder mixed farms.