Wildlife Trends

COMACO sampled areas

Aerial transects were flown in 1999, 2002 and 2006 to evaluate population trends for 10 wildlife species for the COMACO core area [Chikwa, Chifunda, Chanjuzi (Kazembe and Chitungulu combined) and Mwanya, see map below]. These species are higher risk, lower density species. Munyamadzi Game Management Area was included in the sample because of comparable food security efforts undertaken throughout the same period as the core area, although COMACO market support was not in place in that area. Fixed, parallel aerial transects were flown with sampling intensities varying from 10.2% to 19.8%. Selection of the species censused was based on several criteria, including large body size and vulnerability to snaring because of frequent use of waterholes during the dry season. Use of snares around waterholes is a common practice to take advantage of predictable animal movements. Those species judged to be most at risk to poaching, at both population and individual levels, because of low local density, included wildebeest, waterbuck, eland, hartebeest, roan and kudu. For the purpose of this summary, we term these the poaching susceptible species. Elephants were included in our census because of their conservation value. Buffalo were initially included because of their status as the preferred target of illegal gun hunters (data not shown), but an accurate estimation of their populations is proving difficult due to their heterogeneous distribution (i.e. clumping), which leads to extremely high variation. Puku were included to represent one of the small-bodied species with a relatively high reproductive rate.

In addition to a species-by-species analysis of the combined area, the study also looked at population trends within each area surveyed to determine if there were local differences in response. Because of their high variance and small sample size within the individual surveyed areas, data from the poaching-susceptible species excepting wildebeest were combined for some analyses. Wildebeest had relatively large group size variation as compared to the other species, which generally occurred in smaller size groups of 5 to 12. Therefore, inclusion of this species with the other poaching susceptible species would have skewed the data.

The Tables below show the combined surveyed area data and the combined population trends respectively, for the high-risk, low-density species group,. For the combined surveyed area data, only hartebeest showed a significant increase from 2002 to 2006.

Most species showed positive, upward trends, but these were statistically insignificant. Wildebeest and eland numbers showed negative, downward trends, which again were insignificant. Buffalo were excluded from the analysis because of extreme heterogeneity in spatial sampling. In the analysis of population trends for individual surveyed areas for the combined poaching-susceptible species, two of the four areas from the COMACO core areas, Chanjuzi and Chikwa, showed significant increases (p < .05 ) from 1999 to 2002 and from 2002 to 2006. All the other areas showed positive but statistically insignificant trends, with the exception of Mwanya, which showed a negative but statistically insignificant change.

In summary, population estimates from the 2006 survey showed a preliminary indication of population change for selected species relative to population sizes prior to COMACO (1999) and at the beginning of the COMACO intervention (2002). No significant population decreases were found for any of the analyses. Significant increases were found for hartebeest for the combined surveyed area sample and for two of the four COMACO core areas for the high-risk, low-density species group. Of interest, almost every species showed a downward trend from 1999 to 2002, suggesting that populations were in decline at the beginning of the COMACO program. Increases in all populations would be unlikely within 4 years, given the likelihood for firstly a stabilization of species occurrence, followed by stronger evidence of recovery of population numbers.

Population trends observed throughout the total Core COMACO program area. D-test comparing 2006 to 1999 and 2002 shown, with significant values shaded (p<0.05, ~25 df).

Data combined for the poaching-susceptible (large-bodied, low density) species (waterbuck, roan, eland, kudu and hartebeest). Comparisons between 1999 and 2006, and 2002 and 2006 are shown, with significant values shaded (p<0.05, ~25 df)

Comparison with non-COMACO areas

Other regions of the Luangwa Valley were not included in the original aerial censuses because focus was on alleviating the heavy poaching pressure around the two most significant national parks, the South and the North Luangwa National Parks. COMACOs focus is on the human populations in the game management areas, which serve as buffer zones around these precious resources. However, from a scientific perspective, comparable population trends between areas experiencing the COMACO intervention, and areas outside of COMACOs activities would call into question the degree of COMACOs effects on wildlife populations. In 2006, three different control areas where COMACO is either not active, or does not have a strong presence, were censused and results were compared with those from the COMACO core area.

Absolute animal count for species surveyed in 2006. All areas were sampled during the second half of September, using the same observers and sampling procedures.

As can be seen from the above table, the COMACO core area has much higher animal counts than any of the other areas. Perhaps most disturbing of these findings is that the Lukusuzi National Park had so few animals, and such limited species represented. Clearly, these studies need to be repeated at different times of year to factor out potential effects of seasonal animal movement. However, several other possible explanations exist in addition to animal movements. These areas might never have had significant wildlife populations because of differences in habitat or human populations, and might therefore not be similar enough to function as acceptable controls. Arguing against that possibility is the fact that the three control regions are geographically found to the north, south, and east of the core. Regarding habitat and access to water, they include the same mixed mopane woodlands and two borders or include the Luangwa River. In addition, the third control is a protected National Park, without human dwellings within it. In light of Lukusuzis proximity to so many individuals practicing outgrower farming schemes to produce cash crops, such as cotton and tobacco, one can understand why a good number of the individuals who hunted illegally as a profession reside immediately around this park to secure inputs and possibly food if less time was devoted to food crops in comparison to cash crops. BACK TO TOP