Thumbnail Image

Improving Conservation Efforts Through a Better Understanding of Forest Elephant Ecology, the Impacts of Threats on Elephants and Freshwater Fisheries in Northern Congo

IMPROVING CONSERVATION EFFORTS THROUGH A BETTER UNDERSTANDING OF FOREST ELEPHANT ECOLOGY, THE IMPACTS OF THREATS ON ELEPHANTS AND FRESHWATER FISHERIES IN NORTHERN SEPTEMBER 2016 ROGER PATRICK BOUNDJA, B.Sc. FORESTRY, MARIEN NGOUABI UNIVERSITY, BRAZZAVILLE MSc. UNIVERSITY OF CAPE TOWN PhD. UNIVERSITY OF MASSACHUSETTS AMHERST Directed by: Professor Curtice R. Griffin Catch data, including fish numbers, length-weight were collected during a multi-mesh Gillnet fisheries-independent survey in 2007-2008, and fisheries-dependent mixed gear surveys in 2009-2010 and 2015 across 400km stretch of the Sangha River located in the Sangha Tri-National. Overall, very high species richness (Chao 2 mean=250, SD=16.15) and diversity index (Simpson Inverse Mean=43.72, SD=0.02) estimates suggesting that the Sangha River mainstream could host one of the highest freshwater fish diversity across the region. Overall, Canonical Correspondence Analysis (CCA) suggested a high total variance in species data, of which significant proportions explained by environmental variables. Latitude and distance to park borders had the strongest importance in ordination, and fishes tended to avoid high human-disturbed areas near large towns and high water turbidity. A special focus was further put into 7 fish species, recognized for their economic value in local fisheries. Catch Per Unit Effort (CPUE) and contribution in local fisheries were computed for the 7 species for main fishing techniques. Overall, significantly highest average CPUE was found in wet season months, and set Gillnets appeared the most efficient but were mainly targeting small-size and immature fishes and therefore not a sustainable fishing technique. Extrapolations from otolith readings suggested that an overwhelming majority of harvested fishes were estimated to be only a few months old, under their age-at first maturity. Finally, Classification And Regression Tree (CART) of the same 7 focal species’ CPUE suggested that Citharinus gibbosus was the most important species, predicted with at least 55% chance in Set Gillnets, followed by Mormyrops anguilloides with at least 64% chance in basket-traps. Overall, during a 5-year seasonal camera survey carried out at the core-area of a nearly pristine Nouabalé-Ndoki National Park in Northern Congo, camera trap sites were occupied by elephants nearly 82% of weeks but elephant occupancy was not affected by any of environmental variables in the model. However, detection probability was 0.182, and significantly affected by percentage of Gilbertiodendron Dewevrei mono-dominant forest, mixed closed canopy forest, and number of sampling days. Large elephant social groups were detected near forest clearings or “bais”.