Conservation monitoring: estimating mammal densities in woodland habitats

Abstract

Conservation and ecological monitoring programmes often estimate animal densities over time, but in wooded and forested areas practical techniques are still poorly developed. Here I have examined five simple methods of deriving densities of large and medium-sized mammals using line transects driven through miombo woodland habitat in Africa. These methods calculated area by dividing the number of individuals seen by (i) an average of each species' sighting distances, (ii) a fixed 200 m belt width, (iii) the area visible from the centre of the transect, (iv) visible area weighted by species ' vegetation preferences, and (v) by dividing the number of groups seen by area visible from the transect. Individual-based methods produced quite divergent estimates of species' densities and overall biomass with belt transects giving the lowest values. The group method and corresponding individual-based method gave similar values, however. Most of these calculations yielded considerably higher density estimates than aerial surveys conducted in the same area over a similar time period. Sightings of species were not distributed evenly across vegetation types although the majority of mammal species were observed more often in open habitat. These findings indicate that ground-based conservation monitoring programmes should set transects through several vegetation types, restrict comparisons to those studies that use the same methodology, and refrain from comparing ground and aerial surveys.

Cost and Efficiency of Large Mammal Census Techniques: Comparison of Methods for a Participatory Approach in a Communal Area, Zimbabwe

“…The heavily bushed or wooded savannah which covers much of our site may require a low speed, and faster methods such as car counts may be less effective in this context. The vegetation constraint is even more problematic in aerial surveys, where the classical undercounting bias of aircraft counts (Caughley 1974;East 1998) is especially pronounced in areas with high vegetation cover (Bayliss and Yeomans 1989;Southwell 1996;Caro 1999b;Jachmann 2002). This is consistent with the low detection efficiency we found with aerial censuses of our study site.…”

“…Density of vegetation cover will modify the detection ability of methods, through visibility bias (Caro 1999b;Jachmann 2002). Vegetation type may set a threshold in cruising speed during a count for detection to be efficient.…”

Monitoring large herbivore diversity at different scales: comparing direct and indirect methods

“…Aerial counts have indeed been shown to have an equally low sampling effort as dung counts (Jachmann 1991). Several studies, however, pointed out that aerial censuses hugely underestimate abundance and are strongly biased towards the largest species especially in forest or woodland habitat (Caughley 1974;Caro 1999;Barnes 2001;Jachmann 2002;Gaidet et al 2005). Since the major part of African reserves is covered by these habitats we argue that aerial censuses are unsuitable to monitor mammal species diversity.…”

Overcoming the challenges of measuring the abundance of a cryptic macropod: is a qualitative approach good enough?

“…Indirect methods of surveying population size are often considered a more cost effective and practical alternative to mark-recapture studies, despite weaknesses documented in the literature relating to known biases, inconsistent detection and an inability to meet model assumptions (Nichols and Pollock 1983;Montgomery 1987;Slade and Blair 2000;McKelvey and Pearson 2001;Hopkins and Kennedy 2004). Indirect monitoring techniques such as the use of photographic captures (Karanth and Nichols 1998), driving transects (Caro 1999;Olson et al 2005), walked line transects (Short and Turner 1991;le Mar et al 2001;Poole et al 2003;Wayne et al 2005) and dung-pellet counts (Johnson and Jarman 1987;Allen et al 1996;Buckland et al 2001;Hayward et al 2003) have been widely used to provide a quantitative estimate of population size for small and mediumsized mammals. More recently, these have been coupled with sophisticated models that can actively account for changing detection probabilities (e.g.…”