Estimating Absolute Densities of Elephant Populations Using Dung Counts Along Line Transects: Field Methods

Citation

Hedges S, Maisels F & Blake S (2012) Estimating Absolute Densities of Elephant Populations Using Dung Counts Along Line Transects: Field Methods. In: Hedges S (ed.) Monitoring elephant populations and assessing threats: a manual for researchers, managers and conservationists. Himayatnagar, Hyderabad, India: Universities Press, pp. 172-213. http://www.fws.gov/international/pdf/Monitoring_Elephant_Populations_and_Assessing_Threats_to_press_version.pdf

Abstract
First paragraph: As we saw in Chapter 4, there are three essential components to a dung density based population survey: estimating dung pile density, estimating dung decay (dung disappearance) rates and estimating defecation (dung production) rates. The first two of these components requires dung piles to be classified into stages based on their state of decay, and so unambiguous dung classification systems are also of great importance. The method of estimating dung pile density recommended in this manual is the line transect [Buckland et al. 2001; Buckland et al. 2004; Chapters 3 and 4]. In recent years, a number of so-called ‘recce' (reconnaissance) survey methods have been used in conjunction with line transects in an attempt to reduce the logistic challenges posed by line transect based methods and improve the precision of dung count surveys [Walsh and White 1999; Chapter 4]. The recce transect method, as described by Walsh and White, involves walking along a ‘path of least resistance' through the forest and counting all dung piles found but not measuring perpendicular distances to these dung piles. The recce data set can then, in theory, be used to estimate dung pile density providing the functional relationship between encounter rates on recces and line transects was derived from a subset of recces matched with line transects. However, the use of these recce methods to estimate elephant dung pile density is considered problematic (Section 4.2.2) and is not therefore recommended in this manual.