Dung surveys are extremely time- and labor-intensive, and can be expensive if the survey area is large. This method assumes a stable elephant population creating stable rates of dung accumulation and decay during the study period. Because dung decay times are long, more rapid changes in the elephant population may lead to inaccurate estimates. The rate of dung decay and defecation can also vary with habitat and season.
While less clumped than elephants, dung piles are typically not randomly distributed. To avoid inaccurate estimates due to the clumped distribution, the transects need to be arranged throughout all vegetation types in a manner that is representative of the different habitats. It is often advisable to stratify the study area by vegetation type and/or by poaching intensity. A minimum of 30 transects per stratum is advised to improve the precision of estimates.