Prunus Africana is a cherry tree that occurs in montane forest regions of Africa, for example on the slopes of Mount Cameroon. Its bark is used for medicinal purposes and is harvested commercially for pharmaceutical companies. The extent of the harvesting meant that there were concerns of over-exploitation and so it was listed on CITES Appendix 2. Consequently the government of Cameroon needed to set a national quota for the amount of bark, and therefore the number of trees, which could be harvested each year.
Previous forest inventories of P. africana came under scrutiny because the species seems to occur in relatively sparse clusters. Traditional sampling techniques, such as strip transects, tended to find very few clusters, and even then only part of the cluster would be included in the sample.
Adaptive cluster sampling (ACS) was developed for estimating the abundance of species that occur in rare and clustered populations. Essentially the area to be surveyed is divided into a number of units. An initial random sample of some units is taken. If the number of individuals found in any one unit is more than a pre-specified threshold then all the surrounding units (using a pre-defined definition of surrounding) are also sampled. If any of these meet the threshold then the surrounding units are sampled. This continues until no more units meet the threshold. Methods for estimating the total number of individuals in the survey region can be calcuated together with an estimate of how precise this estimate is.
It was decided to carry out a pilot field trial of ACS on Mount Cameroon and to compare the efficiency and cost with conventional strip sampling. Field protocols were developed based on the methodology,
The pilot study obtained an estimate of the number of exploitable trees on Mount Cameroon. The cost of obtaining an estimate of equal precision using strip transects would be 70% higher. More P.africana were observed than under strip transects. Field workers preferred using ACS because they saw more of the species of interest although they found the methodology quite difficult to implement – in particular keeping track of which units they needed to sample.
Recommendations of how to apply this method for the national inventory were made, including local capacity building. and some ideas for extensions and improvements of the method were provided.
My role was to take the academic literature describing the methodology and (1) translate this into practical actions for the field work; (2) produce tools to take the data and obtain estimates of the total number of trees with a measure of their precision.
I went into the field, up Mount Cameroon, to understand how it was working in practice and working with a colleague analysed the field data to obtain the final estimates and produced a report describing the methodology and made recommendations.