| Abstract |
Knowledge of population size is essential for the management of animal and plant populations. Direct count methods such as strip counts provide accurate estimates of objects abundance, but such methods are only applicable in more open ground areas, where the objects can be easily seen [1]. An alternative approach which reduces any bias and improves the precision of abundance estimates is the use of distance sampling methods. One main approach of distance sampling is point transect sampling [10]. In the conventional method of point transect sampling, the number of points and the time spent at each point are fixed, with the number of distances detected being random. Conceptually, it is possible to do the reverse, that is, fix the number of distances detected and stop counting at a point when the predetermined number is reached, without any significant difference in their statistical properties [3]. In [9], Opao attempted to apply sequential estimation methods to point transect sampling and successfully developed a fully-sequential sampling procedure for point transect sampling. Opao [9] came up with comparisons between the performances of the conventional method of point transect sampling versus his fully-sequential sampling procedure in terms of sample size, accuracy and precision. However, in his paper, the conventional method and the sequential estimation method appeared to perform just as the same on the average in terms of sample size, accuracy and precision with consideration of different values of cost. Motivated by the study of Opao, this paper developed a two-stage sequential procedure for point transect sampling with the hope of improving Opao’s [9] fully sequential method. |
Mindanao State
University - Iligan Institute of Technology