Census and Estimates

Census and Estimates

There are several methods of obtaining Population Census and all of them give rough estimates of the actual count. There is an inherent problem with animals estimates: "All animals are secretive and mobile".

To overcome such barriers we must conduct our investigation in a very short period or accommodate mobility into the study area. As its almost impossible to obtain an accurate census, rough estimates are taken which are then extrapolated to form a census. If direct observation is not available, indirect observation is used to obtain rough estimates of animal populations (Ex: scratch marks on trees, feces counts, calls)

If the population being studied inhabits an area too large for a complete inventory then sampling becomes necessary. There are several types of sampling techniques that can be used to obtain a census. Basically these are conducted on only a part of the space or time dimension of the population. The section sampled is assumed to represent of the combines population. Once this is obtained a statistical estimate of the entire population can be made from the sample.

There are 5 basic sampling techniques that are practices.

• 1. Simple Random Sampling requires that the population be subdivided into equal space or time units. Each sub unit should have an equal probability of being selected for sampling. As this is the most easiest method, a random number is picked (using a dice) and the corresponding grid location is quantified and the overall census estimated.

• 1. Stratified Random Sampling requires prior knowledge of the population so it will be possible to subdivide the area into separate homogeneous sub-populations. For example: We might know the Elephant population at Udawalawe to be the highest, while the Yala population to be much lesser. Using the above mentioned Simple Random Sampling method, we can get a rough estimates from both parks to obtain a total population census. Stratified Random Sampling has the advantage of providing data on subdivisions of the population thereby giving more accurate estimate in animal populations.

• 1. Systematic sampling is a method used when sample grid units are spaced at regular intervals. This method is used when the study area is marked off in grids and every third or fourth grid is sampled. If the animal distribution is homogeneous, this method becomes highly accurate.

• 1. Two-stage sampling is used when very large populations are to be censused such as the Sri Lanka pelican population. The Simple Random Sample Method could be used to determine which districts would be sampled. Within the chosen districts a random sample would then be taken of a certain number of smaller land units. Two-stage sampling is basically a means of working down to a sampling unit of manageable size.

  2. Double sampling is a modification of the Two-stage Method whereby on the county level one might count the number of raccoons seen along the road as roadkills. This extensive sample would then be strengthened by intensive sampling on smaller land units to determine what type of population density contributed to the associated count of road killed raccoon.

Population variability

Populations are subject to change based on 4 known factors.

• 1. Natality (rate of birth)

  2. Mortality (rate of death)

  3. Emigration (Local population moving out of the designated areas)

  4. Immigration (Foreign population moving into designated areas)