The usefulness of a number of parameters is investigated for describing the seasonal variability in insect populations, especially in relation to sample size. It is found that Seasonal Niche Breadth (B) and its maximal value (BM) are useful only for large samples. For smaller samples one tends to find the same values for these parameters whatever the distribution they are supposed to describe. The “standardized” ratio of these two parameters (B/BM) is completely useless. The Season Length (S. L.), defined as 52 minus the longest series of weeks in which the species was absent, or not observed, was found to be quite useful, especially at smaller sample sizes. For species which occur over a large part of the year and have been observed in large numbers, the ratioMin/Max is found to be quite useful. This is the number of individuals during the four weeks at the low point of the season divided by the number observed during the four weeks in the peak of the season. It is suggested that the seasonal distribution be estimated usingS. L. or, when appropriate,Min/Max in order to estimate the Seasonal Standard Deviation (SSD), which assumes a normal seasonal distribution. With thisSSD the 99 per cent range (Seasonal Range) can be calculated to characterize the seasonal distribution. With this Seasonal Range different species or different areas can profitably be compared in their seasonality characteristics even if there is a large difference in sample size.
- 1 - sciences appliquees, technologies et medecines
- 2 - sciences biologiques et medicales
- 3 - sciences biologiques fondamentales et appliquees. psychologie
- 1 - Life Sciences ; 2 - Agricultural and Biological Sciences ; 3 - General Agricultural and Biological Sciences