page 97

A Five-Day Week Calendar for Use in Southern Hemisphere Field Studies

by E. C. Young
Zoology Department, University of Auckland

Students doing field studies in ecology or life histories in the Southern Hemisphere are inconvenienced compared with their northern colleagues in not having a tidy relation between the biological annual cycle and the common calendar. Especially important is the lack of a precise start point, corresponding to January 1, from which to tabulate observations on seasonal species.

This problem has been felt particularly in a study of the relation between skuas and penguins in Antarctica which has necessitated the grouping of sparse data over several days. For convenience five-day runs were selected, giving fair numbers of observations in each interval and providing a mid point for recording the mean and variance. In a single year the start of this grouping series can be located at any convenient date, perhaps at the first observation, but difficulties arise when several years are to be compared. Where, then, should the first interval grouping be begun? At the first of all the observations; about the obvious groupings of records, allowing conscious bias; or simply at a selected fixed point, say November 1? If the latter, then what happens if later studies provide even earlier observations; or work on other species with an earlier cycle is to be compared? To overcome these problems a five-day week calendar has been devised (not for the first time one imagines), beginning on June 24. Starting at this date (instead of the actual shortest day) brings November 1, December 1 and March 1 to the start of week periods; giving the closest correspondence possible to the common calendar. No allowance is made for variation in the length of February so that on leap years the last week in this month has six days.

Using this calendar, observations are assigned to the five-day interval in which they occur. Unconscious biassing of results is prevented and the problem of where to begin grouping is obviated.

This calendar has advantages over the simpler one which breaks each month up into six five-day weeks irrespective of actual month length, in having only a single variable week, in February.

Similar tables can be worked out for seven-day weeks. One beginning on June 20 brings July 1 and April 1 to the start of a week interval; the best that can be managed in this system to relate it to the common calendar. Shorter intervals, say three-day periods which are also commonly used, are less likely to cause problems of data grouping.

It is not intended that this system should replace the normal calendar in the present studies except for the special situation

page 98

Five-Day Week Calendar

page 99

outlined above. The difficulty of needing continually to relate these ‘weeks’ to those of the calendar months argues against its widespread use in descriptive work except where grouping of data is required.