[Received by the Editor, July 6, 1959.]

Abstract

Riblet frequency in single populations of Charopa coma (Gray), Fectola tapirina (Hutton), Cavellia biconcava (Pfeiffer), Ptychodon hunuaensis Suter, Allodiscus tullia (Gray), Suteria ide (Gray), and in two widely separated populations of Phenacohelix ponsonbyi (Suter), has been studied to ascertain the extent of variation within post-nuclear whorls and levels of correlation between such whorls (Charopidae and flammulinidae). In all species studied the standard deviation is lowest in the first post-nuclear whorl and rises sharply in the second. With one exception there is a further but often less pronounced rise in the third whorl. Significant correlations between whorls are most frequent in the case of whorls two and three. The second and third post-nuclear whorls appear to be of comparable value in determining population statistics, but there are other factors which favour the use of the former.

Introduction

Riblet orientation, form, definition, and frequency, both in the protoconch and post-nuclear whorls are variable characters of major importance in the systematics of our terrestrial Mollusca. Frequency is commonly employed at the species and sub-species levels (Suter, 1913).

The present investigation was carried out to ascertain whether any general principles apply with regard to the relative variability of riblet frequency in postnuclear whorls. The use of whorls based on post-eclosion growth (especially of the second post-nuclear whorl) has a number of advantages, but before accepting this it seemed desirable to examine riblet frequency variation in representative species of a number of the genera concerned.

Method

Riblet counts in post-nuclear whorls were made on single populations of 3 flammulinid and 4 charopid species belonging to 7 genera. Delimitation of whorls was determined by tracing ribs between sutures rather than by strictly radial means. The selection of the species depended solely on the adequacy of samples available in the author's collection.

Results

The information in respect of each species may be summarized as follows:

Allodiscus tullia (Gray) (Flammulinidaa) Rimutaka Range, 28. 1.40.

detected and separated are often not evident at this early stage. The increased variability in succeding whorls may be due to the inverse of that stated for the first post-nuclear whorl, that is, the ecological conditions are not so stable, whorls are more time-consuming, and the species is losing the stability associated with youth.

It is possible that in many species the completion of the third post-nuclear whorl signifies maturity but not old age, and that had samples with a larger number of specimens been available and counts made on whorls situated 2½ to 3½ revolutions from the protoconch, then increased variability associated with old age would have been more evident. In Ptychodon hunuaensis there is possibly some evidence of this In this species it would appear that specimens rarely proceed far beyond the third post-nuclear whorl, so that here we have full expression of increasing variability associated with old age. This species besides being the smallest studied is also a fairly robust one which features may contribute to higher survival rates of older individuals.

This aspect may appear to minimize the value of the significant correlations occurring above between whorls two and three, but if the species is to be characterized by the form of the mature individual which most commonly occurs rather than by the occasional old specimen, then the value of the correlations is still real.

In view of this high degree of correlation in frequencies between second and third post-nuclear whorls in species which produce 3 to 4 post-nuclear whorls, there seems to be no reason why the second whorl should not be selected for use in statistical studies, for it does possess certain advantages over other whorls. Where 4 to 5 post-nuclear whorls are usual, the third whorl may possess similar advantages. As shells become older, the riblets on younger portions often become worn away and difficult to count. In older portions of the shells the riblets often become so closely spaced that they present similar difficulties. Most samples of populations contain a fair proportion of younger shells in which the second but not the third whorl is available for counts. Less counting is involved in the second than in the third whorl.

It has been customary in the past to draw riblet frequency data from the socalled “body-whorl”. This is that portion of the shell which runs from the lip back and around to that point exactly adjacent. This procedure introduces a variable which may be eliminated by substitution of frequencies strictly related to the post-nuclear whorl. It is desirable that species definition contain statistics pertaining to all post-nuclear whorls to assist the identification of young specimens.

Acknowledgment

The author gratefully acknowledges the assistance of Mr. A. C. Glenday, of the Applied Mathematics Laboratory, Department of Scientific and Industrial Research, who made the analyses of riblet counts and provided useful suggestions in the presentation of results.

Reference

Suter, H., 1913. “Manual of New Zealand Mollusca.” N.Z. Govt. Printer, Wellington 1120 pp.

R. A. Cumber,