No abstract

Some factors which affect the aerial counts of dugongs and sea turtles were examined experimentally. There was no significant difference in the observed density of dugongs when survey height was doubled from 137 m to 274 m with an accompanying doubling of transect width on either side of the aircraft from 200 m to 400 m. In contrast, a significantly higher density of turtles was observed at the lower heightharrower transect width. The higher level of glare on the exposed side of the aircraft, the time of day and the time from high tide made no significant difference to the observed densities of dugongs or turtles. The survey crew included a tandem team of two observers on each side of the aircraft, who reported their uncolluded observations into separate tracks of a two-track tape recording system. This allowed the reports of tandem observers to be compared in order to assess observer reliability. Overall, observers missed over 40% of dugong groups and over 80% of turtles visible within the transect including groups of more than 10 dugongs. The chance of observers missing a group of dugongs was independent of group size. There was little disagreement between tandem observers about the identification of animals, or the position of animals in the water column. However, observers differed markedly in their categorisation of dugong behaviour and in their counts of animals (particularly dugong calves) in larger groups.

No abstract

The phenotypic affinities of 11 species of Australian mullet are examined using, principally, canonical variate analysis of 32 external body characteristics measured on individuals of each species. The body forms of four species of Liza and two of Mugil are compared within genera. All species proved to be distinct on the basis of the measured variables and the major discriminators proved to be the locations and relative sizes of the caudal, pectoral and first dorsal fins. Some degree of sexual dimorphism was evidenced in the analyses. In a combined analysis of all 11 species, no generic groupings of species were noted. One species, Rhinomugil nasutus: proved to be substantially different in body shape from the remainder. The Mahalanobis distance between Myxus elongatus and Liza argentea was not significant although the distances between all other species were. As shown by a minimum spanning tree analysis, M. elongatus occupies a central position and its relationships require further elucidation. The major discriminators in the combined analysis were again fin locations and relative sizes although some other body depth and width characteristics served to separate R. nasutus from the remaining species.

A test is proposed to determine whether a presumed environmental gradient 1s accompanied by a change in species composition. A measure of asymmetry of the relation between each pair of species leads to the data belng summansed as a skew-symmetric matrix. The colun~n sums of this matnx reflect specles sequencing. A Monte Carlo test procedure is described. The method is shown to be an improvement over a technique proposed recently in this journal (Bunt et al. 1985). Two sets of mangrove distnbution data from northern Austrahan rivers are used for illustrahon.