Other formats

    Adobe Portable Document Format file (facsimile images)   TEI XML file   ePub eBook file  

Connect

    mail icontwitter iconBlogspot iconrss icon

Proceedings of the First Symposium on Marsupials in New Zealand

Analytical Methods

Analytical Methods

Only some of the conventions adopted in the analysis are outlined here; others will become evident in the Results section that follows.

Definition of seasons

Throughout this paper the following groupings of months into seasons are used: December-February = summer; March-May = autumn; June-August = winter; September-November = spring.

Age classes and age determination
Four main age-groups are differentiated in the analysis:
(i)pouch-young;
(ii)yearling;
(iii)two-year-old;
(iv)adult.
(i)

Pouch-young

For analyses of pouch-young survival, the arbitrary end to pouch life was assumed to be 170 days, (although emergence from the pouch is a gradual process, Crawley (1970) noting ages of 137 and 175 days for earliest pouch emergence and latest pouch occupation respectively). While many young were seen with females beyond 200 days, the more page 93 conservative limit of 170 days was chosen to minimise misinterpretation of records of absent young: through most of the pouch-life this was confidently assumed to be due to mortality, but increasingly during the 'back-young' phase absent young were found later again with their mothers, some even with foster-mothers (Bell 1974). The 170 days figure is based mainly on data from the study area, especially Crawley (1970), supplemented with those of Dunnet (1956), Lyne and Verhagen (1957), Gilmore (1966), Smith et al. (1969), How (1974) and Kean (1975).

Birth dates were estimated by aging pouch-young on head length and/or tail length measurements using linear regression lines from a sample of newly born young measured at intervals during their pouch-life (Crawley 1970, Bell, unpubl.). For head length and tail length the length (Y) on age (X) regressions1 were, respectively: Y = 0.3430X + 7.1736 and Y = 0.7984X + 2.6113; age (X) was determined using the reciprocal of Y on X following Simpson, Roe and Lewontin (1960).

(ii)

Yearlings

These are animals in the first calendar year after their year of birth, representing ages from about 7½ to 19½ months. Each year newly caught yearlings were identified by comparing their body measurements with those of known-age yearlings known as pouch-young the previous year. The term 'juvenile' was also occasionally used to designate those young caught alone at the end of their year of birth prior to their entering the 'yearling' age-class in January.

(iii)

Two-year-olds

These are animals in the second calendar year after their year of birth representing ages from about 19½ to 31½ months. Most (89%) were formerly trapped as yearlings; the rest were aged by comparison with body dimensions of known two-year-olds.

(iv)

Adults

Adults comprise all possums approx. 2½ years or older, i.e. those in the third calendar year after their year of birth onwards. This amalgamation of all adult year classes was necessary since in the initial phases of the page 94 study none could be aged more precisely. Twelve females caught as either two-year-olds or adults could not be aged with confidence on body measurements so were excluded from analyses involving age.

Female breeding histories

Analysis of life-history records enabled investigation of female breeding histories. The amount of information per female per year depended largely on her capture rate. Overall, breeding histories of 268 area A females and 108 area B females were checked. Since many possums were caught over successive seasons a combined total of 1,067 breeding histories was compiled, comprising 854 from area A (1966-75) and 213 from area B (1971-74). Higher capture rates enabled good documentation of breeding for 421 histories (49.3%) in area A and 101 (47.4%) in area B; less complete breeding data were obtained for a further 230 (26.9%) histories in area A and 51 (23.9%) in area B; for the remaining cases (203 area A, 61 area B) the breeding performances could not be assessed since captures were too infrequent.

Females of known breeding history were categorised as:

(i)non-breeders (those evidently having no pouch-young);
(ii)those having young but losing them before 170 days;
(iii)those successfully rearing pouch-young over 170 days.

1 For head length on age regression: sample size 23, standard error of estimate 2.0879, correlation coeficient 0.9941 (p<0.001); for tail length on age: sample size 18, standard error of estimate 2.9731, correlation coefficient 0.9948 (p<0.001). The head length regression was based on young up to 170 days old, the tail length regression on young up to 100 days old.