The behaviour of six geographic races of three Poephila species of Australian grassfinch was compared in order to study the microevolutionary changes in behaviour within the genus. Most of the observations were made on wild-caught birds but some observations were made in the field. Five of the six subspecies bred in captivity. Their behaviour was recorded on movie film and their vocalizations on sound tape. Films were analysed on a single-frame analysing projector and the tapes analysed on a sonagraph machine.
All subspecies produced complex sonagraphed sounds. Each species had four juvenile calls and one song; P.a. hecki had 14 other distinct adult call-notes, P.a. acuticauda, P.c. cincta and P.c. atropygialis had 12 and P.p. personata and P.p. leucotis had 11. The notes of the first two subspecies were relatively pure, with few harmonics and high fundamental frequencies; the third and fourth subspecies had more noise and harmonics and a lower fundamental frequency whereas the last two subspecies had harsh nasal notes with low fundamental frequencies and numerous harmonics. Conspecific races differed in the same way but to different degrees. These differences were exemplified in the loud, distant-attraction Long Call. Statistical analysis indicated that each species had its own fundamental frequency broadcast range within which each subspecies had its own distinct range; individuals from the same geographic locality were more similar in frequency than those from different localities. Each individual bird had its own uniquely structured Long Call; variations in the fundamental frequency, the sonance and timbre characterized the call. In some subspecies the Long Calls were consistently different between the sexes. Males of each species had a long, complex series of notes - the song phrase, which they performed in two situations: (1) in front of the female at the end of precoital courtship and (2) during non-courtship situations at breeding when the female was not in sight (Undirected Song). Song phrases lasted approximately 2.5 seconds in P. acuticauda and P. cincta and less in P. personata. The first two species had successive phrases separated by pauses; each phrase had a soft introductory series of notes followed by a main section of louder, more musical notes. The three species differed in the number, type and amplitude of the notes. Races in each species differed in their phrase structure; the differences were large in P. acuticauda and small but consistant in P. cincta and P. personata. Each individual male had his own unique song phrase but males from the same geographic region were more similar than those from different regions - that is, there were song 'dialects'. Sons learnt their fathers' songs. The development of the adult song took nearly a year in captivity. During the evolutionary development of the song in the Poephila group there are tendencies to increase! structural segmentation, phrase and note stereotypy, simplification of notes, loudness, duration and musicality. During ontogeny the call-notes developed from the food begging calls.
Seventy maintenance behaviour patterns of the individual bird were observed; details of the sequences of bathing and sunbathing movements are given. Poephila are highly social throughout the year and breed in colonies of 20-30 individuals. They Head Jerk on greeting, occasionally clump and mates allopreen. They hold Social Meetings where the colony members congregate and exchange Mandibulation Stretch greeting displays; P. personata also Tail Quiver and exchange allopreening partners. Only flock members greet in such a way, strange conspecifics are first courted and raped, then attacked and later ignored. There are few agonistic signals; supplanting and beak-fencing are the main types of fighting. Birds fight mostly in the defence of the mate or nest. Mates never fight; they may incite one another to attack a third bird. Mates may greet each other with a special type of peck, the Low Twist Peck.
Poephila pair for life; mates remain in close association throughout breeding and non-breeding seasons. The bond is maintained by many behaviour patterns but formed by stage 1 and 2. courtship and nest solicitation displays. Both sexes are equally active in bond—formation and maintenance. The bond has its own motivation and is not predominantly sexual. Courtship has three stages; stage 1 consists of the courtship initiation followed by the long sequence in which both birds hop to and fro about each other on the perch; stage 2 begins when the male stops and sings directly at the female; stage 3 follows when the female solicits copulation and the male mounts; a stage 4, a post-copulatory greeting, occurs in two of the Poephila species. Courtship of the subspecies was analysed in detail as its use at pair formation would result in strong selection pressures for divergence. In stage I the form of the displays differed only slightly between species but not between subspecies. A computer-based classification and principal co-ordinates ordination of rates of performance of stage 1 displays revealed species groups but no infraspecific groupings. The sequence of displays, not their progressive tempo or dance pathways, was found to differ between the subspecies. In stage 2 some subspecies differed in their rates of singing. For courtship to result in copulation both birds must achieve a certain rate of displaying; they mutually arouse and stimulate each other but there is no definite one to one response relationship. The function and evolution of courtship in Poephila is discussed.
Other breeding behaviour is described in detail for captive birds and for free-living P.c. atropygialis. Analysis of mouth-markings of first generation aviary-bred nestlings indicates large variation in this taxonomic character. Undirected Singing is frequent at the start of breeding. The sight, not the sound, of the female inhibits her mate from singing; flock members but not strange conspecifics also inhibit it. Undirected Song aids in maintaining vocal contact between the female in the nest and the male out of sight nearby; this occurs when the pair bond is being loosened at the start of breeding.
The microevolutionary changes in the behaviour of Poephila seem to be greatest in the vocalisations and the colour patterns of the plumage and beak; courtship was unexpectedly similar. Selection has probably been strongest on these long distance signals. The main function of courtship is to arouse and synchronise the mates and is not to act as an isolating mechanism The differences between the behaviour of subspecies were quantitative and probably due to differences response threshold. Similar trends are found in other species of birds.