Mating behaviour and reproductive success in the reed bunting Mmberiza schoeniclus.

The reproductive behaviour and success of a number of male reed buntings were studied at Rutland Water (Leicestershire) during the the breeding seasons of 1988, 1989 and 1990. A major constraint on reproductive success was the level of nest predation. Only 42% of nests survived to fledging, representing 41.5% of males' territories. The annual mean productivity for each male was estimated to be 0.77 offspring surviving to the following season. With these large constraints on male reproductive success, extra-pair paternity is seen as potentially highly advantageous to males. Females are constrained in their mate choice indicating a potential advantage in seeking sires from surrounding males of higher quality than their mate. Rates of extra-pair paternity as derived from DNA fingerprinting were found to be high: 50% of chicks and 69% of broods. Approximately one-third of males were responsible for 51 cases of extra-pair paternity. However, 70% of males lost some paternity in their nests to other males with no correlation between levels of extra-pair and within-pair paternity. Copulations were observed to be frequent (estimated to be 4.06 per day) during the prelay period and continued through to the day of the third egg. A large between male variation in copulation rates was observed, with a diurnal peak immediately following the laying of an egg. Observed extra-pair copulations were rare (4.39%) with a significant disparity compared to extra-pair paternity levels, indicating that females may seek furtive extra-pair copulations. Mate guarding as measured by time spent less than 10 m from the female, flights orientated towards the female and vigilance were observed to peak during the prelay period. These behaviours declined dramatically on the day of the first egg, indicating that peak fertility probably occurs prior to egg laying. No increase in mate guarding activity was observed to occur in response to an increased neighbour density, whilst neighbour status did not influence guarding. A diurnal pattern in guarding was recorded for vigilance and all three behaviours showed significant between male variation. Territorial intrusions were mainly by neighbours; they peaked significantly during the prelay period and were responded to by an attack from the defending male. The seeking of extra-pair copulations was observed as excursions into neighbouring territories, followed by an attack from the territory holder. Excursions occurred primarily when males were not mate guarding, during egg laying and incubation, and were significantly directed towards territories with nests in the prelay stage. Song was found not to function as a mate guarding behaviour, with a significant decline in output during the prelay period. Song output increased significantly in the presence of a neighbour with a "fertile" female, supporting the hypothesis that song functions as a measure of male quality to fertile females Song of unpaired males was significantly different from that of paired males and is hypothesised to function in territorial maintenance (and to indicate male status), whereas the more complex song of paired individuals also functions to display male quality to their partners. An analysis of the relationship between behaviour and reproductive success when analysed indicated that males which guarded most were more subject to losing paternity. Acquisition of extra-pair paternity was found to be significantly enhanced through increased song output and excursions. Song output was also recorded as significantly increasing total reproductive success, indicating that females may choose males on the basis of their song to obtain furtive extra-pair copulations