The ecology of the skylark Alauda arvensis L. on the Canterbury Plains, New Zealand

This thesis investigated habitat associations of the skylark on the Canterbury Plains, New Zealand, a lowland farming area characterized by extensive pastoral and mixed cropping/arable farming. Findings were discussed in relation to recent skylark population declines in comparable agricultural habitat in Europe. Skylark territories were mapped and territory densities and territory sizes were estimated on three farms representative of the study area. Winter occupancy and densities were measured in 1998 and 1999 on representative field types. Experimental investigations into the role of predation on facsimiles of skylark nests and into the potential role of cereal stubble fields as wintering habitat for skylarks were carried out. Skylark territory densities during the early part of the breeding season (October/November) were significantly higher on the extensively managed pastoral sheep farm (1.13 territories/ha) compared with the dairy (0.66 territories/ha) and mixed cropping arable farms (0.49 territories/ha). Lucerne (Medicago sativa) pastures held the highest skylark territoe role of predation on facsimiles of skylark nests and into the potential role of cereal stubble fields as wintering habitat for skylarks were carried out. Skylark territory densities during the early part of the breeding season (October/November) were significantly higher on the extensively managed pastoral sheep farm (1.13 territories/ha) compared with the dairy (0.66 territories/ha) and mixed cropping arable farms (0.49 territories/ha). Lucerne (Medicago sativa) pastures held the highest skylark territoe between 21 and 50% and vegetation height of between 31 and 45 cm were estimated as providing optimum conditions for breeding skylarks. Territory shifts to adjacent crops with a more suitable vegetation structure were evident, indicating that habitat diversity at the farm scale was advantageous to breeding skylarks. Productivity at the level of the individual nesting attempt was compared between intensively managed dairy pastures and extensively managed lucerne pastures. A total of 84 nests were found during the 1997/98 and 1998/99 breeding seasons. Mean clutch size was 3.2 in lucerne pastures and 3.4 in dairy pastures. The nest success rate at the egg stage was significantly lower in lucerne than in dairy pastures. This was due to differences in predation rates between the two habitats. On average 0.23 and 0.66 chicks were produced per nesting attempt in lucerne and dairy pastures respectively. Three artificial nest experiments were carried out to determine the identity of nest predators and to investigate the importance of various habitat variables, including field boundary type, in influencing rates of predation. Evidence from time-lapse video monitoring in conjunction with predator identification from signs left at the nest confirmed hedgehogs (Erinaceus europaeus) as the main predator in lucerne and other habitats. Australian magpies (Gymnorhina tibicen) and Australasian harriers (Circus approximans) were also recorded while predating nests. The difference in predation rates by avian predators between fields bounded by shelterbelts and fields bounded by post and wire fences approached significance, with higher predation in fields bounded by shelterbelts. Predation by hedgehogs occurred significantly earlier near the boundary than in the centre of fields. A separate experiment showed a significant difference in predation rates between four crop types with the highest rate of predation in 'new lucerne' and the lowest in dairy pasture. High predation rates were correlated with more open and shorter vegetation structure. Skylark densities during winter were highest in lucerne and other extensive pastures compared with all other crop types during both years (7.59 and 14.91 skylarks/ha of lucerne; 6.00 and 6.02 skylarks/ha in mixed grass during 1998 and 1999 respectively). Variation in skylark densities was related to field vegetation and field boundary characteristics, field size and seed availability. In 1998 differences in skylark densities between fields were independently explained by field size, with larger fields supporting greater densities. Lower densities were also associated with increased proportions of tall boundary vegetation, confirming the species' preference for open spaces. In 1999 there were significant differences in skylark occupancy between field types with skylarks most often encountered in winter cereals and extensively managed mixed grass and lucerne pastures. An experimental investigation into the importance of surface grain to wintering birds showed that surface grain at a density to be expected in post-harvest cereal fields led to increased skylark numbers in the experimental area for approximately six weeks after the spread of the grain. This thesis demonstrated the importance of extensively managed pastures, including lucerne, as favoured habitats for breeding and wintering skylarks, compared with other intensive pasture and arable habitats. Conservation measures leading to the extensification of pastoral and arable farming and to greater habitat diversity at the farm scale are likely to benefit skylark populations. The establishment of managed lucerne crops could prove especially valuable in the conservation of skylark populations. However, conservation measures should take into account the skylark's preference for open, unconfined spaces and they should aim at providing breeding habitat in which the risk of nest predation is minimized. Future research needs to be primarily directed at collecting long-term demographic data to be able to build complete population models of skylarks.