Our study found that both skylark and yellowhammer densities were significantly higher during the winter in New Zealand than in Britain, providing further evidence that these species are examples of introduced species experiencing enhanced success in their new environment (Thomsen et al., 2001; Donald, 2004; MacLeod et al., 2005a). However, it is important to establish that our study sites were representative of national trends in each country. Our study sites in Britain were located in an area which supported some of the highest skylark and yellowhammer densities in the country (Lack, 1991), suggesting that the difference in skylark and yellowhammer densities between countries cannot be explained by selection of sites with low bird densities. Density estimates for yellowham-mers and skylarks at the national scale are not available in New Zealand, although similar densities of both species were detected in a larger scale project in the Canterbury region, suggesting that the high densities found in this study are representative of bird densities, at least at the regional scale (MacLeod and Drew,
To determine whether enhanced winter food resources in New Zealand may explain the success of these introduced bird species in their new environment, we compared seed densities in fields in their introduced and native ranges. We found that small seed densities were significantly higher in fields in New Zealand than in Britain. However, contrary to our prediction, large seed densities were lower and more aggregated in New Zealand. In addition, winter seed densities were only significant predictors of skylark densities in Britain. A key question is, therefore, why is seed density a poor predictor of bird densities in New Zealand during the winter?
In our study, skylark densities in Britain were highest on stubble fields where large seed densities were also highest. Skylarks in Britain are known to select stubbles, which have relatively high densities of spilt grain, because the birds are able to feed more efficiently than in areas with low seed densities (Robinson and Sutherland, 1999; Robinson, 2004). Other studies have shown that skylarks in Britain tend to avoid grassland areas during the winter (Gillings and Fuller, 2001; Donald, 2004). The shift toward more intensive farming practices in grassland systems may have reduced seedbanks associated with this habitat, thereby, reducing their value as winter foraging habitats for granivorous farmland bird species (Vick-ery et al., 2001; Robinson et al., 2004). Indeed, numbers of seed-eating bird species in grassland-dominated systems in Britain are positively associated with the area of available arable habitat (Robinson et al., 2001). Furthermore, breeding densities for seed-eating birds are higher in areas with stubbles during winter, indicating their importance in maintaining Britain's farmlandbird populations (Gillings etal.,
Grassland habitats in New Zealand supported the highest small seed densities. Although skylark densities were also lowest in grassland fields in our New Zealand study site, they were still relatively high compared to stubble fields in Britain. Elsewhere, we found that skylarks at our New Zealand study site avoided grassland fields in the mid-winter and that crop type was a poor predictor of skylark distribution later in the season (MacLeod and Till, in press). However, another New Zealand study found that skylarks congregated on Lucerne Medicago sativa and ryegrass pastures and growing cereals during the winter, suggesting that grassland areas may provide important winter foraging habitats in some areas (Thomsen et al., 2001). The differences in habitat selection and relationships between bird and seed densities between the two countries may reflect, therefore, differences in the availability of the different habitats and the food resources associated with them. For example, in New Zealand, stubble fields were found on only 39% of 19 1-km squares surveyed during the winter with an average field size of 7.2 ha
(MacLeod and Drew, 2004), whereas in Britain ~61% of 601 sites 1 km2 in size had stubbles, with 35% having < 10 ha of stubble, and 26% with > 10 ha of stubble (Gillings et al., 2005). Although skylarks have access to more abundant large seed resources in Britain, birds in New Zealand may still have access to better winter seed resources because, at large landscape extents, grassland habitats are more abundant and support high densities of small seeds. The percentage of land classified as grassland or arable land covered by grassland in England in 2000 was 57% of 7.2 million ha (source: Defra); in New Zealand in 2001: 97% of 12.1 million ha was covered by grassland (source: Statistics New Zealand). Seed densities may be poor predictors of bird abundance in New Zealand because seed resources are not a limiting factor.
In Britain, grain is an important component of skylark diet over the winter (Donald, 2004; Robinson, 2004). However, weed seeds are also present but usually compose only a small proportion of the diet (Wilson et al., 1999; Donald, 2004). This pattern may reflect the decline in weed species in farmland habitats. Because small seeds are more abundant in New Zealand, the foraging efficiency of the skylark may be similar or greater than that of skylarks feeding on supplies of large seeds in their native range. To understand the relative importance of small and large seed resources for skylarks in New Zealand, detailed studies of the weed seed composition and distribution are required as well as studies of skylark foraging behaviour and diet composition.
Milder winter conditions may facilitate a longer growing season for grass and weed species in New Zealand, increasing seeding opportunities and hence higher densities of small seeds. Differences in management practices may also explain the differences in seed densities between countries. For example, although there is an increasing trend for use of artificial fertilisers in New Zealand, application rates are generally lower than in Britain (Jarvis and Ledgard, 2002; MacLeod and Moller, in press). In addition, clover Trifolium mixtures are still widely used as a part of the crop rotation for nitrogen fixing in New Zealand (Langer, 1990) and, therefore, may play an important role in maintaining winter seed resources. In addition, livestock husbandry practices differ between countries. In Britain, livestock are usually housed indoors over the winter period, whereas in New Zealand livestock are over-wintered outdoors and provided with supplementary food stuffs that may also provide birds with an additional winter seed supply. Crops in New Zealand may also support higher weed densities because the herbicide application rate is lower.
In Britain, skylark distribution is not only determined by the availability of seed resources. This species is also known to preferentially select large, open fields (avoiding fields that are less than 2 or 3 ha) without boundary features such as hedgerows and trees, which may harbor predators. They take to flight to avoid predators, so they need plenty of time to detect and react to an approaching raptor (Cresswell, 1994). Although predation risk is relatively difficult to quantify, in Britain the density of avian predators is relatively high, with several species (Merlin Falco columbarius, Sparrowhawk Accipiter nisus, and Peregrine F. peregrinus) being significant predators of farmland passerines. However, skylarks in their introduced range were utilising much smaller fields; average field size on our New Zealand study sites was only 3.28 ha. Skylarks may use smaller fields in New Zealand because of a reduced risk of avian predation. Only one avian predator, the Australasian Harrier hawk Circus approximans is present at high densities. Reduced predation risk may create increased niche opportunities in New Zealand by increasing the total available food supply, as individuals can (safely) forage in areas that would otherwise have had too high a predation risk. It may also result in higher survival rates too.
The temporal variation in food abundance on our study areas in New Zealand and Britain was broadly out of phase, with greater breeding season invertebrate densities in Britain and more abundant winter food supplies (seeds) in New Zealand. In New Zealand, an increase in winter resource abundance and a reduction in predation risk appears to have resulted in more available foraging habitat. However, although there was some evidence that local climatic conditions may be less variable during the breeding season in New Zealand, we found no evidence that temporal variation in invertebrate food resources through the breeding season was reduced. This suggests that enhanced niche opportunities, if they exist, are not due to an increased abundance or temporal reliability of invertebrates, but rather enhanced niche opportunities may exist because of a warmer, more stable climate during the breeding season and lower energetic costs associated with foraging activities. Introduced birds in New Zealand are able to maintain high population densities despite relatively poor breeding food resources. Enhanced winter food resources may be a key factor contributing to their success. It appears that winter survival rates of individual yellowhammers and skylarks, mediated via reduced food supplies, are a significant factor in the decline of these and perhaps other species in Britain.
More detailed comparative studies of the diet, foraging patterns, and population dynamics of bird species that inhabit both native and introduced ranges are required to understand better the influence of temporal variation in winter food resources and climate on breeding bird populations. Well-designed studies would aid management programs aimed at either controlling pest bird populations or enhancing declining native bird populations.
Acknowledgments. We are grateful to the landowners in both countries for allowing us access to their farms for fieldwork and to Andreas Till for doing the winter fieldwork in New Zealand. The Esmee Fairbairn Foundation, the Game Conservancy Trust, a Royal Society Post-doctoral Travelling Fellowship, the Carnegie Trust for the Universities of Scotland, New Zealand's Foundation for Arable Research and a Landcare Research's Hayward Post-doctoral Fellowship helped fund this research.
Was this article helpful?