Recommendations for depletion modelling of granivorous birds
|
|  Yüklə 6.33 Mb.
|
5.1.3. are there similarities in dietary composition between bird species?Correspondence analysis was used to investigate whether there were similarities in dietary composition between bird species (CANOCO 4.02, GLW-CPRO 1997-1999). A list of standard invertebrate taxa was defined in order to overcome differences in taxonomic resolution between studies. Birds that had fewer than two invertebrate taxa in their diet were excluded from the analysis. The percentages of invertebrate taxa in the diet were not used because of the difficulty in comparing studies that use different measures of proportion in the diet (i.e. % of items, % occurrence, % biomass). For each bird, the mean percentages of invertebrate taxa were converted into four abundance categories: 1 = present (mean is 0-10% of the diet) 2 = present (mean is 11-20% of the diet) 3 = abundant (mean is greater than 20% of the diet) and these were used to perform the multivariate analysis. Analyses were carried out separately on bird diet in the breeding and non-breeding seasons, however only the results for the breeding season are presented here, as only five birds had sufficient data for the non-breeding season. 5.1.3.1 Diet of adults and young during the breeding season The invertebrate component of the diet of adults and chicks during the breeding season was examined using correspondence analysis. The first two ordination axes explained 30.1% of the variation in the data, and a plot of the sample scores is shown below (Figure 5.1). This reveals that closely related bird species and members of the same family often had similar dietary composition, as they appear near each other on the ordination diagram. For example, members of the Emberizidae (the three buntings and yellowhammer) are all situated near the centre of the diagram, while grey and red-legged partridge chicks (Phasianidae) also appear to have similar diets. The plot also reveals that for many bird species, adults and young have different dietary composition e.g. chaffinch, house sparrow, lapwing, and tree sparrow. The exceptions to this were the rook and stone curlew, where adults and young had relatively similar diets. The first ordination axis was dominated by a strong negative score for Ephemeroptera (mayflies), and a positive score for Myriapoda (Diplopoda/Chilopoda, millipedes and centipedes). This means that birds at the lower end of axis one had a high proportion of mayflies and a low proportion of Myriapoda in their diets, i.e. chicks of the linnet and chaffinch (Fringillidae). In contrast, stone curlew adults and chicks had a relatively high proportion of Myriapoda in their diets and a low proportion of mayflies. The second axis is responsible for separation of quail and lapwing chicks from the rest of the birds, and this axis was associated with a strong positive score for Amphipoda, indicating that these chicks had a high proportion of amphipods in their diets. 
Figure 5.1 Correspondence analysis of farmland bird diet (adults and young) during the breeding season (30.1% variance explained). Chaff = chaffinch, cirl = cirl bunting, corn = corn bunting, grey = grey partridge, house = house sparrow, red-leg = red-legged partridge, reed = reed bunting, tree = tree sparrow, yellow = yellowhammer. AB = adults in the breeding season, C = chicks. 5.2 List calorific value of invertebrates and digestibility of different parts. There is very little information on the calorific and nutritional values of farmland invertebrates in general, and virtually nothing on the digestibility of different body parts. It was not possible to assign calorific values to invertebrate taxa because of differences in taxonomic resolution between diet and energetics studies. Calorific information is usually determined for individual invertebrate species, whereas most dietary information is only given to the level of Order. 5.3. Relative frequency of occurrence of invertebrates in the field under various farm management regimes, also consider spatial and temporal variation in abundance. The relative abundance of invertebrates in the farmland environment is important for estimating the amount of food available to farmland birds. Farm management practices (e.g. crop rotation, insecticide and tillage regimes) have an effect on invertebrate abundance and diversity, and some of these factors were taken into consideration. There may also be spatial variation in the density of invertebrates within a field (Moreby, Southway & Boatman 1999, Holland et al. 2002), and the mean abundance of invertebrates was calculated separately for centre and edge positions. The most commonly used sampling method for estimating invertebrate abundance was D-vac suction sampling, although pitfall trapping and sweep netting were also used. Data was compiled from several projects carried out in the United Kingdom (see Appendix 3, Section 3.3, Table 3.7), and this was made available to the modellers at UEA to facilitate estimates of the typical invertebrates densities in the farmland environment. 5.4 Spatial patterns of bird foraging.Bird foraging patterns were presented in Objective 3 (see Appendix 3, Section 3.4, Table 3.8). 5.5 Compare presence in diet with availability in the field according to feeding location (edge or centre). There is very little information on the relationship between the abundance of invertebrates in the field and bird dietary preferences, and very few studies compare the presence in the diet with availability according to bird foraging locations. As with weed seeds, it is difficult to measure the availability of invertebrates as food items because of the need to sample the areas where birds are actually foraging, and this may not be known. Even for well-studied species such as the grey partridge “the real availability of the invertebrate taxa for grey partridges is difficult to verify, since very little is known about how efficiently chicks can find and catch different prey items in different habitats” (Itämies et al. 1996). Twelve studies measured the abundance of invertebrates in the field as well as their proportions in the diet (Southwood & Cross (1969), Davies (1977), Vickerman & O’Bryan (1979), Green (1984), Galbraith (1989), Jenny (1990), Beintema et al. (1991), Brooks et al. (1995), Itämies et al. (1996), Aebischer & Ward (1997), Poulsen, Sotherton & Aebischer (1998), and Stoate, Moreby & Szczur (1998), although only six of these explicitly examined the relationship between these variables and demonstrated clear dietary preferences for birds. Galbraith (1989) reported that “on Scottish agricultural land, the diet of lapwing chicks is varied and is influenced by prey availability and habitat differences in the invertebrate communities.” Differences in diet between habitat types reflected differences in the actual prey abundance as shown by pitfall trapping and soil sampling. For example, in the rough grazing areas where soil invertebrates (earthworms and leatherjackets) were less abundant, the chicks ate more surface species (mainly beetles and their larvae). “Although soil invertebrates comprised the major part of the prey biomass, they were taken in comparatively small numbers, which suggests that the chicks were feeding opportunistically rather than selectively.” Grey partridge chicks appeared to have a preference for Collembola and Homoptera (particularly aphids), but avoided parasitic Hymenoptera and Diptera (Southwood & Cross 1969). Similarly, Itämies et al. (1996) found that Delphacidae (Hemiptera, Homoptera), Coleoptera and Aphidoidea were selected for, whereas Araneae and Diptera were consumed less than expected on the basis of their abundance in the environment. Green (1984) showed that grey and red-legged partridge chicks generally consumed invertebrates in proportion to their abundance in vacuum samples, but appeared to show a preference for aphids (Hemiptera: Aphididae) and carabid beetles (Coleoptera: Carabidae), as the proportions of these in the diet were high compared with their availability. However, they also discovered differences between the numbers of invertebrates caught in vacuum and sweep net samples, and concluded that the disparity between diet and availability might be attributable to biases in the sampling techniques. This point was also re-iterated by Poulsen, Sotherton & Aebischer (1998) who stated that the efficiency of the suction sampler varied in different crop types. Davies (1977) revealed that the percentage of Drosophilidae (Diptera) in yellow wagtail diet was not related to the numbers available in traps (r2 = 0.176, p>0.05), but was inversely correlated with the numbers of Chironomidae (Diptera) caught (r2 = -0.830, p<0.01). The percentage of Chironomidae in the diet was positively correlated with the number of Chironomidae caught in the traps (r2 = 0.843, p<0.01). He concluded that “when chironomids were abundant the wagtails fed on them almost exclusively, but as numbers decreased, more and more Drosophilidae were incorporated into the diet.” Two of the smallest prey, Staphylinidae (Coleoptera) and Sphaeroceridae (Diptera), were abundant in dung pats but were not selected by foraging wagtails, and Sepsidae (Diptera) also appeared to be distasteful. Skylarks generally ate invertebrate taxa in proportion to their abundance in vacuum samples, however Lepidoptera and spiders (Araneae) were eaten more often than expected, whereas beetles were avoided (Jenny 1990). 5.6 Rank invertebrates in terms of their value As with plant food items, a final ranking of the importance of invertebrates for farmland birds could not be determined, as there was insufficient information available in the literature. APPENDIX 6. Identify sampling and data recording techniques applicable to questions concerning farmland bird food resources. |