The diet of a declining Red-billed Chough Pyrrhocorax pyrrhocorax population on Islay, Scotland

ABSTRACT

Capsule: We quantify Red-billed Chough Pyrrhocorax pyrrhocorax diet from faecal samples collected on Islay. Dung invertebrates formed the majority of prey biomass in dune pasture and Tipulid larvae in mixed pasture but Aphodius larvae were scarce in the diet, whereas they had been a major component in the 1980s. There are management implications from the indication of a reduction in availability of preferred food.

The Scottish population of the Red-billed Chough Pyrrhocorax pyrrhocorax (hereafter Chough) is confined to the Hebridean Islands of Islay, Colonsay and Oronsay. Although the UK population is stable (Eaton et al. 2015), the Scottish population has declined over a protracted period. There were 105 pairs in Scotland in 1986, of which 95 pairs were on Islay (Monaghan et al. 1989), compared to 46 pairs in 2014 (Hayhow in prep.). Unprecedentedly low first-year survival in 2007–10, which occurred primarily during the post-fledging period (July–October), was found to be threatening population viability (Reid et al. 2008, 2011) and undoubtedly driving the current population decline. As a result, supplementary feeding started in 2009, with a mix of meal worms (Tenebrio molitor larvae) and pinhead oatmeal provided before the birds went to roost (Bignal & Bignal 2011). This initially led to increased survival of first-year birds and suggested that food availability was limiting survival (Bignal & Bignal 2011). Choughs have a greater reliance on invertebrate prey than other UK breeding corvids, and local agricultural management practices and seasonal invertebrate availability are linked to diet (Roberts 1982, Bullock et al. 1983, Warnes & Stroud 1989, McCracken et al. 1992, McKay 1996). Direct knowledge of Chough diet on Islay is crucial to inform current conservation management. Bullock et al. (1983) presented evidence of Islay Choughs relying more on dung invertebrates than Choughs in other parts of the UK. There are published diet data, from when the Islay Chough population was healthy (Warnes & Stroud 1989, McCracken et al. 1992), but methods of collection, sampling and analyses differ. However, it is useful to make broad comparisons to look for the absence or considerable decline of whole prey groups from Chough diet now, compared to before the current decline. In the light of the current population decline, we describe Chough diet as quantified from faecal samples collected on Islay in 2011–13, focussing on dune pasture habitats during the period when post-breeding flocks of Chough are present. Given our expectations of Chough diet on Islay from published studies, we looked for obvious absences of expected dung and soil invertebrate prey groups, and for the presence of cereal grains and earthworms. We compared a small sample of our results with a methodologically comparable unpublished data set from 1988 (McKay 1996).

Chough faeces were collected from key feeding locations along the west coast of the Scottish island of Islay; in 2011–13 these were: Ardnave, Kilchoman, Smaull and the Oa; and in 1988 the South Rhinns. Faecal samples collected from the South Rhinns and from Smaull and the Oa were from Chough feeding on a range of agriculturally improved, semi-improved and permanent pasture habitats (mixed pasture), and samples from Ardnave and Kilchoman were mostly from dune pasture systems (online Table S1). Each area had livestock present (sheep and cattle), providing both short-cropped pasture and dung. Samples were collected in the summer and autumn months when first-year Chough mortality was found to be unusually high (Reid et al. 2011). Samples from 2011 to 2013 were collected using the same protocol as in 1988 (McKay 1996), associated with standardized Chough feeding observation transects, covering available potential feeding habitats. During these observations, if birds were observed to defecate the observer located the dropping once the birds had moved away from the location. If a faecal deposit could be identified with complete certainty, it was collected and single or multiple droppings from each feeding bout were stored in one container equal to one sample. There was one difference between sample treatment in the 1988 and 2011–13 studies; the 1988 samples were frozen, defrosted, and then examined, whereas the 2011–13 samples were stored in 70% alcohol, then exposed to the air for 5 minutes prior to examination. One gram, randomly selected, was analysed from each sample, if faecal remains in a sample were less than 1 g then all of that sample were analysed, except those with a weight of less than 0.7 g which were omitted. None of the samples were sieved or washed. Further protocol detail (based on Green & Tyler 1989) is described in the online supplementary material (Appendix S1). A collection of fragments and images was saved to aid future identification; these were referenced against samples described by Moreby (1988) and identification assisted by Skidmore (1991) and Jessop (1986). Identifiable body features of each species or group identified are listed in online Table S2. These were translated into numbers of individual prey items by: head fragments for dung beetles, mandibles and jaws (2 = 1 beetle); fangs (2 = 1 spider); cerci (2 = 1 earwig); leg fragments (2 fore-leg = 1 individual, e.g. grasshopper); fly larval spiracles (2 = 1 individual). There were few clear remains evident from some soft-bodied prey items (e.g. Diptera larvae) which are probably under-represented by this type of analysis. Twenty-three 2011–13 samples from August and September contained some soft, indistinguishable items, probably the remains of soft-bodied larvae which could not be identified. The timing of faecal sample collection and Chough behaviour was such that we assume food items obtained at the supplementary feeding trial site did not dominate the diet, nor detract Chough from foraging normally during the day. Pinhead oatmeal has a different appearance from barley cereal, and mealworm larvae are readily identifiable (Table S2). Fresh weight equivalents listed in Table S2 were taken from McKay (1996) or otherwise gained from prey items sampled from pitfall traps, soil and dung sampling on Islay (2012–14). The number of individuals identified per sample was recorded and these were amalgamated into prey groups by sample. Groups refer to taxa/family groups: larvae of family groups are considered here as separate from adult forms, e.g. all Scarabaeidae adults are one group, their larvae a separate group due to their differing ecology/form. Adult Lepidoptera, earthworms (Oligochaeta) and cereal or seed fragments were recorded as present/absent rather than quantitative numbers of individuals converted to a weight. Their percentage frequency of presence within samples was similarly recorded in other studies (Warnes & Stroud 1989, Meyer et al. 1994, Kerbiriou & Julliard 2007).

The number of breeding pairs of Chough on Islay declined by half between 1988 and 2013. Within the sampling areas, the largest declines were in the mixed pasture areas in the South Rhinns (30 pairs in 1986, 9 in 2013) and the Oa (22 pairs in 1986, 3–4 in 2013). Eighty-two faecal samples were analysed from 171 faeces collected during 2011–13, comprising a total of 1488 individual prey items from 39 species/taxa groups (online Table S2). The seventy-four samples from dune pasture habitats were composed of invertebrate prey, which had equivalent weight biomass dominated by coprophagous invertebrates from livestock dung (61.1%), 4.8% Diptera larvae (other than Tipulids) also likely to have been extracted from dung, 19.6% Tipulid larvae most likely extracted from the soil and 14.5% biomass of other soil and surface invertebrate species (Figure 1). Warnes & Stroud (1989) found dung invertebrates to be important in Chough diet throughout the year, whereas McCracken et al. (1992) found seasonal changes to diet with dung-associated invertebrates most abundant in spring, summer and autumn, with Aphodius larvae taken in large numbers in late summer and autumn. The 2011–13 dune pasture samples had only 3.5% Aphodius larvae, with most of the dung species biomass comprising Geotropes adults (50%) and Aphodius adults (25%) (Figure 1). Tipulid larvae featured less in the diet with the progressing season (Figure 1), as the larvae of the most common pasture species Tipula paludosa, pupate and emerge as adult craneflies from July onward (Coulson 1962). McCracken et al. (1992) and McKay (1996) found Tipulid larvae taken in large numbers until July. Compared to the dune pasture samples, the small number of samples from 2011 to 2013 mixed pasture habitats (n = 8), had contrasting composition, with the invertebrate biomass dominated by Tipulid larvae (72%). The 8.1% biomass of dung specialist species was mostly Geotrupes (5%) and Aphodius adults (3%), with only 0.1% Aphodius larvae. There was also 8% of other Diptera larvae likely to be from dung and 10% of other soil and surface invertebrates (Figure 1). Aphodius larvae, which provided the largest biomass of any prey group (37.1%) in the 1988 mixed pasture samples, were almost absent in 2011–13 (Figure 1). In all 2011–13 samples, cereals were present with increasing frequency from August (Figure 2). In 1988, there were no cereal grains present July to September (n = 15), but 27% of October to December samples (n = 15) contained cereals. McCracken et al. (1992) found cereals to be extremely abundant in the diet during October to April. Warnes & Stroud (1989) found oats present in July; in 100% of October samples and then present through the winter until April. Earthworms were present in all months except July in samples from May to October 2011–13. Most samples were August to October, in which up to 25% contained earthworms (Figure 2). In 1988 samples, earthworms were present in 7% of July to September samples (n = 15); and 20% October to December (n = 15), yet McCracken et al. (1992) and Warnes & Stroud (1989) found no evidence of earthworms in the diet. Sandhoppers and Kelp fly larvae (online Appendix S1) found in beach sand and seaweed were only recorded in the 2011–13 dune pasture samples (2%), perhaps due to predictable large seaweed wrecks being uncommon around the Islay coast.

Figure 1. Invertebrate composition (by percentage of equivalent fresh weight) of Chough faecal samples, using the same invertebrate groups for samples collected from: (A) mixed pasture habitats in 1988 and 2011–13 and dune pasture habitats 2011–13; and from (B) dune pasture habitats in 2011–13.

Figure 2. The percentage frequency of presence of prey groups in Chough faeces in 2011–13. These prey groups were identifiable but only detectable in small parts and not able to be reconstructed into definite numbers of individuals. The number of samples in each month were 27 in August, 25 in September and 16 in October.

Our results show that the biomass of prey taken by Chough feeding on dune pasture habitats, during July to October 2011–13 was dominated by dung invertebrates. Dung invertebrate biomass was mostly adult beetles, both Geotrupes and Aphodius sp. Geotrupes were not present in the 1988 samples, when samples were from mixed pasture habitats rather than dune habitats, and were only found sporadically in other studies (Warnes & Stroud 1989, McCracken et al. 1992). It is a large chitinous beetle whose remains are very obvious within faeces, primarily associated with dung within the wider grassland habitat, but Chough may feed on Geotrupes either when it is in or under dung or crawling on the soil surface. There is an indication of reduced availability of Aphodius dung beetle larvae, a prey group previously found to dominate Chough diet in the post-fledging period. If this is a true decline, then it is significant for Chough. At this time young Chough are seen to feed more in dung than are adults (Bignal & McCracken 1996, McKay 1996). It has been suggested that this may be due to inexperienced young Chough being better able to find food by digging apart dung (requiring little skill) than, for example, finding and extracting Tipulid larvae from pasture (McKay 1996) further exacerbated by declining availability of T. paludosa larvae at this time of year (due to their pupation and emergence as adults). Aphodius larvae are present in dung for much longer than adult beetles and are at their most profitable size when dung has dried and become a more friable and easier substrate from which to extract food than the wet, fresh dung in which adult beetles are more commonly found.

Abundance of dung fauna has been found to be affected by livestock treatments (McCracken & Foster 1993, Beynon et al. 2012) and this should be part of further investigations. Relative to other corvids, Chough are less well adapted to digesting cereals (Soler et al. 1993). However, there is evidence that they make up a significant proportion of the diet, particularly during the winter months. Warnes & Stroud (1989) noted that soon after dawn and in late afternoon, when flocks formed, leaving/returning to roosts, Chough fed on stubble fields and around areas where whole crop silage oats had been fed to out-wintered cattle. Oats as a crop has become a rare feature over the last thirty years on Islay as its use as a supplemental cattle feed has declined. Elsewhere cereals are also found in the diet of Chough, for example, in west Wales Meyer et al. (1994) and they are still a regular part of Islay Chough diet at a time when invertebrates are more difficult to find. Our results show that earthworms form a small but regular element of Chough diet on Islay during July to October, more so than in 1988 and more than Meyer et al. (1994) found in west Wales. Earthworms and molluscs are plentiful on Islay, and it is perhaps surprising that they do not form a larger part of this soil specialists’ diet. However, both groups act as intermediate hosts of the parasitic nematode gapeworm Syngamus trachea, which can cause respiratory infection and death (Meyer & Simpson 1988). Choughs are susceptible to gapeworm (Bignal et al. 1987, Meyer & Simpson 1988) and young Chough in particular (Scottish Chough Study Group pers. comm.). It is possible that the higher frequency of cereals, worms and molluscs in the more recent sampling period indicates reduced abundance or availability of preferred invertebrate prey items.

Acknowledgements

Sincere thanks to Nick Wilkinson for data collection and Adrian Plant for help with taxonomic nomenclature of invertebrates. Thanks to the RSPB staff, farmers and landowners on Islay for allowing access to collect to samples.