Nest sites
Most starlings nest in holes, in trees, cliffs, river banks, rock screes, and also in holes that become available in man’s artefacts, such as house roofs, walls and other structures, and in nestboxes. Some forest species nest in cavities found in the tangles of epiphytic ferns and other plants. Many species often nest in holes made by other species, especially woodpeckers (Picidae), barbets (Capitonidae) and kingfishers (Alcedinidae), but also other hole-nesting species like Sand Martins Riparia riparia; previously used holes may be adopted by starlings, or owners may be usurped from freshly made ones. In Africa, old nests of White-billed Buffalo Weavers Bubalornis albirostris may be used by Chestnut-bellied Starlings, while Superb Starlings sometimes breed in old nests of Red-billed Buffalo Weavers B. niger and Rufous-tailed Weavers Histurgops ruficauda. A few starlings make their own holes; Bank Mynas and African Pied Starlings do so regularly and Common Starlings have been recorded doing so on rare occasions, in areas where other cavities are scarce and where substrates are sufficiently soft. The Grosbeak Myna of Sulawesi bores its own holes in dead trees, woodpecker fashion, but as this species is densely colonial the occupied parts of trees become riddled with holes, weakening the trees (Wiles and Masala 1987). The Magpie Starling, of northeastern Africa, is unusual in nesting inside termite mounds.
A few species have foresaken hole-nesting. Black-collared and Asian Pied Mynas and Brown-winged, Wattled, White-crowned, Fischer’s and Chestnut-bellied Starlings build bulky domed nests in trees and bushes, and Common Mynas and Superb Starlings sometimes do so. These nests usually have side entrances. Common Myna also sometimes builds cup-shaped nests, but this species most frequently nests in holes. Shining and Yellow-eyed Starlings, of New Guinea and Queensland, are unique among the starlings in building pendulous nests that hang from the branches of trees and are reminiscent of the nests of weaver birds (Ploceidae). Some of the red-winged starlings Onychognathus build cup-shaped nests on ledges, usually on cliffs but Red-winged and Tristram’s Starlings are increasingly nesting on buildings, and Slender-billed Red-winged Starling achieves extra protection by nesting behind waterfalls.
Nest
Although most starlings nest in cavities, the majority of species build bulky nests inside the cavities, suggesting that hole-nesting is a secondary acquisiton in starlings. The nest is normally constructed from dried grasses, fine twigs, leaves and sometimes man-made products, often with a lining of finer materials including hair, feathers and soft vegetation. The amount of material used in nest construction is related to the size of the cavity to be filled, and also to the requirement for insulation; nests in colder regions tend to be larger than those in warmer regions, and some tropical species use very little nest material. Those species of red-winged starling that build cup-shaped nests incorporate mud into the structure.
In addition to these basic constituents, some species add other materials whose function is uncertain. Many Acridotheres mynas place pieces of sloughed snake skin in the nest, as do a few African glossy starlings, and ungulate dung is included in the nests of Red-billed Oxpeckers, Amethyst and Cape Glossy Starlings. Several species, both in Africa and Eurasia, add fresh green leaves, and sometimes flowers, to the nest lining and Clark and Mason (1985) proposed that Common Starlings selected plants that contained aromatic chemicals with insecticidal properties for this purpose. There has been much debate over the efficacy of this tactic in defence against insect parasites in the nests, and in Common Starlings the behaviour ceases before egg-laying, while nests are most susceptible to a parasite fauna when chicks are in the nest. The carrying of fresh green plant material into a nest by a male of this species appears to stimulate a female to follow, and this behaviour may thus also play a role in pair formation.
Eggs
The background colour of the eggs of most species is a shade of blue, and in most genera the eggs are marked with spots and blotches of a darker colour. The eggs are thus typical of birds which nest in exposed positions and this provides further evidence that hole-nesting in starlings is a secondary development and that starlings are derived from birds that built open, cup-shaped nests. In the genera Acridotheres, Leucopsar, Gracupica, Sturnia, Temenuchus, Pastor, Creatophora and Sturnus, however, the eggs are generally unmarked and resemble more closely birds with a more obligate hole-nesting habit.
Parental care of eggs and young
From those species for which there is adequate information, it appears that both members of the pair participate in rearing the young to independence. Both sexes participate in nest construction and in feeding the young, but it is in incubation that there is a dichotomy between Asian and African starlings. Apart from in Wattled Starlings and the two oxpeckers, in all African Starlings the female alone incubates, the male perching nearby, often warbling for long periods. In only a few African starling species has the male been seen to feed the female on the nest. In Asian Starlings, on the other hand, males participate in the care of the eggs by covering them for a part of the day, while the female incubates for most of the day and at night. In Common Starlings, and possibly others, the male does not incubate sensu strictu and has a poorly developed incubation patch; his role in sitting on the eggs is to reduce the rate of heat loss of eggs, rather than to apply heat to them (K Westerterp, in Feare 1984).
When feeding young, the limited information available indicates that Asian frugivorous starlings deliver food both by carrying food items in the bill and by regurgitation. The frequency of the latter may be underestimated, however, as detailed observations of food delivery have been made on only a few species. For example, casual observations indicate food carrying by Atoll and Micronesian Starlings, while more detailed observations of Shining Starlings showed that adults both regurgitated and delivered food in the bill. Common Hill Mynas usually feed nestlings by regurgitation but occasionally carry food in the bill. In the genera Acridotheres, Sturnia, and Sturnus, however, most food is carried in the bill although Red-cheeked Starlings sometimes regrugitate food for nestlings. In Africa, all starlings whose provisioning of nestlings has been observed carry food in the bill and do not appear to regurgitate. Starlings do not possess a crop and may not have a mechanism to prevent digestion of food items in transit, so that those species whose nestling diet is almost entirely insectivorous may be forced to carry food in the bill to prevent damage to such easily digested prey.
Starlings in most genera have been recorded removing the faeces of nestlings from the nest. This behaviour may be important from two standpoints, firstly to ensure that nest material remains dry and retains its insulating properties, especially before the young have developed homoiothermy, and secondly to avoid the accumulation in the nest of droppings that could provide a growth medium for pathogens and coprophagous infauna. The widespread occurrence of this behaviour in the family is doubtless related to nesting in holes, whose availability may be limited and which are therefore often re-used for successive broods both within and between breeding seasons. Starlings also frequently remove old nest material before starting to build a new nest and this behaviour may also help to prevent the build-up of potentially harmful organisms.
Social behaviour during breeding
Coloniality
Most starlings breed in colonies or in clusters of nests. In some species, such as Grosbeak Myna, Shining, Rose-coloured and Wattled Starlings, colonies are obvious due to the close proximity of nests, which in Rose-coloured Starlings may coalesce. This sometimes applies to other species, for example when several pairs of Asian Pied Mynas or White-crowned Starlings nest together in one bush; in the latter species, nests may even be joined together. Where the distribution of suitable nest cavities precludes breeding in dense colonies, colonial behaviour may nevertheless be evident. This is seen in Common Starlings, where pairs that breed in tree holes or nestboxes that are widely distributed in open woodland display a remarkable synchrony in the initiation of their first clutches each year (Feare 1984); this synchrony is presumably achieved through the birds’ social behaviour although the exact mechanism is not understood. Nevertheless, such synchrony indicates that the birds are breeding as a colony, and when the opportunity arises both Common and Spotless Starlings can form dense colonies, as seen in the pantiled roofs of buildings in villages of northeast Spain.
Although breeding in groups is the norm in starlings, some species do breed solitarily. Shelley’s and Brown-winged Starlings are examples, and Mountain Starlings, Golden-breasted and Yellow-faced Mynas may also prove to be solitary. Common Hill Mynas usually nest c. 1 km apart, but occasionally several nests are found in a single tree. This illustrates the flexibility exhibited by starlings, for many species demonstrate wide ranges of sociality in breeding, a feature that must have facilitated the expansion of starlings into new areas, both under natural circumstances and following introductions.
Pair bonds
The duration of the pair bond has received little study in most starlings but there are indications of differences within the family, and these differences have allowed the evolution of different reproductive strategies. In some Asian species, detailed study and casual observation suggest that birds may pair for life. The retention of a pair bond over successive seasons has been demonstrated in Common Hill Mynas and in Common Mynas (Bertram 1970, Sengupta 1982); in both these species paired birds remain together throughout the year. This may be widespread in Asian starlings as many species are reported to be normally found in pairs. This is especially noticeable in the genus Acridotheres, where even in large feeding flocks and in large night roosts, pairs are readily apparent. However, although pairs are also normally apparent in Asian Pied Mynas, one limited ringing study has suggested that pair bonds may not be long-lived (Tyagi and Lamba 1984). In these species, birds are thought to be monogamous but most studies have been insufficiently detailed to reveal polygyny and extra-pair copulations.
In Africa, pairs of Red-winged, African Pied and Cape Glossy Starlings may persist for more than two years.
In the genus Sturnus, however, pair bonds are frequently of short duration and mates may even be exchanged within a breeding season (Feare and Burnham 1978). Failure to retain a long-lived pair bond may be associated with long-distance migration and a with a higher annual mortality than in tropical species, although data on the latter are lacking in tropical Asian starlings.
The existence of a prolonged pair bond, together with limited opportunities for breeding if nest sites are limited, facilitates the evolution of breeding strategies that depart from the simple situation of a pair of birds raising its own broods. Similarly, the lack of a strong pair bond allows birds to practice breeding strategies that depart from strict monogamy, and both possibilities have been exploited during the evolution of the starling family.
Cooperative breeding
Grimes (1976) reviewed the occurrence of cooperative breeding in African birds, and listed four starling species for which this had been reported. Current information suggests that cooperative breeding occurs regularly in eleven African starlings, but only two, African Pied Starling in southern Africa and Chestnut-bellied Starling in West Africa, have been studied in any detail to date. In several cases only anecdotal reports are available, but certain patterns do emerge.
In most cooperative breeding birds, helpers assist in feeding the young, but play no part in nest construction or incubation (Brown 1987), and this appears to be true for all the African starlings except oxpeckers, in which helpers may bring in some nesting material. In a few cases, the helpers have been reported to feed the breeding female. Helpers are most often subadult birds which have yet to breed themselves, and commonly assist their parents in rearing siblings. However, since they may feed young at several nests, they are clearly not only assisting close relatives. Jamieson and Craig (1987) have suggested that attempts to find adaptive explanations for helping behaviour may be misguided: the strong motivation to feed begging young (as is shown by the readiness with which brood parasites such as cuckoos are fed by many birds) can by itself lead to feeding the offspring of other birds, when dispersal does not occur and not all members of a group are nesting.
There have been many attempts to establish a relationship between ecological conditions and the occurrence of cooperative breeding. Brown (1987) concluded that there are no general rules, but there may be a predisposition towards cooperative breeding in some bird groups. Since helping instead of breeding independently implies deferred breeding, the life expectancy of such species should be sufficient to offer a good chance of breeding in the future. This does seem to be supported by available ringing data, in particular for African Pied Starling (AC, unpub. data).
In the African starlings, helpers have been reported only from savanna species, many of which live in highly seasonal and semi-arid environments. In all cases the species are not sexually dimorphic in appearance, while immature birds may retain distinctive features. Nesting may be solitary or semi-colonial, and the birds associate in resident flocks throughout the year. These characteristics suggest that cooperative breeding may be found to occur in the following additional species, whose breeding biology is little known at present: White-crowned Starling, White-winged Babbling Starling, Emerald Starling, Bronze-tailed Glossy Starling and Shelley’s Starling.
Among Asian species, Rowley (1976) suggested that Shining Starlings might breed cooperatively, and Draffan (1977) recorded three Yellow-faced Mynas carrying nest material into a cavity and thought that all three birds assisted in feeding the young. These are the only examples reported, suggesting that in Asian species cooperative breeding is rare and, in contrast to the African species, these two species are forest frugivores, rather than savanna species.
Polygyny and intra-specific nest parasitism
Among the Asian species that have adapted to more open habitats, different departures from the monogamous pair breeding system have been recorded. Polygyny has been described in Red-cheeked, Common and Spotless Starlings, and intra-specific nest parasitism, where a female lays one or more eggs in the nests of other females of the same species, has been recorded in Common and Spotless Starlings, and is suspected in Purple-backed Starlings. These behaviours may be found in other species, e.g White-cheeked Starling, when these receive appropriate study but at present most of our knowledge is derived from investigations of the behaviour of Common Starlings. In addition to polygyny and intraspecific nest parasitism, Common Starlings also demonstrate frequent exchange of mates (Feare and Burnham 1978), indulge in extra-pair copulations (Hoffenberg et al. 1988, Pinxten et al. 1993, Smith and Schantz 1993), and occasionally two females may nest communally. These departures from strict monogamy are all facilitated by the synchrony of behaviour that accompanies the onset of each breeding season.
Polygyny is widespread in Starlings and Pinxten et al. (1989a) found that 20–60% of males were polygynous each year in a Belgian colony. While most polygynous males have two mates, up to five females may be mated to a single male (Merkel 1978, Pinxten et al. 1989b). Polygynous males produce more fledglings than monogamous males (Pinxten et al. 1989a), but females of polygynous males produce fewer young than monogamous females (Pinxten and Eens 1990). This may result from a difference in the behaviour of monogamous and polygynous males, for polygynous males devote less time to the care of their eggs and young than do monogamous males (Pinxten et al. 1993b). The primary female of a polygynous male can compensate for the reduced contribution of her partner by increasing her own contribution of food to the nestlings, but secondary females appear unable to do this and they fledge fewer young than primary and monogamous females (Pinxten and Eens 1994).
Intra-specific nest parasitism was first described in Common Starlings by Yom-Tov et al. (1974) but this behaviour has also proved to be common and widespread. Up to 37% of first clutches have been found to contain an egg that was not laid by the female who owned the nest (Evans 1988, Karlsson 1983, Lombardo et al. 1989, Pinxten et al. 1991a, Romagnano et al. 1990) but estimates of the proportions of parasitized nests are likely to be under-estimates in view of the techniques used to identify parasitism (Feare 1996). The frequency of this behaviour suggests that it is a strategy adopted by some females, but an interpretation of the function of the behaviour requires identification of the females which parasitise, and this has proved difficult. Females may become parasites following disturbance during the laying period (Feare 1991). Parasitism may also be stimulated by competition for nest sites (Evans 1988), and some parasitic females might be unmated young birds who had been fertilised by an already paired male, but which did not possess nests of their own (Pinxten et al. 1991a). Females might also operate a mixed strategy, rearing a clutch of their own but also depositing an egg in another Starling’s nest (Evans 1988, Feare 1996), thereby spreading their potential offspring and insuring against total failure of their own clutch. Evidence for these suggestions is difficult to obtain.
Intra-specific nest parasitism is accompanied by behaviours that help to ensure the success of the parasitic female’s egg, and which also help potential hosts to guard themselves against parasitism. Some parasitic females appear to avoid being close to hosts’ nests at the time when hosts are most likely to be in the vicinity of their nest (Feare et al. 1982) and parasitic females lay their eggs remarkably quickly (Pinxten et al. 1991b). Some parasitic females remove a host egg when they lay their parasite egg (Lombardo et al. 1989, Pinxten et al. 1991a).
Communal breeding has been recorded twice in Common Starlings. Both instances involved two females laying in one nest, with both females and a single male assisting with incubation and feeding the young (Stouffer et al. 1988, Pinxten et al. 1994). Communal breeding in Common Starlings is probably rare, but unusually large clutches, assumed to have been laid by two females, have also been reported in Rose-coloured Starlings.
Brood parasites
The Great Spotted Cuckoo Clamator glandarius has been recording as parasitizing eleven African starling species, although starlings are not its only hosts, and there is no evidence that a high proportion of their nests are affected (Irwin 1988). Adult cuckoos are attacked by nesting starlings, but there is presumably no discrimination by adult starlings against cuckoo eggs, which do not match those of the host species. The egg pattern will only be relevant in the case of hosts such as Red-winged Starling, which builds an open nest not situated in a dark cavity. There are also records of three hole-nesting species parasitized by the Greater Honeyguide Indicator indicator, and two by the Lesser Honeyguide Indicator minor (Short and Horne 1988). However, many other bird species are parasitized by honeyguides, and brood parasitism is unlikely to represent a major source of breeding failure for most African starling populations.
In Eurasia there appear to be no records of brood parasites laying eggs in the nests of starlings.