Our results indicate that European Starling breeding attempts in a recently invaded natural landscape of South America had a probability of success of ~ 38%. Though lower than the reported for its original distribution area (Korpimäki 1978; Mazgajski 2007), this breeding success is considerably higher than that of native cavity-nesting passerines in the same area. Particularly, it is higher than the ~ 25% of Lepidocolaptes angustirostris (Jauregui et al. 2019) and Troglodytes aedon (Llambias and Fernandez 2009) and doubles the ~ 19% of Agelaioides badius (De Mársico et al. 2010). In addition, the success rate here reported is similar to the 37% and the 45% of Colaptes campestris and Colaptes melanochloros, respectively (Cockle et al. 2015; Jauregui 2020), two native woodpeckers breeding in the area. Nevertheless, as the starling is double-brooded (Korpimäki 1978), its population growth rate doubles that of native woodpeckers, which are single-brooded. This breeding output suggests that this population is recruiting new individuals at a concerningly high rate compared to native cavity-nesting birds. Moreover, as the starling invasion in South America is recent, starlings may still be able to develop strategies to face a rather new predator, competitor and parasite communities after some adaptation time. If this is the case, breeding success could still reach even higher levels.
In addition to the high nest success rate found, the time elapsed between first and second broods was relatively short (~ 12 days), even for successful nests for which adults must attend recent fledglings. Although we cannot confirm broods in the same cavity belong to the same breeding pair because we did not band adults, cavity reuse by the same breeding pair for second clutches is frequent for the species (Royall Jr. 1966; Cabe 2020). Short time periods between clutches could be a way to avoid cavity occupation by other cavity-nesting species. Moreover, fledglings tend to form flocks shortly after leaving the cavity (Cabe 2020), which allows reduced parental care by the adults and favours energy investment in a new brood (Russell 2000). The period between broods was also short for failed first attempts. As an alternative explanation, reducing time between clutches (either failed or successful) prevents starlings from breeding during January and February, when bird nest success is considerably reduced in these latitudes (Segura and Reboreda 2012; Jauregui 2020).
The starling clutch initiation peaked in September, indicating a population synchronic start as previously reported for the species (Korpimäki 1978; Ojanen et al. 1979; Flux and Flux 1981). This is early compared to other cavity-nesting passerines within the area that start in late September and early October (Llambias and Fernandez 2009; de La Peña 2016; Jauregui et al. 2019). Hence, the early start in the breeding season could favour the starling breeding success by avoiding the competition with native species for cavity use. In addition, starlings used natural and woodpecker cavities for nesting, which are subject of competition between cavity-nesting species (Cockle et al. 2011). If cavity availability is a limiting factor, starlings can overcome this drawback as they have proven to be considerably successful cavity usurpers through aggressive behaviors (see Frei et al. 2015 and Jauregui et al. 2021 and articles there cited).
Starlings’ nest survival was negatively influenced by cavity-tree diameter. We were surprised as this pattern contradicts the general idea that larger trees increase nest survival chance (Zhu et al. 2012; Nyrienda et al. 2016). Mature large trees have been in the ecosystem for a longer time, compared to smaller trees. Consequently, predators may search and find food in large old trees because they are used to it, similar to what was suggested for old vs. new cavities (Wiebe et al. 2007). Alternatively, among cavity-trees, the smallest individuals (i.e., with lower DBH) were mostly Populus alba. The trunk of this tree has few lateral branches and smooth-surfaced bark compared to Celtis tala trees (frequently used by starlings), which could make it inaccessible to terrestrial predators such as snakes or small rodents (AJ, pers. obs.). Nevertheless, during data exploration we were not able to find a significant relationship between cavity-tree species and nest fate. Whichever is the case, due to the contradiction of our result with previous ideas and the lack of clarity in our explanation, we suggest taking these results with caution and encourage the development of experimental studies that aim to clarify this pattern.
This study provides the first breeding data of the starling in South America, based on a population inhabiting a southern temperate forest of Argentina. Our results suggest that the starling exponential growth and expansion (Zufiaurre et al. 2016; Codesido and Drozd 2021) is favoured by a concerningly high breeding success. Information here reported regarding the starling nest-site characteristics can help to design management actions to mitigate the advance of this alien species (see management strategies in Feare et al. 1992 and Williams et al. 2019). If the patterns here reported (i.e., early and synchronic start, occupancy of natural and woodpecker cavities, and high breeding success) replicate in areas with threatened cavity-nesting species, consequences may be regrettable.