Ecosystem engineers are organisms that impact their environment and co-existing species by creating or modifying habitats, and thus they play important roles as drivers of community assembly. We examined whether cavity characteristics and/or habitat attributes associated with cavities provided by four ecosystem engineers, influence the presence of nests of three secondary cavity-nesting birds [Aphrastura spinicauda (Thorn-tailed Rayadito), Tachycineta leucopyga (Chilean Swallow), and Troglodytes aedon (Southern House Wren)], and whether these variations influence their reproductive success. We tested this by: i) assessing nest presence in cavities supplied by ecosystem engineers and ii) quantifying the reproductive success of secondary cavity nesters as a function of cavity characteristics and habitat attributes supplied by ecosystem engineers. Between 2009 and 2022, we recorded 757 cavities in 546 trees in old-growth and second-growth forests in a Global Biodiversity Hotspot in the Andes of southern Chile. Insect/fungi and Pygarrhichas albogularis (White-throated Treerunner) play a key role as the primary producers of cavities. Insect/fungi generated the cavities for 82% of A. spinicauda nests and 95% of T. aedon nests; in contrast, 57% of T. leucopyga nests were cavities excavated by P. albogularis. Characteristics of cavities (size of cavity entrance, volume, and height above ground) were associated with nest presence of secondary cavity nesters and with reproductive success of A. spinicauda and T. aedon. Habitat attributes such as tree density and canopy cover influenced nest presence for A. spinicauda and T. leucopyga, but did not correlate with the reproductive success of any secondary cavity nester. Bamboo density and forest type were related to reproductive success of A. spinicauda and T. aedon. Diameter at Breast Height of trees was related to the reproductive success of T. leucopyga. This study contributes to understanding the importance of various ecosystem engineers for the conservation of secondary cavity-nesting birds in temperate forests and beyond.

Study area and focal species

We conducted our study in Andean temperate rainforests of South America in the La Araucanía Region, Chile (39°16’S-71°48’W). The main weather characteristics of the area are cool summers and average annual precipitation >2000 mm distributed throughout the year. The surveyed sites included eight old-growth forests (>200 years) within public and private protected areas and seven second-growth forest sites (40-100 years) on private land (Ibarra et al. 2012, Caviedes and Ibarra 2017), each with a mean area of 40 ha (Fig. 1). Old-growth forest sites were mixed conifer-broadleaf forests dominated by Saxegothaea conspicua, Laureliopsis philippiana, and Nothofagus dombeyi (500-1000 masl) or by the conifer Araucaria araucana and Nothofagus pumilio (1000-1400 masl). Second-growth forest sites were dominated by broadleaf species including Nothofagus obliqua, N. dombeyi, and Laurelia sempervirens (Díaz et al. 2005, Altamirano et al. 2017). The understory in both old-growth and second-growth forest sites was mostly dominated by bamboo species (Chusquea spp.), Rhaphithamnus spinosus, different species of Azara and Berberis, and tree saplings of species found in both types of forest (Díaz et al. 2006, Altamirano et al. 2017).

Aphrastura spinicauda, Tachycineta leucopyga, and Troglodytes aedon are strictly dependent on cavities for nesting, and they are among the most abundant secondary cavity nesters in the temperate forests of South America (Ridgely and Guy 2009, Altamirano et al. 2017). A. spinicauda and T. leucopyga are forest specialists and typically prefer to use large trees for nesting, while T. aedon is considered a generalist and can be found in both forests and shrublands (Díaz et al. 2005, Cornelius et al. 2000). The diet of all three species consists mainly of insects (Arayo and Chester 1993, Altamirano et al. 2012). Nests of these species are typically located at heights ranging from 0 to 29 meters above the ground (Cornelius 2008, Altamirano et al. 2012). These species exhibit the highest degrees of habitat specialization among the avian community in southern temperate rainforests, and they are known to be strongly affected by habitat loss and degradation in this region (Ibarra and Martin 2015).

Nest searching, monitoring and cavity characteristics

Between 2009 and 2022, we conducted systematic nest monitoring from October to February throughout the breeding season. We worked 6 hours per day and 6 days per week in 15 stands encompassing both old-growth and second-growth forests. We found and monitored as many active cavity-nests as possible (Fig. 1). The authors and field assistants searched for nests mostly from pre-existing trails in each stand. We located active bird nests by stopping frequently to observe the behaviour of adult birds using binoculars. We were alerted to potential nests if adults visited the same tree or flew out of a tree suddenly, by evidence of recent wear at the entrance of the cavity, or if adults perched near or entered/exited cavities (Martin et al. 2004, Cockle et al. 2011b, Altamirano et al. 2017). Nest-searching was alternated between stand and days. We checked lower cavities (<2 m in height) directly using a flashlight with a mirror. For higher cavities (>2 m in height), we checked the interior using a wireless monitoring system with a telescopic pole reaching up to 15 m high (Martin et al 2004, Huebner and Hurteau 2007, Cockle et al 2011b). We considered a cavity as active when at least one egg or nestling was present inside. We identified nests >15 m high as active by observing adult nesting behaviour (e.g., active feeding nestling or removing fecal sacs). We assigned a unique number code for each nest, cavity, and nest-tree and recorded the nesting bird species, origin of the cavity (excavated or produced by fungal/insect decay) and excavator species (in the case the cavity was excavated). We monitored each occupied cavity every 3-4 days to determine nest fate (number of hatched eggs, number of fledglings) and to record when cavities were available again for additional nesting attempts. For all surveyed cavities that were accessible, we recorded cavity characteristics: cavity volume (cm³), size of cavity entrance (cm) and cavity height above ground (m). As a proxy to assess the cavity volume, we use a cylindrical-shape volume formula: V = pr²d, where r is half of cavity width and d is cavity depth (Koch et al. 2012, Robles and Martin 2013). 

Habitat attributes

At the end of each nesting season, we quantified habitat attributes through vegetation plots (0.04 ha, radius = 11.2 m) around each nest-tree found in old-growth and second-growth forests (Ibarra et al. 2014, Caviedes and Ibarra 2017). We established 0.04 hectare plots with a radius of 11.2 meters around each nest tree, under the assumption that secondary nesters would concentrate their activities, such as foraging and territory defence, within this area (Ibarra et al. 2014, Altamirano et al. 2015). Within each plot, we measured both site-level and tree-level attributes. Site-level attributes included: 1) tree density: the density of all trees with a diameter at breast height (DBH) greater than 12.5 centimeters; 2) Tree diameter at breast height (DBH): the diameter of trees measured at breast height; 3) Canopy cover: the proportion of the sky covered by canopy, estimated from the center of the plot; and 4) density of bamboo understory: the density of bamboo vegetation up to 3 meters in height. For tree-level attributes, trees were assigned to one of five decay classes: Class 1, live healthy trees; Class 2, live unhealthy trees; Class 3, recently dead trees; Class 4, long-dead trees; and Class 5, naturally fallen trees (modified from Edworthy et al., 2012).