Ecological Features of the Exotic Wasp Vespula Vulgaris L. 1758 (Hymenoptera: Vespidae) Invading the Southernmost Unesco Biosphere Reserve

Invasive alien species may cause substantial changes and damaging impacts in the ecosystems they invade. Here, we document the current distribution and ecological interactions with native biota of the relatively recently introduced wasp, Vespula vulgaris, in the southern part of the Cape Horn Biosphere Reserve. We conducted surveys and eld studies in four different habitats on Navarino Island: evergreen, deciduous and mixed forests, and shrublands. The spread of V. vulgaris through the island since its rst detection has led to it occupying suitable habitats for nesting and foraging, both in urban and rural settings. The presence of V. vulgaris on islands and remote areas of the CHBR is likely the result of marine human movement through the channels withing the reserve. Wasp foraging composition was different on each of the four studied habitats, yet strongly linked to each habitat’s resources, indicating the inherent ability of V. vulgaris of exploiting surrounding resources eciently. The lack of natural competitors and availability of multiple resources as allowed V. vulgaris to rapidly become a common pest in urban and rural settings in the southern extreme of South America. In this context, we believe that eradication would be impossible, although it would be possible to control populations at localized scales with proper planning and long-term management. Our results can act as a base for management planning, and we strongly recommend social engagement and dialog with pertinent governmental institutions to achieve this challenging task.


Introduction
For hundreds of years, humans have been responsible for introducing plants, animals, and other organisms into new regions, slowly homogenizing the world's biota (Hulme 2009, Pysek et al. 2020). In recent decades, in concert with the globalization of human activities, the pace of this process has accelerated rapidly, facilitated through commerce, transport, tourism, and global environmental change (Blackburn et al. 2011). These activities enable the accidental or intentional introduction of non-native species in new environments, with examples being present in almost all ecosystems today (Pysek et al. 2020). It is estimated that around 5% of introduced non-native species will establish and have a signi cant impact on the receiving community, with social insects (ants, wasps, bees, and termites) accounting for a quarter of these species (Beggs et al. 2011). Global and regional 'horizon scanning' assessments have regularly highlighted the potential threats posed by these groups (Garnas et al. 2016, Roy et al. 2014). The success of social insect introductions is likely due to a combination of associations with human goods, multiple introduction events, and a low probability of detection by standard biosecurity protocols (Brenton-Rule et al. 2018).
Wasps of the genus Vespula (Vespidae) have been introduced and become established in several countries in the Southern Hemisphere (Lester and Beggs 2019). Social vespids have high rates of reproduction and dispersal capabilities, as well as exploit resources (habitat and food) more e ciently than other insects (Moller 1996). The use of visual, auditory, and olfactory cues in resource exploitation by these wasps has led to novel biotic interactions in the newly occupied regions, such as removing competing ants during foraging (Grangier and Lester 2011) or primary nectar robbing (Rendoll-Cárcamo et al. 2017a). Vespula wasps can live in high densities in both urban and rural areas, and their effects on natural and agricultural environments, as well as urban areas, are generally negative and well documented. (Sackmann andCorley 2007, Gardner-Gee andBeggs 2013). Furthermore, they can represent a nuisance to human society and a threat to health (Rust and Su 2012). Vespula wasps, also known as "yellow jackets," are eusocial insects native to the Holarctic region that have become successful invaders in regions as distant as Australia, New Zealand, and South America. They are notoriously di cult to manage and control. (Moller 1996, Beggs et al. 1998). In South America (Argentina and Chile), as elsewhere, the introduction and expansion of Vespula species is likely to be an inadvertent result of human movement of goods and transport of mated queens (Beggs et al. 2011).
The rst record of a yellow jacket in central Chile was documented in the 1970s (initially recorded as V. maculinfrons, later con rmed as V. germanica) (Peña et al. 1975), with the species reaching the south of the country only 25 years later (Pérez 2000). Barrera The UNESCO Cape Horn Biosphere Reserve (CHBR) is located in southern Chile and is the southernmost biosphere reserve declared worldwide (Rozzi et al. 2020). The CHBR is located within the Magellanic Forests sub-Antarctic ecoregion, which is largely undeveloped and is recognized as one of the world's 24 last wilderness areas due to its high proportion of native vegetation and low urban and industrial development. (Mittermeier et al. 2003, Rozzi et al. 2020. Three Chilean National Parks are located within the CHBR (Cape Horn, Alberto de Agostini, and Yendegaia), each with a core, transition, and buffer zone. vulgaris and V. germanica in the CHBR, with a primary focus on Navarino Island. In addition, we document V. vulgaris foraging composition, nest activity, and associations with native ora and fauna in the various habitats where this species has been found.The annual dynamics of V. vulgaris life cycle on Navarino Island are described.

Materials And Methods
Area and study sites Navarino Island is located south of Tierra del Fuego and the Beagle Channel within the CHBR (Fig. 1). Nothofagus forests, coastal shrublands, wetlands, peat bogs, and high-Andean tundra are among the island's native ecosystems (Pisano 1977). The island remains largely undeveloped today, with a single vehicular road running through the length of the island's north coast, from west to east, covering 81 kilometers. This road served as a transect for determining wasp distributions. We also visited Yendegaia National Park during April 2017, where three locations were surveyed on foot, not nding nests but identifying wasp individuals. Hoste Island, speci cally the location Kanasaca, was visited by personnel of the Agricultural and Livestock Service of Puerto Williams during February 2019.

Compilation of distribution records
Between 2015 and 2021, we conducted systematic waspcensuses and collections along the north coast of Navarino Island during the austral summer (late December to mid-March). We recorded each sighting, capture or nest detection in a corridor of 15 m to each side of the road (wherever accessible). We recorded the habitat type, vegetation and any ecological interaction observed with native biota. Additionally, throughout the months of February 2016, 2017, and 2018, we registered opportunistic sightings, catches, and nest records in Puerto Toro, an isolated settlement on Navarino Island's east coast. During April 2017, we sampled in Yendegaia National Park and during the summers of 2019, 2020 and 2021 in four different watersheds on Navarino Island (Fig. 1). The only record from Kanasaca (Hoste Island) was documented in 2019 by personnel of the Agricultural and Livestock Service of Puerto Williams. We georeferenced each location examined with a GPSmap 78sc Garmin© unit.

Foraging associations
During February 2018, we sampled foraging items taken by V. vulgaris to their nest (n = 32 nests) along the north shore of Navarino Island. We identi ed four main habitat types in which V. vulgaris constructs nests: shrublands, deciduous forests, evergreen forests, and mixed forests. To date, no nests of V. germanica have been located. With an aerial net placed 20-50 cm from the nest entrance, we captured foraging workers returning to nests (n = 30 individuals per nest). Due to the stress of capture, workers frequently released foraged objects into the net, which we gathered with ne-tipped forceps; if objects were not released, we kept both the wasp and the foraged object in a 50 mL ethanol container. To avoid discoloration, we stored wasps and released items in individual vials of 60% ethanol and then transported them to the Wankara Laboratory in Puerto Williams, where they were identi ed using a stereo microscope (Leica S6D). We identi ed prey items to order or family level when possible and later classi ed items as follows: prey type = complete animal or fragments, fruit type = complete or fragments of berries, liquids = colored or colorless uids, wood pulp = wood fragments or woody agglutination, and non-apparent load when wasps returned without items. This proved to be a simple and effective alternative to Harris's (1991) suggestions, aided by the wasps' lack of aggression. We captured returning wasps around noon when they were most active (Malham et al. 1991) and made additional observations of their interactions with native biota on an opportunistic basis.

Nest activity
We counted the number of wasps that entered each nest (n = 32, eight in each habitat) to compare nest activity across habitats. In this context, we recorded the mean number of individuals entering the nest (n = 5 counts) and those leaving the nest (n = 5 counts) over a 60-second period, with each count separated by identical-length non-counting intervals (Harris 1991, Kasper et al. 2008). We assessed activity near midday, as Malham et al. (1991) had done, and during the same week (Kasper et al. 2008). The tra c rate provides a useful estimate of worker abundance for management planning purposes (Malham et al. 1991).

Data analyses
To analyze differences of foraged item composition between habitats, we conducted a Permutation Analysis of Variance (PERMANOVA), using PRIMER-E v7 with Permanova+ add-on package (Clarke and Gorley 2015). Prior to PERMANOVA, a PERMDISP test was performed to evaluate data homogeneity within and between groups (Anderson, 2005). For the composition analysis, data were square root transformed and Bray-Curtis dissimilarity matrices were calculated between pairs of observations. To visualize foraging item composition between habitats a CAP analysis (Canonical Analysis of Principal Coordinates) was performed. Nest activity differences between habitats were assessed by performing a Kruskal-Wallis analysis and post hoc Wilcoxon comparisons. The signi cance level was set at α = 0.05. Kruskal-Wallis analyses were performed in Statistica 7, PERMANOVA, CAP and SIMPER were performed in Primer 7.0 (PRIMER-E, Ltd. Plymouth, UK).

Vespula wasp distribution
The rst non-native wasps collected on Navarino Island in 2015 were identi ed as V. vulgaris, with all records being in the urban environment of Puerto Williams. During 2016, the presence of V. vulgaris was documented far from the main town (~30 km to the east and 10 km to the west), although nests were found predominantly in peri-urban settings with evident anthropogenic activity impact (non-authorized waste collection sites and forestry management depots) and inhabited rural settings (farms, greenhouses). Much wider distribution expansion in the natural environment remote from human settlements became apparent in 2017, during which V. vulgaris was recorded along virtually the entire north coast of Navarino Island, approximately 90 km in length (Fig. 2). In the same year, the species was recorded in the isolated Puerto Toro (a village on the east coast of Navarino Island), and in Yendegaia National Park (Fig. 2). In the summers of 2018 and 2019, V. vulgaris was present throughout the north coast of the island, in Puerto Toro and had reached the southwest of the island (Wulaia), as well as being documented for the rst time in Hoste Island. In addition to being present in almost all terrestrial ecosystems, a V. vulgaris nest was recorded at 450 m a.s.l., close to the tree line on Navarino Island. During the summers of 2020 and 2021, V. vulgaris was found off the forest margin of Windhond lake in the southern part of the island, south of the Dientes de Navarino mountain range, and towards the southeast. In addition, V. vulgaris workers were found foraging in different habitats through valleys (forests, shrublands and peatlands), from low elevations up to the tree line (500-600 m a.s.l.). Nests were recorded along hiking trails, in dead and living trees, old growth and secondary forests, peat bog margins and in riparian habitats close to streams and lagoons.
The rst specimens of V. germanica were recorded and collected in 2016 at four different locations within Navarino Island (queens n = 2, workers n = 10, Fig. 3). During 2017, workers (n = 15) and a single drone were collected at the same locations. During 2018 and 2019, V. germanica workers were recorded at 12 locations, the majority of these being new (Fig. 3). During 2020 and 2021, no records of V. germanica were noted. All records to date have been along the north coast of Navarino Island, and no nests have yet been located.

Foraging associations
Vespula wasps have a broad association with the biotic components of the habitats they nest in, and workers are commonly seen exploiting resources for food and nest maintenance (Table 1). Signi cant differences in foraged item composition were found between all habitats (PERMANOVA, F = 10.188 p = 0.001, Table 2, Fig. 5). The most frequent prey items were dipterans (mainly Tipulidae and Syrphidae), particularly in deciduous, evergreen, and mixed forests (Fig. 6). Occasional prey such as beetles, moths, spiders and dragon ies were less frequently observed in the surveyed nests and related to their presence in a particular habitat (Table 1, Fig. 6). The main non-insect source foraged by workers were berries, mainly Berberis microphylla, which are common through the island. Other berries (Rubus geoides and Ribes magellanicum) were foraged in the speci c habitats in which they grow (Fig. 6). Items that are related to nest maintenance (liquids and pulp) were common and regularly foraged in all habitats (Fig. 6).  Su 2012). In sub-Arctic Alaska, the thermal tolerance of V. vulgaris has been studied by Barnes et al. (1996), nding that it can supercool to below -10 °C. These wasps, while freeze intolerant, can avoid freezing and survive by supercooling and hibernating in litter or nding refuge in human facilities (Barnes et al. 1996, Lester and Beggs 2019). This physiological capacity to survive winter conditions in the sub-Arctic ecoregion, which experiences colder temperatures than those characterizing the sub-Antarctic regions (Convey 1996, Danks, 1996, is likely to be an important factor facilitating the species' spread. Observed differences in foraged item composition between habitats (deciduous forest, evergreen forest, mixed forest and shrubland) in our study are likely linked to the availability and quality of food sources in each habitat (Beggs 2001). Vespula wasps are known for high carbohydrate intake (Harris 1991), which in the study regions is mainly foraged from native berries such as the Magellan barberry (Berberis microphylla) and rainberry (Rubus geoides), and nectar robbing from the Chilean rebush (Embothrium coccineum) during its owering period (Rendoll-Cárcamo et al. 2017a). Our data also highlight the opportunistic nature of the diet of Vespula wasps in newly invaded habitats (Richter 2000). Our ndings are consistent with numerous previous studies of the consequences of introductions of these wasps (e.g. , whose impacts may rapidly become practically irreversible, especially in a largely inaccessible and uninhabited region such as the CHBR. In the context of management options available on Navarino island, the next steps are to identify urban and rural inhabited areas suitable for control actions aiming to minimize wasp presence, impacts and ability to rely on domestic crops, greenhouses, recreational spaces, tourism activities and human health. We recognize that eradication would be impossible yet control at local spatial scales can be achieved if well planned and executed (Beggs et al. 2011). The results presented here can act as a base for management planning and, as suggested by Rendoll-Cárcamo et al. (2017a), we strongly recommend social engagement and dialog with pertinent governmental institutions in order to achieve this challenging task.

Declarations
Funding. No funding was received for conducting this study.
Con icts of interest/Competing interests. The authors have no con icts of interest to declare that are relevant to the content of this article.
Availability of data and material. The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.