This species is assessed as Endangered based on an inferred reduction of the population of 54% in the last ten years, and a strong and continuing decline in the population size in the last ten years as a result of known, ongoing and irreversible causes (A2) according to observed declines (a), an index of abundance appropriate for the taxon (b), a reduction in the extent of occurrence (c), and as a result of the effects of introduced taxa (e) as several published reports suggest that pathogens introduced from non-native introduced bumblebees is an ongoing and increasing threat to this species.

Bombus dahlbomii is distributed in the southern tip of South America, which comprises two countries: Argentina and Chile, (Abrahamovich and Diaz 2001, Montalva et al. 2011). In Argentina, this species has been recorded in the following provinces (subcountry units, from North to South): Mendoza, Buenos Aires, Neuquén, Río Negro, Chubut and Santa Cruz (Abrahamovich and Diaz 2001). More recently it has been recorded for first time in Tierra del Fuego Province (J. Montalva, pers. comm). In Chile, B. dahlbomii has been recorded in the following regions (subcountry units, from North to South): IV (Coquimbo), V (Valparaíso), Metropolitan Region, VI (OHiggins), VII (Maule), VIII (BíoBío), IX (La Araucanía, including Mocha Island), X (Los Lagos, including Chiloé Island), XI (Aisén) and XII (Magallanes, including the Chilean portion of Tierra del Fuego Island), (Ruiz 1936, Toro and Chiapa 1997, Montalva et al. 2011, Díaz Tavie et al. 2015).

The northern limit of the species’ geographic range is located in the IV Region of Chile (Ruiz 1936). In Argentina, the northern historic limit lies in Mendoza Province (Abrahamovich and Diaz 2001). However, the occurrence of the species in Mendoza is based on a single recorded queen (collection date unknown, deposited at La Plata Museum Argentina—hereafter MLP—and published >15 years ago), and bumblebee surveys in this province (Schmid-Hempel et al. 2014) have failed to find this species. Moreover, consulted pollination ecologists and bee biologists who have worked intensively in the field during the last 13 years never recorded this species in this Province (N. Chacoff, G. Debandi, J.P. Torretta, D. Medan, and D.Vazquez pers. comm). Thus, an eventual subpopulation surveyed in the past in Mendoza may be probably extinct, and the current northern limit in Argentina is the Province of Neuquén.

Although the easternmost specimens have been collected near the Atlantic coast in the Argentinian Province of Buenos Aires (Abrahamovich and Diaz 2001), its presence in this province is based on only two workers collected in 1977 (deposited at the at the La Plata Museum MLP). Bumblebee surveys in this province (Plischuck et al. 2009) have not reported this species. Consulted pollination ecologists and bee biologists working intensively in the field in the last 23 years have never seen this species in the province of Buenos Aires (J.P. Torretta, D. Medan, pers. comm). Thus, subpopulations surveyed in the past in this province may be possibly extinct.

The scarcity of records north from the Rio Colorado in Argentina, suggests that this river may be the natural northern limit of distribution of the species in Argentina, and  that  the few records in Buenos Aires and Mendoza may have resulted from small and isolated subpopulations in microhabitats suitable for this species (Abrahamovich et al. 2007). Furthermore, the few records at the Atlantic coast of the Patagonian region (only seven specimens in four localities of the provinces of Rio Negro, Chubut and Santa Cruz, collected prior to 1981), and the lack of records in the Patagonian arid steppe, suggest that this species is associated with forest and humid habitats, more typical of the Patagonian Andes of SW Argentina.

This species reaches the Pacific coast along all the Chilean regions (IV-XII) where it has been recorded. The global western limit of B. dahlbomii is in the Pacific Ocean Islands (Chiloé Island) in the X Chilean Region.Thus, the distribution of B. dahlbomii extends westwards beyond the continental limit of South America.

The southern limit of this species’ geographical range  is located in the Navarino Island, ca. 55°S (a small islet in the Beagle Channel south of Tierra del Fuego Island), in the Chilean XII Region, (queen deposited at  the American Museum of Natural History, AMNH_BEE00147245; Isla Navarino, Chile; Septiembre 1935; J. Bird). It is noteworthy that this location is the global southernmost limit of the Bombus genus.

Finally, the report of this species' presence in Peru by Peat et al. (2005) seems to be the result of a misidentification of the species or the site of collection being incorrectly labelled. Paul Williams (pers. comm.) suggested that queens of B. rubicundus recorded in the High Andes of Peru might pose a misidentification with B. dahlbomii. No additional pinned specimens were found in the Museum of Natural History, London (MNHL), nor any specimen labelled as collected from Peru (Mr. David Notton pers. comm). Rasmussen (2003) did not include B. dahlbomii in a key for identifying bumblebee species of Peru either. Therefore, we can discard Peru as a country of occurrence of B. dahlbomii and therefore uncertain data from Peru has been excluded for this assessment. Finally, this species has been reported as mislabelled specimens from Panama by Milliron 1973a:165, and wrongly reported from Santos, Brazil by Handlirsch (1888:236), both quoted by Abrahamovich and Diaz (2002).

Bombus dahlbomii was reported as abundant in the past, both in Argentina (Abrahamovich and Diaz 2001) and Chile (Montealegre 1927). However, there is a strong consensus, based on independent studies (Morales et al. 2013, Schmid-Hempel et al. 2014), and reports and observations (see personal communications and observations in this assessment) on a consistent decline of the species' abundance in both countries, including the extinction of local subpopulations in particular in the north-eastern part of its geographic range (see also the Geographic Range section). This retraction may have led to a range loss of >54% of the extent of occurrence (EOO) in the last 10 years.

In Argentina, there is evidence of a strong decline in the northern portion of its range, where this species has been almost completely replaced by the invasive European Bumblebee (Bombus terrestris), and to a lesser extent B. ruderatus (Madjidian et al. 2008, Arbetman et al. 2012, Morales et al. 2013, Schmid-Hempel et al. 2014, Geslin and Morales 2015). Since the sequential invasion of these introduced bumblebee species, B. dahlbomii has become much harder to find. Bombus ruderatus was introduced to Chile in the 1980s and arrived in the Patagonia region of Argentina around 1994 (Roig Alsina and Aizen 1996); B. terrestris was introduced to Chile in 1997, and advanced eastward arriving in the Argentine range of B. dahlbomii around 2006 (Torreta et al. 2006).

A broad scale survey during January-February 2011 (austral summer) along the eastern slope of the Andes from North Neuquén to South Santa Cruz (Arbetman et al. 2012, Morales et al. 2013), failed to find B. dahlbomii in areas where it was reported as abundant until 2004 (e.g., Challhuaco Valley and LlaoLlao Forest, in Río Negro Province, Aizen 2001, Morales and Aizen 2002, 2006, Vazquez and Simberloff 2003, Madjifian et al. 2004) and 2006 (e.g., near San Martin de los Andes, in Neuquén Province, Torreta et al. 2006). A comparison of the results of that survey with published maps of historical distribution based on museum collections (Abrahamovich and Diaz 2001, see also the Geographic Range section), shows that many subpopulations of the northern part of its historic geographic range (Neuquén and Río Negro and to a lesser extent Chubut) have suffered drastic reductions, or even disappeared, being replaced by invasive Bombus terrestris and to a lesser extent by B. ruderatus (Morales et al. 2013).

In 2011 only the southernmost subpopulations in southern Santa Cruz were still numerous, which coincided with the only locations in Argentina where B. terrestris had not yet arrived (see Fig. 2 and supplementary materials in Morales et al. 2013, see also Fig. 2 in Schmid-Hempel et al. 2014). However, an even more recent survey performed in December 2013 shows that these locations have been recently invaded by B. terrestris (Geslin and Morales 2015).   By February 2014, M.A Aizen recorded similar abundances of B. dahlbomii and B. terrestris visiting flowers of Gaultheria spp. at the coast of Lago Argentino right in front of Perito Moreno glacier, Los Glaciares National Park (pers comm). Thus in just three years since first observed, B. terrestris has reached similar or higher (Geslin and Morales 2015) abundance than B. dahlbomii, which may vary among sites.

A two decade long study (1994-2013) at the Challhuaco Valley (Nahuel Huapi National Park, Río Negro, Argentina), where B. dahlbomii was the most frequent pollinator species of the lily Alstroemeria aurea in a large monospecific stand, reveals that since these surveys started in 1994 the abundance of this B. dahlbomii subpopulation declined until complete local extinction. Although B. dahlbomii was very abundant and common in this site during the mid-1990s (Aizen 2001 and references therein), the species started to decline after B. ruderatus invasion (Morales 2007, Madjidian et al. 2008, Aizen and Feinsinger 2003). This negative trend was exacerbated with B. terrestris invasion in 2007; in fact, since 2008 no B. dahlbomii has been recorded during comprehensive pollinator censuses (Arbetman et al. 2012, Morales et al. 2013). Therefore, the collapse of this once very large subpopulation was concurrent with the sequential invasion of the introduced European species B. ruderatus and B. terrestris (Morales et al. 2013).

In Chile, a rapid displacement of B. dahlbomii by B. terrestris has been reported by Schmid-Hempel et al. (2014), following a similar pattern of replacement from north to south as that described in Argentina by Morales et al. (2013). For instance, in 2004, the native B. dahlbomii and the introduced B. ruderatus were abundant around the Chilean Lake district of Villarrica–Pucon (IX Region). On the other hand, by 2010, B. terrestris had become the dominant species while B. dahlbomii was no longer found (Schmid-Hempel et al. 2014). During their 2010/2011 surveys, Schmid-Hempel et al. (2014) noticed clear boundaries between the advancing B. terrestris and the presence of the native B. dahlbomii in Southern Patagonia and around Lake General Carrera (XI Region, Schmid-Hempel et al. 2014). Bombus dahlbomii was clearly still abundant below that latitude in the XII Region (See Fig. 2 Schmid-Hempel et al. 2014). However, recent unpublished reports suggest that B. dahlbomii has drastically declined in this region after the recent invasion of B. terrestris (see below and the Threats section).

There have been reports of decline of B. dahlbomii throughout most of its historical range; however, this decline seems to be more drastic in the IV, V and Metropolitan Regions of Central Chile (Ruz and Vivallo 2005, Montalva et al. 2011; see per-region account below). Below, we provide an account of the available information region by region (from North to South). This information although scattered and preliminary, confirms a north to south declining trend.

In the IV Region (Coquimbo), (i.e. the northern limit of this species' range) the species was never abundant and always was constrained to relictual mountain forest areas (Toro and Chiappa 1997). This species was still extant, but very scarce, at  Los Vilos in January 2013 (P. Novoa, pers. comm).

In the V Region (Valparaiso), although a recent survey of entomological collections (Smith-Ramirez, pers. obs.) revealed that the last specimens of this species were collected in 1974, there have been more recent reports of B. dahlbomii (Montalva, pers. comm). However, all of these partial observations coincide in the declining trend experienced by B. dahlbomii in this region. A study of bees in localities of Central-South (V Region) and South Chile (VIII Region), showed that in 2000 B. dahlbomii was still the most abundant species recorded (50%, n=812 bees), whereas the invasive B. terrestris accounted for only 10% (n=170 bees) of the apoidea (Ruz and Herrera 2001). Moreover, this invasive species was present in  Central-South, but not in South Chile.

Overall, in 2010 the relative abundance of B. dahlbomii compared to that reported by Ruz and Herrera ten years ago (2001) had decreased in the V region. In 2010, E. Chiappa (pers. comm.) collected only one B. dahlbomii out 1,115 bees (0.09%) in Quebrada La Horquilla (QuebradaVerde) and only one out of 249 bees (0.4%) in Quebrada El Sapo (El Salto). Similarly, according to collecting records, in 2002 B. dahlbomii was abundant in sites like Laguna Verde, La Ligua, Palmar El Salto and Quebrada Alvarado. However, in 2009-2010 B. dahlbomii was already scarce in the Botanical Garden of El Salto, whereas B. terrestris was abundant. Bombus dahlbomii is no longer found in most of these areas, which are now dominated by B. terrestris (Montalva, pers. comm).

In the Metropolitan region, B. dahlbomii was reported as abundant in the Cerro San Cristobal (Santiago de Chile City) during the first quarter of the twentieth century (Ruiz 1923), but the last collections from the Metropolitan Regions are from 1996 (Smith-Ramirez, pers. comm). Currently the species has been reported for some cordilleran sectors like Yerba Loca, Valle Nevado and Farellones (J. Montalva, pers. comm., S. Rodríguez, pers. comm.), but it is no longer observed in the valley of the Metropolitan Region, which is dominated by B. terrestris (Montalva et al. 2011). In the VI Region, similar to the Metropolitan Region, B. dahlbomii seems to be extirpated in the central Valley, but still present in the Cordillera. In the VII Region, B. dahlbomii is very scarce, and only can be observed in protected areas of cordilleran valleys, whereas is completely absent from urban areas where B. terrestris is abundant (C. Avendaño, pers. comm).

In the VIII Region, in Cobquera near the Pacific coast, B. dahlbomii was the most abundant apoide in 2000 (>50%) and the only bumblebee visiting Eryngium paniculatum (Ruz and Herrera 2001), while in 2005 B. dahlbomii was absent in the same area and the main pollinators of E. paniculatum were B. terrestris and B. ruderatus (L. Ruz, pers. comm). Moreover, in 2013, C. Tobar (pers. comm.) did not observe any B. dahlbomii, or B. ruderatus, but only B. terrestris. Thus, B. dahlbomii seems to have been progressively replaced by B. ruderatus and B. terrestris, with the later finally replacing both other species, as has been reported for Challhuaco Valley, in Argentina (Morales et al. 2013).

In the IX Region, B. dahlbomii was recorded only twice in a recent survey on blueberry orchards, being very scarce in agricultural areas, and slightly more abundant in cordilleran areas with native forests. The whole area is largely dominated by B. terrestris and B. ruderatus (L. Vieli, pers. comm). In addition, in the Nahuelbuta National Park, where around 2006 B. dahlbomii was abundant, L. Packer (pers. comm.) did not observe any individuals in 2012-2013, in contrast seeing only B. terrestris.

In the X Region, in Chiloé Island a long-term survey since 2000 revealed that until 2008 this species was abundant, and declined after that, coinciding with the arrival on this island of the invasive B. terrestris (Smith-Ramirez, 2014). Similarly, in the XI Region, in Coyhaique B. dahlbomii was very abundant in the past. However, around 2008 this species began to decline and now B. terrestris is very abundant (Piere Barattini, pers. comm).

Finally, B.dahlbomii is one the few bee species in the XII Region (Magallanes) of Chile, and also present in Tierra del Fuego Island (Diaz Tavie et al. 2015 and references therein), where it has recently been recorded in Parque Karukinka, Tierra del Fuego (R. Muza, pers. comm). This species was still very abundant near Punta Arenas, Chile until 2013 (Leah Dudley, pers. comm., Díaz Tavie et al. 2015), and the XII Region was considered to host the only populations of B. dahlbomii still not overlapping with the invasive species B. terrestris (J. Montalva, pers. obs., see also Fig. 2 in Schmid-Hempel et al. 2014).

However, since the recent naturalization of B. terrestris in the XII Region of Chile around 2012 (see the Threats section) a rapid and constant decline in the population of B. dahlbomii has been observed (Diaz Tavie, pers. comm. 2016). By the late summer and early fall of 2016, B. terrestris was more abundant than B. dahlbomii in Punta Arenas ,where T. Salazar (pers. obs.) only saw a handful (~5) of B. dahlbomii individuals during the months of March, April, and May 2016. Moreover, B. dahlbomii was not observed on a full-day trip in Parque Nacional Torres del Paine in February 2016, but a handful of B. terrestris workers were recorded (Morales, pers. obs).

The citizen science campaign “Salvemos Nuestro Abejorro” (http://salvemosnuestroabejorro.wordpress.com) was established to encourage volunteers to submit photographic records of this species. It has been instrumental in proving that this species is still extant throughout most of its range, despite a severe reduction in relative abundance. However, the many independently published peer-reviewed studies, photographic records and personal observations summarized here point to a substantial decline in B. dahlbomii abundance and range, as well as local extirpations throughout Argentina and Chile in some sites where it was formerly common.

This species occupies the southernwest most portion of continental South America. As a consequence, it is the southernmost bumblebee species in the world. This is a typical species of the broadleaf and mixed forest biome (according to Olson 2001 classification of global biomes), and in the near past it was very common in the following terrestrial ecoregions: Valdivian Temperate Forests, Magellanic Subpolar Forests (both ecoregions characterized by broadleaf Nothofagus spp. forest types), more or less common in the ecoregions of Chilean Matorral (associated to remnant and relictual hygrophyllous forests) and less common in particular suitable habitats in the Southern Andes Steppe and Patagonia Steppe. It has also been frequently recorded in open vegetation communities of lake margins (C. Morales pers. obs.) and mountain meadows. Although there have been a few records in the ecoregions of Argentine Monte, Argentine Spinal and humid Pampas, it is probably extirpated from all these ecoregions.

This species nests in cavities in the soil, below trunks, sphagnum, soil and roots, and nests are small compared with those of other Bombus species (Claude-Joseph 1926, Montealegre 1927, Ruiz 1939, Montalva et al. 2011). This species was commonly observed foraging in gardens of small towns, although nests have not been reported in such urban areas. Regarding foraging ecology, B. dahlbomii is a long tongued bumblebee, that visits flowers of a broad spectrum of colours, sizes and shapes of a high diversity of plant families, including Fabaceae, Asteraceae, Myrtaceae, Lamiaceae and Rosaceae among other common families (Montalva et al. 2011, Abrahamovich et al. 2001, Morales unpublished). Overall, this species has been recorded in more than 90 native and introduced plant species of  greater than 40 families (Abrahamovich et al. 2001, Morales and Aizen 2010, Montalva et al. 2011).

Bombus dahlbomii has been shown to play a key role in the structure and function of plant-pollinator networks of many communities (Vazquez and Simberloff 2003, Morales and Aizen 2006, Aizen et al. 2008, Ramos-Jiliberto et al. 2009). For instance, model simulations from a recent analysis of plant-pollinator networks revealed the importance of hymenopterans as pivotal groups that maintain the core structure of the pollination network and guarantee overall species persistence in the temperate forests of Chiloé Island. Among them, B. dahlbomii exerted a disproportionately large influence on the preservation of network structure. Thus, we foresee its extinction would seriously harm the reproductive success of plants in this temperate rain forest, and therefore recommend that B. dahlbomii should be considered as focal species for conservation programs given current threats from selective logging and habitat loss (Ramos-Jiliberto et al. 2009). In addition, B. dahlbomii is a key pollinator of many endemic plants like the lily Alstroemeria aurea (Aizen 2001).

With respect to relevant aspects of bee biology and genetics, in colonial bees only queens and drones are reproductive. In addition, B. dahlbomii, as with the rest of the social bees, can show a reduced heterozygosity due to the low effective population size. However, to estimate the effective population size in haplo-diploids is a complex task as a result of atypical and unknown sex ratio effects. No work has been done on this topic for this species, but it is safe to assume that effective population sizes are smaller and more vulnerable in most of the haplo-diploids than for diplo-diploid organisms with similar sex ratios and apparent population sizes.

Commercially Introduced, Non-native Bumblebees

The most important and documented threat to the species at present is the invasion of European Bumblebee species introduced for crop pollination, in particular of Bombus terrestris (Morales et al. 2013). The driver of this threat is the global trade of non-native bumblebee species for pollination of agricultural field and greenhouse crops (Morales 2007, Stout and Morales 2009, Dafni et al. 2010). This commerce triggers the introduction of commercially reared bumblebee colonies outside their native range, along with their associated pathologies, and the additional threat that newly produced non-native queens will escape, find a mate, and become established. The pathogens introduced by commercial bumble bees, and the competition for resources from established feral colonies have become direct threats to the continued existence of B. dahlbomii.

The invasion and rapid spread of B. terrestris in Chile is continuously aided by permanent annual importation of thousands of commercial colonies. From 1997 to 2016, a yearly average of 16,651 colonies, 37,900 fertilised queens and 4,200 ovipositing queens have been imported to Chile, from different companies and geographic origins, including Belgium, The Netherlands, Israel, Slovakia, and Spain, reaching 39,000 colonies, 160,000 fertilised queens and 10,000 ovipositing queens in 2015 (Source: SAG Chile, date of data access: 14 June 2016).  Moreover, with the current existing regulations and recent requests for import authorized (Source: SAG Chile, Resolución N 2099/2016) these introductions will continue in the future. Strong propagule pressure (that is the combination of a high number of individuals of a non-native species involved in any one release event and the number of discrete release events) is one of the key factors favouring biological invasions (Colautti 2006). Here, the high number of colonies introduced on each shipment and the high frequency of introductions per year (Source: SAG Chile Date of data access: 14 June 2016) might subsidize the invasion of B. terrestris and its competitive and sanitary effects on B. dahlbomii.

This process has been documented in the The XII Region of Chile. Bombus terrestris was introduced in the XII Region during three consecutive years there (2011-2013) as part of a Project conducted by  the “Instituto de Investigaciones Agropecuarias” (INIA) to evaluate their pollination efficacy of Red Currant (Ribes rubrum) (Perez 2013, Estay and Mc Leod 2014). The first colonies were used under glasshouses conditions at INIA facilities in Kampenaike, Punta Arenas (Estay y Mc Leod 2014). Since then, commercial colonies have been continuously introduced by local growers for tomato pollination (Díaz Tavie, pers. comm., T. Salazar, pers. comm).

Feral colonies of B. terrestris have recently become established in the XII Region, including Tierra del Fuego, probably subsidized by these intentional local introductions. The first sightings of B. terrestris were documented in 2012 (Torres del Paine National Park, Diaz Tavie, pers. comm.) and the first photographic record in this Region was taken in 2014 (Puerto Natales, Diaz Tavie, pers. comm). These records coincide with observations of local owners of ornamental plant nursery, who agree that B. terrestris seems to have become established since at least 2014 near Punta Arenas (T. Salazar., pers. comm). Recent observations made by C. Morales (February 2016) and T. Salazar (March 2016) in the XII Region of Chile suggest that B. terrestris may be currently more abundant than B. dahlbomii, despite a relatively recent invasion.

Recently published studies confirm the rapid advance and current widespread distribution of B. terrestris. Schmid-Hempel et al. (2014) report that from field surveys in 2004 and from 2010-2012, by early 2010 B. terrestris was by far the dominant species in Central Chile from Santiago to Puerto Montt. In February 2011, no B. terrestris were found along the R.N. 26, road crossing the Patagonia steppe from the Andes to the Atlantic Ocean, however, in October 2011, just eight months later, this species was detected throughout this region.

Throughout the entire B. dahlbomii range, in sites where this species was formerly abundant and now is no longer observed or is very scarce, B. terrestris is extremely abundant (Morales et al. 2013, Schmid-Hempel et al. 2013, Geslin and Morales 2015, J. Montalva, L. Ruz, pers. obs., C. Avendaño, L. Vieli, Dr. Packer, pers. comm). All these studies together document that in less than 20 years the invasive bumblebee B. terrestris range has completely overlapped with B. dahlbomii (see the sections Population and Geographic Range).

Thus, the threat posed by the B. terrestris invasion can certainly affect the global population of B. dahlbomii. Hypothesized mechanisms underlying this observed ecological replacement may include pathogen transmission (Arbetman et al. 2013, Schmid-Hempel et al. 2014, Arismendi et al. 2016), although competition for resources cannot be discarded (Morales et al. 2013, Schmid-Hempel et al. 2014).


Pathogen Transmission and spread

Recent studies report that B. dahlbomii harbours various bumblebee internal pathogenic parasites (namely, Crithidia bombi, Nosema bombi, Locustacarus buchneri and Apicystis bombi). Overall, the timing of detection and patterns of prevalence suggest a role of invasive bumblebees in either co-introducing diseases or novel variants of pre-existing diseases or amplifying their spread, two hypotheses that need to be tested. Schmid-Hempel et al. (2014) performed various large scale surveys of the bumblebee fauna and bumblebee parasites in 2004 and from 2010 to 2012 covering the whole geographical range of B. dahlbomii, from Central Chile to the southern continental tip of South America. They reported infections of Crithidia bombi, a trypanosome protozoa which parasitizes bumble bees, in B. dahlbomii (18.5% prevalence) as well as in the two invasive bumblebee species (B. ruderatus and B. terrestris) in southern Chile. Genetically, the populations of C. bombi sampled in 2004 in B. dahlbomii and B. ruderatus were less diverse, and distinct from the ones sampled later, which were genetically similar among the three bumblebee species; supporting the hypothesis that Crithidia bombi was first introduced by B. ruderatus and then enriched by the later introduction of B. terrestris (Schmid-Hempel et al. 2014). Moreover, infection of C. bombi in B. terrestris was very high near Bariloche, Argentina in 2009, while the pathogen was absent from contemporaneous samples of native bumblebee species collected in other regions of Argentina not invaded by this species (Plischuk and Lange 2009).

Schmid-Hempel et al. (2014) also found that both invasive species carried the microsporidian Nosema, a pathogen  known to affect the survival and fecundity of bumblebees, although in low prevalence (B. ruderatus = 6.5%, n=77, B. terrestris = 2% n = 993), but this parasite was not detected in B. dahlbomii (n=128) nor in the native species more typical from the eastern side of the central Andes in Argentina, B. opifex (n=28) (Schmid-Hempel et al. 2014). However, Vallejos (2013) found Nosema bombi, both in B. dahlbomii (3.4% prevalence, N=176) and in B. terrestris (13,6%, N=88) sampled in South Chile in 2011. In our interpretation, the higher prevalence of N. bombi found by Vallejos (2013) may relate to the fact that he collected both B. dahlbomii and B. terrestris simultaneously in the same flowering plants in Valdivia city in 2011, thus increasing the chance of horizontal transmission. On the contrary, Schmid-Hempel (2014) sampled B. dahlbomii during a longer time lapse (2004, 2010-12), and over vast regions, including many sites still free of B. terrestris (see Fig. 2 and Table 2, Schmid-Hempel et al. 2014). This interpretation is consistent with a potential co-introduction of N. bombi by introduced bumblebees, and/or an increase of N. bombi infections in B. dahlbomii in the presence of B. terrestris, two hypothesis that remain to be tested.

Locustacarus buchneri, a parasitic mite known to affect bumblebees, was found in B. dahlbomii and in invasive bumblebees by two independent studies. Vallejos (2013) found L. buchneri in B. dahlbomii ( 0.6% prevalence, n=176) and B. terrestris (14%, n=88), in Valdivia city in 2011. More recently, Arismendi et al. (2016) found this pathogen at high prevalence in B. dahlbomii (23% of the 52 samples infected), as well as in the invasive B. terrestris (41% n=232) and B. ruderatus (30% n=94). The sequences of the parasite from the three hosts shared 100% identity among themselves and >98% identity with other sequences of L. buchneri reported in Belgium, the Netherlands, and >99% in Japan. Moreover, the L. buchneri infesting B. dahlbomii and invasive B. terrestris and B. ruderatus in Chile are more closely related than any other haplotypes reported in the Netherlands and Japan suggesting a potential horizontal transmission of this pathogen from the invasive bumble bees to B. dahlbomii (Arismendi et al. 2016). These authors suggest that L. buchneri may spread from commercial bumble bees (especially B. terrestris) to the native B. dahlbomii, since non-native and native species overlap and interact extensively throughout Chile (Montalva et al. 2011). The overall high prevalence of L. buchneri in B. dahlbomii compared to other native species whose ranges do not overlap with invading B. terrestris and B. ruderatus is remarkable, and suggests that the invading bumble bees may be the source of this pathogen, though a causal effect has not been clearly documented.

However, Plischuck et al.(2013) did not find infections of L. buchnerii in B. dahlbomii (n=2), B. ruderatus (n=8), or B. terrestris (n=704) sampled in Argentina. Although a low sample size may explain the lack of detection in the two former species, the lack of infected B. terrestris samples in Argentina contradicts the findings of Vallejos (2013) and Arismendi et al. (2016) in Chile, suggesting that further research on the epidemiology of this disease and the role of invasive species on its introduction and spread is needed.

Finally, there is evidence that the highly lethal bumblebee pathogen Apicystis bombi, detected in B. terrestris populations of southern Argentina but not in other native bumblebees of other regions of Argentina (Plischck and Lange 2009), shares the same haplotypes with Apicystis of bumblebees collected in Europe (Maharranov et al. 2013). This pathogen was further detected in B. dahlbomii and B. ruderatus sampled after but not before B. terrestris invasion, suggesting that the pathogen might have been co-introduced with invasive B. terrestris, from which it might have jumped to B. dahlbomii and B. ruderatus (Arbetman et al. 2013).

Although the pathogenic effects of all of these parasites on B. dahlbomii remain unknown, C. bombi has been shown to sterilize founding queens of B. terrestris (Brown, Schmid-Hempel andSchmid-Hempel 2003), leading to failure during colony founding and severely compromised reproductive success. This effect might be stronger in a species like B. dahlbomii that may not have encountered this parasite prior to invasion by B. terrestris and B. ruderatus (Otterstatter and Thomson 2008); a hypothesis that remains to be tested.

It is likely that only a novel, rather abundant, sufficiently virulent parasite should have the rapid effects that could have induced the observed population decline of B. dahlbomii in such a short time (Schmid-Hempe et al. 2014). The pathologies and known virulence of A. bombi and C. bombi would likely qualify for this. Likewise, the mite L. buchnerii invades, parasitizes, and reproduces in the tracheae and air sacs of adult bees (Yoneda et al. 2008, cited in Arismendi et al. 2016), affecting the host’s physiology, lifespan and behaviour (Otterstatter and Whidden 2004, Otterstatter et al. 2005). Urgent research is required to determine the direct effect of C. bombii, A. bombii, L. buchnerii and N. bombi on B. dahlbomii, as well as their combined effects, along with the interaction with other potential stressors like the direct and indirect competition for resources with invasive congeners.


Competitive effects of B. terrestris on B. dahlbomii

Bombus terrestris has larger colonies, emerges earlier from hibernation, has a longer season of activity than B. dahlbomii (Morales et al. 2013). Thus competitive displacement seems plausible (Schmid-Hempel et al. 2014), although it has not been experimentally demonstrated, partly because of the rapid collapse of B. dahlbomii subpopulations after B. terrestris invasion.

Bombus terrestris puts competitive pressures on B. dahlbomii. For example, presence of B. terrestris alters B. dahlbomii’s behaviour. Bombus terrestris primarily nectar robs Fuchsia magellanica flowers, and when it does, it precipitates a shift in B. dahlbomii's behaviour from legitimate visitation to secondary nectar robbing (Combs 2011, N. Rosenberger, pers. obs). All across their overlapping range, B. dahlbomii legitimately visits F. magellanica’s flowers, and it is one of the most frequent pollinators of populations growing in the area of Puerto Blest, Parque Nacional Nahuel Huapi: the eastern limit of the Valdivian Temperate Forest (Morales and Aizen 2002). However, a recent study in this area shows that after B. terrestris arrival to this area, B. dahlbomii frequently robs F. magellanica flowers using holes made by B. terrestris (N. Rosenberger, pers. obs). Preliminary comparisons of visitation to F. magellanica flowers before (2000-2001, Morales and Aizen 2002, 2006), and after (2015-2016, N. Rosenberger, in preparation) B. terrestris’ arrival does not show evidence of a local negative impact on B. dahlbomii’s overall visitation frequency to F. magellanica (Morales and Rosenberger, unpublished), however, the impacts of this phenomenon on diseases transmission through corolla holes, and on plant reproduction remain unknown.

On the contrary, a recent study focused on Vicia nigricans, a typically bumblebee pollinated plant, whose main pollinator was B. dahlbomii (Morales and Aizen 2002, Vazquez and Simberloff 2004), is increasingly visited by B. terrestris who chews through the corolla tube to access the floral nectaries thereby not providing any pollination services (Graham 2015). Bombus dahlbomii visitation frequency to V. nigricans flowers decreases in sites strongly invaded by B. terrestris. As a consequence, comparisons of populations trends across a broad geographic region suggest a consistent decline in visitation frequency to V. nigricans by B. dahlbomii, and a reduction in seed output after B. terrestris invasion (Chalcoff et al. 2015, Chalcoff et al. in prep).

Other focal threats that can affect local subpopulations are the increasing habitat loss and urbanization rate (in particular in Central Chile, Arroyo et al. 1999). Habitat fragmentation and disturbance seems to indirectly affect the species by favouring the spread of B. ruderatus (Aizen and Feinsinger 2003, Morales and Aizen 2002), and B. terrestris (Morales, pers. obs). Finally, global warming may negatively affect and probably accelerate the observed southward retraction observed at the northern part its range (Morales et al. in prep).

This species has been experimentally reared in Chile for a couple of decades (Estay et al. 2001, Olavarria Morales 2003), after which the rearing methodology has been recently patented by the Ing. Patricia Estay, head of the Entomology Laboratory of  the National Institute of Agronomic and Food research of Chile (INIA, La Platina), (http://www.inia.cl/blog/2014/09/01/inapi-otorgo-patente-a-invencion-inia-para-domesticacion-y-crianza-de-abejorros-para-polinizacion/). However it is uncertain whether hives of this species are being commercially produced or subject to trade.

Many of the remaining Argentinian subpopulations as well as many of the Chilean subpopulations are located in National Parks and other Natural and Wildlife Reserves. However, if invasion of introduced bumblebees are the main drivers of its decline, solely protecting habitat will probably not be effective enough to revert this declining trend, because these invasive species also thrive in these habitats.

In addition there have been a lot of outreach activities (radio interviews, talks, etc.) and publications (Morales 2006a,b, 2007, 2009; Arbetman and Morales 2014) aimed to enhance the public awareness on this species.

In Argentina an educational brochure "El mangangá y la polinización de nuestrosbosques" has been designed by Carolina L. Morales and published with the support of the Cannon National Parks Science Scholar Program, the National Parks Administration of Argentina and the Universidad Nacional del Comahue.

In Chile a targeted campaign called "Salvenos Nuestro Abejorro" (Save our bumblebee) has been launched by Jose Montalva: https://salvemosnuestroabejorro.wordpress.com/. In addition, a proposal to classify and protect this species under the National Red Listing Scheme in Chile has been recently approved by the Ministry of Environment of Chile (http://www.mma.gob.cl/clasificacionespecies/listado_especies_12o_pac.htm). The final resolution will be officially published during 2016, and this classification is expected to trigger practical conservations measures.

Some proposed realistic actions needed to mitigate major threats to the taxon are to stop the ongoing importation of B. terrestris, control and eradicate the currently established populations of B. terrestris and B. ruderatus in areas where subpopulations of B. dahlbomii are still healthy (manual removal of colonies/mated queens), identify subpopulations of B. dahlbomii which are immune or resistant to pathogenic diseases that can be used for captive breeding and reintroduction programs, study rearing techniques specific to the native species, genetic analysis of populations where still abundant, to inform management programs. A possible action to reduce populations of invasive Bombus spp. might be to release sterilized drones or males to reduce the fecundity of the species as has been successful in other taxa.