Evidence of endozoochory in upland geese Chloephaga picta and white‐bellied seedsnipes Attagis malouinus in sub‐Antarctic Chile

Abstract Birds are known to act as potential vectors for the exogenous dispersal of bryophyte diaspores. Given the totipotency of vegetative tissue of many bryophytes, birds could also contribute to endozoochorous bryophyte dispersal. Research has shown that fecal samples of the upland goose (Chloephaga picta) and white‐bellied seedsnipe (Attagis malouinus) contain bryophyte fragments. Although few fragments from bird feces have been known to regenerate, the evidence for the viability of diaspores following passage through the bird intestinal tract remains ambiguous. We evaluated the role of endozoochory in these same herbivorous and sympatric bird species in sub‐Antarctic Chile. We hypothesized that fragments of bryophyte gametophytes retrieved from their feces are viable and capable of regenerating new plant tissue. Eleven feces disk samples containing undetermined moss fragments from C. picta (N = 6) and A. malouinus (N = 5) and six moss fragment samples from wild‐collected mosses (Conostomum tetragonum, Syntrichia robusta, and Polytrichum strictum) were grown ex situ in peat soil and in vitro using a agar Gamborg medium. After 91 days, 20% of fragments from A. malouinus feces, 50% of fragments from C. picta feces, and 67% of propagules from wild mosses produced new growth. The fact that moss diaspores remained viable and can regenerate under experimental conditions following the passage through the intestinal tracts of these robust fliers and altitudinal and latitudinal migrants suggests that sub‐Antarctic birds might play a role in bryophyte dispersal. This relationship may have important implications in the way bryophytes disperse and colonize habitats facing climate change.

, it is important to understand the dispersal mechanisms and potential for bryophytes to colonize new habitats.
According to Urban (2015), if the Earth's temperature increases by 3℃, South America will be one of three regions where extinction risks of species will be highest (23%), followed by Australia and New Zealand (14%). Additionally, climate change is causing an upslope shift in montane plant and animal communities (Elsen & Tingley, 2015;Freeman et al., 2018) that is driven by niche conservatism, which is the retention of ancestral ecological characteristics, such as a habitat, by a species. Faced by climate change, species are more likely to respond by "following" their niches or ancestral climate regime rather than adapting their climatic tolerances (Wiens & Graham, 2005), which may represent a challenge for sessile organisms such as mosses.
Bryophytes, as most land plants, are sessile organisms depending on biotic or abiotic vectors to disperse propagules to other areas.
Their movement is made possible via wind dispersal facilitated by sexual reproduction morphological features such as exposed spores, tall sporophytes, or the production of a large number of spores (Barbé et al., 2016;Muñoz, 2004); or via water dispersal, such as gemma or splash cups (Glime, 2017a(Glime, , 2017bZanatta et al., 2018).
On the other hand, some species of the Splachnaceae family, such as Tayloria dubyi, have even been known to have sticky spores and brightly colored sporophytes that emit a strong odor, mimicking decomposing organic matter, to attract flies as potential dispersal vectors (Jofre et al., 2011). Bryophyte dispersal and colonization are also facilitated by bryophyte cell totipotency, an asexual reproduction mechanism able to regrow the entire gametophyte from tissue fragments (Anderson, 1963;Cleavitt, 2002;Longton, 1997;Proctor et al., 2007;Zhang et al., 2003). However, because some birds and bryophytes share the same habitats, these birds may serve as animal vectors that allow these small plants to reach areas that they would not reach otherwise, or help reach them quicker and more frequently. The behavior of birds can aid in directed and longdistance movement of bryophytes as they may act as dispersers in local sites through foraging and nesting (Amélio et al., 2017;Calvelo et al., 2006;Parnikoza et al., 2012Parnikoza et al., , 2018, and transcontinentally through migratory movements (Chmielewski & Eppley, 2019). There has been evidence of migratory birds carrying bryophytes via ectozoochory, by external transportation (Lewis et al., 2014). Given this information, birds may be able to do the same via endozoochory, through internal ingestion, as has been shown with ferns and other herbs (Blanco et al., 2019;Hervías-Parejo et al., 2019;Lovas-Kiss et al., 2018;Silva et al., 2020).
The upland goose (Chloephaga picta, order Anseriformes, family Anatidae (Carboneras & Kirwan, 2020)) and the whitebellied seedsnipe (Attagis malouinus, order Charadriiformes, family Thinocoridae (del Hoyo et al., 1996)) are two herbivorous bird species that inhabit sub-Antarctic South America and could be potential vectors for bryophyte dispersal (Russo et al., 2020). A. malouinus is an altitudinal migratory shorebird that moves downwards from its upland habitats to lowland flats during the harsh winters (Fjeldså & Krabbe, 1990;del Hoyo et al., 1996;Jaramillo et al., 2003;"eBird: White-bellied Seedsnipe Attagis malouinus," n.d.) and has occasionally been known to leave Patagonia to the Falkland Islands (Hayman et al., 1986). C. picta is a migratory goose that is known to have larger movements through South America, migrating between breeding and wintering grounds. Pedrana et al. (2015) tracked the migratory route of a male C. picta and found that he migrated a minimum distance of 1,485 km from Buenos Aires (the wintering grounds at 34°S, 58°W) to Santa Cruz Province, Patagonia (the breeding grounds at 51°S, 69°W). This species also migrates altitudinally, as the same study found that the individual moved to lower than 100 masl on the wintering ground and between 1000-1500 masl on the breeding ground.
Upland geese tagged on Navarino Island have been observed some 400 km north near Rio Gallegos in Argentina and some geese perform daily altitudinal migration from sea level up to over 700 masl there (J. Jiménez, 2020, pers. comm.).
Three moss species that are within the ranges of C. picta and A. malouinus, and that we hypothesize might be fed upon by them, are Polytrichum strictum, Brid., family Polytrichaceae, also known as "pigeon wheat"; Syntrichia robusta (Hook. & Grev.) R.H. Zander, family Pottiaceae; and Conostomum tetragonum (Hedw.) Lindb., family Bartramiaceae. P. strictum and C. tetragonum are bipolar species, meaning that they are present in high latitudes of both the Northern and Southern Hemispheres, whereas S. robusta is considered endemic to Patagonia (Pereira et al., 2006). According to the Tropicos database of the Missouri Botanical Garden  http://www.tropi cos.org, visited on 4 April 2021), in the Southern Hemisphere, P. strictum can be found widely distributed from Bolivia to the Magallanes region of Chile and Tierra del Fuego, Argentina (Ireland et al., 2006). C. tetragonum extends from Costa Rica and Colombia (Ceballos & Rangel, 2008) to the Magallanes region. Finally, S. robusta was described between 35°S (Maule and Bío-Bío, Chile) and 55°S (Magallanes, Chile, and Tierra del Fuego, Argentina). It can also be found within 40°S and 43°S in New Zealand and has been recorded in the Falkland and South Georgia Islands (Moraga, 2020;Pereira et al., 2006).
Previous research suggests that birds, such as mallard ducks (Anas platyrhynchos) and sub-Antarctic geese and shorebirds, might be capable of dispersing bryophytes through endozoochory (Russo et al., 2020;Wilkinson et al., 2017). Additional evidence of bryophyte dispersal through endozoochory has been reported in spectacled flying foxes (Pteropus conspicillatus) (Parsons et al., 2007) and Even though viable bryophyte gametophyte fragments have been recovered from avian and mammalian species that feed on these plants, previous research has attempted to cultivate or regenerate these fragments with little success (Parsons et al., 2007;Russo et al., 2020;Wilkinson et al., 2017). Given that the dispersal and establishment of plants involve sequential and interdependent steps to be successful under field conditions, it is inappropriate to assume that finding bryophyte fragments in bird feces is directly correlated with successful dispersal. Thus, we cannot conclude that these fragments successfully propagate after passing through the bird's digestive system. For these reasons, together with the search of viable fragments in bird feces, it was necessary to first test whether the bryophyte fragments found in bird feces were capable of regenerating when grown under laboratory conditions. Here, we report observations of the potential role of two herbivorous bird species, C. picta and A. malouinus, as endozoochorous bryophyte dispersers by testing the viability and regenerative capabilities of fragmented bryophyte gametophytes retrieved from their feces. We propose that both avian species have the potential to serve as dispersal vectors for bryophytes in the sub-Antarctic through endozoochory. Our prediction is that after being ingested, defecated, and cultivated under the proper conditions, the bryophyte fragments will have the totipotence to regenerate a new individual. Endozoochory is likely to be a widespread phenomenon. Therefore, our findings could be applicable to regions beyond sub-Antarctic Chile, like the Arctic, where birds also feed on mosses (Fox & Bergersen, 2005) and climate change is altering the vegetation communities and plant-herbivore relations (Bjorkman et al., 2018;Klein et al., 2008).

| MATERIAL S & ME THODS
The research was conducted on Navarino Island (54°S, 67°W), Magellanic region, in sub-Antarctic Chile, at the southern end of the Americas. The island has a rugged topography with marked, but still moderate seasons. At the lower ranges, the forest is covered by a mix of southern beech species (Nothofagus betuloides and Nothofagus pumilio) growing in krummholz formation at the tree line. Above the tree line, a rich community of small plants, including cushion plants, some graminoids, lichens, and mosses, thrive in a Magellanic tundra environment (Méndez et al., 2013). We sampled C. picta feces at sea level and C. picta and A. malouinus feces at ca. 700 masl, in open meadows some 100 m above the tree line. Fieldwork and laboratory work were conducted over the course of five weeks, and the growth of recovered fragments was monitored over an additional 13 weeks in a growth chamber in Navarino Island, from December 2018 to April 2019.
In the field, fresh C. picta and A. malouinus fecal samples were collected from six locations, focusing on upland and lowland sites. We sampled opportunistically two sites near sea level, Lake Zañartu (19F 587,064.87E 3,911,888.71S) and Robalo Bay (19F 585,969.41E 3,911,044.12S), and four sites above the tree line, near streams, and large ponds on the Cerro Bandera summit (19F 587,490.15E 3,908,430.45S, 19F 586,779.97E 3,907,237.18S, 19F 586,790.75E 3,907,270.36S). Fecal samples of both birds were identified in the field through our personal experience observing these birds in situ. Multiple fecal samples that visibly contained bryophyte fragments were collected from each site and later only one was selected from each site for processing (n = 6), three from each bird species. Of those six samples, there were two C. picta feces from lowland sites, one C. picta feces from an upland site, and three A. malouinus feces from upland sites. Additionally, three moss species (P. strictum, S. robusta, and C. tetragonum) were sampled near fecal sample sites to test for regeneration capabilities compared to fragments from fecal samples under the same growing conditions. Bryophytes collected in the field were identified by RM using the preliminary key to mosses of Isla Navarino (Buck & Goffinet, 2010).
In the laboratory, C. picta and A. malouinus fecal samples were stored in paper bags at room temperature before processing. We used a disinfected precision knife to remove the outer layer of the feces to eliminate the possibility of contamination by wind-or soilborne spores and fragments. Two thin disks were then sliced from each fecal sample (approximately 1 mm in thickness). Each disk was dissolved separately in clear dishes using filtered water collected from the Robalo River (0.22μm PVDF sterile syringe filters, Millipore, Cork, Ireland). The disk contents were observed under a Leica ICC50 HD compound microscope at 40x and 100x magnification. Bryophyte fragments were picked out from the dishes using a probe and forceps, placed into Eppendorf tubes corresponding to each disk sample and partially filled with filtered river water, following Russo et al.'s (2020) protocol. These samples were stored for a few days at 4℃ until inoculation. Additionally, the wild-collected mosses sampled were also cut into fragments.
The six fecal samples (three from each bird species) resulted in 11 disk samples analyzed, and 138 bryophyte fragments recovered and inoculated (106 belonging to C. picta and 32 to A. malouinus samples). The treatments tested were either a culture container, consisting of two plastic cups sealed together at the openings with parafilm to obtain a greenhouse effect, with commercial sphagnum peat soil (Kekkilä Professional Substrate, Vantaa, Finland) or a microplate for in vitro growth. The in vitro treatment was made using an agar Gamborg mixture basal salt (B5 salts) medium (G768, Phytotechnology Laboratories) with a pH of 5.8, a specific medium for moss growth (González et al., 2006). Once all fecal samples were processed, we randomly assigned the fragments from each fecal disk sample and from each moss species sample to either treatment to be propagated, with the exception of one A. malouinus sample, which only had one disk that contained one bryophyte fragment, and was therefore placed only in the agar Gamborg treatment. For the peat soil treatment, the fragments were placed on the soil, thoroughly misted with filtered water, and then covered and sealed for the remainder of the experiment with occasional misting throughout the observation period to keep the culture containers humid. High inner humidity and condensation were observed in the sealed containers throughout the experiment. For the agar Gamborg treatment, the samples were placed in a 28-well agar plate and then covered and sealed with parafilm. A total of 18 culture mediums were placed into a growth chamber, with the temperature fluctuating between a minimum of 5 ºC and maximum of 25 ºC throughout the entirety of the experiment. Bryophyte cultures were grown on red and blue LED lights in a 4:1 ratio, at a day:night cycle of 16:8 hr. The wild moss sample fragments inoculated in the peat soil containers were placed in the growth chamber on 8 January 2019, followed by the fecal sample fragments inoculated in the peat soil containers on 9 January 2019, and fecal sample and wild moss sample fragments inoculated in the agar Gamborg on 11 January 2019. All the samples were left in the growth chamber for a total of 91 days.

| RE SULTS AND D ISCUSS I ON
A total of 8 culture mediums of 18 (44%) produced growth in the treatments tested. One of five culture mediums with A. malouinus fragment samples (20%), three of six with C. picta fragment samples (50%), and four of six with wild moss fragment samples (67%) produced new growth (Table 1)  Previous research has attempted to cultivate fragments found in feces in vitro with some success (Russo et al., 2020;Wilkinson et al., 2017). In this study, we partially followed Russo et al.'s (2020) fragment-processing methods. However, we decided to set two growing conditions for the fragments recovered from the feces to increase the probability of bryophyte growth. These consisted of an ex situ growth condition using commercial peat soil and an in vitro condition using agar Gamborg medium for both fragments from fecal and wild moss samples taken from the field. We predicted that the fragments from both wild moss and fecal samples would regenerate. Previous research described growth from bryophyte spores after 60 days (Proctor, 1961) and from bryophyte fragments after 11 days (Wilkinson et al., 2017). However, we did not observe signs of regeneration from our cultures until 42 days of growth.
Our reasoning for the chosen treatments was to test which, if any, would allow bryophyte fragment recovery and growth under laboratory conditions. Although the sample size was small, we observed clear evidence of bryophyte regeneration from feces of both bird species in the peat soil treatment (Table 1). These results reveal that a small fraction of moss diaspores remains viable following the passage through the intestinal tract of both birds and is capable of regeneration in suitable conditions. Although small, our cases are Note: Growth results for fragments from feces and mosses (N = 17). Regeneration presence and absence of fragments recovered from upland goose (Chloephaga picta) and white-bellied seedsnipe (Attagis malouinus) feces and fragments of wild-collected mosses (Syntrichia robusta, Polytrichum strictum, and Conostomum tetragonum) grown under controlled light regimes and two types of substrates in a growth chamber.
F I G U R E 1 Moss regeneration from bryophyte fragments. Detailed image of green shoots and a moss bed that regenerated in the peat soil treatment from the bryophyte fragments extracted from an upland goose (Chloephaga picta) fecal sample

(a) (b)
It is plausible that the in vitro (agar Gamborg) conditions could have yielded false-negative results from the fecal sample fragments tested of both bird species (see Table 1), as moss diaspores may take longer to germinate in sterile conditions due to the necessary acclimation to the agar Gamborg substrate and near-neutral pH conditions in the medium (Sabovljevic et al., 2014). However, there was evidence of bryophyte regeneration in fragments recovered from feces in the peat soil substrate. On the other hand, wild moss gametophyte growth was observed in samples inoculated both in agar Gamborg medium and peat soil. Therefore, these results support the hypothesis of bird endozoochory of these bryophytes in the sub-Antarctic environment.

| CON CLUS IONS
We confirmed the hypothesis that fragmented bryophyte gametophytes retrieved from the feces of herbivorous birds could regenerate in laboratory conditions. Consequently, our research shows that it is plausible for fragments to be dispersed through endozoochory by these herbivorous birds in the sub-Antarctic, possibly further aiding bryophytes in this region to disperse beyond wind or rain, and increasing their capability to reach specific habitat types ( Figure 2).
As both bird species are robust fliers, altitudinal migrants for short distances and latitudinal migrants for long distances, their role as dispersers might not only occur at a local scale within high-latitude regions, but also at a broader scale to similar habitats across latitudes where these bryophyte species are still found. This is especially important as with warming conditions, organisms such as mosses would have to move to higher altitudes or latitudes to maintain viable populations. By serving as dispersal vectors, birds would be able to aid in this process in short time scales in the vertical and horizontal axes of the landscape. This process is especially critical in maintaining the viability of tundra bryophytes in Navarino Island, which contains an ecosystem of high ecological importance (Goffinet et al., 2012) and holds many native species (Méndez et al., 2013), and just as critical globally where bryophytes are also being affected by changes to their ecosystems.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests. F I G U R E 2 Schematic diagram of native bryophyte endozoochory by sub-Antarctic birds. Schematic illustration of the possible endozoochory mechanism of bryophyte dispersal by the upland goose (Chloephaga picta) and the white-bellied seedsnipe (Attagis malouinus)

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data used in the present study are publicly available in the