Long-distance dispersal of the beach strawberry, Fragaria chiloensis, from North America to Chile and Hawaii

Abstract

Background and Aims

The beach strawberry, Fragaria chiloensis, is found in a narrow coastal band from the Aleutian Islands to central California and then jumps thousands of kilometres all the way to Hawaii and Chile. As it probably had a North American origin, it must have been introduced to the other locations by long-distance dispersal. The aim of this study was to determine which agent carried the beach strawberry to its Pacific and South American locations.

Methods

A deductive framework was constructed to separate between the possible modes of long-distance dispersal involving animals, wind and ocean currents. Bird migration was subsequently identified as the most likely scenario, and then the routes, habitats, feeding preferences and flight distances of all the shorebird species were evaluated to determine the most likely carrier.

Key Results

Six species migrate between North America and Chile and feed on the beaches and rocky shores where F. chiloensis grows naturally: Black-bellied Plovers, Greater Yellowlegs, Ruddy Turnstones, Sanderlings, Whimbrels and Willets. Of these, only two eat fruit and migrate in long continuous flight: Ruddy Turnstones and Whimbrels. Two species travel between North America and Hawaii, eat fruit and forage on the beaches and rocky shores where F. chiloensis grows naturally: Pacific Golden-plovers and Ruddy Turnstones. Ruddy Turnstones eat far less fruit than Pacific Golden-plovers and Whimbrels, making them less likely to have introduced the beach strawberry to either location.

Conclusions

We provide evidence that F. chiloesis seeds were probably dispersed to Hawaii by Pacific Golden-plovers and to Chile by Whimbrels.

Introduction

The beach strawberry, Fragaria chiloensis, is found in a narrow, continuous coastal band from the Aleutian Islands to central California, and then its range jumps all the way to the central coast of Chile and the mountains of Hawaii (Staudt, 1962, 1999; Hancock, 2020). Somehow seeds of F. chiloensis were carried many thousands of miles from North America to these distant destinations, as there is strong phylogenetic data that its progenitor arose in the region of the Bearing Strait, ~0.5–2.5 Mya (Njuguna et al., 2013; Liston et al., 2014). There is no direct ancestor of F. chiloensis now living wild in Eurasia or Alaska, but a number of its more distant relatives are dispersed in this region (Edger et al., 2019a, 2020; Liston et al., 2019).

Herein, we describe a deductive framework that can be used to identify the key players in long-distance dispersal of plants. We then use this framework to find that F. chiloensis was probably introduced to Chile by Whimbrels (Numenius phaeopus) and to Hawaii by Pacific Golden-plovers (Pluvialis fulva). It has long been speculated that shorebirds were the pivotal players in the long-distance dispersal of strawberry (Fosburg, 1951; Staudt, 1999; Sears, 1993; Finn et al., 2013), but no one has provided a detailed analysis of which species it might have been.

Long-distance dispersal

The whole topic of the origin of disjunct species distributions has engaged botanists for many decades (Raven, 1963; Cruden, 1966; Carlquist, 1966, 1967, 1981). Over 200 plant species have been described with discontinuous North and South American distributions (Simpson et al., 2017) and more than 100 Hawaiian angiosperms are thought to have transoceanic North American origins (Baldwin and Wagner, 2010). While there are not many other berry species like F. chiloensis that have been shown to have such discontinuous distributions, a few have been described including species in Empetrum, Gaultheria, Rubus and Vaccinium (Baldwin and Wagner, 2010; Simpson et al., 2017).

Probably the most in-depth analysis of what are called ‘amphitropical disjuncts’ of berry crops has been made on the crowberry (Empetrum), which has a bipolar disjunction much like F. chiloesis between the Arctic and southernmost South America (Popp et al., 2011). Using sequence data from two nuclear low-copy and two plastid DNA regions, the authors found that the red-fruited southern hemisphere crowberry, referred to as E. rubrum, has as its closest relative black-fruited plants of E. nigrum from north-western North America. They suggested that the disjunct distribution of crowberries was probably the result of a single dispersal by a bird from north-western North America to southernmost South America, taking place in the Mid-Pleistocene.

The most popular hypotheses explaining discontinuous distributions of species in North and South America are long-distance dispersal (LDD) and island hopping (Raven, 1963; Cruden, 1966; Carlquist, 1981). In the LDD hypothesis, it is predicted that diaspores were directly dispersed to suitable habitats on the other side of the equator. The island hopping hypothesis predicts that there was a corridor of temperate microhabitats running along the mountains of the two continents on which a species could have ‘hopped’ in stages before it arrived at its current location. Most researchers believe that LDD is much more likely, as there is little empirical data to support the island hopping proposal (Simpson et al., 2017; Villaverde et al., 2017).

LDD occurs through several mechanisms (Cruden, 1966; Gillespie et al., 2012): zoochory (dispersal by animals), anemochory (dispersal by wind) and hydrochory (water dispersal). Zoochory can occur through epizoochory (propogules attach to the animals) and endozoochory (an animal ingest propagules at one site and defecates them at another). Birds are often implicated in LDD because their migrations can be intercontinental and they are in contact with plants through foraging and nesting. The majority of the birds that migrate from the United States to temperate South America and Hawaii are in the order Charadriiformes containing the shorebirds, gulls and auks (Cruden, 1966).

The dispersal distance through epizoochory is determined by the propagules’ adhesiveness and the animal’s ability to detect and remove them, while dispersal distance through endozoochory is determined by gut retention times and how far and quickly the animal travels. Most researchers believe that epizoochory is much more likely than endozoochory in bird-driven LDD, as most gut retention times in birds are less than a day, although there are some examples where seeds have been carried for much longer periods (Viana et al. 2016). Viable seeds of Celtis, Convolvulus, Malva and Rhus were regurgitated up to 340 h after consumption by Killdeer (Charadrius vociferus) and Least Sandpipers (Erolia minutilla) (Proctor, 1968).

No one really knows when birds actually began to migrate between North America and Chile or Hawaii, but the migrating habit itself arose long ago. Various authors have speculated that birds began migrating no later than in the late Mesozoic (252–66 Mya) or early Tertiary (66–2.6 Mya) (Cruden, 1966). Paleotringa (Scolopacidae), a genus of extinct sandpipers, is known from the Eocene, and fossils of the genera Charadrius (plovers) and Limosa (godwits) are known from the Miocene (Wetmore, 1940). Molecular dating has suggested that the three suborders of shorebirds originated in the late Cretaceous between 79 and 102 Mya (Baker et al., 2007). When the extant species of shorebirds emerged is not known; however, most scientists feel they arose during the glacial periods of the Pleistocene (Conners, 1983; Tan et al., 2019) and have undergone only limited amounts of population divergence since then due to their long-distance migratory behaviour and polygamous breeding structures (Belliure et al., 2000; D’Urban Jackson et al., 2017).

Dispersal agent carrying F. chiloensis to Chile and Hawaii

It is highly unlikely that F. chiloensis was dispersed to Chile and Hawaii through anemochory and hydrochory. Anemochory can be eliminated as strawberry seeds are imbedded in relatively heavy fruit, have no dispersal structures and their seeds are too large to be carried by wind. Hydrochory of F. chiloensis fruit and seeds to Chile is also highly unlikely, because although the fruit of F. chiloensis can float, there are no ocean currents that would take them further south than Mexico (Gillespie, 2012), and it is highly unlikely that the fruit would not decay and disintegrate even across that distance. There are currents that could potentially push fruit down the coast of California to southern Mexico and then to Hawaii, but again it is highly unlikely that the fruit would stay intact for the months that it would take to get there.

The possibility of human transport of seeds to Chile and Hawaii can also be largely excluded. Most scholars agree that people entered the Americas through Beringia at least 16 000 years ago and arrived in Monte Verde Chile about 2000 years later (Dillehay et al., 2008; Goebel et al., 2008; Bodner et al., 2012). These migrants would have been initially in contact with F. chiloensis and could very well have consumed its fruit; however, these hunter-gathers would have had to carry strawberry seeds for many thousands of miles outside their native range to introduce them to Chile. It is also unlikely that people carried F. chiloensis seeds from the coast of North America to Hawaii, as the evidence of pre-Columbian contact across those thousands of miles of open ocean is quite limited (Arnold, 2007; Thompson et al., 2014).

There is a possibility that Polynesians introduced the strawberry from Chile to Hawaii. Polynesians are known to have been wide-ranging explorers, established a settlement on Rapa Nui (Easter Island, a remote island far off the coast of Chile) and may have introduced the South American sweet potato throughout the Pacific, including Hawaii. However, the sweet potato of Polynesia probably came from the region of Peru/Ecuador where F. chiloensis is not native (Roullier et al., 2013), and the evidence of Polynesian and South American contact remains controversial (Gonpora et al., 2008; Lawler, 2010; Gonçalves et al., 2013; Fehren-Schmitz et al., 2017).

The most likely explanation for the disjunct distributions of F. chiloensis is zoochory via bird migrations. Birds have been migrating from North America to South America and Hawaii for probably hundreds of thousands of years, and a few could have carried viable seeds the requisite distance. Birds that consume fleshy fruits could transport strawberry seeds in their digestive tract or on their bodies (Staudt, 1999). Carlquest (1983) placed F. chiloensis in the class ‘viscid or gelatinous seeds carried on the feathers or feet of birds’ and suggested this category represents 20 % of the plant species subject to LDD.

Strawberry seeds are well adapted for LDD. They have a tough seed coat and can remain viable for months or even years (Scott and Draper, 1970; Reed and Hummer, 1995). They also can germinate on a wide array of soil types if provided with ample moisture (Hancock and Bringhurst, 1979a). Strawberries reproduce asexually through stolons and once a mother plant is established, dozens of daughter plants can be produced within a year (Staudt, 1999; Hancock and Bringhurst, 1979b).

To successfully introduce F. chiloensis seeds via LDD, a bird would have had to make the trip in a relatively short period of time with minimal stopovers, if they are to retain seeds in their gut or on their bodies when they arrive. The probability that a single individual might carry a seed might be quite remote, but the movement of tens of thousands of these birds a year over hundreds of thousands of years would make an occasional introduction likely.

MATERIALS AND METHODS

The analysis was begun by determining which shorebird species migrate from the geographical range of Fragaria chiloensis (L.) Duch in North America to Chile and Hawaii. The Atlas of Bird Migrations (Elphick, 2011) and The Cornell Birds of the World (https://birdsoftheworld.org/bow/home) were first searched to discover which birds were most likely to make long-distance migrations from North America to Chile and Hawaii. The Birds of Chile (Jaramillo, 2013) was then used to determine how common these birds were in Chile and these observations were further verified by an older report in Bulluck (1949) and more recent eBird reports (https://ebird.org/map/). The migrants regularly travelling to Hawaii were identified by consulting species lists of the birds in Hawaii (Pyle and Pyle, 2017) and again examining the available eBird records. The eBird data were also used to determine the degree of overlap in the migration routes of individual bird species and the geographical range of the beach strawberry. Among those shorebird species which were found to migrate from the northern geographical range of F. chiloensis to Chile and Hawaii, it was then determined which utilize habitats where F. chiloensis is known to grow wild. The notes of foraging behaviour found in the Audubon Guide to North American Birds (https://www.audubon.org/bird-guide), the Cornell Guide to North American Birds (https://www.allaboutbirds.org/guide/search and Bent (1927, 1929) were consulted to obtain this information. Fragaria chiloensis is commonly found in high density on sandy beaches, rocky shores and scrub lands in a narrow band adjacent to the ocean extending from the Aleutian Islands to Central California (Hancock and Bringhurst, 1979b; Staudt, 1999; Daubney, 2003).

Next, the Audubon and Cornell guides were consulted to determine which of the shorebirds that forage in the habitats where F. chiloensis grows are also known to eat berries. The birds most likely to carry strawberry seeds during migration are obviously those that eat fruit and carry seeds either through their digestive systems or on their feathers or feet after preening. Strawberries are ripening all along the range of F. chiloensis from late July to early September, at the same time shorebirds are migrating to their wintering grounds.

Finally, published literature was searched to find records of the distances flown by those shorebird species that come into contact with the fruit of F. chiloensis and fly to Chile and Hawaii. Considerable work has been done using telemetry to document the long-distance flights of land-based birds (Gill et al., 2009; Hedenström, 2010; Sokolov, 2011). To carry F. chiloensis seeds from North America to Chile, a bird would have had to carry them about 9000 km from the southern extent of the F. chiloensis range in central California to Region VI (Región del Maule) in Chile. To introduce seeds from coastal Alaska to Hawaii, a bird would need to carry F. chiloensis seeds about 3500 km.

RESULTS AND DISCUSSION

Shorebirds most likely to have carried F. chiloensis seeds

There are 17 shorebird species that now migrate regularly from Alaska to Chile and there was another one that made the trip until the early 20th century when it was driven to extinction (Eskimo Curlew) (Table 1). All utilize the Pacific Flyway along the western coast of North America where strawberries grow wild, except the Eskimo Curlew which probably headed east from its nesting ground to the Delaware Valley and then south, missing the northern coastal range of F. chiloensis (Bent, 1927). Of these, ten are commonly found in Chilean Regions (VII–XII) where F. chiloensis grows natively. However, only six of these feed on the beaches and rocky shores in North America where F. chiloensis is found: Black-bellied Plovers, Greater Yellowlegs, Ruddy Turnstones, Sanderlings, Whimbrels and Willets. Of these six, three are also reported to eat berries: Greater Yellowlegs, Ruddy Turnstones and Whimbrels. Between these three, Whimbrels are the only ones known to commonly consume berries, and as such are most likely to migrate carrying seeds of F. chiloensis (Audubon and Cornell Guides). The Whimbrels in Chile represent the race Numenius phaeopus hudsonicus, which nests in Alaska and migrates across the entire range of F. chiloensis along the Pacific Flyway.

Table 1.

Range and feeding preferences of North American shorebirds commonly found in Chile [sources: Jaramillo (2013); Audubon Bird Guide. https://www.audubon.org/bird-guide; The Cornell Lab: All about birds. https://www.allaboutbirds.org/guide/search]

Species	Common name	In Regions VII–XII?	Forages on beaches?	Consumes berries?
Actitis macularius	Spotted Sandpiper	Uncommon	Yes	No
Aphriza virgata	Surfbird	Yes	No	No
Arenaria interpres	Ruddy Turnstone	Yes	Yes	Limited
Calidris alba	Sanderling	Yes.	Yes	No
Calidris bairdii	Baird’s Sandpiper	Yes	Uncommon	No
Calidris canutus	Red Knot	No1	Yes	No
Calidris fuscicollis	White-rumped Sandpiper	No1	No	No
Calidris melanotos	Pectoral Sandpiper	Uncommon	No	No
Calidris semipalmatus	Semipalmated Plover	No	Yes	No
Charadrius vociferus	Killdeer	No	No	No
Limosa haemastica	Hudsonian Godwit	Yes	No	No
Numenius borealis  2	Eskimo Curlew	Yes	No	Yes
Numenius phaeopus	Whimbrel	Yes	Yes	Yes
Pluvialis squatarola	Black-bellied Plover	Yes	Yes	No
Pluvialis dominica	American Golden-plover	Rare	Yes	Yes
Tringa flavipes	Lesser Yellowlegs	Yes	No	No
Tringa melanoleuca	Greater Yellowlegs	Yes	Yes	Limited
Tringa semipalmata	Willet	Yes	Yes	No

¹Tierra Del Fuego in Argentina.

²Now extinct. Travelled east on its migration route from its breeding grounds and did not pass over coastal Alaska.

Sixteen North American shorebirds regularly travel to Hawaii on the Nearctic–Hawaiian Flyway, but only four forage on beaches and consume berries: Bristle-thighed Curlews, Pacific Golden-plovers, Whimbrels and Ruddy Turnstones (Table 2). Of these, only the Pacific Golden-plover and Ruddy Turnstone probably come into contact with native populations of F. chiloensis. Bristle-thighed Curlews migrate directly from their breeding grounds to the Pacific Islands and overfly the Aleutian Islands completely (Marks et al., 2020). The Whimbrels in Hawaii represent the race Numenius phaeopus variegatus, which breeds in north-east Siberia and travels across the Bay of Bengal to Melanesia, Micronesia and Australasia (Cramp and Simmonds, 1983; Skeel and Mallory, 2020). Between the Pacific Golden-plovers and Ruddy Turnstones, Pacific Golden-plovers are the most likely to pick up F. chiloensis seeds during migration, as they are reported to eat many more berries (Bent, 1929; Marks et al., 2020, Audubon and Cornell Guides), and they make pre-migratory stopovers of up to 40 d at sites located south of their nesting grounds where F. chiloensis grows (Marks et al., 2020). We are not alone in speculating that the Pacific Golden-plover played a significant role in the dispersal of North American plants into Hawaii (Hillebrand, 1888; Carlquist, 1967; Price and Wagner, 2018).

Table 2.

Range and feeding preferences of North American shorebirds commonly found in Hawaii [sources: Pyle and Pyle (2017); Audubon Bird Guide. https://www.audubon.org/bird-guide; The Cornell Lab: All about birds. https://www.allaboutbirds.org/guide/search].

Species	Common name	Forages on beaches?	Consumes berries?
Pluvialis squatarola	Black-bellied Plover	Yes	No
Pluvialis fulva	Pacific Golden-plover	Yes	Yes
Calidris semipalmatus	Semipalmated Plover	Yes	No
Numenius tahitiensis	Bristle-thighed Curlew	Yes	Yes
Numenius phaeopus	Whimbrel	Yes	Yes
Limosa lapponica	Bar-tailed Godwit	No	Limited
Arenaria interpres	Ruddy Turnstone	Yes	Limited
Calidris alba	Sanderling	Yes	No
Calidris alpina	Dunlin	Yes	No
Calidris minutilla	Least Sandpiper	No	No
Calidris melanotos	Pectoral Sandpiper	No	No
Calidris mauri	Western Sandpiper	Yes	No
Limnodromus scolopaceus	Long-billed Dowitcher	No	No
Gallinago delicata	Wilson’s Snipe	No	No
Tringa incana	Wandering Tattler	Yes	No
Tringa flavipes	Lesser Yellowlegs	No	No

Flight distances of shorebirds potentially carrying seeds of F. chiloensis

Of the four species that are most likely to carry seeds of F. chiloensis between North America and South America and/or Hawaii (Greater Yellowlegs, Pacific Golden-plovers, Whimbrels and Ruddy Turnstones), all have been documented to undertake very long continuous flights during migration except Greater Yellowlegs (Elphick and Tibbitts, 2020). Based on fat content, the flight ranges of Greater Yellowlegs have been estimated to average 1378–1600 km, with maxima of 2864–3032 km (McNeil, 1970; McNeil and Cadieux, 1972). These distances would require them to stopover multiple times on route to Chile, greatly reducing the probability that F. chiloensis seeds would remain on them when they reached their wintering grounds.

Pacific Golden-plovers have been shown to routinely fly continuously from their staging grounds in Alaska to Hawaii (Johnson et al., 2020). Johnson et al. (2011) found that on average, the northward passage of the Pacific Golden-plover on the Nearctic–Hawaiian Flyway took about 3 d and covered 4800 km, while the southward trip took around 4 d and covered 4900 km. In another study, Johnson et al. (2015) found the Pacific Golden-plover’s transoceanic migrations between non-breeding grounds in Kwajalein Atoll, Japan and Hawaii and breeding grounds in Alaska and Chukotka were non-stop, ranging from 4100 to 9370 km over periods from 3 to 8 d. Their close relative, the American Golden-plover (Pluvialis dominica), is known to migrate 4000 km in the Atlantic flyway between Nova Scotia and South America (Stoddard et al. 1983).

Ruddy Turnstone have been documented to routinely fly 3600 km in continuous flights from Alaska to Hawaii in 3 d (Thompson et al., 1974; Davidson and Gill, 2008). Their flights on the Pacific Flyway have not been measured, but on the East Asian–Australasia Flyway flights have been documented of at least 7600 km over 6 d (Minton et al., 2010, 2011). If they fly the same distances on the Pacific Flyway, it would take only one stopover to make the journey from central California to Chile.

The length of Whimbrel flights has not been measured from North America to Hawaii or Chile, but flights in other directions have been recorded that exceed the distance between Alaska and Hawaii and fall only 1000 km short of the distance between central California and Chile. Johnson et al. (2016) followed the migration of nine Whimbrels breeding in the eastern Canadian sub-Arctic and found that three of them made non-stop flights of ~8000 km from Churchill to South America down the Central Flyway. The other six stopped at two staging sites on the mid-Atlantic seaboard of the United States along the way. A Whimbrel named Hope staged on South Hampton Island in upper Hudson Bay and left on a non-stop southern flight of more than 3500 miles over the open Atlantic to St. Croix in the U.S. Virgin Islands (Smith et al., 2011). Three birds breeding in the Mackenzie Delta, extreme north-west Canada, migrated east and south across the Atlantic to north-east South America, one flying ~7000 km in non-stop flight over 6 d (Watts and Smith, 2012). Their travel time in a single flight lasted 3.5–4 days. Alvies et al. (2016) found that Icelandic Whimbrels (Numenius phaeopus) cover a round-trip of 11 000 km in two non-stop sea crossings, flying at speeds of up to 24 m s⁻¹; this is one of the fastest recorded for shorebirds.

Conclusions

Three shorebird species are most likely to have introduced F. chiloesis seeds outside of North America: Pacific Golden-plovers to Hawaii, Whimbrels to Chile, and Ruddy Turnstones to both Chile and Hawaii. These three species all migrate across the range of North American F. chiloensis, forage in the habitats where F. chiloensis grows wild, eat berries and often migrate in flights over long continuous distances, which are far enough to reach the two destinations with no more than one stopover. Of these three species, the Pacific Golden-plover to Hawaii and the Whimbrel to Chile are the most likely to have been the carriers, as Ruddy Turnstones browse far less frequently on fruit. Pacific Golden-plovers are quite common in their wintering grounds in Hawaii (Andres et al., 2012), as are Whimbrels in southern, coastal Chile (Andres et al., 2009). The first migrations to Chile would have been relatively recent, as the southern half of the country was covered by ice until 12 500 years ago (Heusser, 1974).

It is not surprising that a few seeds of F. chloensis became established in Chile after their delivery by Whimbrels. The seeds of F. chiloensis have a thick seed coat and would have remained viable during the trip. The Whimbrels would have arrived in Chile in the late spring when temperatures and moisture levels would be optimal for germination and growth. Strawberries are clonal in nature and a vigorous colony could have rapidly become established through stolon production. Perhaps most importantly, the environments where strawberries are found in Chile are almost identical to those in central California (Mooney et al., 1970) from where the Whimbrels were most likely to have carried seed. We wonder if the reverse has also occurred: that Whimbrels have carried F. chiloensis seed back to California on their return migrations.

In Hawaii, F. chiloensis appears to have moved into a different habitat than in North America, as it found in the mountains at 4000–6000 m altitude rather than on beaches (Staudt, 1999). However, the mountains of Hawaii may actually be closer to the original habitat of the beach strawberry in North America, as the beaches of Hawaii are much hotter than those of the Aleutian Islands where the Pacific Golden-plover probably picked up seed. The Pacific Golden-plover ranges up to at least 2500 m elevation in Hawaii, where suitable habitats (pastures) occur on mountain slopes (Van Riper et al., 1978; Johnson et al., 2020).

In a previous study, Popp et al. (2011) showed that the crowberry (Empetrum) has a bipolar disjunction much like F. chiloesis between North and South America and they suggested that the disjunct distribution was probably the result of dispersal by a bird from north-western North America. Based on the analysis we present here, that species could well have been the Whimbrel, particularly because crowberries have been specifically mentioned as one of the berries that Whimbrels often consume (Marks et al., 2020).

The approach outlined in this paper could be used to identify the agent of LDD of plants in many other amphitropic disjunct and island distributions. The flyways of most bird species have been characterized through eBird observations, along with the starting point and destination of their migrations. By identifying the habitats and feeding behaviour of the species utilizing each flyway, associations could be made to explain many more disjunct distributions than that of F. chiloensis.