Seeds provide the vital genetic link and primary dispersal agent between successive generations of plants. They are produced and packaged in botanical structures called fruits, and come in an endless variety of shapes and sizes. Tropical drift seeds and fruits are especially remarkable because they can survive months or even years at sea. They are very buoyant, with thick protective shells which are impervious to salt water. In some drift fruits, such as the coconut, the seed embryo and fleshy white "meat" (endosperm) is enclosed within a hard, bony layer (endocarp) surrounded by a thick fibrous husk. Other drift seeds have thick woody seed coats and internal air cavities which produce their buoyancy. During their long voyages they often cross entire oceans, perhaps colonizing the shores of a coral atoll or isolated volcanic island. Sea dispersal is a hit-or-miss method; after a long, perilous journey, other dangers await the vulnerable seedlings. They may perish in the parched sand without adequate moisture or be quickly devoured by ravenous land crabs. Drift seeds carried by the Gulf Stream from the Caribbean to beaches of northern Europe may find themselves in another precarious situation--aliens in a hostile climate much too cold for them to survive.

The tropical islands of Indonesia, Polynesia and the Caribbean probably have the greatest variety of drift seeds. They are especially abundant along wave-swept beaches following the hurricane season, and may be carried inland by tidal waters. The Cayman Islands and Yucatan Peninsula lie directly in the path of the Equatorial Current, bringing thousands of unique drift disseminules from the Amazon and Orinoco deltas and across the Atlantic from river deltas of tropical west Africa. In the Pacific region drift seeds may be deposited well above the inhospitable beach zone by enormous tidal waves known as tsunamis. The tsunami emanates from an earthquake or volcano epicenter and moves at great speed across the Pacific Ocean.

Plant dispersal by ocean currents has fascinated many famous explorers, including Charles Darwin and Thor Heyerdahl. Sea currents have also been studied by the U.S. Coast and Geodetic Survey using stoppered bottles containing a numbered postcard. When a bottle is found on a beach the finder fills out the card and drops it in the mail. It takes about one year for a drift bottle to float from Yucatan to Ireland. A bottle launched near Caracas, Venezuela reached the Florida Keys four months later, traveling at an average speed of 16 statute miles per day. It is estimated that tropical seeds found on European shores probably have been adrift for a year or longer.

During his famous voyage around the world on the H.M.S. Beagle, Charles Darwin championed the idea of drift seeds and fruits colonizing distant islands, particularly isolated volcanic islands which have never been connected to the mainland. Darwin studied the role ocean currents played in the flora of Cocos Keeling Islands in the Indian Ocean, and concluded that most of the endemic vascular flora was derived from drift seeds and fruits. His original article published in 1836 was reprinted in chapter 20 of Journal of Researches, D. Appleton & Company, New York, 1883.

After he returned to England, Darwin conducted flotation experiments with cultivated plants. In the Journal of the Proceedings of the Linnaean Society (Vol. 1, 1857) Darwin stated: "I soon became aware that most seeds, in accordance with the common experience of gardeners, sink in water; at least I have found this to be the case, after a few days, with the 51 kinds of seeds which I have myself tried; so that such seeds could not possibly be transported by sea-currents beyond a very short distance." Darwin also mentioned rafting as a dispersal mechanism for seeds that generally don't float well in sea water. In addition, he stated that seeds contained within pods, capsules and the heads of Asteraceae may be carried by ocean currents and washed ashore on distant beaches. In his Origin of Species, 1859 (Chapter 12 Geographical Distribution: Means of Dispersal), Darwin summarized his experimental data on seed dispersal in salt water, and expressed a higher confidence in dried seeds: "Therefore it would perhaps be safer to assume that the seeds of about 10/100 plants of a flora, after having been dried, could be floated across a space of sea 900 miles in width, and would then germinate."

Of all the 250,000 species of seed plants on earth, only about 250 species (0.1 percent) are commonly collected as drift disseminules on tropical beaches; and only about half of these are known to produce seeds that can float in seawater for more than a month and still be viable. This relatively small number of drift seed species does not include seed plants which are dispersed on vegetation rafts, drift garbage from ships, or true marine seagrasses which live totally submersed in seawater. Although the total number of drift seed species with long viability periods may be relatively small, they nonetheless form a floral flotilla comprising countless thousands of individuals riding the ocean currents of the world.

The Hawaiian archipelago has been isolated from continental land masses during the past 30 million years, and yet the 1,000 species of indigenous Hawaiian angiosperms are believed to stem from natural introduction by long-distance dispersal of 280 ancestral plant colonists (Wagner, Herbst and Sohmer, 1990). According to Sherwin Carlquist (Hawaii: A Natural History, 1980), only about 14 percent of the original flowering plant immigrants to the Hawaiian Islands are clearly adapted to oceanic drift. If dispersal by birds and air currents are ruled out, it appears that seeds were carried thousands of miles to these islands, possibly by rafting or within protective capsules and pods. For example, California tarweeds are not included with tropical drift seeds, and yet an ancestral tarweed traveled at least 3,000 miles to the Hawaiian Islands where it gave rise to a remarkable group of endemics known as the "Silver Sword Alliance." The small seeds from ancestral members of the lobelia family (Campanulaceae) also reached these islands giving rise to an unusual group of Hawaiian lobelioids. One of these is the pachycaul Brighamia insignis that grows on steep sea cliffs on the island of Kauai.

An interesting coral tree endemic to dry, leeward slopes of all the main Hawaiian islands is called wiliwili (E. sandwicensis). This Hawaiian endemic belongs to a group of closely related passerine species in the section Erythraster, including E. tahitensis, endemic to Tahiti; E. variegata, a widespread species in tropical Asia and the South Pacific; and E. velutina, another widespread species in northern South America, the West Indies and the Galapagos Islands. Although the seeds of wiliwili sink in water, it may have originated from an ancestral species with buoyant seeds that reached these islands many thousands of years ago. In fact, a related tropical Asian species (E. variegata), which is commonly planted as a street tree in Hawaii, has seeds that float in water. The seeds of this species sometimes show up along Hawaiian beaches and tidal waters; however, they are usually of local origin. Trees of E. variegata are fairly easy to identify because the pods are longer than most cultivated coral trees, often up to 12 inches or more in length.

According to Sherwin Carlquist (1980), the seeds of wiliwili (Erythrina sandwicensis) have gradually lost their ability to float in seawater. Although the reasons are not identical, Carlquist compares this with the loss of flight in certain insects and birds. "If a plant shifts its ecological preference, it will tend to lose contact with the agent responsible for its dispersal."

The Hawaiian islands are thousands of miles from the nearest continent or island mass. Even though they include some of the most isolated islands on earth, they contain numerous endemic species of plants, many of which evolved from ancestral seeds that reached the islands by drifting. Some striking members of the sunflower family evolved on the Hawaiian Islands from an ancestral California tarweed (Hemizonia) that colonized these isolated Pacific islands millions of years ago. This group of plants, called the Silver Sword Alliance, includes three genera and about 30 endemic species, an excellent example of adaptive radiation. One of the most amazing of all these Hawaiian endemics is the silver sword (Argyroxiphium sandwicense ssp. macrocephalum) that only grows in the cinders of Haleakala Crater on the island of Maui. The rosettes of sword-shaped leaves are covered with silvery hairs that reflect light and heat and provide insulation against the intense solar radiation and extreme aridity of this 10,000 foot (3,000 m) volcanic mountain.

Some of the most ubiquitous drift seeds that readily germinate on tropical beaches are the gray and yellow nickernuts, Caesalpinia bonduc, C. ciliata, C. major. All are climbing, sprawling beach shrubs armed with vicious, recurved thorns and prickly pods bearing gray or yellow seeds. The smooth, marble-like seeds are commonly strung into necklaces and bracelets, often mixed with other colorful seeds. In Guayaquil, Ecuador, drilled nickernuts are sold by street vendors as amulets to ward off evil spirits. On some Caribbean islands nickernuts are used in a strategy board game called "Island Waurie." Two players sit on opposite sides of a hardwood board with six depressions (wells) on each side, each well with four nuts. Through a series of carefully planned moves, players attempt to empty all six of their wells before their opponent can do so. Legend has it that Island Waurie was originally introduced to the Cayman Islands by Blackbeard the Pirate on one of his voyages from South Africa. The game became popular with Cayman Islanders and was enjoyed by Ernest Hemingway whenever he visited the islands. Know in Africa as "mancala," the games were played for thousands of years in Egypt, where boards have been found carved into the stone of the pyramid of Cheops and the temples at Luxor and Karnak.

Probably the best known of all plant drifters is the coconut (Cocos nucifera). In fact, it is hard to imagine a tropical beach without coconut palms. The origin of the coconut and whether it floated to islands of the tropical Pacific has been debated for more than a century. T.H. Everett (Encyclopedia of Horticulture, Vol. 3, 1981) flatly states: "...it has never been found truly wild, every coconut palm is planted by man or derived from such a planting." According to the world authority on palms, E.J.H. Corner (The Natural History of Palms, 1966), "There is no island or shore where its presence is not due directly or indirectly to its having been planted by man." Probably under most situations, coconut seedlings require watering and other attention from humans; however, J.V. Dennis and C.R. Gunn (Economic Botany 25, 1971) cite several localities where coconuts appear to have seeded themselves naturally, including cays in British Honduras (Belize), rocky islets in the Fiji group, the east coast of Trinidad, Cocos-Keeling Atoll in the Indian Ocean, and Krakatau and adjacent islets following the catastrophic eruption of 1883. In addition, I have photographed coconut palms sprouting on Tetiaroa Atoll in French Polynesia. Whether these would survive the ravaging effects of land crabs and intense sunlight and grow into mature palms is hard to say, but there appeared to be palms of different ages along the beaches. I have also observed self-seeded coconut palms growing among mangrove thickets on cays off the coast of Belize.

Hugh C. Harries (Botanical Review Vol. 44, 1978) argues convincingly that coconuts have naturally established themselves on beaches of the tropical Pacific. According to Harries, there are many varieties of coconuts, but they all belong to either of two major types known as niu kafa and niu vai. The niu kafa types have an elongate, angular fruit, up to 6 inches in diameter, with a small egg-shaped nut surrounded by an unusually thick husk. Niu vai types have a larger more spherical fruit, up to 10 inches in diameter, with a large, spherical nut inside a thin husk. The niu kafa type represents the ancestral, naturally-evolved, wild-type coconut, disseminated by floating. The niu vai type was derived by domestic selection for increased endosperm ("meat" and "milk") and is widely dispersed and cultivated by humans. Both types of fruit can float, but the thicker, angular husk adapts the niu kafa type particularly well to remote atoll conditions where it can be found today. The presence of "undesirable" wild-type coconuts growing in mangrove swamps is clear evidence that they were self-sown and not planted by farmers. In two fascinating papers by Harries and his colleagues, W.S. Gruezo and R. Buckley (Biotropica Vol. 16, 1984), wild-type, self-sown coconuts have been documented in the Philippines and as far away as Australia. In addition, throughout the humid tropics intermediate types have arisen by hybridization with the commonly cultivated niu vai coconuts.

The argument for a New World origin of the coconut centers around the presence of coconut palms on Cocos Island off the coast of Costa Rica and parts of the Pacific Coast at the time of Columbus. One of the proponents of the New World origin is Thor Heyerdahl, citing his own experience on the Kon Tiki and legends handed down by early Polynesian voyagers. In his book, Sea Routes to Polynesia (1968), Heyerdahl expresses great confidence in Polynesian sailors who crossed vast stretches of the Pacific Ocean carrying staple foods from the New World, such as sweet potato, plantain and the coconut.

A major factor in Hyerdahl's argument comes from the distribution of the sweet potato (Ipomoea batatas). Archaeological evidence shows that sweet potatoes were cultivated in South America by 2400 B.C. and fossilized sweet potatoes from the Andes have been dated at 8,000 to 10,000 years old. Although the sweet potato is clearly native to South America, it was cultivated in Polynesia as early as 1200 A.D. In fact, the sweet potato had already become the principle food of the Maoris in New Zealand by the time of Captain Cook's historic voyage to that part of the world in 1769. It is interesting to note that the sweet potato is known as "kumar" or "kumal" in the Lima region of coastal Peru, and it is called "kumara" by the Maoris of New Zealand. Heyerdahl (1968) postulated that sweet potatoes were carried across the Pacific by Peruvian Indians before Europeans began to sail the world's oceans. He tested his hypothesis in 1947 by sailing a balsa wood raft, the Kon-Tiki, fashioned after the reed rafts of the Oru Indians living on Lake Titicaca in Bolivia. Although Heyerdahl's hypothesis about the transoceanic exchange of sweet potatoes by skillful pre-Columbian sailors remains an enigma (at least to some skeptics), his New World origin for the coconut has been rejected by most botanists.

Muñoz-Rodriguez et al. 2018. "Reconciling Conflicting Phylogenies in the Origin of Sweet Potato and Dispersal to Polynesia" Current Biology 28: 1–11 (April 23, 2018). https://doi.org/10.1016/j.cub.2018.03.020

According to J.V. Dennis and C.R. Gunn (Economic Botany 25, 1971), the coconut is indigenous to the Indo-Malaysian region. It spread by sea currents to many Pacific Island groups with adequate rainfall and moderate temperatures where the seedlings became established along beaches. Coconuts prefer the well-drained coral sand beaches of tropical islands and atolls, and are poorly established along shores of continents. In a recent article by F. Rosengarten, Jr. (Principes 30, 1986), naturally dispersed coconuts can withstand occasional brief salt water flooding. Developing coconut palms obtain fresh water and mineral nutrients from a lens (soil layer) of fresh water (derived from island rainfall) which literally floats above the denser salt water beneath the beach sand.

According to Dennis and Gunn, 3,000 miles seems to be the average maximum distance that a coconut will remain afloat and still remain viable. These limitations greatly diminish the chance of a coconut reaching the New World, let alone sprouting on a continental shore. Most authorities agree that the coconut was introduced to the New World by Portuguese and Spanish traders. Although the coconut is the best known drift fruit it is also one of the world's most valuable trees, providing thatching, food, drink, coir, copra, oil, and hard endocarps which are fashioned into all sorts of decorative and useful articles.

Coconut on a beach in Belize.

The origin of the generic name for coconut "Cocos" may be traced to the three germination pores on the endocarp layer surrounding the seed. According to R. Sokolov (Natural History Oct. 1989), Portuguese and Spanish traders introduced the coconut into West Africa after 1500. They called it "coco" from the Portuguese or Spanish slang word for monkey face, supposedly because of the eye pattern on the endocarp and the brown, fibrous hair (husk). Coconuts were later introduced into the Americas by these early traders. The center of origin for ocean-dispersed coconuts appears to be the Indo-Malaysian region.

Perhaps the best physical evidence for an Indo-Pacific origin of the coconut comes from New Zealand and India, where fossil specimens of a species of Cocos have been dated to the late Tertiary, at least two million years ago. According to Hugh Harries (Principes Vol. 36, 1992), fossil coconuts of an extinct Miocene palm (Cocos zeylandica) have been found in North Island, New Zealand. Fossil evidence from such far-flung areas suggests that the coconut had ample time to disperse naturally in the Indo-Pacific, before humans came along to speed up the process.

But of all the arguments for a transoceanic dispersal of coconuts, perhaps the most interesting comes from the widespread distribution of the coconut crab (Birgus latro). According to Harries (Principes Vol. 27, 1983), this large land crab could not have achieved its present widespread inter-island distribution with only a 30 day aquatic larval stage unless they rafted to distant islands. Harries postulates that the tiny post-larval (glaucothoe) stage was spent in the moist husk of floating coconuts. Ancestors of today's coconut crabs may have literally migrated on floating coconuts to remote islands and atolls of the South Pacific.

The Galapagos Islands, always a provocative place for students of natural history, offer a wide variety of examples of drift dissemination, including the independent transport of seeds and fruits floating in the ocean and the transport of seeds and animals on vegetation rafts from the mainland. Several species of mangroves grow on these islands, including the ubiquitous red mangrove (Rhizophora mangle) with its conspicuous prop or stilt roots descending from its limbs. The cigar-shaped disseminule, known as a "sea pencil," is unusual because it is a germinated seedling and not a seed or fruit. The seedling, with its long, pendant taproot (mostly hypocotyl), drops from the parent plant, where it may be automatically planted in the mud of shallow water or washed away to a distant shore. Beach vines, such as beach morning glory (Ipomoea pes-caprae), gray nickernut (Caesalpinia bonduc) and beach bean (Canavalia maritima), occur on several of the islands. Their durable seeds ride the currents of the world's oceans, resulting in enormous pantropic distributions.

Ian Thornton, a zoologist at La Trobe University in Melbourne, Australia, has calculated that, based on the speed of the currents and winds of the area, it would take about two weeks for a floating log or vegetation raft to reach the Galapagos Islands from mainland Ecuador. Considering that the archipelago is over 3 million years old, the drift seed scenario offers a plausible explanation for how some plants made the 600-mile (965-km) journey.

Successful dissemination by drifting would not have to occur often for it to have a profound effect on the ecology of the islands. Taking into account all methods of dispersal, including transport by wind, birds, and drifting on the ocean surface, Duncan Porter, a botanist at Virginia Polytechnic Institute and State University, estimates that 378 original introductions could account for the 522 indigenous plant species known to grow on the islands. Of these natural introductions, 59% were a result of transport by birds, 32% by wind transport, and 9% by drifting on the ocean surface. If only one species successfully drifted to the islands and germinated there every 50,000 years, this would still account for the introduction of roughly 10 percent of the islands's flora. According to D.M. Porter (Nature Vol. 264, 1976), for 378 introductions to account for the present flora, it would require only one species to arrive and become established every eight to ten thousand years.

Ancestors of present-day Galapagos reptiles, such as the remarkable marine iguanas, may have also drifted to the islands. In the swift current of Ecuador's Guayas River, trees, branches, and large mats of vegetation broken loose from riverbeds are commonly carried out to sea. Reptiles now on the Galapagos may have ridden such rafts from the mainland many thousands of years ago. For example, in 1827 a large boa constrictor reportedly arrived on the island of St. Vincent in the West Indies, wrapped around the trunk of a floating tree.

Mangroves include several dozen species of trees and shrubs that grow along tidal swamplands throughout tropical regions of the world. Since their roots are emersed in water-logged silt and mud, often deficient in oxygen, the buttressed trunks and aerial roots of mangroves are typically dotted with numerous lenticels--small pores which provide gas exchange between the roots and the atmosphere. Lenticels are especially conspicuous on the prop roots of the ubiquitous red mangrove (Rhizophora mangle) and on the deeply-fluted trunk of the tea mangrove (Pelliciera rhizophorae), an interesting swamp tree of the tea family (Theaceae) along the Pacific coast of Costa Rica. The lateral roots of some mangroves produce upright, slender outgrowths called pneumatophores which extend above the mud and water like snorkel tubes. The porous pneumatophores are covered with lenticels and provide additional aeration for the root systems. The black mangrove (Avicennia germinans), of Central America and the Caribbean region, produces literally hundreds of pencil-like pneumatophores around its base.

Mangroves survive in seawater with a salinity that would be lethal to most trees and shrubs. Like celery or carrot sticks placed in saltwater, the roots of most plants rapidly lose water if they are suddenly emersed in seawater. Halophytes (salt-loving plants), such as mangroves, generally have a lower concentration of water molecules (lower water potential) in their root cells so they can take in water. They maintain lower water potentials in their roots by having higher internal salt concentrations than seawater and by losing water at the leaf surface. Since high internal salt concentrations can be lethal to plant cells, some species such as the black mangrove and white mangrove (Laguncularia racemosa), can excrete excess salt through special glands in their leaf blades and petioles. Red mangroves have root cell membranes which prevent the absorption of excess salt.

The seeds of mangroves are especially remarkable because they commonly germinate within their fruit while still attached to the parent plant, a condition known as "viviparous seeds." Having their embryonic root (hypocotyl) already elongated gives them a better chance of establishing themselves in soft mud during low tide. Called "sea pencils," the cigar-shaped seedlings (disseminules) of red mangrove may have an elongate taproot up to 10 inches long when they drop from the parent tree. The unusual, top-shaped fruit of tea mangrove contains one of the largest seeds in the world (excluding palms). It floats with the elongate, embryonic root pointing downward, and readily becomes implanted in soft mud. Germinated seedlings and sprouted fruits of the black mangrove and white mangrove are also dispersed by ocean currents.