Plants, Man and Life, pp. 136-151 (1952)
Edgar Anderson

Dump Heaps and the Origin of Agriculture

WHEN I FIRST WENT to live in San Pedro Tlaquepaque, a small pottery-making town in western Mexico, I was under the mistake impression that my Mexican neighbors had nothing but dump heaps and a few trees in the yards behind their homes. As I lived there longer and came to know more about the life of the village, I realized that many of these dump heaps were carefully managed gardens and orchards. I saw enough so that several years later in highland Guatemala I played hooky from the cornfields I was supposed to be studying, and spent an afternoon in an Indian village with a Spanish-speaking youth who belonged to one of the two or three Ladino families in that town. Again I came back home with the little more I had learned and thought it over for another two years. By the time of my next visit to Guatemala I was certain that the simple facts concerning these gardens were something worth careful study; and in the few days at my disposal I again visited the village of Santa Lucia on the little height of land between Antigua and Guatemala City and again prevailed upon Senor San Salazar to let one of his sons serve as a guide and interpreter. This time I spent a good part of one day visiting and photographing Indian gardens and finally mapped and measured one of them in detail.

The garden I charted was a small affair about the size of a small city lot in the United States. It was covered with a riotous growth so luxuriant and so apparently planless that any ordinary American or European visitor, accustomed to the puritanical primness of north European gardens, would have supposed (if he even chanced to realize that it was indeed a garden) that it must be a deserted one. Yet when I went through it carefully I could find no plants which were not useful to the owner in one way or another. There were no noxious weeds, the return per man-hour of effort was apparently high, and I came away feeling that as an experienced vegetable gardener (I am one of those strange people who would rather hoe vegetables than play golf or go to the movies) I had gotten more new ideas about growing vegetables than from visiting any other garden anywhere.

The garden, like most of those in Santa Lucia, was rectangular and much longer than broad. On three sides it was surrounded by a low fence made from the local cornstalks, which are so big and so durable that I suspect some of the Guatemalan corns must have been deliberately selected for such purposes. Along the other side was a pruned hedge of chichicaste, a rough-leaved shrub which the Mayas used for various purposes and which has given its name to a well-known Guatemalan town, Chichicastenango, literally, "place of the chichicaste." Growing along the fence and the hedge were several varieties of squash and pumpkins as well as some squash relatives not so well known in the Temperate Zone. There was the perennial chayote whose pale-green, large-seeded fruits are now becoming a common vegetable in our own markets, the related "caiba," much more piquant in flavor, and the black-seeded Cucurbita ficifolia, which we know from archaeological evidence to have been one of the most anciently cultivated cucurbits in the New World. The squashes grew more rampantly than they usually do with us and the chayote draped itself over everything, garden walls, trees, mature cornstalks, making the whole garden into a picturesque tangled bower.

FIGURES 13 and 14. Diagrammatic map of an orchard-garden in the Indian village of Santa Lucia, Guatemala. The glyphs listed above not only identify the plants as shown in the plan on the opposite page, they indicate by their shapes in what general category the plants belong. Circular glyphs indicate fruit trees (such as plum and peach) of European origin; rounded irregular glyphs indicate fruit trees (such as the manzanilla) which are of American origin. Similarly, dotted lines are for climbing vegetables, small circles for subshrubs, large stars for succulents, and an irregular wedge-shaped figure for plants in the banana family. The long irregular mass at the right-hand side of Figure 14 represents a hedge of "chichicaste," a shrub used by the Mayas.

Though at first sight there seemed little order, as soon as we started mapping the garden, we realized that it was planted in fairly definite crosswise rows. There were fruit trees, native and European in great variety: annonas, cherimoyas, avocados, peaches, quinces, plums, a fig, and a few coffeebushes. There were giant cacti grown for their fruit. There was a large plant of rosemary, a plant of rue, some poinsettias, and a fine semiclimbing tea rose. There was a whole row of the native domesticated hawthorn, whose fruits like yellow, doll-size apples, make a delicious conserve. There were two varieties of corn, one well past bearing and now serving as a trellis for climbing string beans which were just coming into season, the other, a much taller sort, which was tasseling out. There were specimens of a little banana with smooth wide leaves which are the local substitute for wrapping paper, and are also used instead of cornhusks in cooking the native variant of hot tamales. Over it all clambered the luxuriant vines of the various cucurbits. Chayote, when finally mature, has a large nutritious root weighing several pounds. At one point there was a depression the size of a small bathtub where a chayote root had recently been excavated; this served as a dump heap and compost for the waste from the house. At one end of the garden was a small beehive made from boxes and tin cans. In terms of our American and European equivalents the garden was a vegetable garden, an orchard, a medicinal garden, a dump heap, a compost heap, and a beeyard. There was no problem of erosion though it was at the top of a steep slope; the soil surface was practically all covered and apparently would be during most of the year. Humidity would be kept up during the dry season and plants of the same sort were so isolated from one another by intervening vegetation that pests and diseases could not readily spread from plant to plant. The fertility was being conserved; in addition to the waste from the house, mature plants were being buried in between the rows when their usefulness was over.

It is frequently said by Europeans and European Americans that time means nothing to an Indian. This garden seemed to me to be a good example of how the Indian, when we look more than superficially into his activities, is budgeting his time more efficiently than we do. The garden was in continuous production but was taking only a little effort at any one time: a few weeds pulled when one came down to pick the squashes, corn and bean plants dug in between the rows when the last of the climbing beans were picked, and a new crop of something else planted above them a few weeks later.

I was so impressed by the apparent efficiency of the garden that I have since tried out several of its basic principles on my own vegetable plots with considerable success. Instead of putting my sweet potatoes all neatly in one little bed down at the far end of the garden I plant them one row at a time in different places. They now grow out vigorously across the garden by late summer; they keep the ground moist during the dry days of August; and they help keep out weeds. I also plant a few cornfield beans in among the corn plants after they are pretty well up and have a good extra crop of string beans after the sweet-corn season is over. From these experiences I suspect that if one were to make a careful time study of such an Indian garden, one would find it more productive than ours in terms of pounds of vegetables and fruit per man-hour per square foot of ground. Far from saying that time means nothing to an Indian, I would suggest that it means so much more to him that he does not wish to waste it in profitless effort as we do.

This is the most primitive type of garden which I have so far been able to study personally; from the literature I gather that it is characteristic of wide areas in the New World, Asia, and Africa. Under more strictly tropical conditions (because of its altitude, Santa Lucia for all practical purposes is almost in the Temperate Zone) such garden-orchards blend even more closely into the native vegetation. In Malaysia, on more than one occasion deliberately planted native orchards have been mistaken for part of the natural woodland by European and American plant collectors. Varieties of citrus fruits said to be wild were actually collected from native gardens by Occidental botanists so imperceptive that they could not distinguish between man-made orchards and more-or-less natural vegetation. Perhaps a goodly number of the primitive varieties of various cultivated plants which have been collected at one time or another in the tropics by northern botanists and which are said to be the wild ancestors of such-and-such a crop are in the same category. Here, as in many other basic problems of tropical agriculture, we are up against a serious difficulty. Agriculture, as an ancient art, began in the tropics and has various special complexities there by reason of its long persistence in those areas. Agriculture as a modern science developed in the Temperate Zone. Most of our scientific understanding of agriculture comes from our experiences during the last few centuries with the relatively simple agricultural problems of northern Europe and North America. When the average scientific agriculturist goes to the tropics he has much more to unlearn than to teach, but he frequently seems to be unaware of that fact.

These hit-or-miss tropical gardens are of particular significance because they fit in perfectly with a theory of Carl Sauer's that agriculture may have originated among a sedentary fisherfolk. The first definite beginnings of agriculture took place very early, about the time man domesticated the dog, perhaps in those early Neolithic times which many authorities designate as the Mesolithic. As a geographer, Sauer thinks of the Mesolithic as a time in which rapidly melting ice sheets were raising the sea level all over the world, flooding continental shelves which had been bared when a good portion of the world's precipitation was locked up in polar icecaps. It was a time when coastal valleys were becoming estuaries, when there was a rapid building of new deltas, an increase in length and complexities of shore lines and the formation of river swamps and oxbow lakes. In such an environment people living primarily by fishing would have been able to supply themselves with food without shifting from place to place. The archaeological records show that shell middens accumulated of such a size as to indicate centuries of occupation. Excavations indicate that this was an era in which there was an elaboration of fishing gear, boats, harpoons, and fishhooks. The fishhooks are important for our story; they imply fishing lines, and fishing lines indicate the use of plants for cordage.

From the world-wide distribution of fish poisons Sauer infers that this lazy-man's way of stunning your prey with a plant poison is a very ancient trait. He thinks it likely that the first fish poisons may have followed naturally from the making of fishlines. What could be more natural than that in the process of bruising vines and bark with blunt stone instruments and soaking the fibers to rot away the soft parts, man should accidentally have discovered that some of these plants can stupefy fish and make them easier to catch? As a further argument he points out that some primitive fiber plants are known to be still in use as fish poisons. To him it seems likely that such a folk, blessed with abundant food, becoming increasingly skilled in navigation and hence in transport, already using plants for fiber and for poison, might gradually shift from plant gathering, to unintentional domestication, to the purposeful growing of plants. He points out that in one of the most ancient centers for man, southeastern Asia, there was extensive drowning of ancient coast lines and the production of long and complicated new ones. In this area he finds evidence for an early center of gathering and preparing plant products with blunt stone instruments, the grubbing up of tuberous plants from river swamps, the manufacture of nets and cordage, the elaboration of fish poisons, the manufacture of cloth and of bark, and the building of bark houses. He brings out the interesting fact that the making of sago by shredding and soaking and pounding the stems of certain palms and cycads is mostly limited to this area. "The practice of shredding, pounding, washing, and decanting runs through plant uses throughout southeast Asia and seems to tie fiber-making and toughening, poison preparation for fishing, hunting, and medicine, and food preparation, including the coagulation of sago, into one culture complex."

To me this theory is worth careful consideration because such people would have also created refuse heaps and I am even more intrigued than Sauer by the notion that such refuse heaps may have played a key role in the origin of cultivated plants. These ancient dump heaps are of extraordinary interest in connection with the origin of cultivated plants, because they are open habitats.

What do we mean by an open habitat? Well, let me begin my explanation by telling you the story of Mrs Swune and the hollyhocks. Mrs. Swune (that was not her real name, but she was a very real person) loved hollyhocks, and they did well in her garden with very little care. Seedlings came up all through the perennial border and even among the shrubbery, and all during the early summer her place was ablaze with hollyhocks. Naturally, she became very hollyhock-conscious, and she wrote some poems about them and had her friends in for a sedate tea among the hollyhocks, and brought out the poems and read them to her guests. I was invited to one of the teas, and it really wasn't as bad as it sounds: the poems were short and there were plenty of excellent teacakes, and the other guests were swell people and we kind of banded together. At the conclusion of the tea she presented each of us with a large packet of hollyhock seed and wanted to know if we would please scatter them through the Ozarks whenever we drove down that way. Spring bloom in the Ozarks she found satisfactory, but to her the summer roadsides were dull, and she thought if we would just toss the seeds out the windows as we motored along they would just naturally come up all through the Ozarks, and Highway 61 would become a Hollyhock Memory Trail.

Well, I knew the hollyhocks wouldn't come up (even though they do grow along roadsides in the Eastern states), but it is fruitless to argue with people like Mrs. Swune. She was so persistent that eventually I even tossed my seeds out the car window according to her directions, and so, I suspect, did some of the other guests, but the Ozark highways are just as bare of volunteer hollyhocks as they ever were. What Mrs. Swune did not understand is that most plants are very choosy about where they will and will not grow, and that some places, like gardens and dump heaps, are relatively open habitats, receptive to a good many kinds of plants, while other places, such as meadows and mountaintops are relatively closed habitats in which aliens will have trouble getting a footing.

A useful concept in discussing such problems is that every organism has a kind of niche which it has been evolved to fit into (an ecological niche is the precise scientific way of phrasing it); and that many of the things which happen in the domestication of plants and the origin of weeds can best be understood in terms of this concept. It really isn't so complicated when you start thinking it over,. Even Mrs. Swune would have realized that you can't plant orange seeds in Greenland and expect to raise orange orchards. Any person of normal intelligence knows that all plants and animals have different likes and dislikes; some are suited to one place and some to another. You don't try to grow water lilies in a desert or plant cacti in a redwood grove. What such people as Mrs. Swune do not realize is not only that oranges will not do in Greenland, but that the average plant is most awfully finicky about just where and when it will grow, under exactly what conditions it will germinate, under exactly what circumstances it will persist to maturity after it has once come up as a seedling. So choosy are most plants about these matters, so individual are most of their likes and dislikes, that we understand it scientifically in only the crudest sort of way. Precise limits of temperature and moisture have been worked out for some plants but the vastly more intricate business of which plants they will and will not tolerate as neighbors and under what conditions, has never been looked into except in a preliminary way for a few species.

If you don't mind, let us again consider the spiderworts. You will remember that Tradescantia ohiensis and T. pilosa seldom produced hybrids in nature, though in the experimental plot they came up like weeds. We interpreted this as due to the garden being an open habitat. The complex flora of the woodland was made up of species which had evolved together. Natural selection had made them fit into each other's company like the pieces of an intricate jigsaw puzzle. If the plants in this association went to seed, the seedling was likely to find a niche to which it was suited and its chances of survival were good. If a hybrid seed, or an alien seed from some other flora were planted there, even though able to germinate, it did not fit into the strict interlocked economy of that vegetation, or the chances of its doing so were miniscule. The habitat was closed. Start making a refuse heap in the neighborhood of these two species, however, and one might find even more hybrids in it than in a garden. Here is a strange new kind of habitat. Many of the plants in the native flora do not fit into it, some aliens will, and some hybrids. Plants which can grow in such places will have less interference from other plants. Kitchen middens would be likely places in which fruit pits, seed heads, and the like, brought to the village from some distance, might germinate and survive. However, before we come to grips with the dump-heap theory, let us spend a little longer with the concept of the ecological niche.

I learned a great deal about such matters from seeing a fire line plowed across upland meadows in our Arboretum. In the lowlands of this area, meadows and fields are dotted every spring with the green-yellow blossoms of winter cress, a kind of wild mustard. With us, the winter cress is common in the flood plain, but until we plowed the fire lines it had never appeared on our hillsides. Yet for several years after the plowing was done, there were so many winter-cress plants along the fire line that they made a streak of yellow in the landscape. Since then the grass has grown back in and the cress is again restricted to the lowlands. From this I learned that cress grew down in the lowlands not just because it wanted more water, but because the water in the lowlands held in check some of the plants, such as bluegrass, which competed with the cress. Standing water made enough bare spots in the flood-plain meadows So that cress got started there; on the uplands, it required plowing to give cress a start and repeated plowing if it was to persist.

Nor is winter cress the only lowland plant which one finds in the uplands, following a plowing. If we watch deserted fields in the northern edge of the Ozarks, most of the trees which come up spontaneously are flood-plain trees, rather than upland trees. Cut down a woodland of oak and maple, or of oak and hickory. Clear the land and plow, grow wheat, or soybeans, or potatoes for a while, and then leave the field fallow. Even though there are oaks and maples and hickories all around the edge of the plot, they are not the first trees to come up there, Far from it. It is elms and sycamores and honey locusts, all of them trees which belong in the flood plain.

Why do flood-plain trees come up in deserted upland fields? Drive around Missouri in the wintertime and look at the sycamores and you will find your answer. The sycamores are easy to spot in the landscape. Their bark peels off in big flakes and in the wintertime they are nearly as shining white as a paper birch. In Missouri they are one of our commonest trees and one sees them lining the banks of rivers, forming little parks of all one kind of tree on our big gravel bars, or growing in the actual beds of small creeks. Ordinarily they are seen only along these watercourses and you might easily suppose that only there do they find things to their liking. On a day's drive, with sharp eyes, one can find a fair number of exceptions. These are the key to the story. Quite frequently one sees a few isolated sycamores, far from any watercourse, in a feed lot or near a barn. Some few of these may intentionally have been planted there, but surely not all of them, In the abandoned mine dumps and borrow pits of die lead belt and the tiff district one sees much more evidence of this sort. These picturesque small-scale mining operations have been going on for over a century. On brushy hillsides, irregularly dug-over fifty to a hundred years ago, sycamores are nearly as common as they are along the rivers and creeks. The first need of a sycamore seedling is evidently not a high water table; it is open soil. On the uplands under natural conditions, areas of exposed soil are very rare; it is only when a big tree blows over and brings up a mound of soil with its roots that one finds them at all. In the lowlands the river is continually plowing its banks and dumping mud, sand, and gravel in new places. Upland plants have not been evolved to fit into habitats where the topsoil is being churned and rechurned; many flood-plain plants like nothing better. So it was that flood-plain trees came up in the abandoned upland fields. Not until they had grown to some size and made a leaf-cover over the raw soil would the sugar maples and white oak which belonged there spread back in.

Now patches of open soil, like dump heaps, are a part of our story, for these are two of the commonest scars man leaves on the landscape. When he began to spread out of his original corner and into lands previously without human inhabitants, the open habitats which he tended to create, the strange new niches where something different might get a foothold were dump heaps and patches of open, more-or-less eroded soil. In both of these habitats his only natural partner was the big rivers. They too make dump heaps of a sort; they too plow up the mantle of vegetation and leave raw scars in it. Rivers are weed breeders; so is man, and many of the plants which follow us about have the look of belonging originally on gravel bars or mudbanks.

If we now reconsider the kitchen middens of our sedentary fisherfolk, it seems that they would be a natural place where some of the aggressive plants from the riverbanks might find a home, where seeds and fruits brought back from up the hill or down the river might sometimes sprout and to which even more rarely would be brought seeds from across the lake or from another island. Species which had never intermingled might do so there, and the open habitat of the dump heap would be a more likely niche in which strange new mongrels could survive than any which had been there before man came along. Century after century these dump heaps should have bred a strange new weed flora and when man first took to growing plants, these dump-heap mongrels would be among the most likely candidates.

If we look over our cultivated plants with the dump-heap theory in mind, we find that a goodly number of our oldest crops look as if they might well have come from some such place. Hemp is difficult to keep off of modern dump heaps once it is established in a neighborhood; squashes, pumpkins, beans all have the look of such an origin. Among the world's oldest but least-known crops are the grain amaranths. These big coarse pigweeds, or "redroot" as they are sometimes called, have little seeds about the size of the head of a common pin. They are grown as a grain crop by primitive highland peoples both in the Old World and the New. The red-leaved variants are used for food color, or planted in other crops to scare away devils; the young leaves are used for greens; the seeds are made into a gruel or are popped and used in foods and various religious rites. Amaranths are a dump-heap plant par excellence, and are common in barnyards, middens, and refuse dumps throughout the world. The ancient Aztecs in a sort of pagan communion ceremony mixed the popped seeds with human blood, molding the mess into the shape of a god which was sacrificed on the altars and then passed around to be eaten. Of these grain and weed amaranths we can write with some assurance, for Jonathan Sauer has begun their careful study; as the result of his work we have a fairly precise notion of what species of grain amaranth there are, where they are grown, and how they are used. For an understanding of many other dump-heap plants we shall have to wait until other scholars take an equally keen interest in the other humble plants which have traveled with us so long and so far. The history of weeds is the history of man, but we do not yet have the facts that will let us sit down and write very much of it.

FIGURE 15. Amaranthus caudatus. After W. E. Safford