Darwin's Backyard

by James T. Costa

  Why should this be the case so often? Darwin was convinced that it was the result of selection, of course, but it is important to keep in mind that he doesn’t just mean that selection has led to a more diverse roll call of species in a given area. His “ecological division of labor” idea goes beyond that. To him, so many species packed in a small area is an indication of greater overall productivity. This is one expression of “progress,” that slippery concept that dominated (and often still dominates) social and evolutionary thinking from the Enlightenment onward. The idea of progress is linked with improved productivity, efficiency, increased resources. It is inherent in the division of labor promoted by Adam Smith, the physiological division of labor promoted by Milne-Edwards, and it, too, is inherent in the ecological division of labor promoted by Darwin. He was sure that “a greater absolute amount of life can be supported in any country or on the globe; when life is developed under many & widely different forms, than when under a few & allied forms.”17

  Darwin’s plant diversity surveys and Lawn Plot Experiment helped him demonstrate that more times than not small and uniform areas really do support surprisingly diverse forms. The logical next step was to demonstrate that such areas were disproportionately more productive of life, say, or biomass than comparably sized areas dominated by just a few forms.

  In the course of his researches, Darwin came across an account of a most remarkable experiment that had been conducted some decades before at Woburn Abbey, Bedfordshire, just north of London. The Scotsman George Sinclair was gardener to the 6th Duke of Bedford at Woburn Abbey from 1807 until 1825, during which time he oversaw various experiments in agricultural improvement, the most significant of which became known as the Grass Garden Experiment. It was aimed at determining how grass monocultures compared with mixed grass plantings in terms of productivity (reckoned by yield, herbage produced). Some 242 2 × 2-square foot plots were planted with different grasses singly or in various combinations, and with different experimental soil treatments overseen by the renowned Sir Humphrey Davy. It may not have met modern standards of replication and other hallmarks of statistical rigor, but for its time this experiment was cutting-edge scientific agriculture—nothing like it had ever been attempted before.

  Sinclair had published the results of the experiment several times, but Darwin saw Sincair’s paper of 1826 that was published in Loudon’s Gardeners Magazine describing the plan of the garden experiment and its results. In it, Sinclair wrote of how the Woburn Abbey experiments “clearly prove that any certain soil will maintain a greater, and produce more nutritious produce, if cropped with a number of different species of grasses, than it will maintain and produce if cropped with only one or two species.” Sinclair went on to note that “This is a curious and important fact” generally neglected in practice. Then he gave the data: “If an acre of good land is sown with three pecks of rye-grass, and one peck of the clovers . . . 470 plants only will be maintained on the square foot of such land.” However, “if, instead of two species of grasses, from eight to twenty different sorts are sown on the same soil . . . a thousand plants will be maintained on the same space, and the weight of produce in herbage and in hay increased in proportion.” Sinclair concluded that this is why “every variety of soil and situation, from the alpine rock to water itself, is provided with its appropriate grasses, destined for the support of animal life, and for covering the soil with the colour most pleasing to man.”18

  Plan of George Sinclair’s Hortus Gramineus, or grass garden, at Woburn Abbey: “The spaces allotted to the proper grasses are in number two hundred and forty-two, of two square feet each, inclosed by cast-iron frames. Paths of gravel two feet nine inches wide separate the spaces on every side; these are surrounded by a path three feet wide, with a border for the herbage plants . . . a hedge of hornbeam separates the compartment from the rest of the grounds, and an outside border of roses completes the grass-garden.” Each of the 2 × 2-foot plots was either transplanted with natural turf or planted with a single grass species or various combinations of multiple grass species. From Sinclair (1826), pp. 114–115.

  Sinclair’s experiment found that when between 8 and 20 species were simultaneously sown, the number of individual stems supported more than doubled to 1000 per square foot. Monoculture, or near-monoculture, gives a poorer yield than, well, “polyculture” (to perhaps coin a term)—precisely what is to be expected from the ecological division of labor, and precisely, Darwin was sure, what lies behind the surprising diversity found in even the smallest and most uniform of small, uniform lawn plots. Selection had made it so.

  Botanical Arithmetic

  The Lawn Plot experiment was a project that Darwin kept up for 2 years. It dovetailed with a number of other seemingly unrelated interests in the 1850s, the most significant of which could not seem more different in approach: a foray into the ratios and distribution of species and varieties gleaned from stacks of botanical manuals. We saw in the previous chapter how Darwin’s barnacle work gave him a new appreciation for the abundance of natural variation. Darwin always saw this variation as the raw material for selection to act upon, but beginning around the summer of 1855 he began to consider that variation in a statistical way, trying to quantify it to somehow find selection’s signature. He had an idea sparked by an analogy in an article by Swedish botanist Elias Magnus Fries. Darwin had come across a passage where Fries remarked that in genera containing lots of species, “the individual species stand much closer together” than they do in species-poor genera. “Hence,” recommended Fries, “it is well in the former case to collect them around certain types or principal species, about which, as round a centre, the others arrange themselves as satellites.”19 Darwin was struck by this. He later told Hooker that he was led to undertake his analysis of varieties in large versus small genera by Fries’ remark “that the species in large genera were more closely related to each other than in small genera; and if this were so, seeing that varieties and species are so hardly distinguishable, I concluded that I should find more varieties in the large genera than in the small.”20 What Darwin came to realize was that the inverse was also true: small genera with few species, limited in range, comparatively depauperate in variation, are on their way out. Speciation and extinction go hand in hand, with groups diversifying even as others wither. In a telling note to himself, he declared: “We can look far into future by looking to the larger groups.”21 In other words, the large groups of today are likely to be larger still in the future. The reason, he thought, was because the ancestors of species in large, widespread genera were favored by selection in the past, multiplying into the large number of the present. In all likelihood current species in large genera continue to be favored and selection is actively promoting their growth, competitive success, and diversification. Hence you should see more varieties associated with species in such genera. Since varieties are “little species,” with time they become bona fide species that produce more varieties in turn, and so on.

  One way to test this “success breeds success” way of looking at things is to quantify the variation found in different sized groups—an idea that led him to tabulate and compare the amount of variation in genera large and small. Are distinct varieties—which can be seen as products of selection acting upon past variation—more commonly found in genera with large numbers of species? Botanical manuals, he realized, are handy compendia of such information and are available for just about any region or country thanks to the labors of ardent botanists everywhere. Hooker recommended he look at three for British plants: Babington’s Manual of British Botany, Henslow’s Catalogue of British Plants, Watson’s and Syme’s London Catalogue of British Plants. Darwin’s approach was his own version of a by-then well-established line of investigation called “botanical arithmetic.” Coined and pioneered by the polymath Alexander von Humboldt and pursued by other naturalists like Alphonse de Candolle in France and Robert Brown in England, botanical arithmetic was an indispensable tool of early to mid-nine
teenth century geographical botany. With a simplicity that belies its significance at the time, botanical arithmetic involved enumeration of species, genera, etc. of different plant groups according to variables such as elevation, latitude, and continental area in the pursuit of divining underlying relationships that may give insight into the workings of nature. How does the proportion of mosses, ferns, or grasses compare to that of other groups on this or that continent? Does it vary according to elevation? Does the average number of species per genus vary between regions or between continental areas and oceanic islands? The ledgers and tables of botanical diversity worldwide patiently assembled by explorer-naturalists since the seventeenth century held the key.

  From Henslow’s tutelage, Darwin was long familiar with botanical arithmetic, and his transmutation notebooks show that he used the technique in different ways. In the B notebook, for example, in an entry probably made in early 1838, we find Darwin tabulating genus-to-species ratios for the plants of north Africa, St. Helena, and the Canary Islands. He noticed that the “poorness” of the flora of the islands was “in exact proportion to distance” from the African mainland—something he soon came to understand in the context of chance voyagers landing at oceanic islands from the nearest mainland. Now, in the 1850s, Darwin realized that just as facts in the distribution of plant species and genera could be used to shed light on underlying principles governing biogeography, so too could facts in the number, ratio, and distribution of varieties in relation to species and genera be used to test his ideas about the action of selection. If the current geographical distribution of species represents a clue to—or even a diagnostic marker of—geological, climatological, or evolutionary history, Darwin was sure that the frequency and distribution of varieties was diagnostic too. He worked his way through botanical manual after manual doing his own brand of botanical arithmetic. Besides the three on British flora, he looked at about nine more on the flora of the United States, Russia, Germany, France, and Holland.

  In the Weeds

  Even as Darwin was poring over his botanical manuals, he was hard at work on Natural Selection—and experimentising. By mid-October 1856 he finished a chapter on domestication, and by mid-December, just a week after the birth of his 10th child, Charles Waring, he completed another on the crucial importance of intercrossing—two great strands of his thinking that we explored in the last chapter. As 1856 became 1857 he jumped into the all-important chapters on variation (completed near the end of January), the struggle for existence (finished up in early March), and natural selection (done by the end of March). They got him thinking about how struggle and natural selection gave rise to the close coexistence of divergent plant groups he had found in those small lawn plots. That means struggle must be evident even—or maybe especially—in that small spatial scale. He decided to find out with another kind of experimental plot.

  In January he cleared an old planting bed about 2 × 3 feet square that had been used for strawberries for a year or two. He called it his “weed garden.” Given the myriad pressures that organisms face, by dynamically changing over time with age, size, reproductive condition, complex interactions with competitors, predators, and more, he needed to simplify to get at the basic question of mortality at one stage of life. He chose a single kind of organism—plants—at one life stage—seedling—in one small plot where conditions were uniform and no crowding was experienced. The plan was to check the bed daily and mark the position of each weed seedling as it naturally sprouted by sticking a piece of wire in the ground next to it. Then over time he could census the wires and see which weeds were still alive and which had died. He recorded his progress: “Early in March seeds began to spring up: marked each daily. March 31st About 55 marked, of which about 25 Killed already. April 10th. Pulled up 59 wires marking where seedlings before development of two leaves had been devoured, I suppose by slugs . . . April 20th Pulled up 28 wires, dead (I think dry weather is beginning to tell against some) . . . May 8th Pulled up 95 wires (I suspect that some seedlings are Killed by drought) . . . June 1st Pulled up 70 wires.”22 By the beginning of August 1857 he found that just 62 of the 357 plants that came up survived. Slugs and insects had taken their toll. These results were eventually given in Origin as an illustration of the struggle going on all around us that we fail to notice. Tracking seedling mortality over 8 months, he found that of 357 seedlings that appeared, 83 percent of them were destroyed.23

  “I can often gaze for a long [time] at a square yard of turf,” he mused, “and reflect with astonishment at the play of forces which determine the presence & relative number of the 30 or 40 plants which may be counted in it.”24 He saw epic struggle in a tiny plot of land and wanted to find other ways to get this point across. He soon realized that the abundant heathlands and commons he encountered on his countryside rambles told the story too, in a different way. All the better that such spaces were familiar to everyone. Who didn’t know the stark and beautiful heaths, those wide open spaces clothed in grasses and heather? Some might say they know the heathlands like the back of their hand, yet who really knows the back of their hand? The seemingly familiar is really unfamiliar, once you learn to see. Darwin made observations in Surrey, near pine-covered Crooksbury Hill. A scan of the hundreds of acres where cattle grazed revealed not one pine. Yet a closer look, low to the ground, showed trees were everywhere. Free-ranging cattle kept them cropped. These were pygmy trees, he thought, and he estimated the age of one at 26 years. Such is the power of one species—cattle—to keep others in check and shape the landscape. He observed another heath in Staffordshire where pines abounded in a fenced area enclosed 25 years prior. The whole ecology was altered by the exclusion of cattle: he counted 12 plant species in the enclosure not found outside of it, and six insectivorous bird species. These enclosures became a central part of Darwin’s ecological thinking: species influencing species, great and small, in a web of interconnectedness. It was all there, plain as day, evident in a woodlot, heath, or backyard weed garden.

  A replication of Darwin’s “Weed Garden” experiment at Down House. Photograph by the author.

  Disgraceful Blunder Averted

  The distinguished philosopher of science Michael Ruse once referred to Darwin’s chapters on variation, struggle for existence, and natural selection as the “deductive core” of On the Origin of Species. They hang together, building an argument almost syllogistically. In them Darwin almost seems to be saying that if there is abundant variation in nature that can be passed down to descendants, along with constant and intense population pressure and struggle, then survival and successful reproduction based on those variations must follow. He dubbed that process “natural selection,” and it is well known that he first conceived of the idea in 1838. But it is often unappreciated that Darwin came to see his initial concept of natural selection as incomplete: natural selection as such might explain how groups supplant groups, or how species slowly change one into another, but how does it generate the diversity of life?

  A crucial element of these chapters was his case for the mechanism of divergence, and so all the while he was also (among many other things!) making steady progress with his botanical arithmetic. He went through the manuals one after another, counting up varieties that the authors had listed for larger and smaller plant genera. As he expected, larger genera (i.e., those with more species) also had more varieties. On March 31st he recorded in his diary having completed his “Natural Selection” chapter—which included his ideas to date about divergence, or diversification, and the evidence from larger genera. He was writing confidently but soon needed a break. Exhausted, he headed to Dr. Lane’s establishment at Moor Park, in Surrey, the hydropathic spa he now favored. After the death of Annie, the Malvern facility was too fraught with painful memories. “My health has been very poor of late,” he wrote to Lyell in April. Two weeks later he was home—“Returned May 6th Did me astonishing good.”25

  He would be back soon. Fourteen-year-old Etty had been afflicted with a myste
rious ailment for months. Emma had first taken her to the seaside town of Hastings. When that failed, they went to Moor Park. In a long letter to Hooker, Darwin mentioned that he’d be tag-teaming in June with Emma: “My wife Emma & Etty have just started for Moor Park; she [Emma] will stay a fortnight, & then I shall relieve guard for another fortnight.”26 Etty ended up staying through October except for a brief trip home in August, the delicate state of her health a chronic source of worry for her parents.

  Darwin worked away at his “everlasting species-book” as best he could. By the end of September 1857, he had finished chapters seven and eight. But he was practically derailed in mid-July 1857 when he received a letter from his neighbor John Lubbock, informing him that his botanical arithmetic calculations were all wrong. Lubbock pointed out that Darwin had been rather arbitrarily labeling genera as “big” or “small,” but a more rigorous approach would have been to predetermine the size categories and then look at the frequency of varieties found in each. This way he would not be using relative “bigness” or “smallness” of genera, but an absolute size measure that would be used to categorize all genera for each region in the study. Of course, deciding on “large” and “small” categories is itself arbitrary, but at least there is then a consistent standard. Darwin was initially shocked by Lubbock’s revelation. Still, he knew Lubbock was right and he thanked him profusely: “You have done me the greatest possible service in helping me to clarify my brains. If I am as muzzy on all subjects as I am on proportion and chance—what a book I shall produce! . . . I am quite shocked to find how easily I am muddled, for I had before thought over the subject much, and concluded my way was fair. It is dreadfully erroneous. What a disgraceful blunder you have saved me from.”27 He couldn’t resist a frustrated P.S.: “It is enough to make me tear up all my M.S. & give up in despair—It will take me several weeks to go over all my materials. But oh if you knew how thankful I am to you.” He immediately wrote to Hooker too, kicking himself in the pants and crying that he was the “most miserable, bemuddled, stupid dog in all England.”28