A weed is something growing where you don't want it to. At best, the impact of climate change on plants is not well understood.

A common thing heard in these parts is that CO2 will make crops grow faster. Well - if so, it will probably have that impact on weeds, too. Currently, weeds cost U.S. farmers about 12 percent of their harvest, an estimated annual loss of $33 billion.

USDA researcher Lewis Ziska has found that elevated carbon dioxide levels in the atmosphere dramatically increase weed growth. Not only did the weeds grow much larger in hotter, CO2-enriched plots but they also produced more pollen. Even more alarming was the way that the increased heat and CO2 accelerated and perverted the succession of species within the plots.

By comparing three sites — an organic farm in western Maryland, a park in a Baltimore suburb and the one by the inner harbor — Ziska planned to study three circumstances: the present (on the organic farm), the mid-century future as predicted by the climate-change panel (in Baltimore) and something in between (the suburban site). He took soil from the organic farm, which already contained seeds of 35 common weeds, and with it created uniform beds at each of the sites, urban, suburban and rural, so that the growing medium and weed population would be the same throughout. What happened over the next five growing seasons surprised even him.

Not only did the weeds grow much larger in hotter, CO2-enriched plots — a weed called lambs-quarters, or Chenopodium album, grew to an impressive 6 to 8 feet on the farm but to a frightening 10 to 12 feet in the city — but the urban, futuristic weeds also produced more pollen.

Even more alarming was the way that the increased heat and CO2 accelerated and perverted the succession of species within the plots. Typically, a cleared area in the Eastern United States, if left to itself, returns to native woodland. This process varies with the site and circumstances, but in its archetypical form fast-growing annual weeds cover the soil first, playing the role of what ecologists classify as “pioneer plants.” These gradually give way to longer-lived perennial weeds, which are in turn replaced by shrubs and trees.

In the natural version of this process, the pioneers and their successors are species indigenous to the area, and the woodland’s restoration takes decades. But what Ziska observed in his urban plots was ecology on amphetamines, a nearly completed succession to trees by the end of five years, with a domination by invasive weed trees of the most troublesome sort: ailanthus, Norway maples and mulberries...

Weeds are already very well adapted to living with human society, and especially with agriculture. We've even accidentally bred them to be stronger and more resistant to our efforts to eradicate them.

Simply put, any plant, if we dislike it, becomes an intruder in our landscape and so a weed. Arguably, then, there was no such thing as a weed until mankind developed the need to discriminate, which came with the development of agriculture in the Neolithic era, around 9,000 B.C.

In fact, many of the wild grains like red rice or wild oats that are among our most troublesome agricultural weeds today were valued food sources until we graduated from the hunter-gatherer stage of our existence.

Much has been made of our scientific triumph in breeding modern crop plants from those wild ancestors. The transformation of an east Asian wild grass (red rice) into the crop that provides 20 percent of humanity’s caloric intake is extraordinary. What generally goes unrecognized, though, except among weed scientists, is the extent to which we also made weeds what they are.

Coexistence with mankind has given rise to the sort of tough plants that flourish despite the worst we can do — hoeing, pulling, burning and, more recently, spraying the fields with herbicidal chemicals. Weeds have adapted to every means we used to exterminate them, even turning the treatments to their own advantage.

Attacking a Canada thistle (actually of Eurasian origin and a regular entry in “worst weeds of North America” lists) with hoe or plow, for example, may destroy the plant’s aboveground growth but leaves the soil full of severed bits of fleshy root, each of which may sprout a new plant.

As weeds have grown stronger and more genetically diverse, crops have moved in the opposite direction. The more diverse a plant is, the better it will be able to cope with climate change.

A result of this history is that crops and weeds embody diametrically opposed genetic strategies. Over the centuries, we have deliberately bred the genetic diversity out of our crop plants. Creating crop populations composed of clones or near clones was an essential step in achieving higher yields and the sort of uniform growth that makes large-scale, mechanized cultivation and harvesting possible.

Because weed populations live as opportunists, however, they must include individuals with the ability to flourish in whatever type of habitat we make available. They also need diversity to cope with the wide range of punishments we inflict. A patch of Canada thistles, if it is to survive when the farmer switches from hoeing to herbicides, must include individuals that develop a resistance to the chemicals over time.

Weed populations that lacked the necessary genetic diversity faded from our fields, lawns and waste places; historians of agriculture can cite many such casualties.

The survivors are an astonishingly plastic group of plants... “When you change a resource in the environment,” Ziska said recently, sitting in his compact office, “you are going to, in effect, favor the weed over the crop. There is always going to be a weed poised genetically to benefit from almost any change.”

... What he and his colleagues have found, he said, is that weeds benefit far more than crop plants from the changes in CO2 and that the implications of this for agriculture and public health are grave.

Tests with common agricultural weeds like Canada thistle and quack grass found them more resistant to herbicides when grown in higher concentrations of CO2, making them harder to control. Ziska hypothesizes that this may be a result of faster growth; the weeds mature more rapidly, leaving behind more quickly the seedling stage during which they are most vulnerable. This promises to be an expensive problem for farmers, who will have to spend more on chemicals and other anti-weed measures to protect their crops. (Herbicides already cost farmers more than $10 billion annually worldwide.)

But enhancing CO2 levels, Ziska has found, not only augments the growth rate of many common weeds, increasing their size and bulk; it also changes their chemical composition. When he grew ragweed plants in an atmosphere with 600 p.p.m. of CO2 (the level projected for the end of this century in that same climate-change panel “B2 scenario”), they produced twice as much pollen as plants grown in an atmosphere with 370 p.p.m. (the ambient level in the year 1998). This is bad news for allergy sufferers, especially since the pollen harvested from the CO2-enriched chamber proved far richer in the protein that causes the allergic reaction. Poison ivy has also demonstrated not only more vigorous growth at higher levels of CO2 but also a more virulent form of urushiol, the oil in its tissue that provokes a rash.

Ziska says that he worries about mankind’s ability to feed itself in a fast-changing future. Paradoxically, it is weeds, he says, that can provide solutions. They have helped us deal with lesser crises in the past. When diseases and pests overwhelmed our domesticated food crops, it was to their wild relatives — plants that mankind has been battling for millennia — that plant breeders turned. Because weeds have more diverse genomes, it is easier to find one with the proper genetic resistance to a given threat — and then to create a new hybrid by breeding it with existing crops. An answer to the Irish potato blight of 1845-6 was eventually found among the potato’s wild and weedy relatives; a wild oat found in Israel in the 1960s helped spawn a more robust, disease-resistant strain of domesticated oats.