Much tropical forest may not be primary forest, as mentioned earlier in this document. As much as 12% of the Amazon forest (and probably a great deal in Africa and Southeast Asia as well) may have been altered by past agricultural systems of indigenous peoples. Therefore many rainforest ecocommunities are probably the result of centuries of swidden agriculture, which has left forests composed of many stages of regeneration and succession, although we view them today as “pristine” or “virgin” forests. However, modern agriculture does not leave forests relatively intact, as did these earlier agricultural systems. As the human population has grown and the standard of living has improved for some, the demand for the products of agriculture has soared. In addition, improvements in agricultural technology have allowed ever greater areas to be cultivated by relatively few individuals.

The greatest alterations of the global environment have been a consequence of the expansion of agriculture into former grasslands, forests and even mountainous enclaves. The total area of cultivated land has increased more than 450% in the past three centuries, although the rate of expansion has slowed lately due to intensified agricultural management, improved technologies, and the decline in arable land which has not been already utilized for agriculture (Matson, et al., 1997). Thus, while the human population increased from three to five billion (an 80% increase) in the three decades between 1960 and 1993, the global area devoted to cropland increased only 8% (Goklany, 1998). (Since then more than one billion people have been added to the human population.)

Much of the new agricultural land has come from former forest land, particularly rainforests. Over the past century and a half, approximately 40% of the agricultural land in Africa, 40% in Latin American and 70% in Asia has been derived from former tropical forest land. Even so, it has provided only two million km2 of the 15 million km2 of farmland globally (Pimm, et al., 2001). During this time period, the amount of land converted from forest to agriculture was more than double all of the land converted from the earliest origins of agriculture to about 1850. And the trend continues. Already 23% (4,700,000,000 hectares) of the earth’s land area has been converted to agricultural and pastoral use. This represents 45- 60% of the land potentially suitable for agriculture (Dobson, 1995). Some predict that pastureland may increase by more than five hundred million hectares, and cropland by more than three hundred million hectares in the next half century. In that case there will be 18% more agricultural land in 2050 than at present, with a loss of one billion hectares of natural ecosystems (larger than the land area of the United States), mainly in the Neotropics and sub-Saharan Africa (Tilman, et al., 2001b). This represents one-third of the remaining forests, savannas, and grasslands. Lost will be their ecosystem services and many of their species and products. However, most of this presently uncultivated land is little suited to agriculture. This marginal land might produce crops for a few years and then become scrub land or desert, as has happened already in many places where such areas have been co-opted for agriculture. In Africa, the Sahara Desert is expanding rapidly because of overgrazing and overuse of arid lands for agriculture. Declines in agricultural productivity because of land degradation could increase these impacts, driving a demand for yet more land and intensification of pesticide and fertilizer use. These effects could be mitigated by lower per capita consumption of meat, or if global population were stabilized at lower levels than projected (and, conversely, these effects could be exacerbated if the human population continues to increase)

Modern technologies which have allowed agricultural expansion have an ugly face, in that they have contributed heavily to habitat degradation by the excessive use of water for irrigation, the release of excess nutrients which causes eutrophication, pollution from pesticides and herbicides, salinization, and other problems with water resources (see below). It can also be argued that these technological advances, in addition to increasing productivity, have permitted more population growth than would otherwise have been possible, and thus have increased land conversion. Against this one can say that smaller populations lacking the new technologies would still have required much more agricultural land and would not have saved any land for conservation when it could have been utilized for food production.

Humans now appropriate almost 40% of the primary production of terrestrial ecosystems, much of it for agriculture and pastureland (Vitousek, et al., 1986; Rojstaczer, Sterling and Moore, 2001; Field, 2001). As the human population continues to increase (to perhaps nine billion by 2050) and as per capita wealth increases, what can we expect for agriculture in this next century? Will we continue the trajectory of the twentieth century, when global food production doubled between the 1960’s and 2000, and during which there were great increases in global nitrogen and phosphorus fertilization, as well as in irrigation? Tilman, et al., (2001) have made a series of projections based on past trends. Their mean projection was for global fertilization to be 2.7 times present values by 2050, with annual additions of 236 million metric tons of nitrogen to the terrestrial environment, and with phosphorus fertilization rising to 2.4 times present values. Irrigated land would be 1.9 times as great in area as at present. This may be put in perspective if we realize that humans already add as much nitrogen and phosphorus to the land as is supplied by all natural sources of these minerals; thus, the addition of yet higher proportions of them to the soil would be staggering, and would have serious environmental consequences.