Techn%KJ', Vol 6, pp 107-130. 1999Printed In the USA All rights reserved

1072.9240/99 SIO 00... 00Copyright tC 1999 Cognizant Communication Corporation

MEE'-ING GLOBAL FOOD NEEDS: THEENVIRONMENTAL TRADE-OFFS BETWEENINCREASING LAND CONVERSION AND LANDPRODUCTIVITY

Indur M. (30klany*Department of the Interior, Washington. DC

TF 9803-503 M (Received 2 March 1998; accepted 12 June 1998)

Despite this century's dramatic population increase, the global food situation, possibly exceptingsub-Saharan Africa, has improved remarkably-mainly due to the interdependent forces of economicgrowth, technology. and trade. However. low purchasing power and strife keep certain populationsvulnerable. Due to increased land conversion and inputs (fenilizers. pesticides. and water). theimprovements have extracted an environmental price. That price would have been higher absenttechnological change which, since 1961, forestalled additional conversion of 3.550 million hectares (Mha)of habitat globally to agricultural uses, including 970 Mha to cropland. Increasing-and richer-popu-lations may raise food demand 120% between 1993 and 2050, which can only be met by increasingcropland, productivity, or both. The pre~ise combination is critical for global biodiversity. Increasingaverage productivity 1 %/y implies losing 368 Mha of habitat to cropland by 2050; while a 1.5 %/yincrease would reduce cropland by 77 Mha. Either is plausible. given productivity-enhancingopportunities. To the extent productivity increases result from additional inputs. the environmental benefitsof reducing habitat loss may be panially offset. Given the severity of habitat conversion, prudence suggestsincreasing productivity while using inputs efficiently and mitigating their impacts. That requires sustainedcommitment to economic growth, technological change. and freer trade. Otherwise, technologies-whetherto maintain or increase productivity or mitigate impacts-can neither be developed nor afforded, and accessto food will diminish That would-by reducing food supplies, increasing vulnerability of the poorest tohunger. escalating habitat loss, and increasing environmental degradation-deprive humanity and despoilthe rest of nature.

INTRODUCTION

At least since Malthus (1926) wrote his first essayon population in 1798, raising the specter of geo-metric population growth outstripping humanity'sability to expand food supply-which he presumedcould only increase arithmetically-scholars andpolicy makers have debated whether future food.Views expressed in this article are the author's and not the U,SGovernment's

needs can keep pace with population growth. Tothis debate, a new issue should now be added,namely, whether these needs can be met withoutdisplacing the rest of nature (Goklany and Sprague1991; Goklany 1992, 1993; Waggoner 1994). Inthe past two centuries global population increasedfrom under one billion to 5.7 billion, and ever larger

1m

1M GokIany108

shares of the world's land, water, and other re-sources have been expropriated for human needs(Vitousek et al. 1986; Turner et al. 1990; Cohen1995; Goklany 1995a). In particular, conversion ofland to agriculture is the single greatest agent ofhabitat conversion, and associated displacement ofspecies and increasing stress on biological diversity(Goklany and Sprague 1991; Goklany et al. 1992;Goklany 1995a, 1998a). These pressures will onlyincrease if, as many project, the world's populationdoubles over the next 100 to 150 years (WorldBank 1994) and becomes wealthier (Goklany andSprague 1991; Goklany 1996, 1998b).

This paper first addresses whether Malthus'prognostications regarding an increasing imbalancebetween the growth in population and in foodsupplies have been vindicated and, if not, why\not?In this context, it will examine the correspond~nce,or lack of it, between trends in global popurationand cropland over the last two centuries and, over amore lirnited period, trends in agricultural land. Thepaper will then address whether past progress inmeeting global food demand can be continued intothe future and, equally importantly, whether and towhat extent the quest to meet that demand coulddiminish habitat for the rest of nature. Finally,based on the lessons of the past, it will recommenda set of self-consistent, practical policies that wouldhelp meet the increasing food demand of a largerand wealthier population into the middle of the nextcentury, while reducing environmental harm andminimizing land conversion, and associated loss ofbiodiversity .

A HISTORICAL PERSPECTIVE

During the last two centuries, global populationhas sextupled; yet the average person is fed better,more cheaply, and spends less time and effortgetting food on the table. Thus, Malthus' country-man is about three inches taller, health ier, less proneto disease, and lives twice as long today as hiscontemporary of 1798 (Fogel 1994; Wrigley andSchonfeld 1981; UNFPA 1995). The situation else-where around the world has also improved mark-edly, particularly since World War II. Since theirpost war lows, average daily food supplies percapita in China and the Indian subcontinent-42%of humanity-increased about 30%, and 45%,

respectively (Table 1). The number of people inareas with reported famine dropped from over 700million annually (1950-56 average) to 35 million in1992, a drop of95 percent (Chen and Kates 1994).Chronic undernourishment decreased from 941million in 1969-71 to 781 million in 1988-90,declining from 36% of the population of developingnations to 20% (Alexandratos 1995b). In fact, it isa measure of humanity's success that some suggestthe issue is no longer one offood quantity, i.e., foodsecurity, but quality, i.e., nutrition security (Swami-nathan 1989).

Despite increased demand due to both increasingnumbers and more affluent populations, the realprice of food and produce has dropped. Between1950 and 1992, international food commodity pricesdropped 78% in constant 1990 prices (Mitchell andIngco 1993). Following that, prices increased,taking back 8% of the drop by the first half of 1996(World Bank 1996; WRI 1996), mainly due to acombination of several singular and temporaryfactors including reductions of agricu ltural subsidiesin both Eastern and Western European nations; poorwinter wheat prospects in the U.S.; and delayedspring wheat plantings in the U.S., Canada, andWestern Europe in 1996, coupled with historicallylow levels of reserve stocks (F AO 1996b). Despitethese recent increases, food prices are much lowertoday than for much of history. Some have ascribedthese increases to a fundamental turnaround, presag-ing increasing future difficulties (Brown and Kane1994), but there seems to be little evidence of that(Alexandratos 1995a; Harris 1996). In fact, whileone should be cognizant of them, it would be impru-dent to base long-term policies on just short-termtrends-hence the attention in this paper to bothlong- and short-term trends.

For the U.S., from the turn of this century(1897-1902) to 1992-94, retail prices of flour,bacon, and potatoes relative to per capita personalincome dropped 92%, 87%, and 80%, respectively,even as the length of the average work week de-clined 40% (Maddison 1989); bread dropped 78%between 1919-21 and 1992-94 (Fig. I). As relativeprices of food and agricultural products have de-clined, so has agriculture's share of the Gross Do-mestic Product (GDP) and labor force. It nowcontributes about 2% to the U.S. GDP and 3% to itscivilian labor force as opposed to about 75% inI 800 (Bureau of the Census 1975, 1993). Ironically,~

Meeting global food needs: the environmental trade-offs 109

the very success of the agricultural sector, by de-creasing its economic and demographic importancefor the rest of society could, as will be discussedlater, adversely affect the ability to meet future food

needs. Nevertheless Churchill's immortal words,"never ...has so much been owed by so many to sofew," may be just as apt to today's American farmeras it was to the victors of the Battle of Britain.

Table 1. Per capita food supplies (Kcal/Day), selected countries (1934/38 to 1992).

1934-38 1950-51 1961-63 1970 1980 ]990 1992Country

France 2,830 2,827 3,299 3,349 3,455 3,650 3,633

3,O70a.hGennany 2,807" 2,945" 3,217 3,382 3,455 3.344

2,234bPortugal 2,570 2,966 2,925 3,590 3,634

3,367 3,357 3,638 3,156Czechoslovakia 3;330

Yugoslavia

EUROPE

2,400 3,119 3,327 3,594 3,551

3,098 3,255 3,390 3,470 3,410

u.s. 3,150; 3,085" 3,195 3,192 3.333 3,680 3,732

2,6091 2,286 2,638 3,104 2,833Cuba 2,682 2,998

Brazil 2,150; 2,3530 2,315 2,448 2,705 2,731 2,824

2,O77dPeru 2,225 2,294 2,042 1,825 ,882

S. AMERICA 2,395 2,500 2,662 2,621 2,689

2 226ej, 2,1150.£China ,666 2,032 2,332 2,679 2,727

India ,970gk

,635 ,997 2,082 1,959 2,297 2,395

Pakistan ,624I

1,705 2,200 2,114 2,431 2,315k

Bangladesh ,953 2,196 .902 ,994 2,019

2,591Japan 2,180 2,100 2,691 2,758 2,906 2,903

,887 2.544 2,585ASIA 2,144 2,276

2,450 2,342 2,296 2,515 3,119 3,336 3,335Egypt

Zimbabwe 2,224 2,053 2,226 2,292 2,173 1,985

Somalia ,704 1,819 1,788 ,769 1,499

AFRICA 2,086 2,244 2,285 2.294 2,282

WORLD 2,289 2,464 2,559 2,718 2,718k. see IndiaI. see Pakistan

a. West Gennan portion only

b 1949/50

c 1951d 1952

e. 22 provinces onlyf. 1947/48g Includes what is currently

Pakistan and Bangladesh

h. 1935-1938i. 1935-39

j. specified as "prewar'

(FAO 1952. 1954. 1991. 1994)

110I.M. GokIany

---Flour-

Bread..

Bacon"

Potatoes.

NOTE~ are 3-ycar averages centered on the year for which data are shown, except for 1899 and 1909.whIch are 5-year averages centered around those years.price relative to income = I in 1897-1901; ..price relative to income = 1 in 1919-192\

Sources Historical SlDtistics of the United States. various issues ofCPI Detailed Report and Statistical Abstracts.

Fig. 1. Retail prices of various food items relative to per capita personal income (1899-1993).

Still problems linger: as noted, famine and mal-nourishment while reduced, have not been elimi-nated. However, today's problems are due as muchto failures of institutions, policies, and politicalsystems causing greater poverty, civil strife, anddisincentives to food production, as to the inabilityto physically produce food (Sen 1981, 1993; Drezeand Sen 1990; Kumar 1990; Tapsoba 1990). InAfrica, despite a 60% increase in food productionbetween 1969-71 and 1992 (FAO 1994b),availablefood supplies have barely kept pace with populationgrowth rates (Table 1) due, in part, to economicsystems which inadequately reward farmers fortheir investments and risks, or subsidize someconsumers at their expense (Tapsoba 1990). Civilstrife in Ethiopia, Somalia, and Sudan convertedwhat may have been more manageable droughtsinto full-scale famines. Similarly, the problems ofPeru can be traced, in part, to civil strife. Theregression indicated by the 1992 figures for theex-centrally planned economies (ex-CPEs or "transi-

tion" nations) is due to the transition from anunsustainable socio-economic system to a marketeconomy, aggravated in the former Yugoslavia byan uncivil war. With the current peace in Bosnia,food supplies are recovering in that area (F AO1996b).

In summary, the last two centuries' experiencesuggests that so far at the global scale, at least,Malthus' fears regarding the ability of food supplyto keep pace with. population growth have not yetbeen borne out. However, at local to regionalscales, there have been several instances where foodsupplies have effectively diminished, at least tem-porarily, but not necessarily for the reasons heanticipated. Is it possible that, as the world'spopulation doubles over the next century he may yetbe vindicated at the global scale?

The following attempts to shed light on thisquestion by first examining the reasons for pastsuccesses and failures before evaluating futureprospects into the next century.

Meeting global food needs: the environmental trade-offs III

FACTORS CONTRIBUTING TO IMPROVEDFOOD SECURITY: PAST LESSONS

human or animal energy-include scientifically-bredhigh yielding varieties (fNVs) of crops whichmature faster, making multiple crops more possible;improved livestock management practices; fertilizerand pesticide usage; mechanization; refrigeration;canning and other means of preserving food andproduce; and meteorological forecasts (Goklany andSprague 1995). Irrigation-an ancient technology,freshly applied-today has been extended to a sixthof the world's cropland (compared to about 2% in1800 and 6% in 1950), and provides a third of thecrop production, i.e., on average, it increases yields2.5-fold (Crosson 1995a; Rozanov et al. 1990).Finally, an elaborate transportation and distributioninfrastructure makes possible local, regional, andinternational trade in agricultural products andinputs (fertilizers, pesticides, fuels) with greaterrapidity-<:heaper costs and lower losses than everbefore.

Land, water, and technology

Between 1800 and 1993, while global populationincreased about 500%, cropland increased 250%from about 408 million hectares (Mha) to 1,448 Mha(Fig. 2). The difference between these increaseshints at the contribution of technological change toincreasing global food supplies. In fact, technologyreversed the gradual, centuries-long global trend ofincreasing cropland per capita: from 0.43 ha percapita in 1700, it peaked at 0.48 in the first fewdecades of this century; today it is at 0.26 (Fig. 2).

Technologies responsible for this turnaroundaffect each link of the entire food chain, from farmto consumer. Specific technologies-many of whichrely on increased use of inanimate rather than

-population

actual cropland per capita

actual cropland

cropland, if no tech change after 1961

~AB ; incn...t in crop/am( IP61 10 /PP1 -0/1 Bha

BC -/oI.J_djrom conwnio" 10 cropla~. /P61 10 1991

-OP7Bha

.~.u"'., McE..dy & JI1M3 1978. Hough'on,' 01 I98J. WRlI98.. R,c- J9P11. FAD JP96

Fig. 2. Global cropland area, 1800-1993.

112 I.M. Goklany

1961, cropland would expand proportionally topopulation, and the world's 1993 population wouldbe fed no better than the inadequate levels of 1961(particularly considering the unfortunate realities ofunequal distribution). i.e., daily per capita caloric andprotein intakes would be at 2.259 Kcal and 59.5 grncompared to current (1992) levels of2,718 Kcal and70.8 grn, respectively (FAO 1991, 1994a).

Similarly. Fig. 3. which compares the 1961-1993trends in total global agricultural area (i.e., croplandand pennanent pasture), population and agriculturalarea per capita. also shows that under a technologyfreeze at 1961 levels. at least 3,546 Mha (BC}-inaddition to the 375 Mha actually converted(AB}-would need to be converted to agriculturaluses. That would have increased agri~ulture's shareof all land (excluding Antarctica) from the current34% to 61%.

Of course, these technologies also create their ownproblems. The adverse environmental effects ofdi veTting water for irrigation, and inefficient use offertilizers and pesticides are well known (Ehrlich etal. 1993; Pimentel et al. 1994a, 1994b, 1995). Lessappreciated is that technology, in addition to feedinga much larger population better, has also preventedthe conversion-globaIlY-Qf enormous amounts ofhabitat (including forests and grasslands) intocropland or other agricultural uses (Goklany and

Sprague 1991; Goklany 1993).Between 1961 and 1993, population increased from

3.08 to 5.54 billion (F AO 1996a). If technology hadbeen frozen at 1961 levels, then in 1993 the worldwould have needed to convert at least 1,073 Mha ofother habitat to cropland (indicated by line AC onFig. 2), rather than 107 Mha (line AB). This (AC) isprobably an underestimate. This assumes that after

--population

ago area per capita

--a--

agricultural area

ago area, if no tedl change after 1961

~:A8 -irK:1rase in og tlrra, 1961 10 1993 -0.38 8M

8C -laIwi..., co1r-rn.d 10 ag 1961 10 1993

-3.55 8M

.~rrr FAD 1996

Fig 3 Global agricultural area. 1961-93.

Meeting global food needs: the environmental trade-offs

Economic growth/reduction of poverty

Economic growth affects food security and agri-cultural production in several ways. First, as therecent food-related problems in Africa and thetransition nations illustrate, greater wealth translatesinto greater purchasing power which, in turn, in-creases food security for society, families, andindividuals (Goklany 1998a). Table 1 hints at this;the poorer nations seem to have lower food suppl ies.Second, it takes more than the availability oftechnology to have technological progress: technol-ogy, first, has to be created and, then, it has to beadopted (Goklany 1996). That product of economicgrowth-capital, both human and financial-iscritical to both stages. Wealthier nations have morehighly-educated populations and greater resourcesfor research and development (R&D), and povertydirectly reduces the affordability of new and exist-ing technologies (e.g., new seeds, fertilizer, irriga-tion, and transportation systems). Thus, technologi-cal progress and affluence co-evolve (Goklany1995a). In fact, the legal and economic institutionsthat nurture one, sustain the other. Third, affluenceincreases incentives for lowering fertility rateswhich, in turn, would reduce demands on the foodsystem-all else being equal (World Bank 1984;Livi-Bacci 1992).

Finally, consistent with the theory of an environ-mental transition, economic growth enablessocieties, sooner or later, to reverse water pollutionproblems, e.g., phosphorus, nitrate, or dissolvedoxygen levels caused or aggravated by agriculturalpractices (Goklany 1994, 1995a, 1995b). This theoryholds that a society's environmental problems will,in general, first be aggravated by economic growth,then go through a transition after which furthergrowth (and greater affluence) will reduce thoseproblems. The point at which the transition occursdepends upon the magnitude and urgency of thespecific problem and the cost to society of action (orinaction) (Goklany 1 998a). Many economists referto such behavior which results in an invertedU-shape curve if environmental degradation isplotted against affluence (Grossman and Krueger1991; Shafik and Bandyopadhyay 1992), asindicative of a "Kuznets curve"-after SimonKuznets (1955) who showed that economicinequality between the richest and poorest segments

To gain another perspective on the amount ofhabitat "saved" from conversion since 1961, con-sider that: I) globally, forest and woodlands (includ-ing natural and planted tree stands, and logged areasslated to be reforested) declined 144 Mha between1961 and 1993 (FAD I 996a); and 2) all "protected"areas in the world, excluding Antarctica, total 960Mha (WRI 1996), of which 510 Mha are "totally

protected."In the V.S.. the amount of cropland harvested in

1995 was 7% less than in 1910, despite a 186%increase in the population and an increase in exportsrelative to production (Bureau of the Census 1975,1995; USDA 1996). Goklanyand Sprague (1991)calculated that if technology had been frozen at the1910 levels, in order to produce the same quantity offood as in 1988, an additional 370 Mha of croplandwould have had to be harvested in I 988-more thanall the remaining arable and forested lands, com-bined, in the u.S. The resulting land conversion,equivalent to 63% of the U.S. excluding Alaska (ormore than three times all federally owned landsincluding all National Parks, Forests, Grasslands,Wildlife Refuges, Recreation Areas, and multiple-use lands), would have devastated many speciesand rendered irrelevant any discussions in the U.S.about preserving biodiversity in situ. Unwittinglyand virtually unheralded. technological changeleading to increased agricultural productivity,despite its numerous adverse environmental af-fects, seems to have been the single most importantagent for conserving habitat and species (Goklanyet al. 1992).

The development and adoption of new technol-ogy in the U.S. was aided by a high level of generaland technical education; strong commitment toresearch and development; and elaborate systemsfor technology transfer. However, the V .S. experi-ence. con finned by the decades-long experimentsof the CPEs, including China, for example, indi-cate that these conditions are insufficient, bythemselves. to sustain high-yield agriculture and amodern food supply system. The legal, economic,and political framework also has to encourageinnovation and adoption of new technologies by

economically rewarding individuals for theirendeavors and risk taking (Goklany and Sprague

1991).

1M Goklan~114

of society first increased and then decreased as afunction of affluence. Thus, in the long run, wealtheventually increases the overall sustainability of

agriculture.

reinforcing pillars: economic growth, technologicalchange, and trade.

FUTURE POPULATION GROWTH ANDFOOD SUPPLY SCENARIOS

TradeFuture food consumption depends, among other

things, on population size, and the economic cir-cumstances of that population. Annual globalpopulation growth rates peaked in the 1960s and areexpected to decline further, from about 2.0% and1.8% in the 1970s and 1980s to 1.3% (1990-2010),1.1% (2010-2025), and 0.7% (2025-2050) under theWorld Bank's (1994) "standard fertility" projection.This projection suggests that global population willincrease from the current 5.7 billion to 9.6 billion in2050 and 11.4 billion in 2150. The low and highfertility estimates are 8.6 and 10.1 billion for 2050,and 9.7 and 12.9 billion for 2150. Most of thisgrowth is expected to occur in developing nations.Population may not double until the last quarter ofthe next century even under the high fertility as-sumption. Projections for the u.s. are less robustbecause of the added uncertainty of immigrationrates. The Bureau of the Census (1995) estimatesthat U.S. population will grow from 263 M today to300 M (+/-19 M) in 2010, and 392 M (+130/-107 M)in 2050.

If, as is hoped, the world enters an era of sustainedeconomic growth, per capita food demand shouldincrease more rapidly than population-particularlyin developing nations (Poleman and Thomas 1995).The increased demand for meat, fish, and dairyproducts would increase cereal demand even fasterbecause it takes about two to seven pounds of cerealto produce one pound of such products (Brown andKane 1994). Assuming that between 1993 and2050, global per capita food supplies (using cerealsas a surrogate) increase at the same rate as they didbetween 1969-71 to 1989-91, calculated using datafrom the Food and Agricultural Organization (F AO)(1996a), then aggregate supplies in 2050 shouldincrease 121 % over 1993 levels to meet food de-mand for 9.6 billion, the standard World Bankprojection. Per capita food supplies, in terms ofcereal production, would have to increase 28%.(Note: if the past pattern of growth in average dailycalorie intake per capita between 1969-71 to1989-91 is also replicated, then average per capitacalorie intake would increase 35%, i.e., to over3,600 kcal/capita. However, given that much of the

Trade-both within and between nations-is themechanism for allowing non- or low-producers toobtain food. It translates purchasing power (oreconomic growth) into food security. Trade globa-lizes sustainability by allowing one area's shortage tobe willingly made up by another's surplus, i.e., itconverts a locally unsustainable system to one sus-tainable in a wider context (Goklany 1995a). As agauge of its importance consider that in 1993-96,developing nations' cereal imports amounted to 15%of their production (FAO 1996b); and in 1991-93, ofthe 153 countries for which the World ResourcesInstitute (1996) provides data on trade in cereals, 127were net importers. Of the 26 net exporters, only onewas in Africa; six were in Asia; and fourteen, whoprovided the bulk of the exports, were from thewealthy Organization for Economic Cooperation andDevelopment (OECD) block of nations. Similarly,virtually all food aid is donated by the richer nations.Thus, patterns of trade and aid also confirm thesignificance of affluence in increasing the world'sfood security.

Trade also helps reduce the exploitation of mar-ginallands for growing crops-provided neither tradenor production is subsidized. Thus, because ofinternal trade in the U.S., land used for crops andpastures in the Northeastern U.S. declined 59%between 1949 and 1987, despite increases in: 1)regional and national populations of 31 % and 70%,respectively; 2) agricultural exports from 13% ofproduction to 35%, and 3) agricultural subsidies(Goklany and Sprague 1991). Most of the land sofreed up was returned to nature and many species,previously under stress, have rebounded.

Free trade is also an argument against a rationaleoften advanced successfully for subsidies, i.e., foodsecurity demands national self-sufficiency. Free tradewill also increase economic growth and purchasingpower in those developing nations whose economiesdepend heavily on agricultural products or, increas-ingly, on exports (Goklany 1995a).

In summary, past and present success in humanity'squest for food security is founded on three mutually-

Meeting global food needs: the environmental trade-offs 115

additional cereal supplies would probably be usedfor producing animal protein, such replication is

unlikely.)A 12]% increase in aggregate supplies by 2050

could also be viewed either as an increase in percapita cereal supplies of21 % for 10.1 billion people(the World Bank's high projection), a42% increasefor 8.6 billion (its low projection), or a 1] % in-crease for a doubling over] 993 levels.

For the U.S., between] 970 and] 990 the effectivecereal consumption per capita (considering allanimal products) dropped, mainly because of de-creased demand for beef and eggs. Therefore, it isreasonable to assume u.s. food supplies need onlykeep pace with population growth, i.e., increaseabout 50% by 2050.

is limited. Sea level rise due to climatic change andincreased urbanization will further reduce landavailable for agriculture. Fourth, irrigation is un-likely to expand much: the best irrigation sites arealready in use, and new ones will be economicallyand environmentally costly. Moreover, silting,salinization, and water logging will reduce theeffectiveness of existing irrigation projects. Fifth,over-reliance on monocultures makes the entire foodsystem susceptible to catastrophic diseases andpests. Finally, many areas of the world lack suffi-cient fresh water to meet the future needs of agricu 1-ture and other competing demands, which couldseriously limit future productivity increases (Postel

etal.1996;FalkenmarkandBiswas 1995; Engelmanand LeRoy 1993).

Thus, Ehrlich claims the world's carrying capacityhas already been exceeded, which he currentlyestimates at between 2.5 and six billion peopledepending upon whether it adopts a North Amer-ican-style meat-eating, or a vegetarian, diet (Ehrlich1995). Similarly, Brown and Kane (1994) projectthat in 1030, 2.5 billion could be fed at the currentU.S. level or 10 billion at the Indian level, assumingno major breakthroughs in technology.

The Nett-Malthusian case

The projected increases in food demand leadNeo-Malthusians to argue that the earth's carryingcapacity will soon be exceeded, if it has not alreadybeen. In general, they contend that the factorsresponsible for the phenomenal growth in theworld's food supply since 1950 are not sustainableinto the future (Ehrlich et al. 1993; Brown and Kane1994; Pimentel et al. 1994a). First, they argue,technological change can no longer be relied uponto bailout the population. The backlog of unreal-ized yield increases due to existing technologies issmall. Past improvements were due, in large mea-sure, to greater use of fertilizers and pesticides, butthese inputs are at a point of diminishing returns.Also, pesticides are a treadmill; ever more powerfulones are needed as pests adapt to existing chemicals.Similarly, the biological limits for improving yieldsare being reached. Moreover, there are no dramaticbreakthroughs on the horizon; if anything, loss ofbiodiversity compromises the ability to devise suchbreakthroughs. Second, past yield increases havebeen at the expense of future productivity. On land,current practices have increased desertification andsoil erosion to rates exceeding natural regeneration;on the seas, they have resulted in over-harvestingand collapsing wild fisheries. Moreover, futureproductivity will decline further because of in-creased air pollution (including acidic deposition),ultraviolet radiation and, possibly, climate change.Third, the ability to bring new land under cultivation

Non-Malthusian projections

There are, of course, alternative visions of thefuture. Several theoretical analyses suggest thatthe earth could feed between 10 to 40 billionpeople. Cohen (1995) summarizes several of theseanalyses. Most of these studies do not, in general,explicitly address whether environmental con-straints could (or would) be surmounted. How-ever, they implicitly or explicitly assume that withdue diligence and additional R&D, it is possible tocontinue to increase yields and expand cropland tomeet food demand. Waggoner (1994), for instance,estimates that it should be possible to feed) 0 billionon 2.8 Bha of cropland (compared to 1.4 Bha today)with an average daily per capita food supply of3,000-6,000 Kcal. Smil (1994) estimates that 10 toII billion could be fed without relying on biotech-nological advances by increasing cropland by about300 Mha-using and optimizing currently avail-able technologies and farming practices, and re-ducing beef consumption and post-harvest losses.

M GokJan)116

nations, chronic undernourishment would drop from20% to II % afflicting 140 million fewer people(Table 2). The improvement in Sub-Saharan Africawould be small until 20 I 0, but could acceleratethereafter. The ex-CPEs would maintain caloricinput at the 1988-90 level but meat consumptionwould remain lower than their pre-transition levelsdue to loss of subsidies. Food security would, asnow, not pose a significant challenge to the richernations. Despite increased production, developingnations would become net importers of agriculturalproducts as their populations expand and becomericher. Real prices may continue to decline unlesspopulation or economic growth accelerates or yieldgrowth rates decline (Mitchell and Ingco 1995;Alexandratos 1995a).

An elaborate F AO study (Alexandratos 1995a,1995b) concludes that it should be possible to meetglobal food demand to the year 201 {}-and possiblyto at least 2025-assuming evolutionary rather thanrevolutionary changes in technology (i.e., continuingemphasis on R&D and policy reforms to encourageproduction and economic growth) through a combi-nation of increased yields, and modest increases incropland (about 93 Mha in developing countries),irrigated areas, and cropping intensities. Globalcrop production would increase 1.8% per yearbetween 1990 and 2010, lower than historic annualrate increases (3.0%,2.3% and 2.0% in the 1960s,1970s, and 1980-92, respectively) but higher thanthe expected population growth rates noted above.In 2010, compared to 1988-89 levels in developing

Table 2. FAO projections, 2010 and 2025.

Per Capita Food Supplies(Kcal/Day)

1988-90 2010 2025

Chronic Undernutrition(Millions)

1988-90 2010

Population(Billions)

2010

B.EgjQ!!.

1990 2025

3,0005.30 7.07 8.38 2,700 2,880WORLD

1.48 3,400

3,380

3,410

3,380

3,470DEVELOPED COUNTRIES 0.86 0.91

includedabove

includedabove

FORMER CPEs 0.39 0.43

631 (1 (%)6.90 2,470 2.740 2.900 781 (20%)DEVELOPING COUNTRIES 3.90 5.67

2,100 2,170 2,700 175 (37%) 296 (32%)

29 (6%)

195 (12%)

Sub-Saharan Africa 0.49 0.87 1.28

24 (8%)0.31 0.51 0.67 3,010 3.130 3,180Near EastIN Africa

194 2.21S 2.450 2,700 265 (24%)

258 (16%)

South Asia 16 1.68

3,060 77 (4%)

40 (6%)

160 2.01 2,31 2,600 3,045East Asia

59 (13%)0.59 0.69 2,690 2,950 3,030Latin America &Caribbean Islands

044

(Alexandratos 1995a. 1995b)

productive potential can be conserved and whetheradditional suitable lands are available for conver-sion to cropland.

Without judging which camp's projections aremore plausible, the most significant neo-Malthusianarguments will be examined, i.e., that land, waterand technology are running out.

Land conservation technologiesLAND AVAilABILITY

Several technologies are currently available formitigating soil erosion, salinization, and waterlogging (Pimentel et al. 1995; McLaughlin 1993).

Two facets to the issue of future land for agricul-ture are whether existing cropland's long-term

Meeting global food needs: the environmenral trade-offs 117

cultivated or harvested in 1995 (USDA 1996). In1990, based on Bureau of the Census (1993: Table1128) data, domestic needs would have been metwith about 102 Mha. Absent agricultural subsidies,these figures may have been even lower. Clearly,the U.S. has ample potential cropland to meet itsdomestic food needs even if population and amountof developed land (estimated at 37 Mha in 1992)were to double, provided at least current technologyis deployed.

Potential cropland world-wide

Excluding unexploited potential cropland in thedeveloped nations (including transition nations) andChina, there are at least 3,335 Mha of potentialrainfed cropland worldwide (Alexandratos 1995a),about 2.3-times the 1,448 Mha that were arable or inpermanent crops in 1993. However, the 1,887 Mhasurplus in developing nations (excluding China) isunevenly distributed (relative to where populationgrowth may occur). Much of it is in Latin Americaand Sub-Saharan Africa. Second, between 45% and57% of the surplus is in forests or protected areas.Third, it may be inherently less productive thanexisting cropland. About two-thirds of this surpluscropland is constrained by low natural fertility, poorsoil drainage, steep slopes, or sandy and stony soils.Thus, substantial investment will be required to: 1)bring these lands into production~ 2) develop theinfrastructure to sustain agriculture, includingtransportation and distribution networks for inputsand produce~ and 3) ensure that these lands receiveadequate inputs. Compared to these constraints,Alexandratos estimates that the loss of productivepotential due to urbanization will be modest (aprojected 70 MHa by 2010, an increase of20 MHasince 1990).

Controlling erosion involves minimizing soil distur-bance, maintaining vegetative cover, or managingthe slope of the land. Existing technologies includeconservation tillage (e.g., no-till cultivation--effi-ciency, 80% to 99%+), crop rotations with legumes(50%), intercropping with grass (65%), contourplanting (80%), crop residue management (e.g.,stubble mulching), level terracing (95%+), andwindbreaks. Many techniques provide multiplebenefits, e.g., mulching conserves soil moisture andorganic matter; crop rotation with legumes reducesrunoff and fixes nitrogen. Erosion rates on U.S.croplands dropped about 25% between 1982 and1992 due to greater penetration of such technologies(NRCS 1995; Kellogg et al. 1994). Similarly,salinization can be mitigated by selecting appropri-ate salt-tolerant crops (e.g., wheat instead of com orrice) or managing soil moisture through crop rota-tions or surface drainage, and water logging throughwell-maintained drainage systems.

Some techniques entail relatively large initialcosts (e.g., foregoing crop residue as a source ofscarce fuel or fodder, or manpower for terracingwhich requires 750-900 man-days of hand labor perha, or 15 man-days using a 60 h.p. tractor) which maypreclude (or delay) their wider adoption-particularlyin developing nations (Dazhong 1993).

Uncultivated and potential cropland in theU. s.

WATER

As noted previously, irrigation increases yields2.5-fold on average. Nevertheless, competing de-mands (direct human use, industry, and in-streamuses) are increasingly winning out over agriculturefor water. This is virtually inevitable because ascountries get richer and more urbanized, the relativeeconomic and political clout of the agriculturesector diminishes. Moreover, as noted earlier,

An early 1980s estimate placed the amount ofarable land in the U.S. at about 218 Mha (Batie andHealy 1983). More recently, the 1992 NationalResources Inventory of non-federal rural landsestimated that there were 228 Mha in land capabilityclasses I through III which could be cultivated using,at most, "special conservation practices" and/orchoice of crops (NRCS 1995; Kellogg et al. 1994).[This estimate excludes areas classified as "otherrural land," i.e., Conservation Reserve Program(CRP) areas, areas under structures, windbreaks,barren land such as exposed rock or salt flats, andmarshland.] Another 73 Mha of Class IV could beused as cropland with "very careful management"and/or a reduction in the choice of crops. Anadditional 21 Mha of cropland were in Class V toVIII lands or enrolled in the CRP.

These estimates compare with 132 Mha that were

.M. Goklany118

used; and land leveling). Alternatively. water couldbe diverted to higher value crops. Both the amountof irrigation water applied per ha and the economicefficiency of water use vary with the crop, the latterby more than fifty-fold (Cervinka 1989). For exam-ple, rice requires an order of magnitude more waterper dollar of crop than potatoes, and about twice asmuch as wheat.

Fundamental to water trading is the vesting oflegally-enforceable property rights in the waterallocated to the various users. Successful cases offormal and informal water trading have been re-corded in socio-economic milieus as diverse as theU.S., Chile, Jordan, and India (Rosegrant et al.1995). In Chile, trading increased efficiency ofwater use by 22-26% between 1976 and 1992,which effectively expanded the irrigated area by thatmuch. Moreover, water can be, and is often, reused,which itself offers technological opportunities forreducing overall water consumption. In fact, muchof the water used for human consumption andindustrial use-being treated prior to consump-tion-is second or third (or more) hand.

Finally, in the ultimate analysis, if the price isright, one can resort to desalination [97.5% of theworld's water is in the oceans; 69% of the freshwa-ter is locked up in glaciers and permanent snowcover (Engelman and LeRoy 1993)]. Today, desali-nation, because it relies on fossil fuels, is economi-cally and environmentally costly. Its use is re-stricted mainly to relatively rich Middle Easternnations with access to cheap natural gas or oil.However, new technology, in the form of solarpowered plants, could reduce these barriers (TheEconomist 1995).

TECHNOLOGIES TO INCREASE NETYIELDS

agriculture has been a victim of its own success inthat it now commands a smaller share of GDP andemployment than ever before.

Because of its importance, societies have gener-ally been loath to treat water as just another eco-nomic commodity. In fact, most countries subsidizeits usage; when it is in short-supply, it is rationedexplicitly or implicitly by, e.g., limiting its availabil-ity by time of day or week or banning certain uses.Often, even where private entities pay for extractingwater from the subsurface, the pumps and tubewellsare indirectly subsidized through, for example inIndia, low electricity rates. By the law of perverseconsequences, these very sensibilities-by provid-ing direct or indirect subsidies--effectively encour-age wastage, discourage investments in conserva-tion, and result in less-than-socially optimal use ofwater. The reverse side of this coin is that there arenumerous unexploited opportunities for conservingwater. In the U.S., for instance, the efficiency of useof water diverted for irrigation, which accounts forabout 80% of consumptive water use nationally, wasestimated at 47% in 1987-suggesting the highpotential for water conservation (USDA 1989). Infact, a small increase in irrigation water use effi-ciency goes a long way toward meeting water needsfor other sectors. Thus, increasing that efficiencyfrom 41 % in 1975 to 47% in 1982 saved waterequivalent to 80% of domestic consumption. It isencouraging to note that both water withdrawal andconsumptive use per capita seemed to have peakedaround 1975 in the U.S. In 1990, they were down17% and 18% from 197~ levels, respectively; anddespite a 13 % population increase, total withdrawaland consumption in 1990 were below 1975 levels(Bureau of the Census 1995).

Treating water as an economic commodity, i.e.,establishing realistic prices and allowing trading sothat unused water could be sold to other users,would give farmers and other users the neededincentive for conserving water and lead to greateradoption of available conservation technologies. Itwould stimulate research into, and adoption of, forinstance. drought tolerant cultivars. Other availableconservation technologies include methods ofmanaging cultivation and soil moisture (e.g., centerpivot irrigation; drip irrigation; maintenance ofexisting canals, ditches, and on-farm storage facili-ties; optimizing the timing and amount of water

Increasing yields with existing

technologies

The high potential for using existing technologiesto expand yields is suggested by Tables 3 and 4which are based upon FAG (1996a) data. For theten most important types of crops, which cumula-tively account for about 60% of the world'scropland, Table 3 shows the yield ceiling [Y(C)] in1991-93, defined as the average yield for the nation

Meeting global food needs: the environmental trade-offs 119

increased production by 84%, while increasing allyields to Y(C) would have resulted in an increase of154%. Of course, Y(C) is not, by itself, limiting.Yields achieved by research stations or individualfarmers are often higher; at 21 T/ha, the 1992champion U.S. com grower's yield was 136%greater than Y(C) for com, i.e., the average yieldfor the highest nation (Waggoner 1994). Similarlogic applies to all the other crops listed in Tables 3and 4.

with the highest yield, and the average yields in1992-94 for developing, transition and other devel-oped nations [Y(DING), Y(TR) and Y(DPED- TR),respectively]. This Table also shows the area plantedin each of the crops worldwide and the portion of thatin developing and transition nations. Table 4,calculated from Table 3, shows the increases in1992-94 production if various yield gaps wereeliminated. For com, for instance, raising bothY(DING) and Y(TR) to Y(DPED- TR) would have

Table 3 Average yields in developing, transition, other developed and highest yielding nations for major crops (1992-1994 averagcs)

Average Yields (t/ha)

Developing OtherNations Developed

Nations

Yield Ceiling (t/ha)

Highest Yield for a SingleCountry, Y(C), Average for

1991-1993*

Percent of Global CropArea in Developing &

Transition Nations

Global CropArea

(million ha)

TransitionNations

Crop

Y.(IB}.

2.93

1.98

2.96

094

0.70

1.96

/.96

/.64

/2.22

0.90

/.00

Y(DING)

3.56

2.44

2.62

1.10

0.75

1.42

252

0.67

/ /.40

/.70

/.10

YCDPED-TR)

6.15

3.07

7.15

4.01

1.28

3.36

4.16

/..94

31,25

2.51

2.3/

Y.{Q

9.23

8.15

9.87

5.76

1.81

6.05

7-48

4.79

45./7

332

3.82

97

71

72

89

-100

70

80

92

95

58

96

146.1

220.5

135.1

45.1

37.9

73.5

700.0

67./

48.6

60.3

2/.4

Rice

Wheat

Com

Sorghum

Millet

Barley

Cereals. Total

Pulses

Roots and tubers

Soybean

Peanuts

~Australia

Netherlands

Greece

Italy

China

Switzerland

Netherlands

France

BIgmiLurmbg

Italy

Greece

Table 4 Increases in global crop production if yield gaps were erased (calculated from Table J).

Crop Crop Area in Developing and Increase in Global Production (%) ifY(DING) and Y(TR) are Raised to Levels of:Transition Nations ("/0) Other Developed Nations Y(DPED-TR) Yield Ceilings Y(C)

Rice 97 70 155

Wheat 71 22 225

Com 72 84 154

Sorghum 89 184 308

Millet -100 71 142

Barley 70 50 170

Cereals. Total 80 54 /70

Pulses 92 /3/ 472

Roats & tubers 95 /48 258

Soybean 58 24 64

Peanuts 96 86 208

NOTE: Y(C) = average yield (1991-93) for the nation with the highest averagc yield; nation must harvest at least 5,000 ha in 1991-93Y(TR) = avcrage yicld for nations whose economies arc in transitionY(DING) = average yield for developIng nationsY(DPED- TR) = avcrage yield for developcd nations cxcludlng nations whose economies are in transition

120 I.M. GokJan.\

to ripen later (as in the Flavr Savr tomato), biotech-nology can help reduce post-harvest and end-uselosses, estimated at about 47% of global calorieconsumption (Bender 1994).

These yield gaps can be bridged by using moreinputs, particularly fertilizer, in developing andtransition countries; and modifying HYV s so theyare better adapted to specific locations around theworld which would increase their adoption rates.Furthermore, developing location-specific inte-grated nutrient management systems; making wateruse more efficient; and improving extension ser-vices would be useful. There is also a need to helpoptimize the timing and quantities of inputs, de-velop more effective pest control and improve feedsupplies for livestock in pastures and farms using(singly or in combination) the various techniquesmentioned previously for reducing erosion, waterlogging, and salinization, maintaining soil organicmatter, liming acidic soils, and adding micronutri-ents to utilize the full potential ofHYVs.

FUTURE PROSPECTS: THE TRADE-OFFBETWEEN LAND CONVERSION ANDLAND PRODUCTIVITY

Raising the yield ceiling

Maximum theoretical yields [Y(MAX)] exceedwhat can be achieved by even the best fanners. Solong as Y(MAX) is not reached. it should be possi-ble to increase the yield ceilings (Plucknett 1995;Oram and Hojjati 1995). One estimate of thetheoretical maximum global cereal yield for rain-fedagriculture placed it at 13.3 T (grain equivalents )/ha(Plucknett 1995). By comparison. the averageglobal cereal yield in 1992-94 was 2.77 T/ha rang-ing from 0.75 T/ha for millet to 3.89 T/ha for com(FAD 1996a). Yield ceilings may be raised throughcontinued emphasis on plant breeding-bothlocally and professionally. more intensive manage-ment of soils. and development of location- specificintegrated nutrient and pest management systems.

Biotechnology will increase ceilings by. forinstance, allowing breeders to confer resistance tostresses (e.g., droughts, freezes, salinity. pests, andherbicides) and by regulating flowering (Coupland1995). The livestock sector can be made moreefficient by producing growth promotants whichcan improve feed efficiencies, increasing reproduc-tive success, and reducing losses from diseases(aT A 1994). It can also be used to develop plantand livestock breeds with favored characteristics(e.g., plants that produce more oil or starch, ripenlater, or look better; meats with less fat). More-over, by conferring fruits and vegetables the abil ity

The above discussion suggests that it should bepossible for the U.S. to meet domestic food needsrelatively easily in 2050. Global food security,however, is more problematic. In theory, it shouldalso be possible-at an economic and environmen-tal price and through a combination of measuresincreasing cropland and overall productivity-toaugment food supplies by 121 % by 2050 (relativeto 1993 levels) which, as noted, would expand percapita food supplies between 10% and 42% de-pending on whether population doubles or hews tothe World Bank's low fertility projection. In-creases in productivity, as used here, includesanything that augments the amount offood used byeventual consumers without putting additional landinto agriculture. Productivity may be boosted byincreasing yields and cropping intensities, usingfeed more efficiently, increasing pre- and post-useshelf life, or otherwise decreasing post-harvest andend-use losses. Several different scenarios can beconstructed to achieve this. For instance, croplandcould be increased from the 1993 level of 1,448 Mhaby 121% to 3,201 Mha (assuming average produc-tivity is unchanged). Alternatively, in lieu of themassive habitat loss implicit in that scenario (1,753Mha of new cropland), while cropland is keptconstant, productivity could increase 121 % (or anaverage of 1.4% per year for 57 years from 1993 to2050). Thus, there is a trade-off between increasedproductivity and increased habitat loss. Figure 4illustrates this trade-off for average annual increasesin productivity between 1993 and 2050 rangingfrom zero to 2% per year, and the correspondingcumulative increases in productivity.

This trade-off is sensitive to even small changes inthe rate of productivity increase. Thus, acceleratingthe average annual productivity increase from 1.0%to 1.1% (or 76% over 57 years to 87%) would, in2050, reduce net habitat conversion by 100 Mha.

Meeting global food needs: the environmental trade-offs121

2000 240

~1500~0In0N

~1000Q)(/)IVQ)'-(.)C

~ 500cIVc.0

0~ az

180 :;?:~0It)0N>-

120.c c:Om(!)

.2:-:~

60 't):J"00'-a..m"0

0 I-

-net increase in cropland

total productivity growth ("!o)

ASSUMPnONS:a) 2050 population is 96 billIOnb) Cereal supplies per capitaincrease at the 1969n1 to 1989/91rate

-500 -60

0 0.5 1 1.5Annual Productivity Increase (%)

2

Fig 4. Trade-off between productivity growth and habitat loss net conversion of land to cropland from 1993 to 2050.

Productivity increases at an annual rate of 1.0% to1.5% for 1993-2050 are plausible. First, they arewithin the range of historical experience. Between1969- 71 and 1979-81, yields alone, calculated fromchanges in the crop production indices and croplandareas (F AO 1995), increased at an annual rate of2.1%~between 1979-81 and 1991-93, they increased2.0%. They are also below Alexandratos' (1995a)projection of an annual increase of 1.8% in cropproduction for 1990-20 I 0, preponderantly due toincreases in productivity rather than cropland.Second, and more importantly, the correspondingcumulative improvements from 1993-2050 would bebetween 76% and 134% (Fig. 4), which is realizablewith relatively modest technological changes,considering the increases possible if current yield

gaps were closed (Table 4), and considering thepotential for raising yield ceilings and reducingdownstream losses, provided current constraints toincreasing production can be reduced (see below).

The annual 1.0% to 1.5% increase also translatesinto a change in total cropland by 2050 ranging froman increase of 368 Mha to a decline of 77 Mha.Clearly, the precise combination of productivityincreases and new cropland employed to meet futurefood needs will be critical for the globe's biologicalhealth.

To the extent productivity improvements comefrom increased inputs (e.g. fertilizers, pesticides,and water), that will, unless mitigated, increase oneset of environmental and public health-relatedimpacts. For fertilizers and pesticides, these impacts

122 I.M. GokJan~

compared to the impacts of ferti I izers and pesticides.particularly if they are used efficiently and cau-tiously. Second, given time, the effects offertilizersand pesticides seem to be reversible, albeit at somecost and with great difficulty. For instance, thelevels of DDT and other organochlorine residueshave declined in fish, human, and animal tissue inthe U.S. and other OECD nations-in some cases bymore than an order of magnitude, e.g., in Canadaand the Netherlands for some forms of DDT inhuman adipose tissue (Gunnarson et al. 1995;

Goklany 1994, 1998a). Consequently, populationsof several avian species (e.g., bald eagles andperegrine falcons) in the u.S. which were affectedby these residues are now recovering (Goklany et al.1992). Similarly, nitrogen loading in the streamsand water have been reduced and, at long last,dissolved oxygen levels are now increasing. Thenumber of species in some of the major rivers in thedeveloped world are also recovering. For instance,the number of species in the Rhine River which haddeclined to 27 in 1971 had increased to 97 in 1987

(CEC 1992).On the other hand, because of diminishing returns

to yield at higher levels of pesticide and fertilizerusage, it may we" be environmentally unsound toattempt to squeeze out as much productivity aspossible for every hectare of cropland.

The situation is at least as complex if the input inquestion is water: total loss of water can be just asdrastic as loss of habitat. However, in makingchoices regarding such a trade-off it is important torecall that the yield of one irrigated hectare is, onaverage, equivalent to 2.5 un irrigated hectares.Moreover, diversion of water for agriculture doesnot have to be an all-or-nothing enterprise. Also,the adverse effects of water diversions can bemitigated, if not eliminated, by returning or reusingsome of the water or adjusting timings of flows.

Finally, because the choice among the numerousscenarios illustrated in Fig. 4 involves trade-offs,there is an optimization problem which, because ofthe uncertainties surrounding it, is not accuratelysolvable today, though, prima facie, habitat losswould seem to be overall worse than greater use ofinputs-particularly if the latter are used efficiently,with due caution and below the point of diminishingreturns. These uncertainties suggest a researchagenda to help increase efficiency of input use, and~

include surface and ground water contamination,eutrophication, oxygen depletion, build-up of pesti-cide residues in fish and avian species, and inhuman tissue. Where water is the input, diverting itcould have drastic consequences for other existingin-stream uses and for species dependent on thatwater. These environmental impacts must beweighed against the effects of habitat loss that willinevitably result if those inputs are foregone

(Goklany 1998a).Knutson et al. (1990) estimated that foregoing use

of pesticides and inorganic nitrogen fertilizer in theU.S. would reduce U.S. yields for soybean, wheat,com, cotton, rice, and peanuts by 37, 38,53,62,63and 78 percent, respectively. Thus, to make up allthis lost production, the amount of land cultivatedwould have to be increased at least 170% (forsoybean) to 350% (for peanuts). However, it wouldbe uneconomic to do so. Knutson et al. estimatedthat because of higher prices, cultivated area would,in fact, increase only 10%. Nevertheless, because ofthe production shortfall, prices for wheat would stillincrease 24%, while those for other crops would bedoubled (or more). They estimated the lower 20%of the population's outlays on food would increasefrom 38% to 44%. U.S. grain exports would decline50% by volume, and stock carry-overs would bereduced between 42% (for wheat) to 83% (for com).While Knutson et al. did not estimate impacts onglobal food security, clearly the ability of poorernations and peoples to import and purchase foodwould be significantly-and adversely-affected.

Similarly, if pesticide use were eliminated,cropland devoted to fruits and vegetables wouldhave to be increased by 33% to 150% to compensatefor any lost production, based upon Taylor's (1995)estimates of declines in yields of25% to 60%. Theresulting increases in costs would also reduce fruitand vegetable consumption which could itself havenegative impacts on nutrition and public health

(NRC 1996).Thus. the trade-off between increasing productiv-

ity and increasing habitat loss indicated in Fig. 4also involves a trade-off between the environmentaland public health impacts of increased inputsagainst those of habitat loss. Either could lead toreductions in biological diversity. However, itwould seem that the two sets of impacts are notquite equivalent. Habitat loss is much more final

Meeting global food needs: the envIronmental trade-offs 123

make the educated trade-offs necessary to meetfuture global food needs while limiting net environ-mental impacts.

REALIZING THE WORLD'S FOODPOTENTIAL WHILE LIMITING THEENVIRONMENTAL PRICE: POLICYRECOMMENDATIONS

The above discussion suggests that it ought to bepossible to meet the world's food needs in 2050even if the population doubles, but at an environ-mental price. Whichever route is taken to meetfuture global food needs, whether or not it is repre-sented by one of the several scenarios of Fig. 4,underlying each is a basic set of assumptions,namely: ]) there will be sufficient capital to obtainthe inputs, implement technologies needed tomaintain or increase productivity, and to bring anyneeded new cropland into production; 2) R&D,particularly for the scenarios with the higher pro-ductivity increases, will continue to bring some newtechnologies on line and adapt technolo-gies--current and new-to local conditions aroundthe world; 3) knowledge of these adaptations willcontinue to be transferred to farm enterprises, and;4) infrastructure and the trading system will expand,as necessary, to ensure that surpluses are moved toareas of shortage. Therefore, at a minimum, poli-cies are needed that will convert these underlyingassumptions into reality, while simultaneouslyreducing or mitigating any resulting environmentalimpact and minimizing disruption to habitat, for-ests, and biodiversity.

First, complacency because of past successesshould be avoided; and the national, regional, andinternational systems for agricultural research, andthe extension services that transfer technology tofarm enterprises should continue to be vigorously

supported.Second, R&D and extension services should

focus on optimizing location-specific methods ofmaintaining and increasing productivity (includingclosing yield gaps, reducing post-harvest andend-use losses) and reducing or mitigating environ-mental consequences of agricultural technologies,per previous discussions. Such programs shouldinclude developing methods to reduce the amountof inputs that are applied or mobilized into the

environment by, e.g., controlling the timing andratios of inputs so as not to over-apply one inputwhen another may be the limiting factor. In addi-tion to increasing net production, they would alsohelp make agricultural practices more economicallyefficient. Greater effort should also be made todevelop drought- and salinity-resistant crops andcultivars. Such research, by increasing agriculturaloptions, will also help the world adapt to globalchange, whether or not the agents of change arehuman-induced (Goklany 1992, 1995a).

A research program is also needed to evaluate therelative public health and environmental risks andsocial and economic consequences of the differentmethods for increasing net production. Such infor-mation should be incorporated into policy-relatedexercises (e.g., cost-benefit or environmental impactassessments) related to fertilizer and pesticide use,and water and other projects affecting agriculture.Current assessments tend to ignore the effects onland conversion and broad impacts on nutrition andpublic health on various segments of society (Pimentelet al. 1994b, 1995).

Third, the public, environmentalists and policymakers need to be educated about the trade-offsinvolved in feeding the world's billions. The aware-ness of agriculture's role in polluting and divertingwaters, and their environmental impacts, particularlyin richer, developed nations-while necessary togenerate pressure for much-needed improvements infertilizer and pesticide use, water projects, andenvironmental cleanup-unfortunately is not matchedby an appreciation of technology's role, discussedabove, in reducing habitat loss while improving foodsupplies, nutrition, and public health for today'sbillions. Without this, more habitat loss, increasedpollution, less but costlier food, and increasedvulnerability of the poor to malnourishment andhunger could result, short-changing both humanityand the rest of nature. Education is also needed tohelp address "new source bias" evident in thereactions of various segments of the public tobiotechnology-derived products (e.g., the Flavr Savrtomato or milk produced from BST -treated cows)despite their potential role in reducing other inputsand habitat loss (Goklany and Sprague 1991).

Education may also help ensure that there issupport in the richer nations for continued produc-tion of crop surpluses, as dictated by the market

M. Goklan)124

non-producers. Barriers to free, unsubsidized tradeshould be further reduced. Also, removal of agricul-tural subsidies in the richer nations, in particular,would benefit their economies and environment aswell as accelerate economic growth in developingnations, in part, because that would make the latter'sagricultural sectors more competitive (Goklany1995a). Moreover, trading systems allowing thefree movement and repatriation of capital (andprofits) across borders may help generate the capitaldeveloping nations will need to expand cropland andobtain and employ both new and underused existingtechnologies (Goklany 1998a). Despite their imper-fections, the conclusions of the Uruguay Round ofGATT(General Agreement on Trade and Tarrifs)and NAFT A (the North American Free TradeAgreement) negotiations bode well on these scores.

Also, trade-both external and internal-militatesagainst the strife and disruption which have beenresponsible for much of the hunger and malnutritionover the past few decades. It is particularly difficultfor groups involved in trading the most basic ofcommodities, food, to be engaged in hostilities. Forthe seller, it makes little sense to attack and impov-erish one's customer; while for the buyer, thatwould be tantamount to biting the hand that feeds it.The recommendation on trade implies the abandon-ment of the notion of national self-sufficiency infood production. Post-war statistics on food tradeand aid indicate that this goal is unrealistic. More-over, as noted, too often this goal has been used tojustify economically and environmentally unsoundagricultural subsidies. Social and political institu-tions should strive to reconcile themselves to thereality that trade has globalized sustainability

(Goklany 1995a).Economic growth will, as noted, increase the

incentives for smaller families, thereby helpingsecure the population growth trajectory implied inthe World Bank's standard projection, i.e., virtualpopulation stabilization in the 22nd century. Fi-nally, economic growth may reduce pro-natalistsentiments among policy makers as the world'spopulation becomes older which will increasepressures on socia! safety nets (Goklany 1995a).

Fifth, the agricultural sector needs to strengthenits ability to compete for land and water. Increas-ingly, as economies grow, the shrinking ofagricul-ture's share of GDP and total employment will

place (as opposed to subsidizing overproduction).Otherwise, richer nations-untroubled by foodshortages; safe in the knowledge that they canalways purchase, if not produce, food; motivated bythe desire for a cleaner environment and a balancedbudget and, possibly, some new source bias-maydiscourage the production of surpluses, reduceinvestments in agricultural R&D at home andabroad, and decrease food aid. These are essentialfor global food security today, and into the foresee-able future.

Fourth, the institutions responsible for theco-evolution of technological change and economicgrowth should be bolstered. Such support cannotautomatically be assumed even in richer nations.Some transition and developing nations may wellwaver from, if not abandon, the path to long-termeconomic growth. These institutions include marketeconomies, freer trade, conferring and enforcingproperty rights for land, water and innovations, andrequirements for risk analyses to address trade-offsand "new source bias" which support entrenchedinterests or old technologies at the expense ofcompetitors. These will help create new technolo-gies, as well as the necessary capital and otherincentives for adopting new or unused technologiesto raise or maintain yields, conserve soil and water,and control pollution resulting from yield-enhancingtechnologies. As noted previously, the richestnations, in general, have the best environmental

quality.Equally importantly, these institutions would

ensure that farm enterprises are fairly compensatedfor their investments in capital, labor, and technol-ogy, eliminating a major cause of the relatively poorperformance in the agricultural sector in manydeveloping nations. Implicit in this is also thereduction. if not total elimination, of agriculturalsubsidies. This should be accomplished gradually.over five to 1 0 years, so that producers andnon-producers can adjust to a new, subsidy-freeenvironment, without abandoning insurance andother safeguards against catastrophic crop failures.Several strides have been taken in this regard inOECD-and, out of necessity, transition-nations.The passage in the U.S. of the "freedom to farmAct" was a giant step in this direction.

Moreover, as noted, economic growth and tradeincrease food security of both producers and

Meeting global food needs: the environmental trade-offs

about, once again, additional inputs-<:apital andR&D-will be required, and environmental impactson seas may have to be traded-off for additionalhabitat loss and environmental degradation on land.Perhaps if more time is spent in the water, fewerfootprints will be left on the land.

CONCLUSION

Despite a sextupling of population, the foodsituation around the world has improved markedlyover the last two centuries, due primarily to theforces of science-based and market-driven economicgrowth, technological change, and trade. Comparedto the dismal situation immediately following WorldWar II, today famine and malnourishment are downboth in absolute and relative terms; people aregenerally better fed, and they spend much less timeand effort getting food to the table. Nevertheless,hunger and malnutrition, while reduced, has notbeen banished. Most vulnerable to shortfalls arepopulations or areas with low purchasing power;civil strife or dysfunctional relations with the exter-nal world (e.g., Iraq and N. Korea); or some combi-nation thereof. While today's improvements havecome at an environmental price, mercifully thewholesale destruction of forests, habitats, andbiodiversity, which would inevitably have occurredabsent technological change, has been avoided.Since 1961 alone, technology has forestalled theconversion of at least 966 Mha to cropland world-wide (370 Mha in the U.S. since 1910). Another2,580 Mha would have been converted to otheragricultural uses worldwide. Indirectly, and despiteits poor reputation among many environmentalists,technological change is responsible for conservingmore habitat than any other conservation measure.

With respect to the future to 2050, the richestnations should have few problems. As a group, theyare currently running large surpluses. Their popula-tions have, more or less, stabilized, with the notableexception of the U.S., whose population continuesto expand fueled, in part, by immigration. Neverthe-less, it should be able to meet its own foodneeds-even if its population doubles. It has suffi-cient land, water, and the capital to deploy new andexisting technologies, and if it did not, it couldpurchase food security with its wealth-as can otherrich nations.

reduce its economic and political clout. Thus,agricultural interests will be well served if theyestablish property rights to water and land soonerrather than later. Care should be taken that assubsidies' for food crops are reduced, they are notreplaced by subsidies for non-food uses of crop andagricultural land. One can well imagine futureconstituencies for subsidizing carbon sequestrationand fuel fanns similar to the one for ethanol today.

Sixth, whether or not food prices decline anddespite economic growth, there will be some peopleunable to afford an adequate diet. This can andought to be addressed without creating disincentivesfor producers by providing transfer payments, foodstamps and/or food banks for the truly needy.Again, a wealthier society will be more able toafford-and less likely to begrudge-such pro-grams.

Finally, to relieve the pressure on the land, theoceans may have to be utilized more. Most of theworld's food needs are currently supplied by thephotosynthetic product of30% of the world's area,which is land. It would be sensible to obtain morefrom the rest of the globe-but in a manner quitedifferent from what has been used to date. By andlarge, human beings have treated the ocean's bountyas our hunter-gatherer forebears treated the land'sbounty [Serageldin, quoted in CGIAR (1995)]. Thishunter-gatherer mental ity creates havoc, particularlywhen technology is pressed into service for harvest-ing but not for increasing the population of theharvested species. The demise or near-demise ofspecies such as the beaver, American buffalo andvarious whale species are cases in point. In fact,only after conscious and more intensive manage-ment of the land (i.e., agriculture) started to replacehunting and gathering was it possible for the earth'shuman-carrying capacity to go from a few to hun-dreds of millions (Livi-Bacci 1992). In order toallow the billions of humans to co-exist with the restof nature, the realm of conscious management mayhave to be expanded to a portion of the seas. Onthis score, it is encouraging to note that aquaculture(though some of it is land-based or in-shore) is oneof the fastest expanding sections within the food andagriculture sector, growing (for example) 12% justbetween 1993 and 1994 with the growth being ledby the most populous nation, China, and otherdeveloping nations (F AO I 996c). But to bring this~

I.M. GokJal1Y126

(globally) to new cropland or reducing cropland by77 Mha. However, to the extent productivity in-creases are due to additional fertilizer and pesticideuse or water diversions, their environmental impactsmay partially offset the benefits of reducing habitatloss. Given the rather drastic nature of habitatconversion, perhaps the most prudent approach wouldbe to focus on increasing productivity while usinginputs efficiently and cautiously and simultaneouslyworking to mitigate their impacts.

To help assure that the world's food security needsare met without displacing the rest of nature orinflicting irreparable environmental harm, a long-term commitment is needed toward strengthening thevery forces responsible for the current victory, how-ever incomplete or tarnished, over hunger and malnu-trition. Specifically:I. Economic growth is essential. It is needed togenerate the fiscal resources needed for the adoptionand application of new or unused existing technolo-gies to maintain or enhance productivity; to reduce ormitigate environmental effects due to agriculturaltechnology; to bring any new cropland into produc-tion; and to develop the infrastructure for integratingthat cropland into the existing food and agriculturalsystem. Economic growth also helps ensure politicalsupport and funding for social safety nets, as well asR&D, extension services, and food aid. It will aug-ment the purchasing power for non- or low-producersso they can better afford food. Finally, it also helpscreate the conditions for families to voluntarily limittheir sizes, possibly leading to eventual populationstabilization.2. Technological change is needed. Without it,productivity of the land cannot be maintained orincreased, while simultaneously keeping environ-mental impacts in check. This means support forR&D and extension services for agriculture, food,and the environment, embedded within a largereffort to bolster the institutions that stimulate suchchange. This includes greater emphasis on educa-tion in general and science education in particular,protection of intellectual property rights, eliminationof new source bias, and favorable tax treatment forR&D expenditures. The larger effort will helpdevelop a culture conducive to technologicalchange. It is also necessary because the course ofscientific and technological progress is unpredict-able; advances in one field often spread to others

Future food security is more uncertain for the leastdeveloped nations. Despite recent declines, they stillhave the highest population growth rates. Moreover,because of poverty and low incomes, they have lessability to adopt productivity-enhancing technologiesor to purchase food security in the market place.

This paper explored whether global food suppliesin 2050 could be augmented by 120% over 1993levels, an amount sufficient to increase availablegrain supplies per capita even if population were todouble which, under the World Bank's high-endestimate, would occur in the last quarter of the nextcentury. It is, indeed, mathematically possible tomore than double global food supplies by 2050through various combinations of increased croplandand diverse methods of raising productivity. First,the amount of potential rainfed cropland is more than2.3 times what was used in 1993. Second, severaltechniques are available to maintain and increaseproductivity on existing croplands. If current gapsbetween average yields and yield ceilings are erased,global production of virtually all major crops woulddouble without adding a single hectare to netcropland (Table 4). Third, productivity can also beraised by increasing cropping intensity, raising yieldceilings, and decreasing post-harvest and end-uselosses. Fourth, supplies of water-perhaps the mostcritical of all inputs-can, effectively, be stretchedthrough institutional changes which would treat it asan economic commodity and allow trading. Thatwould provide everyone the incentives needed toapply existing. and stimulate R&D of new, technolo-gies to conserve, reuse or increase water supplies.

However, it is not enough to produce sufficientfood. Food security also means ensuring that non-or low-producers have economic and ready access tofood. Moreover, the world's food potential shouldbe realized without undue environmental impacts,with the least disruption to habitat, forests, andbiodiversity .

The precise combination of increased croplandand increased productivity-and the mix of produc-tivity-enhancement measures-used to meet futurefood needs will be crucial for the earth's futurebiological diversity. For instance, average annualincreases in productivity of 1 % versus 1.5% be-tween 1993 and 2050-both plausible given theopportunities to raise productivity-means thedifference between converting 368 Mha of habitat

127Meeting global food needs: the environmental trade-offs

Twenty-First Centuf)': Meeting Future Food Demand of anIncreasing Population. Washington. DC: Int. Food Policy Res.Inst.; 1995b: 25-48.

Batie. S.S.; Healy; R.M. The future of American agriculture.Sci. Am. 248: 45-53; 1983.

Bender. W.H. An end use analysis of global food requirements.Food Policy 19: 381-395; 1994.

Brown, L.R.; Kane, H. Full house: Reassessing the world'spopulation carrying capacity. New York: W. W. Norton; 1994.

Bureau of the Census. Statistical abstract of the United States1995. Washington. DC: U.S. Bureau of the Census; 1995.

Bureau of the Census. Statistical abstract of the United States1993. Washington, DC: U.S. Bureau of the Census; 1993.

Bureau of the Census. Historical statistics of the United States,Colonial times to 1970. Washington, DC: U.S. Bureau of theCensus; 1975.

BLS (Bureau of Labor Statistics). CPI Detailed Repons.Various January issues. Washington, DC: Bureau of LaborStatistics.

CEC (Commission of the European Communities). The state ofthe environment in the European community: Overview.volume \II. Brussels: Comm. Eur. Communities; 1992.

Cervinka.. V. Water use in agriculture. In: Pimentel, D.; Hall,C. W., eds. Food and Natural Resources. San Diego: Acad.Press; 19&9: 141-162.

(e.g., consider the importance of computer chips orsynthetic materials in and out of the food and

agricultural sector).3. Freer, and unsubsidized, trade is critical to ensureefficient movement of food surpluses and capitalacross borders. Developing nations will need toimport more food, as well as capital to open up newcropland or afford modern technologies. Moreover,to pay for food imports, they will need exports (andgrowth) in other economic sectors. Nations shouldeschew the economically and environmentally coun-terproductive notion of food self-sufficiency. Trade,particularly in foodstuff, also gives nations anotherincentive to live in harmony.

Strengthening these interdependent forces will alsoenhance the world's adaptability to global environ-mental change, regardless of its cause (Goklany1995a). In addition, to relieve humanity's burden onthe land-and the effects of technological harvest-ing-a concerted effort may be needed to developsustainable methods of enhancing the oceans' pro-ductivity. Moreover, it is crucial that developednations continue to produce crop surpluses. Conceiv-ably, their desire to maintain and enhance their ownenvironmental quality may lead to policies effectivelyreducing such surpluses. Finally, there has to beacceptance of, and the ability to make, trade-offsbetween imperfect outcomes.

Because they would harness human nature ratherthan seek to change it, the above measures, while nopanacea, are more likely to be successful than ferventand well-meaning calls, often unaccompanied by anypractical program, to reduce populations, changediets or lifestyles, or embrace ascetism. Heroes andsaints may be able to transcend human nature, butfew ordinary mortals can. Sadly, vanity and thepersonal desire for a longer and healthier life havechanged many more diets than urgings to save theglobe. Holding out for perfection and rejecting theadmittedly "second-best" solutions outlined abovewould reduce quantities of, and access to. foodsupplies while increasing habitat loss and environ-mental degradation-shortchanging both humanityand the rest of nature.

Chen. R.S.; Kates, R. W. World food security: Prospect andtrends. Food Policy 19: 193-208; 1994.

Cohen. J.E. How many people can the earth support? NewYork: W.W. Norton; 1995: 532.

CGIAR (Consultative Group on International AgriculturalResearch). From hunting to farming fish-rapid productionincreases are possible. Press release. Washington, DC: Consult.Group Int. Agric. Res.; May 14. 1995.

Coupland. G. LEAFY blooms in Aspen. Nature 377: 482-483;1995.

Crosson, P. Future supplies of land and water for worldagriculture. In: Islam, N.. ed. Population and Food in the EarlyTwenty-First Century: Meeting Future Food Demand of anIncreasing Population. Washington. DC: Int. Food Policy Res.lnst.; 1995a.

REFERENCESDazhong, W. Soil erosion and conservation in China. In:Pimentel. D.. ed. World Soil Erosion and ConservationCambridge: Cambridge Univ. Press; 1993.

Towards

2010.Alexandratos. N.. ed. World agricultureChichester: John Wiley; 1995a: 163.

Dreze, J.; Sen. A.K. The political economy of hunger, volumeI. Oxford, U.K.: Clarendon Press, Oxford Univ. Press; 1990.

Alexandratos. N. The outlook for world food and agriculture toyear 2010. In: Islam. N.. ed. Population and Food in the Early

M Goklan:-128

Ehrlich. P.R. Interview on Diane Rehm show. WAMU.Washington, DC: October 9, 1995. Cassette available fromWAMU 88.5 FM, The American University. Washington. DC20016-8082.

Goklany, I.M. Saving habitat and conserving biodiversity on acrowded planet. BioScience 48: 941-953; 1998a.

Goklany.I.M. Conserving habitaL feeding humanity. Forum forApplied Research and Public Policy 13: 51-56: I 998b.

Ehrlich, P.R.; Ehrlich, A.H.; Daily, G.C. Food security,population, and environment. Popul. Dev. Rev. 19: 1-32; 1993. Goklany, I.M. Factors affecting environmental impacts: The

effects of technology on 'long term trends in cropland. airpollution and water-related diseases. Ambio. 25: 297-503: 1996.Engelman, R.; LeRoy. P. Sustaining water: Population and the

future of renewable water supplies. Washington, DC: PopuloAction Int. 9-11; 1993. Goklany ,I.M. Strategies to enhance adaptabilit}': Technological

change, sustainable growth and free trade. Climatic Change 30:427-449; 1995a.Falkenmark, M.; Biswas. A.K. Further momentum to water

issues: Comprehensive water problem assessment in the being.Ambio. 24 (6): 380-382; 1995.

Goklany, I.M. Richer is cleaner: Long term trends in global air

quality. In: Bailey, R., ed. The True State of the Planet. Ne York: Free Press; 1995b.

F AO (Food and Agricultural Organization of the UnitedNations). F AOST AT database available at http://apps.fao.org/lim500/Agri_db.pl.; 1996a.

Goklany, I.M. Air and inland surface water quality: Long tenDtrends and relationship to aft1uence. Washington, DC: Off. PolicyAnal., Dep. of the Interior; 1994.

F AO (Food and Agricultural Organization of the UnitedNations). Food outlook. June 1996. Available at http://www.f ao.org/wai cent! fao in fol econom i cl g i ewsl eng I i sh/fo/f09606Itables/fot96069.htm.; 1996b.

Goklany. I.M. Climate change and natural resources: Is it toosoon to stan adapting? Climate Change Newsletter 5 (4): 3-6:1993.

FAO (Food and Agricultural Organization of the UnitedNations). Major trends in global aquaculture production:1984-1994. Available at http://www.fao.org/WAICENT/faoinfo/fishery/aqtrend/aqtrend.htrn.; I 996c.

Goklany, I.M. Adaptation and climate change. Prepared for 1992Annual Meeting of the American Association for the Advance-ment of Science, Chicago, February 6-11; 1992. Available fromauthor, Off. Policy Anal.. Dep. of the Interior, 1849 C. St., NW,Washington. DC 20240.

FAO (Food and Agricultural Organization of the UnitedNations). 1995 Country tables. Rome: Food Agric. Org.: 1995:332.

Goklany, I.M. et al. America's biodiversity strategy: Actions toconserve species and habitats. U.S. Department of Agricultureand Department of the Interior. Washington, DC: Off. PolicyAnal., Dep. of the Interior; 1992.FAO (Food and Agricultural Organization of the United

Nations). 1994 Country tables. Rome: Food Agric. Organ.;1994a; 332. Goklany. I.M.; Sprague. M. W. Technological progress increases

food production. In: Barbour. S., ed. Hunger: Current Controver-sies. San Diego, CA: Greenhaven Press; 1995: 118-125.FAO (Food and Agricultural Organization of the United

Nations). FAO quarterly bulletin of statistics 7 (2/3/4): 71-72;1994b. Goklany. I.M.: Sprague, M. W. An alternative approach to

sustainable development. Conserving Forests. Habitat andBiological Diversity by Increasing the Efficiency and Productiv-ity of Land Utilization. Washington. DC: Off. Program Anal..U.S. Dep. of the Interior: 1991.

FAO (Food and Agricultural Organization of the UnitedNations). 1991 Country tables. Rome: Food Agric. Organ.:1991.

FAO (Food and Agricultural Organization of the UnitedNations). Yearbook of food and agricultural statistics: Produc-tion 1954 vol. VIII. part I. Rome: Food Agric. Organ.; 1954:

205-207.

Grossman, G.; Krueger, A. Environmental impacts of a NorthAmerican free trade agreement, discussion paper no. 158.Princeton. NJ: Woodrow Wilson School, Princeton Univ.; 1991.

Gunnarson, J. et al. Interactions between eutrophication andcontaminants: Toward a new research concept for the Europeanaquatic environment. Ambio. 24 (6): 382-385; 1995.

FAO (Food and Agricultural Organization of the UnitedNations). Statistical yearbook 1952. Rome: Food Agric. Organ.:1952: 264-266.

Harris, J.M. World agricultural futures: regional sustainabilityand ecological limits. Ecol. Econ. 17 (2): 95-115; 1996.Fogel. R. W. The relevance of Malthus for the study of mortality

today: Long-run influences on health. mortality. labor forceparticipation. and population growth. In: Lindahl-Kiessling. K.:Landberg. H., eds. Population. Economic Development. and theEnvironment. Oxford. U.K.: Oxford Univ. Press: 1994: 231-284.

Houghton, R.A. et al. Changes in the carbon content of terrestrialbiota and soils between 1860 and 1980: A net release of CO2 tothe atmosphere. Ecol. Monogr. 53: 235-262; 1983.

Meeting global food needs: the environmental trade-offs 129

Demand ofan Increasing Population Washmgton. DC: IFPRl:1995: 167-89.

Kellogg. R.L.; TeSelle, G. W.: Goebel, 1.1. Highlights from the1992 national resources inventory. 1. Soil Water Conserv. 49(6): 521-527: 1994.

Pimentel. D. et al. Environmental and economic costs of soilerosion and conservation benefits. Science 267: 1117-1123:1995.

Knutson, R.D.; Taylor, C.R.; Penson; 1.B.; Smith, E.G. Eco-nomic impacts of reduced chemical use. College Station, TX:K & Associates; 1990.

Pimentel. D. et al. Natural resources and an optimum humanpopulation. Populo Environ. 15: 347-369; I 994a.Kumar, B.G. Ethiopian famines 1973-1985, a case study. In:

Dreze, J.; Sen, A., eds. The Political Economy of Hunger,Volume II. Oxford, U.K.: Clarendon Press, Oxford Univ. Press;1990: 173-216.

Pimentel, D. et al. Environmental and economic impacts ofreducing U.S. agricultural pesticide use. In: Pimental. D.:Lehman, H., eds. The Pesticide Question: Environment.Economics, and Ethnics. New York: Chapman and Hall: 1993.Kuznets, S. Economic growth and income inequality. Am. Econ.

Rev. 45: 1-28; 1955.Plucknett, D.L. Prospects for meeting future food needs throughnew technology. In: Islam, N., ed. Population and Food in theEarly Twenty-First Century: Meeting Future Food Demand ofan Increasing Population. Washington. DC: Int. Food PolicyRes. Inst.; 1995: 207-219.

Livi-Bacci, M. A concise history of world population. Englishedition translated by C. Ipsen. Cambridge, MA: Blackwell;1992: 220.

Maddison. A. The world economy in the 20th century. Paris:Organ. Econ. Coop. Dev.; 1989. Poleman, ToT.; Thomas, L.T. Repon: Income and dietar).

change. Food Policy 20: 149-159; 1995.Malthus. T.R. First essay on population. 1798. In: Bonar. J.. ed.Reprinted for the Royal Economic Society. London: McMillan;1926.

Postel. S.: Daily, G.C.: Ehrlich. P.R. Human appropriation ofrenewable fresh water. Science 271 (5250): 785-788; 1996.

McEvedy, c.: Jones, R. Atlas of world population history. NewYork: Penguin; 1978: 342.

Richards. I.F. Land transformation. In: Turner, B.L., II et al.,eds. The Earth as Transformed by Human Action: Global andRegional Changes in the Biosphere Over the Past 300 Years.Cambridge: Cambridge Univ. Press; 1990: 164.McLaughlin. L. A case study in Dingxi County, Gansu Prov-

ince. China. In: Pimentel. D.. ed. World Soil Erosion andConservation. Cambridge: Univ. Press; 1993: 87-107.

Rosegrant. M W.: Schleyer. R.G.: Yadav. S.N. Water policy forefficient agricultural diversification: Market-based approaches.Food Policy 20: 203-223: 1995.

Mitchell. D.O.; Ingco. M.D. Global and regional food demandand supply prospects. Chapter 4. In: Islam. N.. ed. Populationand Food in the Early Twenty-First Century: Meeting FutureFood Demand of an Increasing Population. Washington. DC:Int. Food Policy Res. Inst; 1995.

Rozanov, B.: Targulian, V.: Orlov, D.S. Soils. In: Turner,B.L"II et al., eds. The Earth as Transformed by Human Action:Global and Regional Changes in the Biosphere Over the Past300 Years. Cambridge: Cambridge Univ. Press: 1990: 203-214.Mitchell. D.O.~ 1ngco. M.D. The world food outlook. Hunger

Notes 19: 20-25: 1993Sen. A.K. Poverty and famine. An essay on entitlement anddeprivation. Oxford. U.K.: Oxford Univ. Press; 1981.NRC (National Research Council). Excess calories pose more of

a cancer threat than natural or synthetic carcinogens on food.News from the NRC. Washington. DC: Nat!. Res. Counc.;February 15. 1996.

Sen, A.K. The economics of life and death. Sci. Am. 268 (5)40-47; 1993.

Shafik. N.: Bandyopadhyay. S. Economic growth and environ-mental quality: Time series and cross-country evidence. PolicyResearch W orki ng Papers. Washington. DC: World Bank; 1992.

NRCS (National Resources Conservation Service). Summaryrepon: 1992 national resources inventory. Natural ResourcesConservation Service. Ames. 1A: Iowa State Univ. Statistica!Lab.: 1995: 3.19-21

Smil, V. How many people can the earth feed? Popul. Dev. Rev20: 255-292; 1994.OTA (Office of Technology Assessment). A new technological

era tor American agriculture Washington. DC: OtT. Technol.Assess.: 1994: 452. Swarninathan, M.S. Agriculture and food industry in the 21st

century. Presented at the Resource Use and ManagementSubgroup. Intergovernmental Panel on Climate Change,Geneva: October 30, 1989. Available from IntergovernmentalPanel on Climate Change. Geneva, Switzerland.

Dram, P.A.: Hoiiati. B. The growth potential of existingagricultural technology. In: Islam. N.. ed Population and Foodin the Early Twenty-First Century: Meeting Future Food

130 I.M. GokJany

Tapsoba, E.K. Food security policy issues in West Africa: Pastlessons and future prospects. A critical review. FAO Eco-nomic and Social Development Paper 93. Rome: Food Agric.Organ.; 1990.

in the United States. analysis of conditions and trends. Wash-ington, DC: U.S. Dep. A~ic.; 1989: 7.

Vitousek. P.M.: Ehrlich, P.R.: Ehrlich. A.H.: Matson. P.A.Human appropriation of the products of photosynthesis.BioScience 36: 368-73; 1986.

Taylor, C.R. Economic impacts and environmental and foodsafety tradeotTs of pesticide use reduction on fruit and vegeta-bles. ES 95.1. Auburn, AL: Auburn Univ.; 1995.

Waggoner, P .E. How much land can ten billion people spare fornature? Ames. IA: Counc. Agric. Sci. Technol.: 1994: 26-27.

The Economist. Sunshine and showers. October 21. 1995: 84. World Bank. The pink sheet. Commodity Price Data. Juneissue, available at http://www.woridbank.org/htmi/ieccp/pink.html; 1996.Tumer, B.L. II et al., eds. The earth as transformed by human

action: Global and regional changes in the biosphere over thepast 300 years. Cambridge: Cambridge Univ. Press; 1990.

UNFP A (United Nations Population Fund). The state of theworld population 1995. New York, NY: United Nations PopuloFund; 1995: 65.

World Bank. World population projections 1994-1995. Wash-ington. D.C.: World Bank; 1994: 13.

World Bank. World development report 1984. New York. NYOxford Univ. Press; 1984.

WRI (World Resources Institute). World resources 1996-97.New York, NY: Oxford Univ. Press; 1996.USDA (U.S. Department of Agriculture). Agricultural statis-

tics 1995-96. U.S. Department of Agriculture, NationalAgricultural Statistics Service. Washington, DC: U.S. Gov.Print. Off.; 1996: IX-16.

WRI (World Resources Institute). World resources 1987. NewYork, NY: Basic Books; 1987.

Wrigley, E.A.; Schonfeld, R.S. The population history ofEngland 1541-1871: A reconstruction. Cambridge, MA:Harvard Uniy. Press; 1981: 529.

USDA (U.S. Department of Agriculture). The second RCAappraisal: Soil, water, and related resources on non federal lands