Acknowledgment

We express our gratitude to Dr. George Blyn, professor of economics at Rutgers, the State University of New Jersey, for assistance in calculating the 1961 data for India.

XI.B.2. Crop Pattern

In an area as climatically varied as South Asia (see plate I.C.1), a diversity of crop patterns is inevitable. Atlas plate XI.B.2 provides in two maps a synoptic view of that diversity, utilizing a variety of district-level data³ on acreage sown and value of crops grown.

Map (a) depicts by colored bars all crops accounting for more than 25% of the total (gross) cropped acreage of each district as of 1961 or, in those relatively rare instances in which no such crop exists, the leading single crop in acreage terms. The widths of the vertical bars representing the various crops indicate the proportion of each crop shown to the total in four class ranges. The major crop regions of the map, bounded by heavy lines, are those throughout which a single crop is dominant (e.g., wheat over much the greater part of Paki- stan and adjacent portions of India). Within the regions so bounded, subregions are delimited according to the specific crop combinations when major second- and third-ranking crops (those with acreages above 25% of the district totals) are also considered.

Far more often than not, particularly within the areas where rice predominates, there is only one crop with as much as 25% of the total acreage. More than any other crop, rice tends to occupy a major fraction of the total cropped area, character- istically more than half, and over a large area of the East Paki- stan (now Bangladesh), West Bengal, and neighboring dis- tricts, the proportion exceeds 75%. Among the other leading crops only wheat and jowar, each in a handful of districts, make up more than three-fourths the cropped acreage. The reason for the strikingly high proportions of rice in the total acreage of those areas wet enough or well enough irrigated to grow it is not far to seek. In yield, measured in calories per acre, rice has no close competitor. This accounts, more than any other single factor, for the exceptionally high population densities of the principal rice-growing areas, as may be seen by comparing map (a) with the various density maps por- trayed on plates XI.A.1, 2, and 8. In drier areas, where rice is of little or no importance, some other staple grain almost al- ways predominates: wheat on the alluvial soils of the north- west; bajra (a hard millet) on the less fertile soils of the Thar Desert and adjacent tracts; jowar (a heat- and drought-resistant sorghum) over much of peninsular India, particularly on the rich regur soils derived from the Deccan lava flows; maize in such mountainous tracts as the Aravalli Mountains and the central Himalayan region. Whatever the specific leading grain may be, a matter depending mainly on the physical environ- ment, typically the main object is to maximize caloric yield per acre—that is, to sustain the basic alimentary needs of the population. Quantity of food produced, more than quality or nutritional balance, governs the choice of crops. Where a grain is not the leading crop, some staple leguminous crop, for ex- ample, gram in north central India or groundnuts in parts of the south, commonly predominates. Nonfood crops, such as cotton or plantation crops, rarely rank first in acreage.

Map (b) relates to the value per male agriculturalist of a number of crops that characteristically are entirely or almost entirely sold off the farm. (For purposes of this presentation it is assumed that the entire crop is sold.) These crops are offi- cially designated in statistical reports as "cash crops" as distinct from "food crops" such as the staple grains with which map (a) principally dealt. The distinction is of course somewhat arbitrary, since much of the total production of food crops is sold for cash, while, on the other hand, a portion of the so- called cash crops is eaten (usually after being sold). It would, however, be exceedingly difficult, if not impossible, utilizing the available published data, to arrive at reliable district esti- mates for the major grains of the value per male agriculturalist of the portion of the total yields sold off the farm. For this reason our presentation is limited to the nine crops specified in the map legend. A few additional crops, such as chillies, pan (betel leaf), mangoes, and potatoes would also qualify locally for inclusion; but published data on those crops are inadequate for our use.

The manner of construction of map (b) was fairly com- plex and entailed consulting a variety of government publica- tions cited in the sources following this text. First, crop-by- crop yield data, by districts, were obtained wherever possible; where such data were unavailable yields were estimated by using published acreage data and multiplying them by regional data on average yields per acre. Second, data were obtained on prevailing wholesale or mandi (agricultural market) prices for each crop, normally by state in India and by province in Pakistan. Third, the number of male agriculturalists reported for each district was multiplied by twenty-five and by one hun- ³ Data are by provinces in Ceylon and groupings of census re- gions in Nepal. dred to derive threshold figures, in rupees, above which the total cash sale value of a particular crop in a given district was deemed important enough to plot on our map, either in a solid tone without stippling (in the range Rs. 25–100/male agricul- turalist) or in a solid tone with stippling (in the range above Rs. 100). Fourth, the known or estimated yield data for spe- cific crops were multiplied by the respective wholesale prices, wherever it appeared likely that the minimum threshold figure might be exceeded. Fifth, those districts in which, for any of the nine crops considered, either of the two value thresholds was exceeded were plotted on our map; where more than one such crop was present they were arranged in rank order of value, the second- or lower-order crops being indicated by a letter abbreviation (within a circle for those in the over-Rs. 100 range).

Surveying the mapped results of the procedure outlined above, one is struck at once by the large portion of South Asia over which not a single one of our nine crops is significant enough in value to be plotted, despite the very low threshold selected. (During the reference years of the map the official exchange rate of the Indian, Pakistani, and Ceylonese rupee was 4.76 to the dollar). The area in question, however, is de- ceptive, since the greater part of it is either desert, mountain- ous, and/or largely tribal in its population makeup. Further, one notes that over approximately half of the area in which cash crops are deemed to be significant, only one such crop is represented; and that over less than half of the area with one or more significant cash crops does the value of any crop ex- ceed the Rs. 100 threshold. Finally, in only twenty districts (counting one province in Ceylon) are there as many as three significant cash crops, and in only two are there four or five; with respect to cash crops in the value range over Rs. 100/male agriculturist only nineteen districts (or provinces in Ceylon) are represented by two such crops and only two districts by three.

Regionally, one discerns a significantly greater relative de- velopment of cash crops in Pakistan, both East and West, and Ceylon than in India, and within India one sees a considerably greater development, in terms of value and diversity of crops, in the south than in the north. Among major Indian states Bihar, Orissa, Madhya Pradesh, and Rajasthan stand out as singularly retarded. Comparing map (b) with map (a), one is struck by the lack of cash crops, other than jute, in the areas where rice predominates, especially in India, as well as in the arid areas where bajra is the leading crop. While there is obvi- ously a significant degree of correlation of cash cropping, as presented on this map, with commercialized agriculture, as depicted on our synoptic economic map (XI.E.1), it is impor- tant to keep in mind that the two are not synonymous, because of the large amounts of grain and other "food crops" grown primarily for sale in many parts of South Asia. If the value of such crops could also be taken into consideration on map (b), the relative standing of certain regions, most notably the Pun- jab and western Uttar Pradesh, would be greatly enhanced.

The graphs on plate XI.B.2 relating to the area in principal crops and the production of food grains and pulses, by coun- try, largely speak for themselves. A general indication of the relative yields of different crops per unit area can be obtained by comparing their proportion of a country's total acreage with their proportion of the total value of food production, making due allowance for the inclusion of nonfood crops in the graphs relating to acreage and their exclusion in those re- lating to production. The graph showing indexes of agricultural production for the first two decades of South Asian indepen- dence shows, on the whole, an encouraging rate of growth. During that period the expansion of production, of food and cash crops combined, grew at substantially higher rates than did the population, while food crops alone more than kept pace with population growth. Despite these facts, the safety margins between sufficiency of production and scarcity have never been great, and the likelihood that sharp falls in pro- duction in a given year might produce acute shortages, even to the stage of widespread famine, must always be reckoned with.

Sources

General References

Ceylon, Statistical abstract . . . (1958); The Ceylon year book . . . 1965; Pakistan yearbook (1970); Times of India direc- tory . . . (1961–62), (1962–63).

Government Documents

Ceylon, Census (1953); Ceylon, Department of Census and Statistics, Statistical abstract . . . 1958; India, Census (1961); India (Republic), Directorate of Economics and Statistics (1961); Pakistan, Census (1961, incl. 62 district volumes for data on crop acreage); Pakistan, Ministry of Agriculture and Works (1962); United Nations, ECAFE, Statistical yearbook . . . (1968); United States AID Mission to Pakistan (1966).

Other Works

G. Blyn (1966); V. G. Kulkarni (1968); Y. P. Pant and J. S. Chandra (1969).

XI.B.3. Intensity of Agriculture

South Asia's ability to cope with the dramatic expansion of its population in the modern period, particularly since the on- set of the population explosion around 1921 (see plate XI.A.3), has depended to a very large degree on the parallel intensifica- tion of its agriculture. This has entailed, where possible, bring- ing more land under cultivation by migration into thinly settled areas, by clearance of forests and wooded tracts in already set- tled locales, by reclamation of lands that, for any of a variety of reasons, had gone out of production, and by reduction of area lying fallow. Further, it has meant expansion of gross sown acreage by double-cropping. Finally, it has involved a massive expansion in area under irrigation, in volume of irri- gation waters utilized, and in augmented use of such additional agricultural inputs as fertilizers, both organic and chemical, pesticides, and improved seed varieties. Plate XI.B.3 relates to several of these factors: cultivated area per capita, area under double-cropping, and area under irrigation.

Maps (a), (b), and (d) indicate, by major administrative divisions, the gross cropped area per capita as of 1900–1901, 1930–31, and 1960–61; map (c) portrays the same variable, by minor administrative subdivisions, as of the crop year 1960–61. Although boundary changes from one period to the next render comparisons over time difficult, it is evident from the accompanying table that the expansion of gross cropped area has failed to keep pace with the expansion of the popula- tion. Both India and Ceylon witnessed a slight decrease over the period 1901–31 in gross cropped area per capita and no- tably sharper decreases from 1931 to 1961. (It must be re- membered that the 1931 area of India included all of what in 1961 was Pakistan, both East and West; the per capita drop from 1931 to 1961 for the combined area of India and Paki- stan was from 0.92 to 0.81 acres.)

As one would expect, the areas on map (c) where one finds a relatively high ratio of cultivated area per capita are on the whole relatively arid, hence relatively unproductive, and of low population density; conversely, most areas with very low ratios (e.g., most of Bengal) are wet, highly productive, and densely populated. Also included in the low ratio category are a number of well-irrigated districts, in the Punjab and else- where, and a number of mountainous tracts, especially in Ne- pal, where despite difficulties of terrain the small, intensively cared for, terraced holding yield sufficient returns to sustain moderately dense populations. One can of course infer nothing about yields of food per capita and the ability of specific areas to sustain their populations from inspecting a map of cultivated area per capita in itself; one must supplement such information with data on population density (plates XI.A.1–3), crop pat- tern and yield (plate XI.B.2), and other aspects of agricultural intensity (maps (e) and (f)), in order that a meaningful pat- tern may emerge.

The area under double-cropping, expressed as a percentage of the net area sown, by minor administrative subdivisions, as of the crop year 1960–61, is depicted on map (e). While the map does reveal a few favored regions where double-cropping accounts for more than 45% of the net sown acreage, mainly in the wet alluvial plains of northeastern India and East Paki- stan, and in a few other localities with ample perennial irriga- tion, one is struck by the territorial predominance of districts where the double-cropping ratio is less than 15%. This is espe- cially true of the inland portions of India to the south of the Indo-Gangetic Plain. Perennial canal irrigation in this region is very limited because of the terrain and the pronounced sea- sonality of rainfall (plate I.C.1), and future irrigation and the expansion of double-cropping there, as well as elsewhere, is likely to depend greatly on the development of tube wells. While it is obvious that there is still scope within South Asia for further expansion of double-cropping, the unit costs of future development are likely to be high relative to those of the past.

Map (f) portrays, by minor administrative subdivisions, the area under irrigation as a percentage of the gross sown area, as of the crop year 1960–61. In some respects the pattern of map (f) diverges strikingly from that of map (e). Over large parts of the northwest of the subcontinent, for example, the ratio of irrigated to sown acreage is exceedingly high; yet within that area, especially to either side of the Indus Plain, double-cropping is not common. This indicates that, within this arid region, irrigation not only provides a supplement to rainfall, as is the case of most of the remainder of South Asia, but is a sine qua non for cultivation to take place at all. A similar situation obtains in what is known as the "Dry Zone" of Ceylon. Conversely, in the northeast large areas character- ized by copious rains and moisture-retentive, alluvial soils pro- vide relatively little incentive for the development of irriga- tion, though it is certainly feasible, yet manage to grow two crops a year over substantial portions of their cultivated acre- age. Between these two extremes, but especially in regions with from 20" to 40" of rainfall per year, irrigation may spell the difference between poor yields and good, or, if it is perennial, as in the Ganga-Yamuna Doab of western Uttar Pradesh,