(3-E) ~ Desertification and Land Degradation Generally ~

(3-F) ~ Antiquities ~
~ (3-F-a) ~ Climatic Changes ~
~ (3-F-b) ~ In the Beginning. ~
~ (3-F-c) ~ Salinization as a soil destroyer ~

Deforestation and over-grazing in the Armenian highlands (eastern Turkey) is believed to be the source of the silt load that is believed to be the root cause of the decline of the Mesopotamian civilization and the irrigation system on which it depended ((74C1), p. 53). The coastline change at the mouth of the Tigris and Euphrates Rivers due to siltation between 700 BC and the 20th century AD is described in Ref. (56D1) (See map, Fig.101). The Persian Gulf has been shortened by 150 miles by the silt load of these rivers (56D1).

The Ahkadian (spelling?) empire (Persian Gulf to headwaters of Euphrates River in modern-day Turkey) collapsed around 2200 BC as a (probable) result of a 300-year drought which drove people out of northern cities in Mesopotamia (93W1). By 2300 BC, wheat was no longer grown in the Sumarian city-state of Agode (Mesopotamia). By 2100 BC Ur also had abandoned wheat production, and wheat represented just 2% of the crop in the Sumerian region. By 2000 BC Isin (spelling?) and Larsa (spelling?) no longer grew wheat, and by 1700 BC, salt levels in soil throughout southern Mesopotamia were so high that no wheat was grown (90P3). In Iraq, the shift from wheat to barley was accompanied by a serious decline in fertility (attributed to salinization). In Girsu, yields were 253.7 m³/ km² in 2400 BC; 146. m³/ km² by 2100 BC, and 89.7 m³/ km² in 1700 BC (58J1).

In southern Iraq, about 3500 BC, wheat and barley were nearly equal in occurrence. In 2500 BC the less-salt-tolerant wheat accounted for 1/6 of the crop. By 2100 BC wheat accounted for 2% of the crop in the Girsu area. By 1700 BC, wheat had been abandoned completely in the southern part of the alluvium (58J1).

Three major occurrences of salinity have been established from ancient records. The earliest, and most serious, affected southern Iraq from 2400 BC to 1700 BC. A milder salinity problem occurred in central Iraq from 1300-900 BC. Finally, the Nahruan area east of Baghdad became salty only after 1200 AD (58J1).

With the converging effects of mounting maintenance requirements and the declining capacity for more than rudimentary maintenance tasks, the virtual desertion of the lower Diyala area of Mesopotamia was inevitable. By the middle of the 12th century AD, most of the Nahrwan region was already abandoned. Mongol horsemen under Hulaga Khan arrived a century later, (but are blamed for the devastation they found ever since) (58J1).

Current spectacular erosion in Jordan is said to be an old problem (Ref.3 of (92D1)). Erosion-induced productivity losses were extensive 1000+ years ago, and certainly reduced yield significantly over large areas in the hills prior to 1900. Many references note rocky soils in places that were once cultivated (92D1).

In Asia Minor, the interiors of the old Roman provinces of Caria (sp?) and Phrygia were completely deforested by the first century AD (90P3).

Extensive cultivation began in Israel between 5000 and 3000 BC (Ref. 36 of (92D1)). Shortly before 0 BC, crop rotation and terracing of hill slopes were practiced widely. After the Muslim conquest in 640 AD, terraces were destroyed through neglect, causing catastrophic soil erosion. Much erosion occurred in this region during Roman times also (Ref.4 of (92D1)). Others believe catastrophic erosion of productive lands occurred after the decline of Roman power in Palestine and Syria in the fifth century (Ref.15 of (92D1)).

Erosion reduces water-availability (85F1), as well as nutrients to growing plants, and diminishes soil organic matter and soil biota (91L3).

Globally, about 10% of presently cultivated land, and 8% of rangeland has probably experienced a severe loss of productivity. Perhaps 65% of rain-fed croplands have sustained a moderate (10-50%) loss. The situation is considerably worse in developing countries (85D2).

A paper ("Soil Erosion Effects on Soil Productivity: A Research Perspective", Journal of Soil and Water Conservation 36(2) pp. 82-90 prepared by the Nat. Soil Erosion-Soil Productivity Research Planning Committee, Science and Education Administration- Agricultural Research) says that damage to soil water-holding capacity is probably the most important mechanism by which soil erosion reduces productivity ((83C1), p. 26). Other mechanisms include nutrient-loss damage to soil structure and farm-management difficulties (83C1). Soil erosion depletes soil productivity, but the relationship between erosion and productivity is not well defined (Refs. 63, 67, 80, 83, 99, 100 of (81N1)). Effects of soil erosion on cropland productivity are discussed in depth in Ref. (89S1) and (89B3).

Some tests on simulated erosion effects are not comparable to actual erosion because "In most instances, instantaneous removal of topsoil would likely affect productivity more drastically than would removal by erosion". Note that removal of surface soil has sometimes increased productivity (expose better soil). Note that fertility deficiency can be corrected, but poor physical condition is very difficult to correct (85B6).

Some studies on the effects of erosion focus only on changes in soil depth, and neglect the importance of biodiversity, organic matter and the other complex of interdependent variables. These studies indicate crop-yield reductions of only 0.13-0.39%/ cm. of soil lost (Refs. 88 and 89 of (95P1)). A reduction in soil depth of 2.8 cm. reduces productivity by about 7% (95P1).

Of all known measurements of erosion's effects on soil productivity, about 60% have been made in the US, and 2/3 since the 1980s (90D1) (Ref. 54 of (92D1)).

Cumulative topsoil erosion has reduced the production potential of US croplands 10-15%. An earlier estimate (35%) by Bennett (Ref. 27 of (76P2)) is generally considered to be an over-estimate (76P2).

Erosion-productivity relations were evaluated on 3 Indiana soils from 1981-86 (6 years). (SCS criteria used for slight, moderate, and severe erosion phases). Corn yields reduced 15%, and soybeans 24%, on severely eroded sites compared to slightly eroded sites (89S5).

Compared erosion-productivity relations on four soils: Otero (eastern Colorado), Athena (eastern Washington), Cecil (Georgia Piedmont), and Marshall (eastern Iowa) (Used EPIC (no gully erosion) to do computations.). If these four soils eroded at 20 tons/ acre/ year (4500 tonnes/ km²/ year) for 100 years (losing 450,000 tonnes/ km² of soil), Otero productivity would decline 35%, Athena 35%, Cecil 20%, and Marshall 20-25% (89B6). Comments: Typical topsoil inventory: 400,000 tonnes/ km². Stripping away all topsoil would reduce topsoil productivity far more than the numbers given above.

Tables 5-1, -2, -3 of Ref. (80C1) give results of numerous studies on the effects of erosion on yield reductions for wheat, corn, and grain sorghum agriculture.

Crop yields drop significantly as topsoil depth drops below 6" (15 cm.) ((76P2), Table 2, p. 153). Comments: Decreasing yields mean: (1) reduced soil cover, and (2) increased pressure on the land to produce at previous levels. These two effects provide positive feedback, i.e. instability that leads ultimately to gullies, which themselves represent an additional positive feedback that increases soil erosion rates.

Some 26 papers on the effects of soil erosion on soil properties and crop yields are collected in Ref. (84U1).

Loess (wind-deposited soil, typically with low organic matter content) may be more than 100 meters thick, so losing several meters would have little effect on productivity (92D1). Comments: Loess soils are low in organic matter and are therefore very erosion-prone and low in productivity.

In Idaho (Kimberly) irrigation removed up to 28 cm. of topsoil at top of fields and deposited up to 28 cm. at the bottoms of fields. Yields decreased sharply at the tops of the fields where sub-soil mixed with remaining topsoil. Yields increased - but less sharply - at the bottoms of the fields as topsoil depth increased due to deposition of as much as 28 cm. to a total topsoil depth of 66 cm. (85C7).

In the Palouse River Basin, average wheat yields basin-wide are 22% less than what they would be if erosion had been controlled over the past 50 years (83F1) (83F2).

In the Pacific Northwest (Palouse?) a 1960 soil survey showed 15% of land had lost all original topsoil since first cultivated. Others had estimated 25% in that condition (85S1).

In the Pacific Northwest soil erosion had robbed wheat farmers of 1/3 of the increased productivity provided by technological improvements in past 50 years (said in 1981). In Whitman County, yields would be 80 bushels/ acre instead of the 55-60 now (Reference apparently lost - could be (85C8) -see below.).

Crop production over 1 million acres in Idaho is probably reduced by 10% over what it would be if there were no erosion (85C8).

Soil erosion has robbed Pacific Northwest wheat farmers of 1/3 of the increased productivity provided by technological improvements in the past 50 years (said in 1981). In Whitman County, yields would be 80 bu./ acre instead of the 55-60 now obtained (this is 33% less based on present yield of 60 bu./ acre, or 25% less based on 80 bushels/ acre yield). Soil survey in 1960 showed 15% of land lost all original topsoil since first being cultivated. Kaiser figured 25% in that condition in 1981 (85S2).

The loss of 650 tons of soil is equivalent to the loss of one acre of cropland productivity (650 tons = 4 acre-inches) ((79S1), p. 91). Comments: This data is highly dependent on how much soil is left after erosion occurs. If this depth is greater than the root-zone, productivity loss per ton of erosion sediments will be very small; otherwise it will be very large. Ignore the data as virtually meaningless.

At 1977 erosion rates, corn- and soybean yields on highly erodible land in the Corn Belt (43% of US cropland) will be 15-30% less in 50 years than they would be if erosion were controlled ((84U2), p. 54).

US crop yields would drop 0.5-1%/ decade from erosion at current rates (84U2). Comments: This statement almost has to refer to erosion on flat bottomland with deep soil.

In the humid eastern US, grain yields decline 30-40% and forage yields drop 20-30% when the "A" horizon erodes away (7 Refs., (81N1)). Because sub-soil contains fewer plant nutrients than topsoil, added fertilizer is needed. Although fertilizer can partially compensate for low crop yields on exposed sub-soil, production costs and susceptibility to drought increase (many refs. of (81N1)). Comments: Fertilizer cannot compensate for the moisture-holding capacity of topsoil, nor for the tilth and erosion resistance provided by organic matter in topsoil - three of the most vital properties of soil.

Wind erosion on sandy soil near Sparta Wisconsin prior to 1938 caused productivity loss in red pine. Original A horizon of 30+ cm. Eroded "A" horizon was about 5cm thick (87F1).

Corn yields were checked annually (1968-1984) for 4 erosion classes (slight, moderate, severe, depositional) in SW Iowa. Erosion classes are those of the Soil Survey manual. Differences in yields between slight and severe erosion were generally small and non-significant. Corn yields from well-managed deep loess soils were not affected by topsoil loss from natural erosion when studied on a field scale (85S3).

Soybean yields were checked at 40 farm sites near Watkinsville (Georgia Piedmont) with varying amounts of erosion (Soil Survey manual). 1982- and 1983 yields on severely eroded sites were 50% of those on slightly eroded sites (85W1).

In the Southern Piedmont most land is classed as severely eroded (Class IV) or moderately eroded (Class II). Difference: 15 cm. of soil eroded away. Average corn yield loss due to 15 cm. lost soil was 42% - same as yield difference found 28 years ago when yields compared then (79L3).

15 cm. of soil lost by erosion caused about 40% reduction in corn yield (Southern Piedmont). The same 40% loss was noted 28 years ago when total yield/ acre was about half of what it was in 1979 (79L2).

Similar test on Grenada silt loam in west Tennessee where soils with varying degrees of erosion compared for crop yields. In Tennessee on moderately- to severely eroded soil, yield reductions were (79L2): fescue 25%; wheat 28%; cotton 33%; corn 42%; soybeans 50%.

(Africa) Crop yield reductions in Africa during 1970-1990 due to water erosion alone are estimated to be 8% ((95L3), p. 666).

(Australia) In Australia (New South Wales) soil erosion of 3 cm. reduced yields 39% in 1980 and 11% in 1981 (90S4). Total dryland cropping area: 121,000 km² (90S4).

(Australia) In Australia's wheat-growing areas, (control yield of 150-220 tonnes/ km²/ year.) loss of 75 mm. of soil caused yield declined 6%-46%. Loss of 150 mm soil caused yield declined 19%-52% (87B3).

(Canada) In Canada's (3) Prairie Provinces 362,000 km² of croplands are being used (88P4). Prairie Farm Rehabilitation Administration estimated in 1983 that a 65-year production period in the prairies would have experienced sufficient wind/ water erosion to reduce long-term crop productivity by 13.7%. Another source says Prairie Farm Rehabilitation Administration estimated 8% yield loss was due to wind erosion, and 5.7% to water erosion. This may be fertility (OM?) loss (88P4).

(Ethiopia) Debre Berhan, 100 km north of Addis Ababa: In old days, grain yields were up to 100-150 tonnes/ km²/ year. Current yields: 20 tonnes/ km²/ year. Reduction is attributed to erosion (87M3). Comments: The reduction might also be due to nutrient mining, due to lack of fertilizer.

(Hungary) On badly eroded fields in Hungary, crop yields are 20-50% less than on flat fields. On degraded (moderately?) fields, yields are 10-30% less. On less eroded soils yields are 0-20% less (92S2).

(Nepal) Crop yields in Nepal have declined 22-30% due largely to erosion (91N2).

(Thailand) Erosion in Thailand is blamed for an average yield decline of 50% for corn and upland rice (Ref. 2 of (92D1).

(Philippines) Some Zamboanga del Sur farmers claim corn yields have dropped 80% in the previous 15 years due to erosion (92D1).

(Ukraine) In Southeastern Ukraine Grain crop (barley, wheat) yield is 40-50% of that on non-eroded soils. At Voroshilovgrad, barley yields on severely eroded land is 57% of that on non-eroded (75D1).

When monoculture is practiced, yields tend to decrease with time regardless of how much fertilizer is applied. A steady annual presence of a particular root system favors a few organisms - bacteria, fungi, nematodes -that are potagenic (spelling?) to plant roots. Changing to a different crop alters the circumstances, and all but the most unspecialized pathogens are unable to thrive in the absence of their usual host (90A1).

Many US farmers have returned to crop rotations, but monoculture is still the norm for 40% of US corn (Ref. 49 of (96G1)).

Background pros and cons related to monoculture are in Ref. (87P1).

Fig. 133 of Ref. (56H2) shows a plot of soil nitrogen vs. time during 50 years of continuous cropping, with and without manure, for timothy, oats, wheat, and corn. Bad effects of monoculture (no crop rotation) are clear. Also see Fig.135 p. 665.

Within 50 years of the introduction of moldboard plowing on the US Great Plains in the 19th century, 60% of the soil's organic matter was washed or blown away (96G1).

Soil degradation in developing countries is growing worse owing to increased burning of crop residues and dung for fuel. This reduces soil nutrients (94M1), (90D4) and intensifies soil erosion.

Soils in China are low in organic been extending into the Adriatic Sea for at least 20 centuries. Old coastal cities are now 6 miles inland ((56D1), p. 511).

The city of Adria (gave its name to the Adriatic Sea) is now 15 miles from the sea, and 15 ft. higher in elevation. It was an Etruscan town built in 550 BC on an island a few miles north of the mouth of the Po River. It was an important Roman seaport. By the end of the Western Roman Empire, Adria was on the mainland. It was an important port on the Adriatic Sea in the 12th century AD, but it was then several miles from the sea and joined to it by a canal ((74C1), p.153).

About 300 BC, Italy and Sicily were still well forested (90P3).

Much of the land of eastern Spain was seriously exhausted during Roman times (500 BC-400 AD). Moors cultivated it intensively from the 8th to 13th centuries AD, so by the 15th century, much of it was severely eroded and exhausted. Thus the possibility of new land across the ocean appealed to Spain before it appealed to Europe that still had adequate good land (due to crop rotation and earlier elimination of the feudal system) ((74C1), p. 174).

Chinese civilization began around 2000 BC in North China, probably in the Loess-hills region of the Yellow River basin. It spread to the coastal plain of North China and, for 15 centuries, flourished there. Then it spread to the Yangtse Valley in central China and Manchuria around 500 BC. By then, the original Chinese civilization was disintegrating, and a dark age ensued. The new civilization shifted to the south, and reached the extreme southern part of China in medieval and modern times. China's best farms are now in the region south of the Yangtse River where civilized people have lived for the shortest time. This region is the most progressive and prosperous ((74C1), p.198). Comments: Note that this history is consistent with the premise that a progressive civilization can survive for only about 50 generations (12.5 centuries) in one place unless there is some mechanism for replenishing the soil.

China's Great Wall (2500 miles long, built to keep out marauding hoards from the north where agriculture had collapsed) (74H1) was built in 221-206 BC (Ch'in Dynasty).

Glover (Ref. 12 of (92D1)) claims that there was no erosion of the uplands in the Punjab region of India and Pakistan before the British pacified the region in the 1800s. (This action stopped wars and encouraged population increase (92D1).)

The central highlands suffered severe erosion with the introduction of tea and coffee plantations in the early 19th century (Ref. 27 of (92D1)). By 1873 erosion had become so severe that the government ruled that no land above 8000 ft. could be sold to private buyers (92D1).

Over-grazing began around 650 BC on the 80% of the land that was unsuitable for cultivation (90P3).

Around 500-300 BC, a Greek writer referred to the erosion and ruin of once-fertile lands along the Turkish coast ((82S1), p. 7).

In Roman times as many as 15,000,000 bushels of grain were transported annually from North Africa (where there is now only desert) and Egypt to Rome (56G2). North Africa had a large, prosperous agriculture during 200 BC-600 AD, as much evidence shows. Erosion is argued to be the reason for the decline (70L1). The land was probably already more-or-less ruined by erosion when Arabs conquered the area near the end of the 7th century AD ((70L1), p. 246). Presumptive evidence suggests erosion in the granary of Rome in the waning days of the empire, but the first real threat seems to have appeared in the 7th century, after the Arab conquest of present-day Tunisia, Algeria, and Morocco (Ref. 26 of (90D1)). By the 10th century the situation was said to have become bad. Deforestation was the principal contributor to erosion (90D1).

No more than 10% of the original forests that once stretched from Morocco to Afghanistan as late as 2000 BC still exist (90P3).

Cotton, like tobacco, a clean-cultivated row crop, and a cash crop, caused southern US fields to surface-wash, especially in winter. The southern US uplands gradually lost their organic horizons, color and protection; gullies were noted even before the Civil War ((56S1), p. 64).

Jappatown MD was a bustling tobacco port in the 18th century. But as the land was cleared and farmed, topsoil eroded into the harbor. Today, two miles of dry land separate early 18th century mooring posts from navigable waters (69T1).

The migrations and decay of the Mayan civilization is often attributed to the soil destruction instability that is inherent in the Ladang (shifting cultivation) system that requires a 15-25 year fallow period ((56G1) p. 338) (Copied to definition list).

Mayans in the Guatemala lowlands expanded continuously over 17 centuries (68 generations) beginning in 800 BC, and reaching a population of 5 million (doubling time: 408 years) by AD 900 when the population density became comparable to that of the most agriculturally intense societies of today. Within decades the population dropped to a tenth of its peak. Core samples from two lakes in the region implicated soil erosion. The area was almost totally deforested by AD 250 (82B4). The earliest Mayan settlements date from about 2500 BC. Within a few decades after 800 AD the whole Mayan society began to disintegrate. The peak population of the Mayan lowland jungle might have been near 5 million in an area that today supports only a few tens of thousands (90P3).

Easter Island, off the coast of Chile, although drier than most of Polynesia, is not as dry as most of Australia, and has more fertile volcanic soils. Archaeological finds and ancient pollen show that when Polynesians arrived on Easter Island, it was covered with a lush subtropical forest of palm trees and giant sunflowers, inhabited by land birds and breeding sea birds (Wilson da Silva, "Long dry spells, outlook gloomy", New Scientist 10/12/96).

Easter Island sustained a population of 58 people/ km². But the population stripped the forests bare and killed the native animals. By 1500, all that was left was grassland. "People turned to the largest protein source around-cannibalism. Easter Island society collapsed in an epidemic of warfare. When the first Europeans reached the island in 1722, two-thirds of its population had died (Ref. Citation above).

In the past few thousand years, 12 other dry land masses in the Pacific subject to the whim of El NiÑo have seen human societies collapse, in some cases leading to the complete disappearance of a people. Polynesian societies on wetter islands with more predictable climates, such as Tonga and Samoa, survived the changes wrought by humans." The main difference seems to be that societies in low-rainfall environments were the ones especially prone to collapse by destroying vegetation. Low rainfall means that vegetation regrows slowly, so it can easily happen that regrowth doesn't keep pace with cutting (Wilson da Silva, "Long dry spells, outlook gloomy", New Scientist 10/12/96).

Population of Easter Island in 1550: 7000. Population in 1850 after deforestation and collapse of natural and agricultural systems: 100 (Clive Pointing, A Green History of the Environment and the Collapse of Civilization, New York, Penguin Books, 1991).

Historical records of the past 60 centuries show that civilized man, with few exceptions, was never able to continue as a progressive civilization in one locality for more than 30-80 generations (7.5-20 centuries). The 3 notable exceptions: the Nile valley (200 generations), Mesopotamia (160 generations), and the Indus Valley (175 generations) ((74C1), p. 7). Comments: Note that the three civilizations (Mesopotamian, Egypt, Indus) that had large river valleys that replenished the soil annually lasted over 150 generations. Civilizations that lacked this feature lasted only 50 or so generations. In these three long-lived civilizations, foreign conquerors, high taxes, etc. came and went. Only when the soil replenishment system failed did progressive civilization end.

Mesopotamia (Tigris-Euphrates River Valley) maintained a progressive civilization from about 4000 BC to the mid-13th century AD (74C1).

A progressive civilization existed in the Indus River Valley (Pakistan) from around 1500 BC until about 1200 AD (74C1). Most of the lower Indus Valley (Pakistan) is now relatively barren country (74C1).

The Indus civilization had its rise in 2600 BC and declined around 1900 BC (John Noble Wilford, NY Times - See Pittsburgh Post Gazette, 2/16/98).

A progressive civilization existed in the lower Nile River Valley for about 6000 years (240 generations) (74C1).

Minoans on Crete lasted 11 centuries as a progressive civilization (2500 BC to 1400 BC) (44 generations) and then declined and vanished in 2 centuries (74C1).

Phoenicians (Lebanon) lasted as a progressive civilization from around 2000 BC until 480 BC (61 generations) ((74C1), p. 74). Deforestation and scavenger goats (over-grazing) bought on most of the erosion that turned Phoenicia (Lebanon) into a well-rained-upon desert ((74C1), p. 72).

Greek civilization lasted as a progressive civilization from 1600 BC until 400 BC (12 centuries: 48 generations) (74C1).

The Roman Empire started around 500 BC and was powerful in 400 AD, but was gone by 476 AD (74C1).

The Syrians lasted as a somewhat progressive civilization from the 16th century BC until the 13th century AD (30 centuries) (It was rarely powerful; usually it was caught in cross-fire between major civilizations.) (74C1).

Ceylon's civilization disappeared around 1200 AD due to deforestation of upland forest that filled irrigation systems with silt ((74C1), p. 202).

The Anasazi Indians (S.W. USA) lasted as a progressive civilization for about 30 centuries (120 generations) until around 1200 AD ((87B2), 74C1?).

The Hohokam Indians (S.W. USA) lasted as a progressive civilization from several centuries BC to 1400-1500 AD (E. W. Haury, National Geographic (year?)).

Teotihuacan (NE of Mexico City) grew rapidly in population during 200-900 AD and then collapsed mysteriously (Ref. 10 of (87N2)). The second population cycle began around 1168 AD - the Tenochas (Aztecs). Hispanics arrived in 1519 and bought smallpox etc. that decimated the population (87N2).

Easter Island's civilization was established by Polynesians around 400 AD in a well forested environment. It was gone (and the island treeless) by the 18th century (50 generations) (87B2).

Marajo Island's advanced civilization collapsed due to soil destruction in a shifting-cultivation system ((56G1) (p. 341)).

Terraced mountain land in Yemen has been continuously cultivated for about 30 centuries. The rich soils were derived from volcanic lava (Ref. 7 of (87V1)).

The Sumerian civilization (in modern-day Iraq) lasted from 2400 BC to 1800 BC (Ref.1 of (96G2)).

The cumulative productivity loss from soil degradation over the past 50 years has been roughly estimated using GLASOD data, to be about 13% for croplands and 4% for pasture lands (98O1) Comments: Rangeland degradation is probably a lot more, since rangeland is more arid, generally, than pastureland, and therefore more susceptible to degradation from overgrazing etc.

In developing countries agricultural productivity is estimated to have declined significantly on 16% of agricultural lands. The GLASOD study estimates that almost 74% of Central America's agricultural land (defined by GLASOD as cropland and planted pastures) is degraded, as is 65% of Africa's and 38% of Asia's ((99S1), p. 18). (la)

Detailed studies based on predictive models for Argentina, Uruguay and Kenya calculate agricultural yield reductions of 25-50% over the next 20 years. ((97M2), p. 39-40) Comments: This seems high, relative to rates in recent decades.

Cropland productivity losses arise from soil erosion, salinization, waterlogging, urbanization, nutrient depletion, over-cultivation, acidification, and soil compaction. These factors decrease soil productivity and substantially reduce crop yields (85M2), (85F1), (90P5), and will reduce crop productivity for the long term (91T2).

Soil erosion is the single most serious cause of degradation of arable land (91E1), (91B4), (93P1).

In Third World, problems of making soil conservation effective are much worse than in the US. Steep land is used more frequently for cultivated crops. Pressure is growing to expand production on to even more marginal lands. Land use rights of villagers so deeply ingrained in some places that change is impossible. Peasants cannot afford to take risks (83H3).

A study by Douglas (67D1) of soil erosion in small catchments in eastern Australia (which had few effects of Man), compared to a 1960 study by Fournier (60F1) of a large, world-wide data-base of small-catchments erosion data, indicate that anthropogenic effects increase soil erosion by about two orders of magnitude.

Clay soils and soils low in organic matter are particularly prone to compaction. Compaction increases risk of erosion, and can reduce crop yields by up to 60% (Ref. 8 of (86D1)).

Sediment yields from construction sites in Baltimore-Washington DC area are 2000-50,000 tonnes/ km²/ year (Wolman and Schick, 1967, in (90M1)). Extra sediment produced by urbanization near Washington was sufficient to double the discharge of suspended sediment by the Potomac River to its estuary (Guy, 1965, in (90M1)).

Soil erosion rates on off-road-vehicle (ORV)-used desert hill slopes are commonly 10-100 times natural erosion rates (Ref. 13 of (81I1)). About 100 years are needed for desert soils to recover from off-road vehicle use (Ref. 17 of Ref. (81I1)).

The 2,484,000 km² of undisturbed forestland in the contiguous US contribute over 100 million tons/ year of sediment to US waterways (36.6 tonnes/ km²/ year). Non-point pollution from undisturbed forestland is 0.001-0.009 tons/ acre/ year (0.2-2.3 tonnes/ km²/ year) in scattered producing areas up to 740 tonnes/ km²/ year on isolated forests in southern and Pacific-Coast areas (81F1). Comments: High forested-land erosion rates along the Pacific Coast are often attributed to steep, unstable slopes resulting from volcanic activity in recent geologic times.

As a result of catastrophic runoff events, lands at the margin of arid- and semi-arid regions typically show the highest rates of mechanical weathering (Ref. 86 of (90S3)), and high concentrations of suspended solids in rivers (90S3).

Globally, 1.6 million km² of hillside land were identified in 1989 as "severely eroded" (96G1).

In Nigeria, cassava on 12% slopes lose 22,100 tonnes/ km²/ year, compared to 300 tonnes/ km²/ year on slopes of under 1% (Ref. 38 of (95P1)).

Southern Ontario-continuous corn: erosion = 1200 tonnes/ km²/ year (0 slope); 4900 tonnes/ km²/ year (10% slope) (Corn-hay rotations cut this to 700 tonnes/ km²/ year (84S2).

Soil loss/ unit-area is plotted against land slope in Fig.2 of (57S1), based on 4 sets of data and studies. Data on four Midwest US soils can be summarized by a table of erosion rate factors normalized to 1.00 at 3% land slope: (57S1)

About 70% of US croplands in 1977 had slopes of 6% or less; 90% had slopes of 12% or less. (See Ken Cook, Journal of Soil and Water Conservation, 37 (1982) p. 92) (RCA appraisal).

The average soil loss from 4 years of corn following 2 years of clover and grasses was 47% of the loss from 6 years of continuous corn (p. 89 of (71R1)). Comments: Erosion from clover- and grasslands is negligible relative to that from cornfields. But the effects of crop rotation are much greater than that computed by merely averaging soil erosion rates as can be seen from the figures above (and much other data). This suggests that crop rotation affects soil chemistry and soil structure in such a way as to reduce soil erosion.

A primary reason why croplands erode so much faster than other lands is that the amount of vegetative cover on croplands is significantly less than the amount on other lands. The phytomass of cultivated lands = 4.5 Gt. carbon in 1981 (3.3 Gt. carbon more than in 1860). The phytomass of the former natural vegetation which was replaced would have been 137+3.3 = 140.3 Gt. carbon (87E1).

Land in fallow is particularly vulnerable to soil erosion and salinization. Introduced in the early 1900s in Canada's prairie provinces, the excessive tillage associated with summer fallow has contributed substantially to reduced organic-matter levels, reduced soil-tilth, nitrogen loss, and less-efficient crop-use of available water (Ref. 7 of (87M1)).

A USDA position paper (1981) is cited as supporting the claim that, where erosion is by water, erosion rates accelerate with increasing erosion. This is attributed to erosion reducing water-infiltration which causes more runoff and hence more erosion. Also, erosion reduces plant-cover and plant-residue cover, and this promotes higher erosion rates (p. 27 of (83C1)). Comments: Storm severity is known to have a very large effect on soil erosion, because after soil becomes water-saturated, water runoff increases dramatically. Note that the effects discussed in (83C1) are positive feedback effects.

Difficulties in detecting soil productivity losses are compounded by the non-linear nature of erosion processes. Erosion increases runoff because of reduced infiltration. Increased runoff reduces available soil water and thus reduces plant growth. Reduced plant growth reduces plant-residue that increases erosion. The process thus advances exponentially, and reversing it may quickly become economically impossible (81N1).

World-wide soil degradation mechanisms: water 56%; wind 28%; chemical degradation 12, physical degradation 4% (from L. R. Oldeman et al, 1990, World Map of the Status of Human-Induced Soil Degradation, An Explanatory Note., revised 2nd Edition, International Soil Reference and Information Center, Wageningen, The Netherlands).

Small particles carried down-wind in suspension account for only 10% (by weight) of wind erosion (89P1). Salinization (lifting and dropping repeatedly) accounts for 50-80% of total wind erosion (Distances moved are under one mile.) (89P1). Surface creep (rolling) represents 7-25% of total wind-eroded soil (89P1).

Most of the soil picked up by wind in one place must come down in another place much like the place of origin (83C1). Comments: or in woodlands or well-vegetated areas that are much better at retaining the deposited dust than a bare, wind-blown field.

Cropland expansion into climatically marginal areas is occurring in the US, Australia, the former Soviet Union, Sahel, Ethiopia, India and elsewhere. As a result, wind erosion has become a greater problem in recent years, even in the absence of droughts (85D2). Comments: Wind erosion predominates on arid lands. These are typically lands that would not be used as cropland unless there was some pressure to do so, e.g. the lack of moister lands.

** (= 37% of 7.48)
US wind erosion (????) = 1.00 Gt. (76P2)
US wind erosion (1982) = 1.80 Gt. (89P1) (1.1 cropland; 0.54 from grazing land)
US wind erosion (1977) = 1.33 Gt. (81B1) (2/3 croplands; 1/3 from grazing land)
US erosion is 3/4 water-erosion, 1/4 wind erosion (76P2)

US rangeland wind erosion (1982 NRI study) = 1.5 tons/ acre (84L1)
Canada's prairie region erosion is 2/3 wind-erosion, 1/3 water erosion (86D1).
Comments: Also see Section (4-D) for wind-erosion data

A good photo of cropland gully-erosion in Tanzania is seen in (94P3).

Sheet erosion is more important, on average, as a sediment source than is gullying (p. 644 of (56L1)). Comments: Gully erosion causes extremely high soil loss rates/ acre. However, very few farmlands are in a state of being damaged by gullies since abandonment often follows soon after gullies render a field non-plowable.

Typical gullied land erosion: 41,000 tonnes/ km²/ year (76E1) (37,000 = 1"/ year)

International Inst. of Tropical Agriculture (Ibadan Nigeria) has shown that applying a crop-residue mulch at 540 tonnes/ km² can control erosion nearly completely on slopes up to 15 degrees (Ref. 51 of (89P2)). In field trials, yields increased over non-mulched plots by 83% for cowpeas, 73% for cassavas, and 23% for maize (89P5).

When corn-residue cover is increased by 10, 30 and 50%, the amount of nitrogen lost in surface runoff is reduced by 68, 90 and 99% (Ref. 99 of (95P1)).

About 60% of crop residues in China are removed and burned for fuel (Ref. 41 of (95P1)). (90% In Bangladesh (95P1)).

In West Africa, on a 10% slope under maize, mulch reduces runoff from 42% of rainfall to 6% - and reduces erosion from 2700 tonnes/ km²/ year to under 50 (Ref. 7 of (87E2)).

Manure reduces soil erosion. 16 tons/ acre onto corn land in Iowa (9% slope) reduces soil erosion from 22.1 tons/ acre to 4.7 tons/ acre (Ref. 53 of (76P2)). Other organic matter would have a similar effect (Ref. 54 of (76P2)).

Some off-farm impacts of soil erosion:

• Sediment-polluted waterways;
• Fertilizer- and pesticide-polluted waterways;
• Siltation of reservoirs and lakes;
• Destruction of breeding grounds for fish;
• Increased harbor- and navigation-channel dredging costs;
• River-bed elevation increases (that cause flooding) (dealt with in Ref. (85C1))

A model developed at RFF (Peshkin and Gianessi) shows that only 40% of the 1.9 billion tons of US sheet/rill erosion in 1977 ended up in waterways (p. 37 of (83C1)). Comments: The remainder presumably winds up on lower slopes and flood plains of the land. (Statement copied to Section (4-G-b)).

Off-site costs of soil erosion are reviewed in Ref. (85C3). Off-farm economic impacts of soil erosion are estimated to be $6 billion/ year in 1980 dollars (85C2). Also see The Economics of Soil Erosion: A Handbook for Calculating the Cost of Off-Site Damage, 135 pp., illustrated, bibliography, (1986) available from AFT01.