Predictive modeling for plant design will soon come
of age. The plant breeder will first identify the ideal
plant type, which may not yet exist, and then breed to
achieve it. Increased nutrient uptake and greater
resilience to environment and biological stresses should
be sought, as well as extended filling periods for grains
and quicker rates of dry-down for forages, feed grains,
cereals and seed legumes. Other possible desirable
characteristics could be lower levels of protein and
higher levels of calories in legumes, and higher levels of
oil and less starch accumulation in oil seeds.
 Multiline selections of cereal grains are gaining in
prominence. Well established networks exist for their
testing and evaluation. A multiline composite is a
mechanical mixture of lines that resemble each other in
height, maturity, grain type, quality and yield but dif-
fer genetically in disease resistance. This is particularly
important for rust resistance in wheat. Much of the early
work in genetic improvement was focused on pest resist-
ance and the stabilization of yield by reducing crop losses
from insects, diseases and viruses. The effort in the two
decades ahead and for the years 2000 to 2030 should in-
crease the genetic yield potential of future varieties and
resistance to environmental stresses, including those from
problem soils. Corn yields have not reached a plateau
and will continue to improve. With wheat and rice, the
severity of periodic catastrophic losses caused by diseases"
and insects has been dramatically reduced.
 An exciting area for the future is genetic alteration
of crops for climatic adaptability and higher yields from
soils that are infertile, too acid, toxic or saline for
varieties now in use. A concerted effort is needed from
plant physiologists, agronomists, geneticists, plant
breeders and agricultural engineers to create new strains
of food crops capable of performing well in hostile en-
vironments, such as a wheat variety suitable for produc-
tion in the lowland tropics.
 Improvement of the nutritional value of crops should
be a research priority. Each of the genetic improvements
for .practical human nutrition-including high-lysine
corn, triticale (a wheat/rye cross), and high-protein
sorghum and barley-has thus far, however, been large-
ly in vain, in spite of early and continuing publicity
campaigns.
 Issues of food acceptability are still important. Food
commodities that differ appreciably in color, taste, tex-
ture, general appearance or storage quality, or that yield
less, are not likely to be accepted. Nevertheless, one of
the promising ways to help solve the future protein needs
of people and to improve the dietary values of cereals
and legumes is to genetically improve nutritional values.
 Plants provide directly or indirectly up to 95 percent
of the world's total food supply. Worldwide, of the
350,000 species of plants, only about 0.1 percent (less


than 300) are important current sources of food. Glob-
ally, 24 plants essentially stand between human life and
starvation (Wittwer, 1983a): wheat, rice, corn; potatoes,
barley, sweet potatoes, cassava, soybeans, oats,
sorghum, millet, sugarcane, sugar beets, rye, peanuts,
field beans, chickpeas, pigeon peas, winged beans,
cowpeas, broad beans, yams, bananas and coconuts.
The cereal grains alone provide 60 percent of the calories
and 50 percent of the protein consumed by people. Inter-
national agricultural research centers have assembled
large numbers of accessions of genetic stocks for major
food crops, including over 65,000 for rice, 26,000 for
wheat, 13,000 for maize, over 14,000 for sorghum,
10,500 for soybeans, 5,000 for pearl millet and mung
beans, 12,000 for potatoes, 2,000 for cassava, 11,500 for
cowpeas, 11,000 for chickpeas, 5,500 for pigeon peas,
and 3,000 each for field beans and peanuts. Added to
these are some 3,000 genetic stocks for barley, 5,000 for
peas, 5,400 for tomatoes, 1,025 for sweet potato, more
than 1,000 for coconuts, and 800 for Chinese cabbage.
Similarly, 12 forest trees are the primary forest species
in the United States. These include loblolly pine, short
leaf pine, Douglas fir, western hemlock, red oak, white
oak, ponderosa pine, sweet gum, aspen, black walnut,
yellow poplar and spruce.
 The establishment in the United States of the National
Plant and Germplasm System (NPGS) was a landmark
in American agriculture for the survival of genetic
resources for future production of food, feed, fiber and
forest products. The NPGS now maintains over 400,000
seed and vegetatively propagated accessions of germ
plasm. These are distributed in various storage
laboratories throughout the United States. Much can be
done to increase food, feed, fiber and forest production
with this vast resource of genetic plant material. The
extensive use of commercial hybrid varieties, however,
now exists only for corn, sorghum, pearl millet, sugar
beets and coconuts.
 Major efforts are now in progress to seek out food and
forage crops and varieties resistant to or tolerant of salin-
ity. There are 3.8 million square miles of soils too salty
to grow crops in the world. Some are important in the
United States. Likely candidates for improved tolerance
to salinity are the wild relatives of barley, wheat,
sorghum, rice, millet, sugar beets, tomato and date
palm. Also included are certain forage crops that do or
can serve the needs of livestock, notably alfalfa, Ladino
clover, creeping bentgrass and Bermuda grass, and
various reeds and rushes. Genetic selections of barley
have already been identified that can be grown with
ocean water once the seeds are germinated (Epstein and
Norlyn, 1977).
 Genetic vulnerability to pests and environmental
 limitations increases with genetic uniformity. Some
 major food crops in the United States are highly