Stumpy cornstalks laden with ears. Square tomatoes that can be crated and shipped easily. Orange trees capable of surviving subfreezing temperatures - without the need for smoldering smudge pots or gas heaters to stave off a deadly frost.
These are among the ''supercrops'' that farmers of the year 2000 may see in their furrowed fields or orchards. They will be the fruits of the genetic engineering of plants - a still-nascent technology that is expected to transform agriculture over the next two decades.
The green revolution of the past two decades was brought about largely by conventional plant breeding. But the big increases in agricultural productivity over the next two decades are expected to come through genetic manipulation of plants.
Recent advances in gene splicing have compressed the time scale that some experts believe it will take for this technology to move from the lab petri dish to the farm.
None of the changes, to be sure, will occur much before the late 1980s. But by the turn of the century, major crops should be genetically altered to resist pests and diseases, survive in salty soils and harsh climates, and perhaps grow without fertilizers.
The main obstacle to getting there remains technical: Knowledge of the cell structure and behavior of plants lags behind that of animal and human cells. But ethical and political factors are also coming more into play, as the field of genetic engineering moves from the lab toward the marketplace. The US Environmental Protection Agency, for one, is grappling with how to regulate the release of genetically tailored organisms into the environment - partly as a result of safety concerns surrounding a California agricultural research project.
Politics notwithstanding, a glimpse of how big the changes on the farm and how long they will take to come about is mirrored in a recent study by L. William Teweles & Co., a Milwaukee, Wisconsin-based international consulting firm specializing in seed and plant science. It projects that plant genetics will add $5 billion a year to the total crop value in major nations by the year 2000 and $20 billion annually shortly thereafter. The new technology is also expected to boost agricultural production by 15 to 20 percent in the industrial world by the turn of the century. Among its other findings:
* The biggest agricultural gains will be in the industrialized world, since that's where most of the know-how is being developed.
* Yield gains will be the most dramatic in 10 major crops, including corn, wheat, soybeans, barley, rice, and tomatoes, as a result of the development of seeds resistant to diseases, insects, and bad weather.
* Other than in some specialized areas, the US is well ahead in research and development in the field.
''The United States is leading the race,'' says L. William Teweles, company president. ''Our Chevrolets may not beat Toyota, but in the new plant genetics we are clearly ahead.''
An early area where genetics will be put to work, still not until the middle to late 1980s, is in herbicide-resistant crops. Companies have long been trying to come up with chemical brews that will selectively kill weeds but not crops. They have been only partly successful. Scientists are slowly closing in on producing genetically engineered plants that will be immune to the sprayer's fog.
It is expected that genetically engineered corn, tomato, and wheat varieties will emerge by the late '80s. Work on these crops, given the market potential, is now drawing interest from food-processing, ag-chemical, and other big companies, as well the small start-ups. Companies as diverse as Campbell Soup and Atlantic Richfield, for instance, are racing to create a genetically engineered ''supertomato.'' The idea: to isolate genes in plants that have desirable traits - resistance to drought and disease and the ability to grow in salty soil, for instance - and implant them in commercial tomato strains.Scientists expect meatier, faster-growing, better-tasting tomatoes to emerge from the lab by the late '80s.
''There has been consistent progress in research,'' says Thomas Hiatt, president of the California-based Sungene Technologies Corporation, a start-up firm. ''It has come time to position oneself to take advantage of the market. The technology is moving toward the applied-research phase.''
With corn, the target is to produce stout stalks that won't fall over, ones with more ears, and plants with upright leaves so they can be planted in denser rows. Sungene, for one, claims its cell- and tissue-culture work has progressed far enough that it will have a commercial high-yielding corn plant by 1987.
Still far off is the ability to engineer plants with higher nutritional qualities and ones that can create their own nitrogen fertilizer. The message, in other words, is that the biggest changes for agriculture are still some time off. ''The potential applications in plants are very good, but I don't look for many of them for 20 years,'' says Dr. Gerald Still, director of plant production in the US Department of Agriculture's research service.