Page Banner

United States Department of Agriculture

Agricultural Research Service

Breeding Methodology
headline bar

Breeding Methods Used in the USDA-ARS Rice Varietal Development Program




About 90% of the crosses made for cultivar development in the breeding program are targeted for the main Southern U.S. market, conventional long grains. The choice of parents that are used in a cross determines the potential of the offspring that can be derived from it through breeding. It generally takes 7-10 years to develop a cultivar through conventional methods. Thus, the crosses that are made in any given year will determine the cultivars that are available to farmers a decade later.

Most of the crosses that we use are "single crosses" that involve just two parents. More complex crossing methods can be used but these take more time to develop and delay the onset of the selection process. We generally use publicly released cultivars and elite breeding lines as the parental starting material. Elite breeding lines are selections that may be close to release as cultivars or may be "near misses", good selections that lack in a few traits. These "near misses" are used in an effort to broaden the genetic base by bringing new gene combinations into the gene pool. Parents are paired together in a cross to complement each other by building on strengths and compensating for weaknesses.


Breeding Method

production of hybrid seed 

Most of the breeding program uses a "pedigree" breeding scheme. When the parents are crossed, F1 (hybrid) seed is produce. Rice is a self-pollinated crop so the F1 seed is planted and allowed to self-fertilize and produce F2 seed. This is the first generation where segregation of the progeny can be observed due to different combinations of genes in individual plants. F2 seed are space-planted so that the characteristics of individual plants can be observed. For each plant that is selected for advancing to the next generation, a single panicle (seed head) is harvested. The process of planting one panicle in a row in the field, selecting among rows, and then selecting one or more panicles from within that row for advancing in each subsequent generation, constitutes the pedigree breeding method.


Selection Process

selection criteria table breeding activities table

With each generation of self-pollination in rice, more and more genes become fixed in a state that they are no longer segregating or recombining. Thus, breeders generally wait to apply selection pressure after a couple of generations of selfing so that some of this segregation has "settled down". However, even in very early generations, breeders can get some sense of the potential of a cross by the variability that they observe. Through the F2 and F3 generations (second and third self-pollinated generations after the F1 cross is made), selection pressure will be applied for easily observed traits like plant type, maturity, plant vigor, and degree of tillering. Panicles from the best appearing plants will be harvested and planted in the next season. At the F4 stage, selection pressure will be expanded to include evaluation of grain shape and translucency, cooking quality and resistance to rice blast disease which is determined in inoculated nurseries. At the F5 stage, enough seed will have been produced to allow for a small yield trial. This will usually be conducted in Beaumont using single field plots that are about 70 sq.ft and is the first opportunity for collecting quantitative data on yield, milling quality, and height in addition to the previously mentioned criteria. There are usually some 600 lines that are tested at this stage and only about 25% are selected for advancement. At each stage of yield testing, the performance of the new breeding lines is compared with the performance of current commercial cultivars. At the F6 stage, essentially the same selection criteria are used but the study will include multiple replications at Beaumont in an effort to reduce the environmental error and collect more accurate data. Lines will also be evaluated for resistance to sheath blight disease in inoculated field nurseries at this stage. The best performing lines from this test will be advanced to the Statewide Preliminary trials which are conducted at Beaumont, Eagle Lake, Bay City, and Ganado, Texas. Testing across this diversity of environments allows for a better assessment of the adaptation of these lines to different soil types and weather patterns. The best performing lines are then advanced to the multi-state regional trial which is conducted in cooperation with other researchers in the Southern rice growing region. Potential cultivars will stay in this trial from 2-4 years to determine their performance across the Southern rice belt. Additional tests for tolerance to insect pests, seedling vigor, and response to fertility management will also be conducted. Samples will be sent to industry users to alert them of a potential new cultivar release. The wealth of information that has been collected by this time is used to determine if the cultivar should be released or not. This information is presented to the State Seed Board for review.

Seed Increase for Release of Cultivars to the Public

seed increase 

During the last couple of years that a cultivar is in the multi-state regional trial, a purified seed increase is produced for the potential new release. Some 600 panicle rows are evaluated for uniformity. This may be done over a couple of seasons to assure that all rows are true to type. A detailed description of the variety is documented and is used by seed inspectors to distinguish it from other cultivars. Once the decision is made to release the cultivar to the public, purified seed is provided to foundation seed programs. These programs increase the seed and make it available to seedsmen and producers.

Last Modified: 2/23/2005