Roguing and Selecting Seeds Based on Desirable Plant Traits


Roguing refers to the process of removing underperforming or off-types plants from a population in order to improve or maintain a variety. This negative-type selection helps to eliminate ill effects resulting from potential and accidental crosses, mixing of seeds, or undesirable genetic variations that may have occurred in previous generations. Roguing is a crucial step that, if done incorrectly, will detrimentally impact your variety over time, which will gradually exhibit unwanted traits.

It is recommended to rogue several times during the season at different stages of the plant’s lifecycle, based on the desired selection criteria: at the early vegetative stage, individuals lacking vigor are removed; later on, plants that have off-type foliage or color are pulled off (and eaten for consumption); plants affected by diseases are also removed; then any individual that bolts prematurely; at flowering, crops are again checked and off-types removed; then same at full seed development (fruit/pod when applicable). Biennial plants that are meant to overwinter outside the garden (more specifically their roots/tubers) should be rogued once again right before they are stored away. Biennials that show any sign of bolting in the first season, as well as plants that bolt so late in the second season that they are unable to mature seeds in a timely fashion should be eliminated. Note that the earlier you determine that a plant (or its fruits) is not true to type, the earlier you can rogue it out and eliminate it as a source of pollen that might contaminate the rest of the plant population. It is important not to select too severely as it might impact the overall population size and therefore make it difficult to maintain a healthy genetic diversity in your variety. All open-pollinated varieties have inherent variations that will inevitably not fit the perfect ideotype (ideal set of morphological and physiological traits of a plant variety).

Roguing may be a challenge for home gardeners that grow plants for food consumption and seed production simultaneously. Letting an off-type plant remain in the garden is usually fine, provided that the plant is harvested for food before it flowers. However, an off-type plant in its reproductive stage that stays too long in the garden will increase the risk of cross-pollination and might compromise the genetic purity of other varieties of the same species that are present. That is usually not an issue with most of the biennials, as they are grown for food during their first season. Seeds from off-type plants that were saved unknowingly or accidentally should be discarded, as they will produce plants and fruits that are not true-to-type.


Selection is done either to preserve–with the help of adequate isolation techniques–the variety’s trueness-to-type and its specific characteristics in their current form while eliminating undesirable individuals (often referred to as mass selection), or to intentionally change the variety in some preferred direction. Roguing goes hand in hand with the selection of individuals based on morphological (i.e., fruit shape, size, color, plant stature uniformity) or biological (plant earliness, productivity, resistance to diseases, adaptation to local environmental conditions) desired criteria within the plant population for seed saving purposes. This type of selection is in fact considered to be one of the most basic and simple forms of plant breeding. Only plants that exhibit good vigor and the chosen criteria should be selected to save for seed. This process requires careful observation throughout the entire growing season, keeping seed selection constantly in mind. Selecting healthy seeds based on a specific trait will lead the variety to gradually develop a genetic strain that will carry this particular characteristic over the subsequent generations. An average of thirty years is required for a cultivated variety to fully adapt to the local conditions of an area, although some cultivars can also evolve and develop new traits in much shorter timeframes.

It is important to note that selecting for too many different criteria simultaneously might prove to be detrimental, as there will be a risk of not being able to select effectively for any of them. A methodical selection process focusing on just one or two characteristics will certainly be more effective. Ultimately and regardless of the criteria chosen, one of the best ways to preserve and fortify the genetic heterogeneity of a variety is to regularly plant it and harvest its seeds. Note that selecting for the preservation of germplasm involves a specific process that strictly avoid adapting intentionally or unintentionally the individuals within the plant population to our personal preferences or particular situation. To this end, seeds are typically saved from all the plants that are true to type, not only the best. Roguing is, however, still carried out to eliminate potential outcrosses.

Is it useful to know how to distinguish when phenotypic variability is caused by genetic or environmental variability and how to select plants accordingly based on your desired criteria. Selection will indeed work only on genetic differences, not environmental ones. Genetic variability is often translated in the emergence of a genetically superior trait that can be passed on to the next generation and improved upon through selection within each subsequent generation (i.e.: pest resistance). Environmental variability can be observed when certain plants perform differently (better) under specific and sometimes very localized climatic conditions and that this advantageous trait is lost in the next generations despite multiple selections. Some plants may thus perform poorly or better due to either favorable or adverse genetic makeup or particular growing conditions. The easiest way to determine whether genetic or environmental variability is responsible is to save both the best and the worse seeds separately and compare them against each other the following year. Note that individuals that perform better than others within a same plant population – especially if they grow in an “insular” fashion from the other plants – does not systematically indicate a genetic superiority. This might be just the result of a favorable microclimate or localized growing conditions in your garden (i.e.: better sun exposure at the end or on the edges of the patch, better access to nutrients in the soil and water, etc.). In order to select adequately and effectively, careful and regular observation and evaluation will therefore be necessary to identify whether one or several of your selection criteria is working in a given situation.

A variety may be quickly improved by further refining the selection process and conducting what we call a family selection. This method consists in saving seeds from each of the best plants from a given population, putting them in individual packets (i.e., 15 plants = 15 seed packets), and replanting them next season in separate rows (one row for each plant you saved seeds from, 15 rows in total). Those rows (called families) are then rogued out, eliminating any entire family that exhibit undesired traits. Any healthy individuals from those underperforming rows should also be removed, as they still share genetics with the unhealthy plants. All remaining families deemed to meet the selected desired criteria (on average) are then further rogued out by individually removing the most underperforming plants as well as any off-types. Finally, only seeds from the individuals left in these best families are saved.