Skip to main content

Seed Savers Garden: A Demonstration of Genetic Variability

David C. Zlesak, Elizabeth Spedaliere, Kathy Bonnett; University of Minnesota Extension Educator and Master Gardeners, respectively

We often hear the saying that saving and starting seeds from hybrids typically leads to inferior plants or plants with too much variability to be worth our while. Even if some variability exists between the seedlings, is it actually so great or the plants so overall inferior that it wouldn’t be worth our effort? To what extent can this be true?

We decided to put this general recommendation to the test and develop a demonstration garden (Seed Savers Garden) at the Master Gardener Education and Research Display Garden at UMore Park in Rosemount, Minnesota to observe what happens.  In the fall of 2007 open pollinated seed was saved off a number of flower and vegetable seed-propagated varieties growing at the University of Minnesota Display and Trial Gardens in St. Paul and the authors’ gardens.  In the spring of 2008 seedlings were started in March in the greenhouse from the saved seed and also purchased seed of the same varieties for comparison.  Some of these seed-propagated varieties were reported to be F1 hybrids and others did not have their breeding methodology disclosed.  In addition to flower and vegetable seed-sold varieties, seed was also saved from some clonally-propagated flower cultivars (‘Lilliput Rose’ penstemon, Mammoth™ ‘Yellow Quill’ mum, and ‘Petite Delight’ monarda) and the seedlings planted out and compared to cutting-propagated plants purchased of the original cultivar.


The plants were planted in rows at UMore Park in June 2008. This fantastic garden is along highway 46 and free and open to the public between sunrise and sunset. Information and directions to this garden can be found at:  We planted the row of the original variety next to the row of the seedlings of that variety for easy comparison. Some varieties had limited numbers of plants and both the original variety and plants from saved seed were planted in a single row (in the following photos the original variety is to the left if there are two rows or to the front of the row if there is only one row and the bars represent mean values with standard deviations). Up to 14 plants each of the original variety and plants from saved seed were planted out. During the growing season when the particular species came either into flower or fruit, data was collected on plant size, flower/fruit number, and other traits that came to mind as being appropriate to document.  For some species data was taken again later in the season because as the plants continued to mature, greater differences were evident or, as in the case of tomatoes, ongoing, weekly fruit yield was recorded and ultimately summed over weeks.

Sometimes, seed packets did not tell us if the variety is an F1 hybrid or an open pollinated variety.  F1 hybrids are developed by crossing two distinct inbred lines that differ from each other and produce a superior hybrid.  They tend to be very vigorous, uniform, and are heterozygous for many traits because the original parents had different alleles for particular genes.  In the next generation, these alleles (forms of a gene) segregate and there can be a lot of variation among the seedlings depending on how these alleles reshuffle in each offspring.  Open pollinated varieties are cheaper to produce as different parental lines are not independently selected and maintained and then crossed to produce seed as in F1 hybrids.  For open pollinated varieties, plantings are made and parents possessing the traits desired are allowed to survive and intermate (those that don’t have the intended traits are pulled out) and seed is saved.  After multiple generations of doing this, the seedlings can come relatively true to type as the genes contributing to undesirable “off-types” have been generally been removed from the population. Following are the summarized results for each variety comparison followed by a general conclusion and some considerations.


Dianthus ‘Dulce White’


Plants from purchased seed performed much better than those of saved seed.  They had greater vigor with fuller, more floriferous plants.  All the plants from the saved seed were uniform for white flowers. Although the seed packet didn’t say, having such a decline in vigor among plants from the saved seed suggests ‘Dulce White’ is an F1 variety.

Dianthus ‘Elation Red’


Plants from purchased seed, again, performed much better than that of saved seed.  They had greater vigor, more floriferous plants, and were uniform for dark red flower color.  Plants from the saved seed segregated for multiple traits including the ability to flower and also flower color.  Only one plant from saved seed had dark red flowers like the parent.  The others had salmon or pink flowers.  Sweet williams are also in the genus Dianthus and do not flower until their second year after they have experienced a cold treatment from experiencing winter.  Having this trait appear among the seedlings was very interesting.  There were no sweet williams or other Dianthus species nearby that possessed this trait from which to pollinate the parents. The inheritance of many traits in the seed crops we purchased have not been disclosed and are kept proprietary by seed companies.  The possibility exists that this variety is an F1 hybrid and is the result of crossing a dianthus that requires a cold treatment with one that doesn’t and the requirement for cold is recessive.  It is not expressed in the F1 hybrid, which is a carrier.  The trait then segregates in the next generation.

Marigold ‘Starfire’


This marigold has petite foliage, is well-branched, and has an abundance of single flowers in a wide assortment of colors.  Plants from the saved and original seed are very comparable.  This variety is likely not an F1 hybrid and is an open pollinated variety.  It appears that breeders selected plants with a generally uniform, compact, mounded plant habit and preferred to keep variation for color within their populations.  The result is mixed colors and relatively uniform plant size both purchased and saved seed.

Monarda ‘Petite Delight’


Monarda ‘Petite Delight’ is a clonally propagated cultivar.  It is patented and the cultivar is propagated from cuttings or crown divisions (it is a perennial).  The unique trait for this cultivar is that it is very dwarf in stature compared to other monardas on the market.  Purchased plants of this patented cultivar were planted that were propagated from cuttings and compared to the seedlings.  Unfortunately, the purchased plants were very weak and stunted, had curled leaves, and did not flower (photo is of the original plant at the St. Paul Display and Trial Garden from which seeds were saved).  Perhaps they were infected with virus. The seedlings grew well and also did not flower.  Seedlings were quite variable for plant size and had normal looking foliage.

Chrysanthemum Mammoth™  ‘Yellow Quill’


Chrysanthemum Mammoth™ ‘Yellow Quill’ is also a clonally propagated, patented cultivar.  Seedlings were variable and segregated for flower color and also flowering time.  Some were clearly earlier and later in flowering (row of three flowers on top are of the original cultivar and the five flowers beneath are of the seedlings that were in flower September 3, 2008).  Although quite a bit of variability was found among seedling, since mums are a perennial, the potential is there to select and multiply very attractive, hardy seedlings through dividing them and taking stem cuttings.  One can use saved seed to develop unique individuals from which to become new clonally propagated cultivars.

Pansy ‘Antique Rose’


This pansy has well-branched, vigorous plants with above average heat tolerance and an assortment of colors.  Plants from the saved and original seed are very comparable.  This variety is likely not an F1 hybrid and is an open pollinated variety.  It appears that breeders selected plants with a generally uniform, compact, mounded plant habit and preferred to keep variation for color within their populations.  The result is mixed colors and relatively uniform plant size from both purchased and saved seed.

Penstemon ‘Lilliput Rose’


Penstemon ‘Lilliput Rose’ is also a clonally propagated, patented cultivar.  Although likely a perennial on other climates, it does not survive Minnesota winters.  There was surprising uniformity for flower color and plant characteristics between purchased plants of this cultivar and the seedlings.  Plants available for purchase were larger than our seedlings at planting time and throughout the growing season continued to remain ahead of the seedlings in development.  The data is a little misleading because of this.  Data was taken in July and then again in late August.  Number of flowering stems were substituted for number of individual flowers for the August data collection date because the number of individual flowers were staggering. Finding such uniformity among seedlings of a patented cultivar propagated by cuttings is very unusual.

Rudbeckia ‘Prairie Sun’


Rudbeckia ‘Prairie Sun’ is a very vigorous, attractive recent All-America Selections winner. Plants from the saved and original seed are very comparable.  This variety is likely not an F1 hybrid and is an open pollinated variety. Two of the eleven seedlings from the purchased seed were solid yellow and did not have the characteristic darker yellow petal base.  Two of the twelve plants from saved seed had flowers with tubular ray petals (typically called quilled petals in the case of chrysanthemums).  These plants are likely chance crosses with the variety ‘Radiance’ of rudbeckia which was planted next to the ‘Prairie Sun’ plants seeds were saved from. If kept in isolation from other rudbeckias, seedlings from saved seed of ‘Prairie Sun’ should be similar to those of purchased seed.

Petunia ‘Blue Wave’


Petunia ‘Blue Wave’ is an F1 hybrid.  Plants from purchased seed were far more vigorous, spreading much further and producing many more blooms than plants from saved seed. Surprisingly, flower color was very uniform among the plants from saved seed and comparable to the original variety, even though the plants the seed was collected from were surrounded by several other petunia cultivars in many different colors.  Although plants from saved seed were smaller, they were about the same height and a person could just plant seedlings more densely for a comparable effect. This may be an attractive option for gardeners, especially considering the price of F1 petunia seed and the fact that a lot of seed can easily be collected off of petunias.

Petunia ‘Baby Duck Yellow’


Petunia ‘Baby Duck Yellow’ is an F1 hybrid.  Just like ‘Blue Wave’, it came relatively true from seed for flower color.  Plants from saved seed were even more comparable to the original variety than ‘Blue Wave’ for plant size and flower number.

Ornamental foliage and fruit

Millet ‘Purple Majesty’


Millet ‘Purple Majesty’ is a very attractive purple-foliaged, ornamental grass.  Plants were very similar from purchased and saved seed.  There are awns that are part of the heads of grain, like wheat.  Among plants of the saved seed there was a single plant, interestingly, that segregated to not have awns. Just like the petunias and considering the cost of purchased seed of this All-America Selections winner, one may find it an attractive option to grow plants from their own saved seed.

Ornamental Chili Pepper ‘Black Pearl’


Ornamental Chili Pepper ‘Black Pearl’ is an F1 hybrid that has very attractive purple foliage and dark purple-black fruits that are round and upward facing.  Among the plants from saved seed, four of the twelve plants had green foliage and downward facing elongated fruit, while the remaining 8 plants had purple foliage and dark rounded, upward facing fruit like the original variety.  The parent plants the seed was saved from did not have other peppers nearby to cross with.  From the unique segregation for green foliage and downward, elongated fruit we can learn some interesting information.  All three of these traits appear recessive, simply inherited traits and since they all segregate together the genes governing them may even be on the same chromosome. One can easily germinate saved seed and save the purple foliaged ones and be pretty certain the fruit will be round and upward facing.  We tasted these chili peppers.  Wow, they were exceptionally hot!


Tomato ‘Juliet’


Tomato ‘Juliet’ is an F1 grape tomato hybrid and a recent All-America Selections winner.  The plants are indeterminate growers (the shoot tips keep growing, producing fruit throughout the season), ripening over an extended period of time.  The fruit is very sweet, firm, and delicious.  The plants from saved seed segregated widely for fruit shape, size, and flavor.  Some were very elongated, while others were round and small like cherry tomatoes.  The seedlings also segregated for indeterminate versus determinate plant habit (upper left and right photos, respectively).  Determinate plant habit means that all terminals end in flower buds and produce fruit that mature at a relatively uniform time, a trait common in commercially grown field tomatoes that aids in mechanical harvesting and processing. With nine indeterminate to three determinate plants from the saved seed, the data is consistent with indeterminate growth being dominant to determinate growth and the original ‘Juliet’ being heterozygous for the trait and a cross of parents differing for these traits.  At the Dakota County Master Gardeners Annual Picnic, we taste tested fruit from each of the individual 12 plants from saved seed, comparing their flavor to that of the original ‘Juliet’.  The flavor of the fruit from the original ‘Juliet’ plants was very uniform and sweet and pleasant. People thought for ten of the plants grown from saved seed the original ‘Juliet’ had better flavor, one plant received a tie, and one plant tasted better than ‘Juliet’. With the segregation for determinate versus indeterminate plant habit and fruit shape in the plants from saved seed it is suspected ‘Juliet’ may be a cross of a sweet cherry tomato (typically indeterminate in growth habit) and a firm fruited, elongated roma tomato (typically determinate in plant habit). With great variability for fruit size and a general reduction in fruit flavor among the plants from saved seed of ‘Juliet’, it may be wiser to purchase new seed in order to enjoy the full benefits of this superior hybrid.

From this demonstration we learned that for many plants, even some F1 hybrids, saving and raising ones own seed produces plants of acceptable or highly comparable quality.  The decision regarding if we save our own seeds or not and for which types of plants depends on multiple factors including: budget, degree of uniformity needed, and how much satisfaction and enjoyment one gets from the process.  For more informal garden designs where variability is actually a desirable feature in especially ornamentals, much more leeway can be had in saving seeds and utilizing the resulting seedlings.  For more formal designs where mass plantings of more uniform plants are needed and high quality vegetables with uniformity for flavor and other important traits, much less variability can be tolerated.  We welcome you to come and visit UMore Park this summer and see new comparisons between seedlings from the original, purchased seed and saved seed of a different set of varieties.

Print Friendly and PDF