Isolation Distance Requirements for Tomatoes

by Jeff McCormack, Ph.D.
Founder, Southern Exposure Seed Exchange

Most seed saving guides lack specific information about the minimum isolation distances for predominately self-pollinated crops such as tomatoes. Some specific guidelines are needed because frequent natural cross-pollination (NCP) of tomatoes may occur when two or more varieties are grown side by side in a garden under certain conditions. Even a small percentage of NCP over a number of years could eventually cause the loss of one or more characteristics which are unique to a particular open-pollinated variety.

Prompted by this concern, I am proposing some guidelines for isolation distances specifically suited to home gardeners who wish to keep their varieties pure. In preparing these guidelines I reviewed the scientific literature on tomato pollination, talked with three tomato breeders in different areas of the country, talked with people who have saved their family heirloom seed for many years, and have made my own observations of bee pollinator activity on tomato blossoms. The guidelines I propose are for both modern and old varieties of tomatoes, and further refinements may be necessary pending the results of experiments I have planned using genetically marked tomato lines.

Isolation requirements may be different for commercial growers than for home gardeners saving their own seed. Commercial growers or breeders of tomato seed may separate varieties by as little as 10 feet, primarily to avoid mechanical mixing of the seed crop. Such plantings are often made in areas which are bee-poor due to pesticide use or lack of suitable habitat. NCP may further be reduced by planting the seed crops in large blocks with a barrier crop in between. Furthermore, seed may be collected from plants in the center of the blocks and not the edges.

In contrast to large block plantings, many home gardeners tend to plant row crops of many varieties in a small space. These crops are frequently visited by wild bees (halictid bees, such as sweat bees) and bumblebees in search of pollen. This situation may contribute to a high frequency of NCP in bee-rich areas in crops that are primarily self-pollinated. Factors leading to higher NCP are explained below.

The amount of NCP of tomatoes is a function of a number of variables: (1) wind movement; (2) variety characteristics such as style length; (3) environmental variables affecting style length such as light density, day length and carbon-nitrogen ratio; (4) type of bee pollinator and its behavior on the blossom; (5) isolation distance; and (6) presence of other pollen-producing plants in the area of the seed crop.

Wind movement of the blossom tends to increase the amount of self-pollination. Even though tomato pollen may be blown some distance, NCP by this method is probably of little significance.

Tomato varieties having long styles (pollen-receptive organs) are more likely to be cross-pollinated by bees than varieties with short styles. If the length of the style exceeds the length of the anther cone (pollen-producing organ), NCP by bees is more probable, and probability increases as style length increases (see Fig. 1). Gardeners attempting to preserve old varieties need to be aware of this point because many older varieties generally have longer styles than modern varieties. Most modern varieties have styles equal in length, or shorter than, the anther cone (see Fig. 2). Our modern varieties were derived originally from wild tomato ancestors (primarily from Ecuador and Peru) which relied on bee pollination to a large degree. As these wild types were transported out of their center of origin to new geographic areas, the absence of their usual bee pollinators resulted in selection for variants which had shorter styles and an increased capacity for self-fertilization. Although style length is genetically determined, environmental conditions may cause style length to increase, thereby affecting the probability of cross-pollination.

Another factor affecting NCP is insect activity. Generally, tomato flowers are unattractive to bees if other pollen sources are available; however, in some bioclimactic regions of the U.S., bee visitation of tomato flowers may be quite common even in the presence of other pollen sources. Such a situation exists in regions of California and parts of the Mid-Atlantic region. For example, in parts of Virginia I have observed and photographed bumblebees (Fig. 3) and halictid bees (Fig. 4) such as sweat bees collecting pollen from tomato flowers. Bumblebees tend to vibrate the flowers while halictid bees appear to chew the anthers to get at the pollen. In terms of their behavior and position on the flower, halictid bees seem more likely to cause cross-pollination than the bumblebees, but this has not been fully investigated.

Close interplanting of two tomato varieties may typically produce 2-5% NCP; however, factors such as long style length, frequent visitation of tomato flowers by bees and suitable environmental conditions may produce much higher NCP values. Various studies have reported values of 12, 15, 26, and 47% NCP values in interplanted tomatoes. The wide range of results reflects the influence of different methods and variables used in these studies; however, it is clear that NCP values can be high under the right conditions.

What does all this mean for gardeners wishing to save their own open-pollinated tomato seed where there is high bee activity on tomato blossoms? Modern tomato varieties (style length equal or less than anther length in most cases) should be separated by a distance of approximately 10 feet to give a high degree of purity. Older varieties may require a 20 to 25 foot isolation distance. In both cases, a tall barrier crop or pollen-producing crop such as squash should be planted between different varieties of tomato to attract the bees away from the tomatoes. (Squash blossoms will present a lot of pollen to bees in the morning, but not in the afternoon.) In most cases, these recommended isolation distances will give average purity values of approximately 99 to 99.5% or better. Because occasional outcrossing may occur at large distances, plants used for stock seed may require an isolation distance of 50 to 150 feet or more in some localities.

The relationship between isolation distance and NCP is geometric, not linear. Thus as isolation distance increases, the amount of NCP falls off rapidly. A study by Currence and Jenkins (1942) illustrates this point very well (Fig. 5). Therefore it is evident that even a separation of a few feet between varieties in a small garden will greatly reduce NCP of tomatoes even though minimum recommended isolation distances cannot be achieved. Also NCP can be reduced or eliminated by taking advantage of different blooming times of early and late varieties.

Gardeners should not be discouraged from saving their own seed because of a small amount of NCP. A small amount of NCP could eventually improve a variety, but it could just as easily cause the loss of quality of a variety. If you are trying to preserve a variety in its purest form, then isolation distance becomes very important. Although a small amount of NCP may not be a problem one year, its effects may be additive and detrimental to preservation efforts in the long run. The goal is not just to save the variety from year to year but for generations to come.

NOTE: This article was previously published by the Seed Savers Exchange and several international periodicals. It has been reprinted here without illustrations. To contact the author send mail to Jeff McCormack.

Selected References

Bennet, J. 1983. A tomato blossom for all seasons. Horticulture 61: 53.

Currence, T.M. and J. M. Jenkins. 1942. Natural crossing in tomatoes as related to distance and direction. Proc. Am. Soc. Hort. Sci. 41: 273-276.

Rick, C. M. 1949. Rates of natural cross-pollination of tomatoes in various localities in California as measured by the fruits and seeds set on male-sterile plants. Proc. Amer. Soc. Hort. Sci. 54: 237-252.

Rick, C. M. 1950. Pollination relations of Lycopersicon esculentum in native and foreign regions. Evolution 4: 110-122.

(The 1984 Winter Yearbook, pages 247-250)