Breeding for Parthenocarpic Fruit Development in Blueberry

Abstract

The USDA-ARS breeding program has identified a recessively-controlled genetic variant in highbush blueberry (V. corymbosum L.) that results in parthenocarpic fruit development. Current studies are underway to refine our understanding of this variant, and to move this trait into desirable and useful genetic backgrounds. In rabbiteye blueberry (V. ashei Reade), a species with critical cross-pollination requirements, significant variation in ability to set fruit without insect pollination has been observed under greenhouse conditions. A current study is evaluating linkages between greenhouse and field behavior of rabbiteye cultivars with regard to fruit set and development, with the long-term goal of enhancing productivity. Breeders often select naturally for cultivars with parthenocarpic tendencies, but improved understanding of the sources of parthenocarpy, and their genetic control will facilitate further advances in this area.

Keywords:

• pollination
• seedless
• fruit set
• rabbiteye
• highbush

INTRODUCTION

Parthenocarpy in any fruit-bearing crop is a very desirable trait because it holds the possibility of reduced concerns about many typical pollination worries. In a cross-pollinated crop like blueberry, parthenocarpy could address at least three pollination concerns: (1) sub-optimal pollination weather during the bloom period, (2) declines of pollinators due to parasitic infections, and (3) desirability of cross-pollination versus self-pollination for optimum yield.

Normal fruit development is initiated by gibberellins or auxins generated by the developing ovule. These compounds stimulate early fruit development and also promote fruit set (i.e., the prevention of abscission) (Crane, 1964). In some cases, parthenocarpy, the development of fruit without seed, may occur. In parthenocarpic genotypes, pollination either is not needed, or if it does occur, it does not result in fertilization and seed development. Natural parthenocarpy has been documented in various crops, including banana (Gustafson, 1939), grape (Wong, 1941), mandarins (Talon et al., 1992), opuntia (Weiss et al., 1993), and tomato (Groot et al. 1987; Wong, 1941). Furthermore, considerable work has been conducted on the induction of parthenocarpic fruit using growth regulators in a wide variety of horticultural crops either to study fruit development or to evaluate potential utilization in production. Among such crops are cucumber (Kim et al., 1994a, 1994b), grape (Fellman et al., 1991), kiwi fruit (Ohara et al., 1997), pears (Yamada et al., 1991), pepino (Ercan and Akilli, 1996), strawberry (Mudge et al., 1981), and watermelon (Pak, 1993).

In blueberry, limited work has been done, but several groups have demonstrated that normal fruit development exhibits high levels of gibberellic acid (GA) production during early stages of development in both highbush and lowbush blueberries (Kender and Desroches, 1970; Mainland and Eck, 1968, 1971). In addition, it has been shown that parthenocarpy may be induced by GA₃ in cases of pollination failure (Cano-Medrano and Darnell, 1998; Mainland and Eck, 1969; Nesmith et al., 1995). Several authors have noted what appear to be tendencies toward parthenocarpy in blueberry (i.e., some parthenocarpic fruit, low seed numbers, high pulp to seed ratios) (Ehlenfeldt, 2001; Ehlenfeldt and Hall, 1996; Harrison et al. 1994; MacKenzie, 1997).

RESULTS AND DISCUSSION

Parthenocarpy Model in Highbush Blueberry

Previously, we developed a genetic model of parthenocarpy in progeny derived from the northern highbush (V. corymbosum L.) clone, G-176 (Ehlenfeldt and Vorsa, 2007). This model postulates that parthenocarpy is the result of a homozygous, single-gene, recessive allele at a locus with incomplete dominance. In the model, PP_ _ genotypes yield normally-seeded fruit, Pppp genotypes yield “diminished seed” fruit (i.e., fewer and smaller seed), and pppp genotypes are parthenocarpic (Fig. 1). The major obstacles to the further use of this germplasm were that the presumed fully recessive types, while parthenocarpic, also appeared to suffer from low overall productivity. The diminished-seed types (simplex heterozygotes), however, were often relatively large-fruited and considerably more productive. We have subsequently endeavored to do two things: (1) to further verify the inheritance of this character primarily by evaluating crosses of diminished-seed x diminished-seed phenotypes (Pppp x Pppp), which should yield high frequencies of both diminished-seed phenotypes (50%) and parthenocarpic phenotypes (25%), and (2) crosses of diminished-seed phenotypes x F₁s (Pppp x PPpp), where the F₁s have been selected primarily for productivity and vigor. These should also yield relatively high frequencies of diminished-seed (42%) and parthenocarpic phenotypes (8%), and, hopefully, some of them will benefit from the parental selection for vigor.

FIGURE 1 Ranges of seed development expressed in families segregating for parthenocarpy: (a) normal, (b) diminished-seed type, and (c) parthenocarpic type (scale in cm). (Source: Ehlenfeldt and Vorsa, 2007).

Since the parthenocarpic allele was initially found in the domesticated gene pool, we believe that this or similar alleles circulate in the highbush gene pool and that breeders often, directly or indirectly, select for these alleles. Several of the hybrids initially produced for the inheritance study had the parentage ‘Duke’ x G-176 (G-176 was the original parthenocarpic clone). Two of these plants have set large fruit under field conditions that (in retrospect) we would characterize as diminished-seed types (Pppp) (Fig. 2). This implies that ‘Duke’ may already carry some dosage of these recessive parthenocarpic alleles and could be of the genotype PPPp, PPpp, or Pppp, and, thus, may have contributed a gamete of the composition Pp in the cross generating these two hybrids. This concept is supported by the finding of Draper and Scott (1969) that large fruit size is inherited recessively.

FIGURE 2 Fruit and seed development in an offspring of ‘Duke’ × G-176 (scale in inches).

Advancing the p Allele in the Hexaploid Gene Pool

A number of years ago, we crossed a synthetic 6x southern highbush (shb) clone, NJ 89–158–6, with G–176 (the original parthenocarpic clone). From this cross, we selected US 1232, a probable pentaploid, that was not notable except that it produced some fruit and was relatively vigorous. Under a conservative assumption, the allelic composition of US 1232 is PPPpp, and it might, if it produces balanced and randomly assorted 3x gametes, produce all of the following: PPP gametes (10% relative frequency), PPp gametes (60% relative frequency), and Ppp gametes (30% relative frequency). US 1232 has been used as a male with clones of rabbiteye (re), V. constablaei Gray (6x con), US 865 and US 872 (6x shb/con), and US 1056 (6x shb/con/re). US 1056 already seems to express parthenocarpic tendencies (high fruit set, high yield, low seediness), although its exact nature is unstudied. Conservatively, the best-case allelic composition of offspring from crosses utilizing US 1232 with any of these selections could be PPP/Ppp (30% frequency). Obviously, further backcrossing or sib-mating among these clones would probably be needed to recover a genotype expressing the parthenocarpic phenotype, but a single generation might be sufficient to recover the desired types, depending upon the relative effects of incomplete dominance within a hexaploid genome.

Parthenocarpy among Rabbiteye Blueberry Cultivars

Under greenhouse conditions, several rabbiteye blueberry cultivars regularly set fruit and develop it to maturity without pollination. Among the more notable cultivars exhibiting this behavior are ‘Baldwin’ and ‘Premier’. Since self-fruitfulness (or lack of) is a critical issue for rabbiteye, and a potential concern for northern-adapted rabbiteye types, we undertook a study of the relative propensity of 48 potted rabbiteye cultivars to set fruit under greenhouse conditions. We measured fruiting interval, fruit set, ripening interval, and fruit size, and also monitored temperature conditions during the study with the goal of not only documenting the performance of different cultivars, but also understanding as much as possible about the conditions leading to parthenocarpy. This study is in its third and (hopefully) final year.

Although the data have not yet been fully analyzed, a few cultivars stand out as having distinct ranges of parthenocarpy when not pollinated. Three older cultivars, ‘Early May’, ‘Ethel’, and ‘Owen’ have virtually no fruit-set without pollination. ‘Baldwin’, ‘Premier’, and ‘Maru’, three modern cultivars, have relatively large fruit at maturity (10–12 mm), but modest overall set (Fig. 3A). In contrast, ‘Chaucer’, a modern Florida cultivar, has extremely high fruit-set without pollination, but its mature fruit are relatively small, typically 5–7 mm (Fig. 3B). Is there an inverse relationship between parthenocarpic fruit size and degree of fruit set? Probably not. ‘Chaucer’ is unique in its relatively high fruit set, but it is easy to find cultivars with low set and small fruit (i.e., not an inverse relationship).

FIGURE 3 Parthenocarpic fruit development in (a) ‘Baldwin’ and (b) ‘Chaucer’ rabbiteye cultivars under insect-free greenhouse conditions.

The most critical issue with this experiment is how greenhouse data will relate to field results. To address this, we are working with Dr. Stephen Stringer of the USDA in Poplarville, Mississippi to compare parameters of field-grown fruit (size, weight, seed/pulp ratios) to see if cultivars with superior parthenocarpic performance in the greenhouse also exhibit less need for pollination to achieve superior yields under field conditions. In a future phase, we expect to also examine self-fertility in relation to parthenocarpy and field fruit development. Prospectively, we have made crosses of the high set/small fruit size cultivar ‘Chaucer’ with low set/large fruit size cultivars (‘Powderblue’, ‘Maru’) to further evaluate inheritance of these two factors.

The Value of Parthenocarpy Beyond the Obvious

Blueberry breeders of several generations have had interest in hybridizing highbush and rabbiteye, and transferring genes between them. This interest has resulted in many southern highbush cultivars with rabbiteye germplasm in their background. A few rabbiteye cultivars possess non-rabbiteye germplasm (‘Snowflake’, ‘Pink Lemonade’), but to date little outside germplasm has been utilized in rabbiteye breeding. Many breeders freely acknowledge that mainstream rabbiteye could benefit from the incorporation of some of the fruit characteristics of highbush and southern highbush (organic acid composition, sweetness). A few cultivars are pentaploid first-generation crosses between highbush and rabbiteye (‘Pearl River’, ‘Robeson’), but breeders will also acknowledge that in general, among such first generation crosses, there is often low productivity, generally attributed to low fertility. Certain rabbiteye-highbush combinations seem to work better than others, but in most cases, ploidy has to be re-established at either the 4x or 6x level for good fertility and fruitfulness. However, imagine a rabbiteye x highbush cross where both parents have pronounced parthenocarpic tendencies. Would this produce hybrids with adequate fruit production in which fertility would be much less of a concern? We don't fully understand the interaction of parthenocarpy with fertility, and a threshold level of fertility would probably be needed for optimum results, but the possibilities are tantalizing.

CONCLUSIONS

The nature of parthenocarpy, as we have deduced it, thus far, for both highbush and rabbiteye blueberries is an imperfect system. Parthenocarpy, however, doesn't have to be a perfect system to be useful. In terms of fruit set and fruit development, even a modest improvement in fruit set and development could have substantial value. Our functional goal is to help the bees, not replace them!

ACKNOWLEDGMENT

I wish to thank Robert B. Martin Jr. for technical support.

Notes

This article not subject to US copyright law.