1. Introduction
Darrow’s blueberry, Vaccinium darrowii Camp, and lingonberry, V. vitis-idaea L., are two Vaccinium species with widely divergent biologies. While V. darrowii is a southern-adapted blueberry species of section Cyanococcus, V. vitis-idaea (lingonberry) is a low-growing circumboreal evergreen species of section Vitis-idaea. Morphologically, these two species differ widely.
V. darrowii is a rhizomatous, colonial species found in scrubby flatlands in a range covering Florida, Georgia, Alabama, and parts of Louisiana.
V. darrowii is evergreen with small waxy leaves and forms small bushes seldom over 90 cm high. Its flowers are urceolate to cylindrical, 4–6 mm long, and typically white but often tinged with pink. Plants often flower profusely and produce numerous small, waxy, and blue berries 8–10 mm in diameter.
V. darrowii is a diploid (2
n = 2
x = 24) and is particularly notable for the role it has played in the development of the southern highbush blueberry [
1].
V. darrowii has no flower-bud chilling requirement, and in the 1950s, a native
V. darrowii clone designated ‘Fla 4B’ was used to introduce lower flower-bud chilling requirements into northern highbush blueberry (
V. corymbosum) [
2]. The
V. darrowii germplasm was introduced via the use of 2
n gametes and resulted in
V. corymbosum-derived cultivars with approximately half the chilling requirement of typical northern highbush blueberry. Along with lower chilling requirements,
V. darrowii transferred plant vigor, an expanded range of fruit volatiles, and increased fruit firmness. Fla 4B has subsequently been determined to have a genetic ancestry that is 68%
V. darrowii and 32%
V. fuscatum [
3].
V. darrowii is considered by many researchers to be the most ancestral North American species [
4] and is possibly the most cross-compatible and most genotypically plastic.
In contrast, V. vitis-idaea (lingonberry) is a low-growing circumboreal evergreen species of the monospecific section Vitis-idaea. Lingonberry is rhizomatous and semi-woody and possesses dark green, leathery, obovate leaves. Its flowers have bell-shaped corollas and are produced singly or as clusters. Lingonberry is highly self-fertile. Plants produce red, shiny fruit, typically 7–11 mm in diameter. V. vitis-idaea is a diploid with 2n = 2x = 24.
Lingonberry fruit is harvested commercially in Scandinavian countries, Canada, and Alaska from both cultivated and wild stands. There has been much interest in growing lingonberry in southern temperate areas of North America; however, such attempts have failed due to a lack of proper adaptation and susceptibility to root pathogens that may kill entire stands (MKE personal observation).
Reports of hybrids of lingonberry with other
Vaccinium species are limited. Natural hybrids of lingonberry and
V. myrtillus (section
Myrtillus) termed
V. intermedium were described by Ritchie [
5,
6]. These hybrids apparently occurred repeatedly; however, their vigor and fertility were reported to be low.
At the diploid level, several successful efforts have hybridized 2
x V. vitis-idaea (lingonberry) and 2
x V. macrocarpon (cranberry) [
7,
8,
9]. Lingonberry was also successfully hybridized with the Hawaiian species 2
x V. reticulatum section
Macropelma (Ōhelo berry) [
10]; however, in both instances, only limited fertility was reported to exist, and no reports were found of advanced materials.
At the tetraploid level, Morozov [
11] succeeded in hybridizing 4
x V. uliginosum (bog bilberry) and a 4
x genotype of lingonberry. Unlike diploids, these F
1 hybrids were fertile and were subsequently crossed to 4
x cranberry, 4
x highbush blueberry, 4
x half-high blueberry, and backcrossed to 4
x lingonberry. Some of these materials were generationally advanced by Marozau and Baranov [
12], who described trispecific hybrids of (
Vaccinium uliginosum L. ×
V. vitis-idaea L.) ×
Oxycoccus macrocarpus (Aiton) Pursh (cranberry). Morosov [
11] also reported successful primary hybridization of 4
x lingonberry × 4
x lowbush blueberry (
V. angustifolium). More recently, Ehlenfeldt and co-workers [
13] reported the production of fertile 4
x hybrids of 4
x V. meridionale (Andean blueberry) with 2
x V. vitis-idaea via 2
n gametes.
In the course of investigating the crossability of multiple Vaccinium species belonging to different Vaccinium sections and different ploidy levels, we succeeded in hybridizing V. darrowii with V. vitis-idaea and generating multiple progenies. We report here on the evaluations of the most vigorous of these hybrids as investigated by post-pollination stylar staining, assessment of pollen quantity and quality, and female fertility determination (based on number of pollinations, fruit set, seed number, and seedling success). Additionally, we report on further hybridization of these two species and extensive evaluations of all generated hybrids, including ploidy determinations, genotyping using SSR markers, and other phenotypic assessments.
3. Discussion
The ability to produce an intersectional hybrid between a blueberry species, in this case,
V. darrowii, and red-fruited non-blueberry species,
V. vitis-idaea, is notable. The fact that numerous hybrids were readily generated in the first cycle of crossing with only modest numbers of pollinations suggests that the effective genetic distance between blueberry and lingonberry is modest. However, the substandard development of the seed from such crosses suggests that embryonic development programs are not fully compatible and are reflective of the evolutionary and geographic divergence between these two species. The small seed seen in these crosses is perhaps suggestive of female-excess type morphology [
15,
16] and is thus suggestive of
V. vitis-idaea having a lower genome strength than
V. darrowii [
18], even though, somatically, these species appear to be compatible at the diploid level.
Many of the true hybrids have appeared to have a juvenile period during which they were slow to develop. The vigor of the US 2535-A, however, suggests that at least some combinations of these two parents have the potential to be significantly heterotic. The repeat of this type of cross on a larger scale with additional female and male genotypes will likely allow the selection of additional types with superior vigor.
To this end, we repeated our initial combination with several other
V. darrowii genotypes and
V. vitis-idaea cultivars. In these crosses, we observed aspects of seed morphology that mirrored the initial cross. It is notable that the pollinations produced both normal-appearing seeds and reduced seeds—the reduced seed being similar to those that generated the initial hybrids. It is likely that some of the normal (‘good’) seeds gave rise to the self-hybrids observed in the second cycle of hybridization (
Table 4), although definitive conclusions cannot be drawn from the available data. In these pollinations, the development of the interspecific hybrid seed may have benefitted from the sharing fruit with the non-hybrid seed in what may be essentially a mentor pollen effect [
19,
20,
21]. In these second cycle crosses, ‘Johnblue’ appeared as the better female parent, and it was perhaps fortuitous that the initial iteration of this cross had been performed with ‘Johnblue’.
The evaluated F
1, US 2535-A, expressed only limited fertility as either male or female. This is not unlike other observations in diploid intersectional hybrids [
10,
22,
23]. The hybrid itself had no apparent self-fertility. The lack of self-fertility is almost certainly due to the combination of low female fertility, low male fertility, and low self-compatibility. The production of a few seeds when using the hybrid as female, however, gives optimism for its future use. As a female, US 2535-A had a success rate of 0.01 seed/pollination. Currently, we have two verified BC
1 plants, one of parentage US 2535-A ×
V. vitis-idaea ‘Sanna’, the other of parentage US 2535-A ×
V. darrowii ‘Native Blue’. Although diminishingly few in number, the success of these two hybrid combinations almost certainly guarantees that more such plants can be generated. We believe that the 75:25 genetic compositions of these backcross seedlings will likely improve both the male and female fertility in this BC
1 generation of plants. The
V. vitis-idaea backcross offspring are likely to allow the recovery of enhanced lingonberry types, with significantly improved heat adaptation and improved root-rot resistance. The backcrosses to
V. darrowii may prove useful in bridging this material to 4
x cultivated
V. corymbosum.
4. Materials and Methods
4.1. Plant Materials
The genotypes used and their origins are listed in
Table 6. All plants used were clones, and three to four plants of each genotype were used. Pollinations were performed on plants approximately five years old growing in 3 L pots in a 50:50 peat:sand mixture. All pollinations were performed in an insect-free greenhouse.
4.2. Crossing
Initially,
V. darrowii ‘Johnblue’ ×
V. vitis-idaea ‘Red Sunset’ crosses were conducted, and based on the success of this cross, other cultivars from each species were selected, and additional crosses were attempted to corroborate the cross-compatibilities (
Table 6). For all materials, pollen was extracted from open flowers by manual manipulation and collected on glassine weighing paper. Pollen was stored for up to a month under refrigerated and desiccated conditions until used for pollination.
Because of the small and somewhat delicate nature of the flowers on many of the parents, no emasculation was used. To perform pollinations, a graphite pencil tip was dipped into the collected pollen and applied to the stigmas of unemasculated flowers in an insect-free greenhouse. Pollinations were made on mature open flowers, and it was presumed that hybrids would be morphologically recognizable.
4.3. Stylar Examinations
To determine the viability of crosses, whole flowers with attached styles from experimental crosses were removed two days after pollination, using pollen as previously described, and fixed in a 3:1 solution of ethanol:acetic acid. After one week of fixation, the whole flowers were transferred to a solution of 90% ethanol until the date of examination. Styles were carefully detached from the remaining flower and stained using a 0.01% decolorized solution of aniline blue in 0.01 M potassium phosphate buffer. Upon preparation, the aniline blue solution was incubated at room temperature for 24 h to complete decolorization. To aid pollen tube visualization and facilitate aniline blue penetration of the inner stylar tissue, a tiny square 50% of the way down the style was delicately dissected on the stalk surface and meticulously removed so as not to disturb the inner contents of the style. An additional shallow and angled cut was made at the base of the style to detect pollen tubes that traveled the entire stylar length towards the floral ovary. After styles were whole-mounted in aniline blue solution and incubated for at least 2 h, slides were examined under ultraviolet light (excitation = 330–385 nm; emission = 420 nm), and tissue was imaged using an Olympus BX60 epifluorescence microscope (Olympus, Shinjuku, Tokyo, Japan) under its 10× objective, with Olympus cellSens Standard software (v1.11, core v3.10, Build 12201) (Olympus, Shinjuku, Tokyo, Japan). The examination of at least 3 styles per cross was determined to be sufficient for the clear assessment of cross success or failure, given that very little variation was observed within stylar samples.
4.4. Ploidy Determination
For flow cytometry, sampled young leaf material (1 cm2/20 to 50 mg) together with leaf material of an internal standard with known DNA content (Zea mays L.) were chopped with a sharp razor blade in 0.5 mL of extraction buffer (CyStain PI absolute P buffer, catalog number 05-5502; Partec, Münster, Germany) containing RNAse, 0.1% dithiothreitol (DTT), and 1% polyvinylpyrrolidone (ice cold) in a plastic petri dish. After 30 to 60 s of incubation, 2.0 mL staining buffer (CyStain PI absolute P buffer) containing propidium iodide (PI) as a fluorescent dye, RNAse, 0.1% DTT, and 1% polyvinylpyrrolidone was added. The sample, containing cell constituents and large tissue remnants of the chopped leaves, was then filtered through a 50 mm mesh nylon filter. After an incubation of at least 30 min at room temperature, the filtered solution with stained nuclei was measured with the flow cytometer (CyFlow ML (Partec) with a green diode laser 50 mW 532 nm (for use with PI); software: Flomax Version 2.4 d (Partec)). The DNA amount of the unknown samples was calculated by multiplying the DNA amount of the internal standard with the DNA ratio of the relative DNA amount of the unknown sample and the internal standard. DNA amounts were measured and compared to a set of standards covering the diploid to hexaploid range (2x V. darrowii ‘Fla 4B’, 4x V. corymbosum cv. Duke, and 6x V. virgatum cv. Powderblue) to determine basic ploidy levels.
4.5. DNA Analysis
In order to substantiate the hybridity of US 2535-A and the remaining hybrids with additional molecular information, the progenitors and putative hybrid seedlings underwent a paternity analysis by amplification of polymorphic simple sequence repeats (SSRs). The progenitor plants and putative hybrid seedlings were grown in a tunnel greenhouse at the P.E. Marucci Center for Blueberry & Cranberry Research in Chatsworth, NJ, in 2022 and 2023. Live cuttings were sent in 2023 by overnight mail to the USDA Cranberry Genetics and Genomics Laboratory (CGGL) in Madison, WI, for genetic testing. At CGGL, plant leaf tissue was freeze-dried using a BenchTop lyophilizer (Virtis, Gardiner, NY, USA), and DNA was extracted from dried leaf material.
A total of 0.03–0.04 g of freeze-dried leaf tissue per plant sample was pulverized to facilitate DNA extraction via a modified CTAB method [
28] with added beta-mercaptoethanol (2 µL in 750 µL CTAB) and incubation at 65 °C for 1 h. Solubilized DNA from each plant sample was retrieved from an aqueous layer after adding 500 µL of chloroform with isoamyl alcohol solution in a 24:1 ratio and centrifuging at 14,000 rpm for 6 min. DNA was precipitated by adding cold isopropanol to the aqueous layers and placing in a freezer overnight at −20 °C. The next day, the DNA solutions were centrifuged at 14,000 rpm for 22 min to form a DNA pellet. Each pellet was washed twice in cold 70% ethanol, then resuspended in 50 µL of 1× TE buffer (10 mM TrisHCl pH 8.0, 1 mM EDTA pH 8.0) plus 3 µL of RNase-A. The final DNA in 1× TE plus RNase-A was incubated for 3 h at 36 °C, then transferred to 4 °C until use.
For this study, a subset of SSR markers originally developed for
Vaccinium macrocarpon [
29] and shown to be cross-transferrable across
Vaccinium [
14] were systematically tested on the progenitor DNA until 6 sufficiently polymorphic SSR markers were found for determining the paternity of our putative hybrids. Polymerase chain reactions (PCRs) were assembled in duplicate for each plant sample, in a total reaction volume of 8 µL per sample. Each reaction was comprised of 5 µL 1× JumpStart REDTaq ReadyMix (Sigma, St. Louis, MO, USA), 1 µL of plant DNA in 1× TE buffer, 0.5 µL of Betaine PCR Reagent (5M, MilliporeSigma), 0.5 µL of 5 µM hexachlorofluorescein (HEX) M13 primer, 0.5 µL of 5 µM forward SSR primer appended with the M13 5′-CACGTTGTAAAACGAC-3′ sequence, and 0.5 µL of 50 µM reverse SSR primer appended with 5′-GTTTCTT-3′. The modifications to the SSR forward [
30] and reverse [
31] primers serve to facilitate fluorescent labeling of PCR fragments and promote non-templated adenylation, respectively.
PCR was completed on S1000 Thermal Cyclers (Bio-Rad, Hercules, CA, USA) on a program of a single melting step at 94 °C for 3 min, followed by 33 cycles of 94 °C, 55 °C, and 72 °C for 15 s, 90 s, and 2 min, respectively—and a final extension step at 72 °C for 30 min. In total, 1 µL of each PCR reaction was added to 10 µL of a pre-mixed solution of formamide and carboxy-X-rhodamine ladder (Custom MapMarker ROX 75-375bp, BioVentures, Murfreesboro, TN, USA) at a ratio of 1000 µL:25 µL. The PCR–formamide samples were sent to Functional Biosciences, Inc. in Madison, WI, USA, for fragment analysis. Samples were run on an Applied Biosystems 3730 fluorescent sequencer with a 50 cm capillary array. The raw results were sent back to the CGGL, and genotyping calls were determined using GeneMarker software version 1.91 (Soft-Genetics LLC, State College, PA, USA).
DNA fragment sizes (in bps) were determined at each marker for every plant in the study, including parentals and putative offspring (
Table 1 and
Table 4). At most, two distinct fragments were observed for a single marker, in agreement with the expectation that all plants in this study are diploid. The 6 markers used to determine hybridity were the following: SCF275d, SCF804, SCF9815, SCF37628, SCF132922, and 172672K70—whose sequences were previously provided [
14]. All 6 SSR markers amplify polymorphic loci on separate linkage groups, according to Schlautman et al. [
29].
4.6. Female Fertility
Cross numbers varied depending on flower availability. Pollinations and fruit set were recorded. Fruits were collected when ripe and measured for fruit size (mm in diameter) at the time of seed extraction. Extraction was performed manually under a dissecting microscope, and the seeds were evaluated for number and quality. For our purposes, seeds were classified as good, good-fair (g-f), fair, fair-poor (f-p), poor, or aborted. ‘Good’ and ‘fair’ described seeds that subjectively ranged from those considered fully normal to those somewhat reduced in size and/or development, but nonetheless were judged likely to be capable of germination. ‘Poor’ described seeds that displayed reduced size and/or development, were often flattened or brown, and were judged less likely to be capable of germination. Intermediate ratings were used as needed. ‘Aborted’ seeds were those that were flat and brown and generally translucent. Subjective observations were made on the size and quality of aborted seeds.
All extracted seeds were germinated directly on a greenhouse mist bench in a soil mix composed of 50:50, peat:sand mixture. At approximately a three true-leaf stage, seedlings were transplanted to 36-cell flats. Subsequently, plants were transferred to larger pots as appropriate for their growth stage.
4.7. Male Fertility
Pollen samples were stained with acetocarmine jelly (75% acetic acid with iron acetate) prepared according to the recipe of Jensen [
32]. Pollen samples were assayed for quantity, stainability, and general condition. A microspore was considered potentially viable if it hydrated to a plump configuration and stained. Pollen was evaluated to determine to what degree perfect tetrads or defective tetrads were formed.