Wild grapevine population nuclei were mapped by a GPS receiver along Rivera de Huelva course. Accessions were assessed at flowering time (May), to determine the sex of the plants. Pollen samples from flowers were obtained by brushing the mature anthers from 10 male and 10 female vines. Grains were introduced in DPX (Fluka) and observed under optical microscope Olympus BX 61 in order to study the morphological structure of the grains.

The phenological development of the vines was followed twice per month from 2005 to 2011. The main ampelographic descriptors were evaluated following the OIV (2009) systematic list on a sample of 25 vines of each sex.

Soil samples were obtained at rooting depths (0-30 cm). The organic matter present in the soil was determined by the Walkley and Black´s (1934) method, organic carbon concentration by dichromate oxidation (Nelson & Sommers, 1982) and calcium carbonate by using a Bernard calcimeter (Álvarez-Iglesias et al., 2003). The pH of the soils was determined in saturated paste extract (1:2.5 v/v) with a portable pH-meter and electrode system (Crison pH/mV p-506, Spain), calibrated in the field.

Samples were dried at 50 ºC for 48 h. Dried soil samples were ground and homogenized by sieving through nylon nets of 2 mm mesh in order to remove large stones and dead material and then ground again to <60 µm. Soil samples of 2 mm fraction were analyzed for size particle distribution by the hydrometer method (Gee & Bauder, 1986), Kjeldahl-N by the method described by Hesse (1971) and available P by the method of Olsen et al. (1954). The extraction was made according to Lindsay & Norvell (1978). Available metals (Al, As, B, Ba, Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, P, Pb, S and Zn) of all samples were measured by inductively coupled plasma (ICP) spectroscopy (Advantage de Thermo, Beverly, MA, USA) (Walinga et al., 1995).

Roots were unearthed up to 40-50 cm of depth to observe possible symptoms caused by subterranean phytophagous and pathogens. Samples of fine roots were observed under binocular to detect possible damages caused by Phylloxera (Daktulosphaira vitifoliae Fitch) (Homoptera, Phylloxeridae), root-knot nematodes and rot fungi.

The detection of possible symptoms caused by phytophagous arthropods, insects and mites, and diseases was carried out in spring and summer, every 15 days along the six years of this investigation, on shoots, leaves and bunches of 100 vines, situated up to 3 m height. In the case of mites 50 leaves/vine were observed to evaluate the intensity of infestation.

Special attention was paid to try to find mealybugs because some of them are vectors of virus, as Planococcus citri (Risso) (Hemiptera, Pseudococcidae) that transmits Grapevine leafroll virus. For this task, 25 vines exhibiting a perimeter higher than 20 cm were debarked from 2009 to 2012 to find the hemipteran.

The degree of infestation/infection caused by mites was evaluated according to the procedures of Ocete et al. (2007) and OIV (2009) in the case of both kinds of mildews. To detect the possible presence of Grapevine fan leaf, Grapevine leafroll and Grapevine fleck viruses, 10 samples from the oldest female and male vines were taken in June 2010. ELISA tests were performed using the Bioreba kit systems.

Together with phytophagous sampling, natural predatory species of Chrysopidae and Coccinelidae were identified. On the other hand, to investigate mite species found inside erinea, 25 available leaves with erinea were randomly chosen from ten individuals in July. The leaves were put in a chamber at 4 ºC for 20 min to slow down the activity of the predatory mites. Immediately after, erinea were studied under the binocular to pick up natural enemies of this mite in action. The Berlese-Tullgren funnel was also used to extract the mites from the erinea. Predatory specimens of Colomerus vitis (Pagenstecher) (Acari, Eriophyidae) were separated and introduced into tubes with ethanol (70%) with 2 drops of glycerin. Mites were cleared in lactic acid and mounted on Heinze-PVA medium. Observations were made using an interference contrast microscope. Generic nomenclature for the Phytoseiidae follows the criteria proposed by Chant & McMurtry (1994) for Typhlodrominae and those referred by De Moraes et al. (1986) for Amblyseiinae.

For microvinification berries from 3 available vines were harvested on 26th September 2012. In total, 10.55 kg of bunches were obtained. The use of the grape destemmer reduced the sample to 8.79 kg. After destalking, berries were pressed by hand, obtaining 1,788 mL of must. The alcoholic fermentation with maceration and daily removing was developed by wild yeasts, at 20 ºC, along 10 days and monitored by density measures. References of the official analytic methods used from OIV (2015) and data obtained are shown in Table 1.

Since wild grapevines are mostly dioecious while cultivated are generally hermaphrodites we have only used male plants for the genetic analysis as they cannot be escapes from cultivated fields nor the result of pollination between wild and cultivated plants. A sample of 63 male accessions were analyzed (Table 2) and compared with a published genotype data, corresponding to Spanish wild grapevine populations (De Andrés et al., 2012).

Total genomic DNA was extracted from young leaves, using the DNeasyTM Plant Mini Kit (Qiagen). Extracted DNA was quantified and used as a working DNA solution of 5 ng/µL. PCR included 11 microsatellites: VVS2 (Thomas & Scott, 1993), VVMD7, VVMD24, VVMD25 (Bowers et al., 1999), VMC1B11 (Zyprian & Töpfer, 2005), VVMD5, VVMD21, VVMD27, VVMD28 (Bowers et al., 1999), VVIV60 (Merdinoglu et al., 2005) and VMC4F3.1 (Di Gaspero et al., 2000). PCR amplifications were analyzed in an ABI 3130 Genetic Analyzer (Applied Biosystem, Foster City, CA, USA) using GeneScan-LIZ 500 as internal marker (Applied Biosystems). Amplicon fragments were sized with GeneMapper 4.0 software.

Allele size and total number of alleles were determined for each SSR (Simple Sequence Repeat). Putative alleles were indicated by their estimated size in bp. Genetic diversity was estimated using the following statistics: number of alleles (Na); mean number of alleles per locus; effective number of alleles (Ne); observed heterozygosity (Ho); expected heterozygosity (He) (Nei, 1973); and fixation index (F), also called inbreeding coefficient. All those statistics were calculated using GenAlex software version 6.0 (Peakall & Smouse, 2006) and the Excel Microsatellite Toolkit (Park, 2001).

Bayesian clustering was applied on the SSR genotype data using the software package STRUCTURE, vers. 2.1 (Falush et al., 2003). Analyses were performed with the total collection (63 unique genotypes). Only accessions with ancestry values superior to 0.7 were included in each population analyses. Admixture model and correlated allelic frequencies were used to analyze the dataset without prior population information. Ten simulations per K value were performed for each number of populations (K) (set from 1 to 10). Burn-in period and Monte Carlo Markov Chain length were set up to 100,000 and 300,000, respectively in each run. To assess the best K-value supported by the data we calculated the second order change of the likelihood function divided by the standard deviation of the likelihood (ΔK) (Evanno et al., 2005).

Micropropagation. From cuttings of the considered population, uninodal explants, 1 cm length were prepared. The explants were disinfected by the following steps: a) washing with water and detergent, gently rinsing with distilled water; b) immersion in 70% ethanol for 45 s; c) immersion in a solution of 45% sodium hypochlorite (3.5 % active chlorine) with some drops of Tween-20 (20 min at 30 ºC with stirring); d) gently rinsing with distilled, sterilized water.

After disinfection, the explants were placed individually in sterile test tubes (25 × 150 mm) with 8 mL of “VID” medium (Sarimento et al., 1992) with 0.13 µmol/L indolebutyric acid (IBA) and 0.32 µmol/L benzylaminopurine (BA) with 0.6% agar, pH 5.7. Each tube was covered with a plastic cap, sealed with parafilm and placed in a growth chamber at 23±2ºC, 30 µmol/m²·s of light intensity and 16 h of photoperiod. Adaptation to outside conditions was carried out according to Cantos et al. (1998).

Conservation in germplasm banks. Sixty explants from the in vitro cultures were chosen and cultivated in the same conditions described above. After 15 days, 30 of these explants were cultured at the same conditions (NC) and the other 30 were placed in a fridge at 4 ºC and darkness to slow the growth (DC). After two and six months, 15 explants of each condition were taken and micropropagated again (1 cm and 1 bud) in De Wit tubes with 8 mL of the same medium and cultured at NC conditions of light and temperature. After 30 days the survival, stem length, bud (nodes), lateral shoots and root numbers, and rooting percentages were recorded in all situations considered.

Descending the river course, the best conserved portion of the bank forest is situated very close to La Minilla reservoir (37º40’2.8’’N, 6º10’47.8’’W - 37º40’27.5’’N, 6º10’5.5’’W) with 87 vines: 53 males and 34 females. Such kind of gallery forest constitutes an azonal formation with deciduous species crossing the Myrtocomunnis-Quercetorotundifoliae series of the climax vegetation of that geographical area (Valle, 2003). Very close to that main populational nucleus there are three more situated along occasional creeks: (37º40’5.7’’N, 6º10’9.5’’W), with 11 vines, 7 males and 4 females; (37º39’34.3’’N, 6º9’25.3’’W) with 7 vines, 6 males and 1 female, and (37º39’50.5’’N, 6º9’47.2’’W) with 10 vines, 7 males and 3 females. In total, the whole population comprises 115 vines.

The approximate phenological calendar was bud swelling: 21 March−3 April; flowering: 15 May−2 June; veraison (onset of ripening): 12 July−10 August; fruit ripening: 16 September−30 September; end of leaf fall: 20 November−7 December.

Main ampelographic descriptors are compiled in Table S1 [suppl.]. The main features found were as follows: A 63% of the vines are males, with a female/male ratio of 0.60. Inflorescences were more abundant in male individuals than in female plants. Female flowers had reflexed stamens while all the male flowers found corresponded to type I (fully developed stamens, no gynoecium).

Young shoots had low intensity of anthocyanin coloration. Mature leaves were bigger in female plants and mainly wedge-shaped while in male plants leaves were more frequently pentagonal. The male grain of pollen is tricolporate, quite similar to that belonging to hermaphrodite cultivars, and the female one is a unaperturate ovoid sac.

Only female plants had berries. All the berries were small, rounded and reddish. The seed morphology is subspherical with a small beak. Wide/length ratio is higher than in the cultivated varieties.

The results of soil analysis are shown in Table 3. According to the percentages of fine sand, silt and clay the soil studied is classified as sandy loam (Haplic Fluvisol). The pH was 7.29, classified as neutral. Electrical conductivity was only 0.052 mS/cm, very low in relation to levels considered normal (< 4 mS/cm) (Báscones, 2004). Level of organic matter was normal and, in consequence, also C and N contents, with a C/N rate of 10.7 considered as medium level (Báscones, 2004). Soil was very poor in available phosphorous. Concentration of available potassium was also very low if compared to the 200 mg/kg of available potassium of normal agricultural soils (Pérez de Mora et al., 2006). Because the percentage of CaCO₃ was very low, the available calcium levels were also low. On the contrary, available magnesium concentration, 274 mg/kg (Table 3), showed higher contents that those considered normal ranging from 180 mg/kg (González et al., 2003), to 120 mg/kg (Báscones, 2004). Na level was within normal values and Fe concentration was slightly high.

Other microelements considered as toxic above certain thresholds as Ba, Cd, Cu, Mn, Pb, S and Zn presented normal values (Table 3).

On roots, no symptoms of infestation or infection attributable to Phylloxera, root-not nematodes or mycelium of rot fungi were detected.

In the vine leaves, damages caused by mites, concretely by the Erineum strain of Colomerus vitis were very frequent in all the vines progressively developing from March to the end of July. The percentage of leaves affected varied usually between 10 to 20%. Symptoms of the presence of Calepitrimerus vitis (Nalepa) (Acari, Eriophyidae) were scarcer, developing from March up to the beginning of June, and affecting from 3 to 18% of the leaves.

Small spots of infestation caused by larvae and adults of Haltica ampelophaga Guérin and Méneville (Coleoptera, Chrysomelidae) were very occasionally found on leaves. Several shoots and leaves showed bites caused by Decticus albifrons (Fabricius) (Ortopthera, Tettigonidae) and Ephipiger ephipiger Fiebig (Ortopthera, Tettigonidae). In June 2012, a very dry year, several specimens of Doctiosaurus maroccanus (Thunberg) (Ortopthera, Acrididae) were also detected in the external part of the gallery forest, but there was no evidence of feeding on vines. On the other hand, some small colonies of Aphis gossypii Glover (Hemiptera, Aphididae) were present only on 7 vines between middle of May to the end of June.

About the origin of the fungal diseases it is necessary to remark that in the surrounding area where this study was carried out there are not vineyards, only some cottages and ancient buildings with very isolated vines. The nearest wine producing area is about 40 km far from the studied zone. The presence of symptoms caused by Erysiphe necator (Schweinitz) Burrill was common in most of the vines, mainly on leaves and shoot axes, affecting between 10-21% of the leaves and around a 15% of bunches. Oil spots on leaves together with other damages on shoot axes and bunches caused by downy mildew (Plasmopara viticola (Berkeley and Curtis) Berlease and de Toni), were also frequent mainly in those vines situated in the shadiest places of the gallery forest, where the level of infection on leaves reached 17% and around 9% on bunches. Symptoms caused by Botrytis cinerea Pearson were not found any year.

Regarding the possible virus infections, all the ELISA tests were negative for Grapevine leafroll, Grapevine fan leaf and Grapevine fleck viruses. No symptoms were detected in the wild vines.

Groups of pedunculated eggs laid by Chrysoperla carnea (Stephens) (Neuroptera, Chrysopidae) are frequent. About coleopterans, two species of ladybirds were identified: Coccinella septempunctata L. (Coleoptera, Coccinelidae), the most frequent species, and Adalia decempunctata (L.) (Coleoptera, Coccinelidae).

The maximum number of mobile forms of predatory mites per erineum of C. vitis, was three. The species found were: Typhlodromus phialatus Athias-Henriot (Acari, Phytoseiidae), Neoseiulella litoralis (Swirski and Amitai) (Acari, Phytoseiidae), Euseius stipulatus (Athias-Henriot) (Acari, Phytoseiidae) and Phytoseiulus persimilis Athias-Henriot (Acari, Phytoseiidae)

Analysis of the microvinification is shown in Table 1. The experimental wine resulted with medium concentration of alcohol, similar to that obtained with traditional cultivars from the region. The total acidity was relatively high, and consequently the pH value was much lower than that obtained with traditional Andalusian cultivars. It is a very important characteristic to highlight, because it would need no addition of tartaric acid to give stability. The volatile acidity was slightly high due to the absence of potassium metabisulphite addition. The color intensity is medium for a red wine due to the irregular degree of maturation of the berries, but it was not possible to delay the harvest due to heavy incidence of berry-consuming birds. The concentration of methanol is correct, and fulfills Spanish legislation, that allows up to 400 mg/L. The dry extract is normal for a red wine and the total polyphenols index measured at 280 nm is also normal.

Genetic diversity of the wild populations. The results of the genetic diversity analysis, according to DNA microsatellite genotyping are shown in Table 2. These genotypes showed a mean value of 6.3 alleles including an average of 0.875 private alleles (alleles found in a single population throughout the study region). Genetic comparisons between the wild grapevine genetic groups from the Iberian Peninsula named here as Northern and Southern Spain wild populations (De Andrés et al., 2012) and La Minilla allowed to find out that the high number of private alleles presented in La Minilla population is similar to the number found in the Northern and Southern wild populations (Table 2). This result pointed out that some alleles in these populations are not present in the other Spanish populations. The mean number of alleles between the different populations ranged from 6.3 to 9.1. The number of effective alleles (Ne) ranged between 3.4 and 4.6. The information index (I) ranged from 1.3 to 1.7; Ho ranged between 0.608 and 0.687, whereas the values of He were slightly higher, ranging between 0.636 and 0.766. The fixation index (F) showed values ranging from 0.012 to 0.121, supporting the existence of inbreeding depression in some wild grape populations.

Population structure between wild grapevine populations from the Iberian Peninsula. The genetic structure analysis included Northern and Southern Spain wild grapevine populations (218 samples) and the wild accessions from La Minilla (63 samples). The choice of a fixed number of populations (K) is arbitrary, each of them characterized by a set of allele frequencies at each locus. Individuals in the sample are probabilistically assigned to genetic groups or jointly to one or more groups. Consequently, we performed an analysis where K varied from 1 to 10. This analysis showed a large increase of likelihood from K=1 to K=3 and smaller increases from K=3 to K=10. Using the methodology of Evanno et al. (2005), this result supports K=3 as the most likely number of genetic groups, which corresponds to wild individuals from Northern, Southern and La Minilla populations (Fig. 1). Then, it may be supposed that La Minilla genetic group comprise a unique population that could be the result of fragmentation due to the recent human alteration of the habitat.

Micropropagation. The in vitro response of the explants obtained directly from the field individuals is indicated in Table 4. After 30 days of in vitro culture, the 64.9% of the explants formed stem and leaf with good growth (averages of 5.5 cm length stem, 9.2 nudes and 1.6 of lateral shoots), and 97.8% of rooting, with an average of 3.5 roots per plant with 7.3 cm of mean length with a very good aspect. In consequence, the acclimation to outside conditions was 80%.

Conservation in germplasm banks. Survival rate of explants was high after 2 months for both normal (NC) and dark cold room (DC) conditions. Although DC treatment showed more survival rate than NC cultures the difference was not significative, ranging from 80 to 100%. After 6 months of culture, survival rate fell to 40% both for NC and DC treatment (Table 5).

The DC treatments reduced the explants growth compared to NC cultures (Table 5). No great differences were observed between 2 or 6 months of culture in DC conditions. In standard conditions, 6 months of culture resulted in lower growth, both in terms of stem length and number of buds (Table 5). The mean number of lateral shoots per plants although higher in the NC culture, showed no significant differences between treatments (Table 5).

Rooting percentage was 100% for both culture conditions independently of the storage period. Mean number of roots was similar for all the conditions except for NC after 2 months that was higher (Table 5).