Two leaves per tree and a minimum of three trees per cultivar were analyzed with the infrared gas analyzer. All leaves were collected and leaf area was quantified with a leaf area meter to normalize the photosynthesis data . Only the most recently fully-formed, sun-exposed leaves were selected for this study. Cuvette temperatures were allowed to vary with field conditions. Leaves were measured in a chamber that provided 1500 µmol m-2 ·s-1 . Instantaneous water-use efficiency was calculated as A/E and intrinsic water-use efficiency was calculated as A/gs. The difference between these two traits is that intrinsic water-use efficiency corrects for differences in vapor pressure deficit while instantaneous water-use efficiency does not. This is important because knowing if climate affects water-use efficiency is useful for cultivar selection. Predawn and midday stem water potential measurements were taken for each day of photosynthesis data collection. For both predawn and midday water potential, nonactively growing shoots were covered with a plastic bag for 10 min before being pruned, placed in a sealed plastic bag and kept in a cooler bag until transferred to a cool, lit indoor environment for plant moisture stress measurements with a pressure chamber . For afternoon stem water potential measurements,raspberry container size canopy-shaded non-actively growing shoots were covered with a plastic bag for 10 min before being pruned.
Flowers were counted during peak bloom, defined as 90% or more of the flowers being open in the grove, which was typically in April and early May of each year . Male flowers, as identified by their signature bell shape in pomegranates, were counted along with the female flowers, which were identified by their easily recognizable vase shape. For reasons unknown, ‘Haku Botan’ flowers had morphological differences between sites that made determining male and female difficult at Somis, the coastal site. Male to female flower ratio was calculated by dividing number of male flowers by the number of female flowers. Total number of flowers was determined by adding male flowers with female flowers. Number of fruit was determined by counting fruit after fruit set, which was determined in the months of July and August. All variables were analyzed with Analysis of Variance . When ANOVA indicated significant differences, post-hoc comparisons were performed utilizing Tukey’s honestly significant difference with an experiment wise error rate of α = 0.05. Relationships between all variables were analyzed using linear regression , with correlations among parameters determined using general regression with Minitab Software, version 16 . For the purposes of this work, the R 2 value is the percentage of variation in one variable that is explained by the variation in the correlated variable. Regression models were fit to determine differences in slope coefficients and constants among variables using regression analysis. Trees at both sites established at varying rates, with low tree mortality for some cultivars and no tree mortality for others . By year four, trunk diameter was significantly different between sites , with Riverside having 7.6 cm trunk diameter on average and Somis having an average of 6.4 cm trunk diameter.
Every year, trunk diameter was larger at Riverside compared to Somis. There were significant differences among cultivars for trunk diameter at Somis, but not at Riverside . ‘Desertnyi’ had a relatively thin trunk at Somis, which was significantly lower than ‘Wonderful.’ Canopy area was larger at Somis , with Riverside having 34.1 m2 canopy area on average and Somis having an average of 37.4 m2 canopy area . There were significant differences among cultivars at each site . At Riverside, Desertnyi, Eversweet, and Loffani had the lowest rate of establishment, with these cultivars having the least vigor as measured by vegetative growth when compared with the other cultivars in the study. At Somis, four cultivars grew at a significantly higher rate than the others. The faster establishing cultivars at the coastal site were Ambrosia, Haku Botan, Proprietary cv. and Wonderful. Of the remaining cultivars at Somis, Desertnyi, Eversweet and Golden Globe were the slowest growing cultivars during establishment. Results for in-row and between-row establishment are similar . There were significant differences between sites and among cultivars for relative chlorophyll content. Riverside had higher leaf relative chlorophyll content than Somis , with average values of 47.7 and 42.4 SPADs for Riverside and Somis, respectively. ‘Haku Botan’ had the highest relative chlorophyll at both sites . At Riverside, Haku Botan had the greenest leaves of all cultivars according to SPAD readings.
Haku Botan was significantly greener than Loffani at Riverside and significantly greener than all cultivars except for Green Globe and Proprietary cv. at Somis. Of all cultivars at both sites, Loffani had notedly low relative chlorophyll content at Riverside. There were significant differences among cultivars and between sites for tree height . Somis had the tallest trees , and had mean tree height 231.4 cm compared to 182.7 cm at Riverside. The tallest cultivars at Riverside were Haku Botan, Phoenicia, Proprietary cv. and Wonderful, which were significantly taller than Desertnyi and Eversweet. At Somis, Ambrosia, Golden Globe, Haku Botan and Proprietary cv. were tallest, and these cultivars were significantly taller than Desertnyi and Eversweet. Results for in-row and between row canopy diameter were similar to those of canopy area. Trees at Somis established and grew larger than trees at Riverside . Values for in-row and between row canopy diameter were similar for a given cultivar, demonstrating uniformity of in-row and between-row growth. There were also significant differences among cultivars at each site, with Desertnyi and Eversweet smaller than most other cultivars at both sites. The pomegranate cultivars were actively photosynthesizing and transpiring during morning hours at both sites during all days of data collection . There were significant differences among cultivars for photosynthesis at Riverside but not at Somis. ‘Haku Botan’ had significantly lower rates of A , gsand Ethan ‘Wonderful’ during the morning hours. There were no other differences detected for photosynthetic variables among the four cultivars investigated. The effect of site on pomegranate physiology was significant. Riverside had significantly higher rates of photosynthesis than Somis, with mean 15.0 and 12.7 µmol CO2 m-2 ·s-1 for Riverside and Somis, respectively . For A, rates at Riverside were about 15% higher on average than Somis . Rates of gs at the two sites were similar, with Riverside and Somis having 0.107 and 0.097 mol H2O m-2 ·s -1 , respectively. Rates of E were about 34% higher on average at Riverside than at Somis . Mean rates of E were 2.45 mmol H2O m-2 ·s-1 for Riverside and 1.61 mmol H2O m-2 ·s-1 for Riverside. There was no difference in intrinsic water-use efficiency among cultivars or between Riverside and Somis. There were significant differences between sites for instantaneous water-use efficiency, with Somis having significantly higher instantaneous water-use efficiency than Riverside . ‘Wonderful’ had significantly higher rates of A·E-1 at Riverside than at Somis, and ‘Wonderful’ had significantly higher rates of A·E-1 than ‘Eversweet’ at Somis and ‘Haku Botan’ at Riverside . There were no significant differences among cultivars or between sites for predawn stem water potential . Predawn stem water potential was significantly less negative than midday stem water potential . There were significant differences between sites and among cultivars for midday stem water potential . Somis had less negative midday stem water potentials than Riverside ,raspberry plant container with Riverside having a mean stem water potential of -2.31 MPa and Somis having a mean stem water potential of -1.53 MPa for midday measurements. ‘Haku Botan’ had less negative stem water potential than ‘Parfianka’ at Riverside . There were no differences in stem water potential among cultivars for midday measurements at Somis. There were positive and negative correlations between physiological variables for morning, afternoon and for data pooled for the two times of day. The relationship between A and gs was positively correlated and statistically significant for photosynthesis measurements for pooled data from both sites at Riverside and at Somis . There was also a positive correlation between A and E for pooled data from both sites , at Riverside and at Somis . There were site differences for precocity . Riverside had significantly higher precocity than Somis , with Riverside having 138.6 flowers on average and Somis having 28.7 flowers on average by year four.
For each cultivar at both sites, number of male flowers was greater than number of female flowers . There were significant differences among cultivars at both sites for number of male flowers and number of female flowers. However, there was no evidence for significant differences among cultivars at either site for male to female flower ratio. There were also significant differences among cultivars at both sites for total number of flowers and total number of fruit. Riverside had significantly more fruit than Somis in the first three years, but not for year four. This investigation provided evidence that some cultivars are more productive than others during the first four years of establishment. In Somis, the most precocious and productive cultivars in this investigation were Green Globe and Wonderful. ‘Wonderful’ was rather productive at both sites. Haku Botan was the only ornamental “double flower” cultivar in the trial and its flowers were difficult to sex at Somis. Observationally, ‘Haku Botan’ had larger flowers at the Somis, the coastal site. This phenotype made it difficult to sex the flowers and thus only total number of flowers was determined for this cultivar at that site. Usually, pomegranate flowers are easy to sex based on their shape, with bellshaped flowers being males and vase-shaped flowers being female. This distinction was not possible at Somis during the four years of this study. The objectives of this research were to evaluate eleven unique pomegranate cultivars for field performance in semi-arid and coastal Mediterranean climates to determine plant establishment rates and site effects on reproductive biology and to evaluate four pomegranate cultivars in coastal versus inland agroecosystems to determine if there were differences among genotypes for physiological traits that would be conducive to commercial crop production in drought conditions. The purpose was to phenotype these accessions for grower suitability and to aide in future breeding endeavors for crop improvement and germplasm diversity enhancement. Our findings indicate that many of the cultivars established at the same rate as Wonderful, but others did not. Namely, ‘Desertnyi’ and ‘Eversweet’ established much slower at both sites. ‘Eversweet’ was known to be less vigorous among rare fruit cognoscenti and ‘Desertnyi’ has been known to have trouble with canopy architecture at the germplasm repository. Our findings provide evidence that these two cultivars may not be the best choices to grow in a conventional spacing for the Californian pomegranate cropping system, unless perhaps rootstocks are employed that alter their growth rate. Literature on rootstocks is limited, so more research is needed to determine if these two cultivars would benefit from being grafted on a more vigorous rootstock. If these cultivars were to be chosen for commercial production due to their soft seeded arils and ornamental fruit, a possible method for commercial production would be to reduce spacing to 2-3 m and/or grow them in a trellis system. For at least one of the two sites, all four cultivars evaluated for physiological traits in these cultivar studies displayed eco-physiological traits suitable for commercial production if the industry standard, Wonderful, is used as the standard for comparison. However, ‘Wonderful’ performed better at Riverside than at Somis for A, which may indicate that ‘Wonderful’ may be better planted for commercial production inland than in coastal regions. Photosynthesis values obtained for ‘Wonderful’ at Riverside were different than those reported in Greece , but similar enough to suggest the possibility that differences in climate or cultural practices between the two sites were the source of this discrepancy. We found evidence that there are differences among cultivars within a site for physiological traits including stomatal conductance, transpiration and midday water potential. Values for physiological traits were similar to previous investigations into pomegranate physiology , a topic which is limited in the literature. Differences were also detected between sites, with higher rates of assimilation and transpiration at Riverside compared to Somis. There was a significant difference between sites for instantaneous water-use efficiency, with Somis having more efficient water use than Riverside. This indicates that some pomegranate cultivars may use less water for commercial production on the coast, but more studies must be carried out on irrigation scheduling to determine if these cultivars can be grown with less water on the warm coast of Southern California compared to hotter, semi-arid climates inland.