Effects of Cluster Size and Shoot Type on Characteristics of Pecan Nuts

Whole fruit clusters of ‘Pawnee’ pecan [Carya illinoinensis (Wang.) C. Koch.] were collected from three shoot types: terminal and lateral shoots without a secondary growth fl ush and shoots that had an early-season secondary growth fl ush. Fruit per cluster were counted and nuts were individually harvested, weighed, shelled and graded. Bloom the following year was determined for the same shoots where clusters were collected. Wafers (cotyledons that failed to develop) were not associated with cluster size or shoot type. When wafers were included in the data, nut weight, kernel percentage and return bloom were not affected by cluster size or shoot type. However, when wafers were excluded from the data there were signifi cant relationships of cluster size and shoot type with the dependent variables. Cluster size on lateral shoots was negatively related to nut weight and kernel percentage. Cluster size on terminal shoots without a secondary growth fl ush was inversely related to kernel percentage, but not related to nut weight. When shoots had a secondary growth fl ush, cluster size was not related to kernel percentage or nut weight. There was a positive linear relationship between cluster size and total kernel weight for the three shoot types. Return bloom of terminal shoots without a secondary growth fl ush was negatively related to cluster size, but cluster size did not affect return bloom of the other shoot types. The number of shoots that developed the following year was positively related to cluster size for terminal and lateral shoots, but not for shoots with a secondary growth fl ush. Shoots with a secondary growth fl ush produced substantially more shoots with larger fruit clusters the next year than the other shoot types. Pecan fruit production is irregular, typifi ed by high production 1 year followed by one or more years of low production (Sparks, 1986). This may best be characterized as alternate bearing with irregular symmetry. Alternate bearing is typically associated with a lack of return bloom rather than fl ower or fruit abortion. Crane et al. (1934) reported a leaf to fruit ratio of 8 to 10 leaves per fruit was required for development of good quality nuts. They also suggested that cultivars with a large cluster size might have a lower percentage of return bloom than those with a smaller cluster size. Reid (1986) speculated that cultivars with large fruit clusters typically had poorer fruit quality than cultivars with small clusters. It is often suggested by scientists working with pecans that quality and consistent production of ‘Desirable’ can be attributed to its large fruit drop, resulting in one or two fruit per cluster by the dough stage. This may be related to the fruit carbohydrate demand or growth regulators that inhibit fl ower induction proportionately to the fruit or cluster size. Return bloom of fruiting shoots is less than that of same season vegetative shoots (Malstrom and McMeans, 1982; Smith et al., tober. Nitrogen was surface-applied in March at 112 kg·ha N from urea plus 142 kg·ha N was applied from NO 3 -contaminated irrigation water. Five foliar Zn applications were made between budbreak and July at 2.4 kg·ha Zn from 36% ZnSO 4 . Pest management followed extension recommendations for a commercial orchard (von Broembsen et al., 1997). Whole fruit clusters from thirty shoots each of three bearing shoot types on each tree were randomly selected, tagged to monitor return bloom and then individually harvested at shuck split. The three bearing shoot types were 1) shoots in the terminal position without a secondary growth fl ush, 2) shoots in the lateral position without a secondary growth fl ush, and 3) shoots, primarily in the terminal position, with a secondary growth fl ush. The number of fruit per cluster (cluster size) was recorded from each shoot, and then nuts were harvested at shuck split, dried, individually weighed, and shelled. In total, 270 fruit clusters were harvested ranging in size from 1 to 11 fruit per cluster for a total of nearly 1300 nuts. The following spring, total current season shoots per 1-year-old branch and the percentage of 1-year-old branches that produced one or more shoots with pistillate fl owers was determined. Various models were fi tted to the data and the most appropriate model was chosen for each variable. Results and Discussion Cluster size was unrelated to nut weight, kernel percentage or return bloom for any shoot type when nuts with undeveloped cotyledons (wafers, testa develops but little, if any, carbohydrate and oil deposition) were retained in the data (data not shown). Percentage of wafers was not associated with cluster size for any shoot type (Fig. 1). Wafers were excluded from the data, and cluster size was recalculated along with weight/nut, kernel percentage and total kernel weight/shoot. The relationship of cluster size for each shoot type was then determined for the dependent variables. The most frequent occurring cluster size on terminal shoots was ≤3 fruit per cluster (Fig. 1). On shoots with a secondary growth fl ush 5 fruit per cluster occurred more frequently than other sizes. Shoot vigor is positively related to the cluster size formed the next year (Dodge and Crane, 1933). If shoots with a secondary growth fl ush were more vigorous the previous year than the terminal shoots without a secondary growth fl ush, then cluster size was probably initially larger on shoots with a secondary growth fl ush. Alternatively, fruit drop may have been less on the vigorous shoots with a secondary growth fl ush. Lateral shoots were intermediate in cluster size with 3 to 4 fruit per cluster dominating. Cluster size was not related to average nut weight on terminal shoots and shoots with a secondary growth fl ush (Fig. 2). However, on lateral shoots average nut weight decreased about 18% as cluster size increased from 1 to 8 fruit per cluster. These data suggest that nuts are more likely to be smaller on lateral shoots than on terminal shoots or on shoots 1986). Return bloom of lateral shoots was substantially more sensitive to over-fruiting than return bloom of terminal shoots (Wood, 1995). Mechanical fruit thinning can be used to manage crop loads to improve fruit quality and return bloom, plus alleviate other disorders (Reid et al., 1993; Smith and Gallott, 1990, Smith et al., 1993; Sparks et al., 1995). Although scientists hypothesized that a small cluster size was desirable to produce high quality nuts and promote consistent production, no data exist relating cluster size to nut quality and return bloom; therefore, this study establishes those relationships. We also determined if shoot position or shoots with an early-season secondary growth fl ush affected nut quality and if these different shoot types had similar return bloom characteristics. Materials and Methods Three 13-year-old ‘Pawnee’ trees growing in a Teller sandy loam (fi ne-loamy, mixed, active, thermic, Udic Argiustolls) in a commercially managed orchard near Charlie, Texas were selected based on uniformity of size, vigor, crop load (85% to 90% bearing shoots), and location within the orchard. Trees were spaced 12.2 × 12.2 m apart and had 19.7 ± 2.1 cm diameter trunks measured 1.4 m above the ground. The entire orchard fl oor was maintained vegetation-free throughout the growing season with glyphosate [N-(phosphonomethyl) glycine]. Trees received 400 mm of supplemental irrigation from micro-sprinklers and 576 mm of rainfall in April through OcHORTSCIENCE 40(5):1300–1303. 2005. Received for publication 3 Dec. 2004. Accepted for publication 19 Feb. 2005. Approved for publication by the Oklahoma Agricultural Experiment Station. Funded by the Oklahoma Agricultural Experiment Station and the USDA Crop Germplasm Committee. Graduate student. Regents professor and corresponding author, e-mail [email protected]. Professor. AugustBook.indb 1300 6/14/05 12:15:16 PM 1301 HORTSCIENCE VOL. 40(5) AUGUST 2005 with a secondary growth fl ush as cluster size increases. The average weight per nut, pooled over cluster sizes, was 8.2, 8.3, and 8.0 g for terminal shoots, shoots with a secondary growth fl ush and lateral shoots, respectively. Thus nuts from lateral shoots averaged about 3% smaller than those on terminal shoots or shoots with a secondary growth fl ush. The leaves on a shoot primarily support fruit development on that shoot (Davis and Sparks, 1974). Few carbohydrates are transported from surrounding shoots to support fruit development; therefore, shoots with more leaves could support larger fruit clusters. Terminal shoots and shoots with a secondary growth fl ush tend to be longer with more leaves than lateral shoots, so nut weight was only affected by cluster size on the shorter lateral shoots. Additionally, because of their location on the branch more shading of lateral shoots occurs, reducing photosynthesis and consequently nut weight. Kernel percentage was negatively related to cluster size on terminal and lateral shoots, but not on shoots with a secondary growth fl ush (Fig. 3). On terminal shoots, kernel percentage was reduced about 5% (from 56.9% to 53.8%) and about 6% (from 56.4% to 52.8%) on lateral shoots as cluster size increased from 1 to 8 fruit per cluster. Kernel percentage of shoots with a secondary growth fl ush averaged 55.4%. Total kernel weight per shoot increased linearly as cluster size increased (Fig. 4). More fruit in a cluster should increase kernel weight per shoot. The linear increase in total kernel weight with cluster size for each shoot type suggests that even at 8 or 9 fruit/cluster, the shoot had not reached its maximum carrying capacity. A curved relationship would suggest that shoots were approaching their maximum fruit carrying capacity with the larger fruit cluster sizes. A reduction in kernel percentage with greater cluster size (Fig. 3) suggests that terminal and lateral shoots were nearer their carrying capacity than indicated by total kernel weight. Alternatively, shell weight may be increased disproportionately to kernel weight when cluster size was increased, and this was refl ected in kernel percentage on terminal and lateral shoots. Return bloom was negatively related to cluster size on terminal shoots, although the correlation was weak, but not on the other shoot types (Fig. 5). This suggests that return bloom on lateral shoots and shoots with a secondary growth fl ush was unaffected by cluster sizes from 1 fruit to 8 fruit per cluster. However, on terminal shoots return bloom was 60% with 1 fruit/cluster and reduced to 17% fruiting with 8 fruit/cluster. Return bloom, averaged over all cluster sizes, was 70% on shoots with a secondary growth fl ush, 33% for lateral shoots, and 43% terminal shoots. Total shoots per 1-year-old branch increased linearly as cluster size increased on terminal and lateral shoots, but not on shoots with a secondary growth fl ush (Fig. 6). Values for total shoots per 1-year-old branch, averaged over all cluster sizes, were 4.2 on shoots with a secondary growth fl ush, 2.4 for lateral shoots, and 2.4 for terminal shoots. Cluster size was not related to total fruits per 1-year-old branch produced the following year for any shoot type (data not shown). However, shoots with a secondary growth fl ush produced over twice as many fruits per branch as the other shoot types. The total fruit per branch produced the following year was 5.7 on shoots with a secondary growth fl ush, 1.7 for lateral shoots, and 2.4 for terminal shoots. These data suggest that a large cluster size on lateral shoots negatively impacts nut weight and kernel percentage. Similarly, kernel percentage was reduced by large fruit clusters on terminal shoots. Shoots with a secondary growth fl ush supported up to nine fruit per cluster with no effect on nut size or kernel percentage. Additionally, shoots with a secondary growth fl ush gave rise to more shoots with larger fruit clusters the next year than other bearing shoot types. It is evident from these data that vigorous trees with substantial amounts of shoots with a secondary growth fl ush are more likely to produce fruit consistently. Also, vigorous shoots can carry larger fruit clusters than less vigorous shoots with little, if any, effect on quality. It must be pointed out that the secondary growth fl ushes in this study began in June, early in the growing season. Thus the additional leaf surface area was present as the fruit developed. Early-season growth fl ushes result in a net carbohydrate gain (Andersen and Brodbeck, 1988), thus there is a positive impact on developing fruit. Pecan trees occasionally make a late-season growth fl ush in September or early October. Trees producing late-season growth fl ushes are typically nonirrigated trees that have experienced summer drought followed by abundant September rainfall. Our observations suggest that a secondary growth fl ush in September or October is detrimental to kernel quality since the rapid shoot growth competes directly with developing cotyledons. Late-summer growth fl ushes consume more Fig 2. Relationship of cluster size and shoot type with weight per nut the same year. Triangles are the average of 3 to 37 observations for each cluster size and shoot type. Terminal: weight/nut (g) = 8.4 – 0.077 (cluster size), r = 0.178, df = 79, p > 0.2374; lateral: weight/nut (g) = 8.8 – 0.221 (cluster size), r = 0.13, df =78, p > 0.0010; secondary growth fl ush: weight/nut (g) = 8.6 – 0.061 (cluster size), r = 0.02, df =79, p > 0.2735. Fig 1. The number of observations for each shoot type and cluster size (vertical bars), and the relationship of cluster size and shoot type with incompletely developed cotyledons (wafers) (scatter diagram). 690-Crop.indd 1301 6/22/05 10:36:00 AM HORTSCIENCE VOL. 40(5) AUGUST 2005 1302 carbohydrates during development than they produce (Andersen and Brodbeck, 1988). We expect early-season and late-season growth fl ushes to affect nut quality and return bloom differently. An early-season growth fl ush was advantageous for current season nut quality and subsequent return bloom. We hypothesize that a late-season growth fl ush negatively impacts nut quality. The weak relationship between cluster size and return bloom for terminal shoots without a secondary growth fl ush and the nonsignifi cant relationships for lateral shoots and shoots with a secondary growth fl ush was surprising. It has been well documented that current-year fruit suppresses return bloom compared to current-year vegetative shoots (Malstrom and McMeans, 1982; Smith et al., 1986). These data suggest that return bloom suppression was only slightly enhanced or not affected as cluster size increased. Thus a small cluster size does not necessarily increase the likelihood of improved return bloom, as has been speculated (Crane et al., 1934; Reid, 1986). Carbohydrate reserves have been implicated in pecan alternate bearing, i.e., large crops deplete carbohydrate reserves resulting in small crops the following year (Smith and Waugh, 1938; Sparks, 1974; Sparks and Brack, 1972; Wood, 1989, 1995; Wood et al., 1987; Worley, 1979a, 1979b). Since larger clusters require more carbohydrates to develop than small clusters, the large clusters should deplete carbohydrates and cause greater return bloom suppression. However, the relationship between cluster size and return bloom was weak or not signifi cant, suggesting that carbohydrate supply on an individual shoot basis was not a major factor controlling return bloom. In fact, another study found that the total non-structural carbohydrates in the fall, during dormancy and at budbreak of current-season fruiting shoots were greater or equal to vegetative shoots, yet return bloom of vegetative shoots was signifi cantly greater than fruiting shoots (Smith et al., 1986). Thus carbohydrates appear to play a minor role in regulating return bloom of individual shoots. A growth regulator(s) produced by the fruit may be involved in determining return bloom (Smith et al., 1986), but in this study return bloom suppression by fruit in the cluster was not additive. The lack of an additive effect by fruit in a cluster might be associated with larger clusters occurring on more vigorous shoots, thus mitigating the effects of cluster size on return bloom. A more likely scenario is that a single fruit produces suffi cient growth regulators to affect return bloom with little additional impact from more fruit in the cluster. Vigorous shoots clearly have a greater return bloom than less vigorous shoots. In this study, return bloom of shoots with a secondary growth fl ush was 70% compared to 43% and 33% for terminal and lateral shoots, respectively. Longer shoots have more leaves that may produce more growth regulators that promote subsequent year fl owering. Longer shoots with more leaves produce more carbohydrates, but the carbohydrate supply does not appear to be the major limiting factor for return bloom of an individual shoot. Carbohydrate reserves appear to be important in determining tree survival (Wood, 2001) and the ability to fl ower (Smith et al., 1986) rather than controlling fl ower induction on individual shoots.