APS_October 2018

J ournal of the A merican P omological S ociety

232

full bloom each year at each site, trees were hand-thinned to obtain a range of crop densi- ties (CD). The target range for CD was 0.5 to 5.0 fruit per cm 2 trunk-cross-sectional area (TCA), but due to spring frost or poor fruit set, the maximum desired CDs were rarely achieved. Each year cooperators recorded TCA, the dates of full bloom and 50% fruit harvest, the number of fruit per tree and to- tal yield (kg per tree). Days from bloom to 50% harvest date, CD and FW were esti- mated for each tree. Daily maximum and minimum temperatures were reported and used to calculate cumulative growing degree days (CGDD) for the 30, 45 and 60 days af- ter bloom using both 4° and 7°C as the base temperature. ‘Crimson Lady’ had inadequate crop loads at most locations in most years, so data analyses were limited to ‘Redhaven’ and ‘Cresthaven’. Data analysis. Various types of scatter plots were generated to evaluate relation- ships between variables, to identify suspi- cious observations, and to visualize potential sources of variation. Initially SAS’s PROC MIXED (Littell et al., 2006) was used to perform analysis of covariance (ANCOVA), where site was included in the model as an indicator variable, and linear and quadratic terms for CD and CGDD were considered regressor variables along with all possible interaction terms. The response variables in- cluded FW and days from bloom to harvest. The most complex interaction term with the highest P-value was deleted from the model and the new model was run. This manual backward elimination continued until only significant ( P =0.05) terms remained in the model (Milliken and Johnson, 2002). Inter- action terms were rarely significant, but scat- ter plots and the analysis of covariance in- dicated that FW was often related to CGDD and CD in a quadratic manner. When the site x CD interaction was not significant, a single model was fit with data pooled over all sites. When site interacted with CD and/or CGDD, PROC REG (Freund and Littell, 2000) was used to fit polynomial regression models for

10 days) declined linearly with increasing growing degree hours during the first 30 days after bloom. When fruit size trends for three major cultivars were analyzed, days from full bloom to reference date was negatively related to cumulative growing degree hours during the 30 days post bloom, and fruit size was negatively related to days from full bloom to reference date (Lopez et al., 2007). Taken together, these data supported the hy- pothesis that trees exposed to high spring temperatures cannot supply assimilates ad- equately to support maximum early-season fruit growth rates. Working with data for a number of peach cultivars in two regions in South Carolina over several seasons, Ke- nealy et al. (2015) validated the concept that cumulative growing degree days during the first 30 days after bloom could be used to predict days from bloom to harvest.  Since early season temperatures and crop load can affect both days from bloom to har- vest and fruit size at harvest, it is important to know if the effect of early season tempera- tures is independent of crop load. Therefore, the objective of this study was to determine the effects of crop density and early-season temperature, alone and in combination on days from bloom to harvest and on peach fruit size at harvest. Materials and Methods General. In spring 2009, under the auspices of the NC-140 multi-state project, a multi-lo- cation peach rootstock trial was established. Trees were planted 5 x 6 m and trained to the open vase form. Ten trees of ‘Cresthav- en’, ‘Redhaven’ and ‘Crimson Lady’ were planted at 14 locations to study the effect of early-season temperature, solar radiation and precipitation on average fruit weight (FW) and fruit soluble solids concentration. All trees for a given cultivar were propagated by the same nursery. Cooperators from ID, KY, MD, NY, and SC participated for four years (2014 – 2017) to study the effects of crop density and early-season temperatures on fruit size. During the first 45 days after