Drupelet Set on 'Arapaho' Thornless Blackberry Flowers Protected and Nonprotected from Cross Pollination and Assessment of Resulting Seedlings

three selected secondary buds of two flowering shoots, and the remaining flowers of those shoots were removed. Cross-pollination was prevented by covering the flowering shoots with No. 535 Lawson bags (Lawson Bags, Northfield, Ill.). Flowers were hand-pollinated once on the day following emasculation using ‘Arapaho’ pollen collected in advance. Bags were placed on protected flowers immediately after emasculation, removed to apply self pollen, and replaced immediately. On 19 June 1997, fully colored fruits from all treatments were harvested, the number of drupelets/fruit was recorded, and the seeds were extracted. Following procedures described before (Lopez-Medina et al., 2000), seed was scarified on 25 Aug. 1997 and stratified until 10 Jan. 1998. On 11 Mar. 1998, newly germinated seedlings were scored for thornlessness and/or thorniness by observing absence/presence of cotyledonary hairs, respectively, with the help of a magnifying glass. The experiment was repeated three times and the repetitions were treated as blocks. For drupelet set, data were subjected to analysis of variance. Bagging prevented drupelet set on emasculated flowers that were not hand-pollinated, while a small number of drupelets developed in similar flowers that were left unprotected (Table 1). Only 25% of the seedlings from nonbagged flowers were thorny, and therefore had developed following cross-pollination. Very similar drupelet set was observed in protected and nonprotected, emasculated, selfpollinated (ESP) flowers, and there was no contamination by thorny pollen in either of Contamination with undesirable pollen via wind or insects is one of the major concerns in a fruit breeding program and dictates that controlled pollinations must be done either in a closed greenhouse or by bagging emasculated flowers in the field. Bagging is the rule in some blackberry (Rubus L. subgenus Rubus Watson) breeding programs (Daubeny, 1996), but is omitted in others (Finn, 1996) because it is time-consuming and tedious. When flower buds are not protected, some degree of contamination can be expected. Finn (1996) demonstrated that nonpollinated, emasculated flowers of several trailing blackberry genotypes set drupelets when not bagged. However, he did not determine whether the drupelets resulted from selfor cross-pollination. The objective of this study was to evaluate the effect on drupelet set of protecting selfand open-pollinated flowers of ‘Arapaho’ thornless blackberry from cross-pollination, and to quantify the degree of pollen contamination by examining the thornless/thorny nature of the resulting seedlings. Any thorny seedling obtained from the cross of two thornless parents can only be the result of contaminated pollen. This study was conducted at the Arkansas Agricultural Research and Extension Center, Fayetteville. The ‘Arapaho’ plants were the same ones used for another study (LopezMedina et al., 2000). Within-row spacing was close (1.22 m) and >98% of the surrounding blackberry plants were thorny. On 8 May 1997, two lateral branches per plant were selected per each of six treatments (Table 1). One branch was sampled from each side of the row at the middle of the plant (0.5–0.6 m). On each branch, one treatment was allocated on