Dataset of Pollinators' contribution to seed yield in two self-fertile almond varieties

Data collection

Bee visitation and pollination limitation. To estimate pollinators’ abundance and visitation frequency, we performed 86 and 39 pollinator censuses in Marta and Marinadablocks, respectively. In each census, we recorded the number of flowers visited by each visiting pollinator to a group of 76 (± 5) flowers for a period of 5 min. (i.e., no. visits . flower-1 . 5 min-1). In total, we observed pollinators visiting almond flowers for ~10 hr. Censuses were performed in experimental Open and Supplemented trees (see below) and inrandomly selected trees within the experimental blocks. Pollinator censuses were performed between 10:00 and 17:00 hours, from February 7 to March 2 in Marta´s block; and from February 26 to March 11 in the Marinada one. During each census, we recorded only flower visitors that contacted anthers and/or stigma. 

To assess whether the abundance of pollinators during bloom was sufficient to maximize ovule fertilization, we tested pollination limitation (i.e., inadequate pollen transfer, either in quantitative or qualitative terms, which limits sexual plant reproduction). For this, six trees of each variety were used as experimental units. Within each experimental tree, six branches of similar size were tagged, and the number of open flowers between the tag and the end of the branch was counted. Tagged branches had, on average (±SE), 25 (±1) flowers. Three of the branches were assigned to the treatment “Control” (i.e., branches were exposed to natural pollination), and the other three branches were assigned to the treatment “Supplemented” (i.e., flowers were supplemented with pollen from the same variety in order to represent a real situation of monovarietal plantation). All experimental flowers under the “Supplemented” treatment were supplemented with fresh pollen (collected the same morning) by using clean brushes. Flowers were supplemented at least twice during their lifespan. Before fruit harvest, we counted the number of developing fruits to estimate the fruit set (i.e., fruits/flowers ratio).

Pollinators´ contribution. To evaluate to what extent pollinators contribute to fruit set and seed production at the tree level, we randomly selected 24 experimental almond trees per variety (i.e., two varieties, totaling 48 experimental trees). Each tree received one of the two treatments: (i) open (i.e., trees were kept open to be visited by pollinators), and (ii) excluded (i.e., trees were excluded from pollinators visitation by covering them with a fine RabitaInsect Cristal mesh, pore dimension 1.5 x 1.4mm). Consequently, each treatment (Open andExcluded) was replicated 12 times per variety. For the Isolation treatment, trees were covered before bloom and the mesh was removed right after the bloom period ended. Additionally, the enclosure of excluded trees was checked throughout the flowering period to certify the absence of pollinators inside. 

Within each tree (Open and Excluded), five branches were tagged, and the number of open flowers between the tag and the end of the branch was counted. Before harvest, we counted the number of fruits to estimate the fruit set. At harvest, all fruits per experimental tree were collected, dried, and weighed with a hand digital scale. From each tree, we randomly sampled 20 fruits, totaling 960 fruits (i.e., 20 fruits/tree, 12 trees/treatment, 2 treatments, 2 varieties). From each individual fruit, we weighed the pericarp, endocarp, and seed. To estimate seed production at the tree level we multiplied total fruit weight times the average proportion of a fruit´s weight represented by the seed.