insects Article Use of a Managed Solitary Bee to Pollinate Almonds: Population Sustainability and Increased Fruit Set Jordi Bosch 1, *, Sergio Osorio-Canadas 2 , Fabio Sgolastra 3, * and Narcís Vicens 4 1 Centre for Research on Ecology and Forestry Application, 08193 Bellaterra, Spain 2 Departamento de Ecología de la Biodiversidad, Instituto de Ecología, Universidad Nacional Autónoma de México, AP 70-275, 04510 Mexico City, Mexico; s.osorio.canadas@gmail.com 3 Dipartimento di Scienze e Tecnologie Agro-Alimentari, Alma Mater Studiorum Università di Bologna, Viale Fanin 42, 40127 Bologna, Italy 4 Diputació de Girona, Servei de Medi Ambient, 17004 Girona, Spain; nvicens@ddgi.cat * Correspondence: jordi.bosch@uab.cat (J.B.); fabio.sgolastra2@unibo.it (F.S.) Simple Summary: Methods to rear Osmia bees to pollinate fruit trees have been developed in various parts of the world. These bees are excellent pollinators but evidence that their populations can be sustained in orchards and their use results in increased fruit production is scarce. We released an Osmia cornuta population at one end of an almond orchard. Then, we surveyed the pollinators visiting the almond flowers and measured fruit set in trees located at increasing distances from the nesting stations. We found that fruit production was higher in the trees that received more Osmia visits. Importantly, this result was obtained against a strong background of honeybees, which were 10 times more abundant than Osmia. The Osmia population obtained at the end of the flowering period was 1.28 larger than the population initially released. Our study demonstrates that Osmia populations can be sustained in orchard environments and that even a small population of a highly effective pollinator may have a significant impact on fruit set. Our results are encouraging for the   use of Osmia populations and for the implementation of measures to promote wild pollinators in Citation: Bosch, J.; agricultural environments. Osorio-Canadas, S.; Sgolastra, F.; Vicens, N. Use of a Managed Solitary Abstract: Osmia spp. are excellent orchard pollinators but evidence that their populations can be Bee to Pollinate Almonds: Population sustained in orchard environments and their use results in increased fruit production is scarce. We Sustainability and Increased Fruit Set. released an Osmia cornuta population in an almond orchard and measured its population dynamics, Insects 2021, 12, 56. https://doi.org/ as well as visitation rates and fruit set at increasing distances from the nesting stations. Honeybees 10.3390/insects12010056 were 10 times more abundant than O. cornuta. However, the best models relating fruit set and bee visitation included only O. cornuta visitation, which explained 41% and 40% of the initial and final Received: 26 November 2020 fruit set. Distance from the nesting stations explained 27.7% and 22.1% of the variability in initial Accepted: 8 January 2021 and final fruit set. Of the 198 females released, 99 (54.4%) established and produced an average of Published: 11 January 2021 9.15 cells. Female population growth was 1.28. By comparing our results with those of previous Publisher’s Note: MDPI stays neu- O. cornuta studies we identify two important populational bottlenecks (female establishment and tral with regard to jurisdictional clai- male-biased progeny sex ratios). Our study demonstrates that even a small population of a highly ms in published maps and institutio- effective pollinator may have a significant impact on fruit set. Our results are encouraging for the use nal affiliations. of Osmia managed populations and for the implementation of measures to promote wild pollinators in agricultural environments. Keywords: pollination service; Osmia cornuta; Apis mellifera; population dynamics; managed pollinators; Copyright: © 2021 by the authors. Li- crop pollination censee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and con- ditions of the Creative Commons At- tribution (CC BY) license (https:// 1. Introduction creativecommons.org/licenses/by/ Approximately three-quarters of the world’s crops benefit from animal pollination [1], 4.0/). and a significant part of this pollination service is provided by wild pollinators [2–5]. How- Insects 2021, 12, 56. https://doi.org/10.3390/insects12010056 https://www.mdpi.com/journal/insects Insects 2021, 12, 56 2 of 11 ever, the current context of agricultural intensification, characterized by low crop diversity, increased crop field size, loss of semi-natural habitats, and increased pesticide use is clearly detrimental to pollinator abundance and diversity [6–9]. As a result, wild pollinators are notoriously scarce in many agricultural landscapes [10–12]. For this reason, the general perception is that wild pollinator populations are insufficient to provide adequate levels of pollination in intensively farmed areas, and populations of managed pollinators are usually introduced to enhance pollination services. Fruit trees are highly dependent on pollinator visitation because they bloom for a short period of time in spring when weather conditions are often suboptimal for insect activity and because most cultivars are self-incompatible. For this reason, honey bee hives are usually introduced in orchards at a rate of 2–6 hives per ha, with each hive containing several thousands of foragers [13–15]. However, honey bees are only fully active at temperatures above 12–14 ◦ C [16]. In addition, because they have long foraging ranges [17–19], and are highly generalist foragers, they often visit other flower species [14]. Finally, honey bees are not very effective fruit tree pollinators, mainly due to their low visit legitimacy (many of the visits result in no contact between the bee and the stigmas; [20–25]). These shortcomings, along with the risks associated with relying on a single species, have prompted the search for alternative pollinators, and methods to manage various Osmia species as orchard pollinators have been developed in different parts of the world [26,27]. These Osmia species are only active for a couple of months in spring and fly under marginal weather conditions [16]. In addition, they have short foraging ranges [28–30] and show a strong preference for fruit tree pollen [21,23,31–35]. Finally, Osmia visit legitimacy on fruit three flowers is close to 100%, and fruit set in flowers receiving a single visit is similar to fruit set in hand-pollinated flowers [20,21,23–25,36]. For these reasons, population densities recommended for orchard pollination with Osmia spp. are as low as 1250–2000 bees per ha (with a 1.6 male/female ratio) [26]. However, even if the pollinating effectiveness of Osmia spp. has been amply documented, evidence that the use of Osmia populations results in increased fruit production is still scarce [11,23,37,38]. The use of a managed pollinator is only sustainable if population levels can be main- tained from year to year. Ideally, populations should be able to grow on site during the flowering period of the target crop. If this is not possible, population losses can be compen- sated by rearing populations under artificial conditions, as done with bumblebee colonies used for crop pollination [39]. However, attempts to mass-rear Osmia populations under ar- tificial conditions have not been successful [40,41]. Therefore, to ensure the sustainability of Osmia spp. as managed pollinators it is essential to understand the dynamics of populations introduced in orchards. There are various factors potentially limiting the growth of Osmia populations released in orchards. These factors include: (a) Winter mortality (some of the cocoons introduced in the orchard contain individuals that are either dead or too weak to emerge); (b) low female establishment (some of the females that have successfully emerged out of their cocoon may be predated, too weak to start nesting activities, or disperse and nest away from the release site); (c) low fecundity (nesting females provision a low number of cells and therefore lay few eggs); (d) progeny mortality (a part of the progeny does not reach the adult stage due to either developmental failure or parasitism); (e) male-biased progeny sex ratio (nesting females produce a high proportion of males, which are less costly to produce than females [42]). In this study, we released a population of the European species Osmia cornuta in an almond orchard. Almond pollination is particularly challenging because almonds bloom very early in the year (February-March) and have an unusually high bearing capacity (as many as 40% of the flowers may bear fruit) [43]. We measured female establishment and population growth, as well as flower visitation rates and fruit set at increasing distances from the nesting site. Our first objective was to assess whether the O. cornuta population had an impact on fruit production. Given the short foraging range of O. cornuta, we expected fruit set to be negatively correlated to distance from the nesting site. Our second objective was to establish whether the O. cornuta population could be increased on site Insects 2021, 12, 56 3 of 11 and, by comparing our results with those of previous studies, to identify the main factors limiting population growth. 2. Materials and Methods The study was conducted in an almond orchard in Vila-Seca (Tarragona, NE Spain). The orchard measured 0.5 ha and had 5 rows of the main cultivar Ferragnes intermixed with 4 rows of the pollinizer cultivar Cristomorto (both cultivars are self-incompatible) (Figure A1). Each row had 16 trees. The orchard was located within a matrix of farmland, including cereal fields, fallow land, and a mixture of olive, carob, and hazelnut orchards. There were no honey bee hives in the orchard or its surroundings (at least within a 500 m radius). In early February 1992, prior to almond bloom, we set up 3 nesting stations for O. cornuta at one edge of the orchard (Figure A1). Nesting stations were made with wooden boxes with the front side open held 1.5 m above the ground on four metal fence posts. Each station contained 15 perforated solid wood blocks with 25 inserted paper straws (length: 15 cm; inside diameter: 8 mm). On 22 February, before the orchard started to bloom, we placed 198 females and 360 males (M/F sex ratio: 1.8) within their cocoons in open cardboard boxes inside the nesting stations. These cocoons had been wintered at 4 ◦ C since October 1991. To determine the number of females that established at the nesting stations, we inspected each nesting cavity with an otoscope. This was done at night, when females were roosting inside their nests, seven times during the nesting period (approximately twice a week). During peak bloom, we conducted pollinator counts on the central row of Ferragnes. An observer slowly walked around each tree and noted all pollinators seen visiting the flowers. Total observation time was 10 h and 40 min (40 min × 16 trees). To assess fruit set (% of flowers that set fruit) at increasing distances (0–80 m) from the O. cornuta nesting shelters, we marked all trees in 3 Ferragnes rows (48 trees), including the row used in the pollinator counts (Figure A1). Before bloom we tagged 4 branches on each tree and counted the number of flower buds on each branch (mean ± SE = 264.4 ± 1.6 flower buds per branch). In April, we counted the number of initiated fruits on each branch (initial fruit set). Following the natural fruit drop, fruits were counted again in June when they had reached their final size (final fruit set). To obtain a measure of maximum potential fruit set, we marked 5 additional Ferragnes trees (Figure A1) on which we tagged 4 branches and counted flower buds as described above. These trees were checked daily and newly opened flowers were hand-pollinated with Cristomorto pollen. Initial and final fruit set was assessed as described above. After petal fall, when O. cornuta nesting activity had ceased, we removed nesting materials and kept them in the laboratory. In September, when development was com- pleted, we pulled out paper straws and analyzed the contents of each nest. We quantified the number of male and female offspring produced (female cocoons are typically larger than male cocoons and are found in the inner cells of the nest [42]). We also recorded offspring mortality. Statistical Analysis Initial and final fruit set were analyzed separately. To analyze whether fruit set declined with distance from the O. cornuta nesting stations, we used Linear Mixed Models (LMMs) with a tree as a random factor. We considered a linear relationship (model with distance as the only explicative variable) and a quadratic relationship (model with distance and distance2 ). We then used a model inference approach [44] to select the best-fit model based on AICc values using maximum likelihood criteria. Models with ∆AICc < 2 were considered equal to the best model [44]. We then run a LMM model with REML to obtain unbiased parameter estimates. We calculated a likelihood-ratio-based R2 of the best models as a measure of explanatory power. Insects 2021, 12, 56 4 of 11 To analyze whether fruit set was related to O. cornuta and/or A. mellifera visitation, we conducted a model selection procedure with LM models, testing all possible explana- tory variable combinations through a multi-model inference approach with the ‘dredge’ function (‘MuMin’ package, [45]). We again selected the best models based on AICc values. Following model selection, we used a model averaging procedure (with averaged variable coefficients) based on AICc. This was done with the ‘model.avg’ function (‘MunMin’ pack- age), which yields model-averaged estimates of variable coefficients and p-values for each explanatory variable. Conditional average and full average approaches yielded almost identical results. We show only conditional average results. Finally, we calculated the adjusted-R2 of the best models (containing significant explanatory variables) as a measure of explanatory power. Percent initial and final fruit set were arcsin-transformed. The distribution of residuals Insects 2020, 11, x was visually inspected for homoscedasticity and the normality 5 ofassumption 13 was tested with the Shapiro test. All analyses were conducted with the ‘nlme’ [46] and MuMin packages in 3. Results R [47]. All means are followed by standard error (SE). Maximum temperatures during the days following the release of theResults 3. O. cornuta population (22 February) ranged between 14 and 18 °C. The first females engaged in nesting activities were observed on 28 Feb- ruary.Maximum Of the 198temperatures females released,during 182 the days following emerged out of theirthe release of the O. cornuta popu- cocoons lation (22 February) ranged between 14 and 18 ◦ C. The first females engaged in nesting (91.9 % winter survival). The maximum number of females established activities werestations in the nesting observedwason 28 February. counted on March Of3the (99 198 females females; released, 54.4% of 182 emerged out of their cocoonsfemales). the emerged (91.9 % winter survival). The maximum number of females established in the nestingWestations recordedwas1114counted on March pollinators visiting3 the (99 almond females;flowers. 54.4% of the emerged females). Alt- hough Wethere were no1114 recorded hivespollinators in sight, thevisiting most frequent pollinator the almond was, Although there were flowers. by hives no far, Apis inmellifera (74.1% sight, the mostof the visits recorded), frequent pollinatorfollowed was, by byfar, O. cor- Apis mellifera (74.1% of the nuta (7.3%). visits Otherfollowed recorded), visitors included various (7.3%). by O. cornuta flies (16.2%), Other hoverflies visitors included various flies (4.11%, Eristalis (16.2%), tenax,(4.11%, hoverflies EupeodesEristalis sp.), andtenax, wild bees (1.1%,sp.), Eupeodes Andrena andni-wild bees (1.1%, Andrena groaenea, Andrena sp., Eucera sp., Bombus terrestris, Xylocopa violacea). nigroaenea, Andrena sp., Eucera sp., Bombus terrestris, Xylocopa violacea). Most (70.4%) O. Most (70.4%) O. cornuta visitation occurred within 30 m from the nest- cornuta visitation occurred within 30 m from the nesting stations (Figure 1). Apis mellifera ing stations (Figure 1). Apis mellifera visitation followed an irregular visitation followed an irregular pattern across the orchard and tended to be higher towards pattern across the orchard and tended to be higher towards the two the two ends of the(Figure ends of the orchard orchard1). (Figure 1). Figure 1. Figure 1. Number NumberApis Apismellifera andand mellifera Osmia cornuta Osmia individuals cornuta recorded individuals visit- visiting almond flowers at recorded ing almond flowers at increasing distances from the O. cornuta nesting sta- increasing distances from the O. cornuta nesting stations. tions. Almond bloom was over by 23 March. At that time, O. cornuta females that were Almond bloom was over by 23 March. At that time, O. cornuta fe- still alive foraged mostly on Diplotaxis erucoides (Brassicaceae), a common weed in the males that were still alive foraged mostly on Diplotaxis erucoides (Bras- surroundings of theweed sicaceae), a common orchard. in theBy 29 March O. surroundings cornuta of the nesting orchard. By 29activity had ceased. The number March O.of nests nesting cornuta produced washad activity 203.ceased. TheseThenests contained number of nests253 female and 653 male cells. pro- Offspring duced wasmortality 203. Thesewas 7.1%. nests Most 253 contained of this mortality female and 653was due male to unknown causes (5.3%), cells. and the rest Offspring to parasitism mortality by the was 7.1%. Mostcleptoparasitic mite of this mortality wasChaetodactylus due to un- osmiae (1.8%). The live female known population recovered causes (5.3%), wasto241, and the rest and theby parasitism live themale population 601 (M/F sex ratio: 2.5). cleptoparasitic mite Chaetodactylus osmiae (1.8%). The live female population recovered was 241, and the live male population 601 (M/F sex ratio: 2.5). Both initial and final fruit set significantly declined with distance from the O. cornuta nesting stations (Figure 2). The quadratic model provided the best fit, with distance explaining 27.7% of the variability in initial fruit set and 22.1% of the variability in final fruit set (Table 1). Insects 2021, 12, 56 5 of 11 Both initial and final fruit set significantly declined with distance from the O. cornuta nesting stations (Figure 2). The quadratic model provided the best fit, with distance explaining 27.7% of the variability in initial fruit set and 22.1% of the variability in final fruit set (Table 1). Importantly, the random factor tree also had a strong effect on fruit set (initial: 38.7%; final: 37.1%). The linear models yielded similar results (Table A1 and Figure A2 in Appendix A). Final fruit set was strongly related to initial fruit set (Pearson’s Insects 2020, 11, x 6 of 13 r = 0.88; p < 0.0001; Figure 3). Insects 2020, 11, x 6 of 13 Figure 2. Initial (A) and final (B) fruit set (% of flowers setting fruit in April and June, respectively) Figure 2. Initial (A) and final (B) fruit set (% of flowers setting fruit in April and June, respectively) at at increasing distances from the Osmia cornuta nesting shelters. Each dot represents a branch. Broken increasing distances from the Osmia cornuta nesting shelters. Each dot represents a branch. Broken lines lines indicate indicate mean mean fruit setfruit set in hand-pollinated in hand-pollinated trees (fivetrees trees;(five four trees; fourper branches branches per tree). The tree). The gray gray bandsbands represent represent 95% confidence 95% confidence intervals.intervals. Table 1. Output of Linear Mixed Model (LMM) relating initial fruit (A) and final fruit set (B) to distance from the Osmia cornuta nesting shelters. Parameter- (t) and p-values are provided for the best-fitted models based on AICc selection. R2 m and R2 c are the marginal and conditional R2 of the best-fitted models. A. Initial Fruit Set B. Final Fruit Set t-Value p-Value t-Value p-Value (Intercept) Figure 2. Initial (A) and final (B) fruit 25.27 set (% of flowers setting fruit in April and<0.0001 June, respectively) at 22.63 <0.0001 increasing distances from the Osmia cornuta nesting shelters. Each dot represents a branch. Broken −2.85 Distance − 3.33 <0.0001 0.0001 Distance2 lines indicate mean fruit set in hand-pollinated trees (five2.12 trees; four branches 0.0395 per tree). The gray 1.75 0.0857 bands represent 95% confidence intervals. 2 2 R m: 0.28; R c: 0.67 R2 m: 0.22; R2 c: 0.59 Figure 3. Relationship between initial (April) and final (June) fruit set (% of flowers setting fruit). Each dot represents a branch. The gray band represents the 95% confidence interval. Table 1. Output of Linear Mixed Model (LMM) relating initial fruit (A) and final fruit set (B) to dis- tance from the Osmia cornuta nesting shelters. Parameter- (t) and p-values are provided for the best- fitted models based on AICc selection. R2m and R2c are the marginal and conditional R2 of the best- fitted models. A. Initial Fruit Set B. Final Fruit Set t-Value p-Value t-Value p-Value (Intercept) 25.27 <0.0001 22.63 <0.0001 Distance −3.33 <0.0001 −2.85 0.0001 Distance2 2.12 0.0395 1.75 0.0857 Figure Figure3.3.Relationship Relationship between initialinitial between (April)(April) and final and(June) finalfruit set (% (June) of set (% of flowers setting fruit). fruit flowers setting fruit). R2m: Each dot 0.28; R2c: 0.67 represents a branch. The gray band R 2m: 0.22; R2c: 0.59 represents Each dot represents a branch. The gray band represents the 95% confidence interval. the 95% confidence interval. The best models analyzing the effect of O. cornuta and A. mellifera visitation Table 1. Output of Linear Mixed Model (LMM) relating onfruit initial fruit (A)set andincluded final fruitonly set (B)O.tocornuta dis- visitation, which ex- tance from the Osmia cornuta nesting shelters. Parameter- plained(t)41%andofp-values are fruit the initial provided for the set and 40%best- of the final fruit set (Table 2; fitted models based on AICc selection. R2m and RFigure 2c are the marginal and conditional R2 of the best- 4). Even then, the levels of fruit set obtained across the orchard fitted models. Insects 2021, 12, 56 6 of 11 The best models analyzing the effect of O. cornuta and A. mellifera visitation on fruit set included only O. cornuta visitation, which explained 41% of the initial fruit set and 40% of the final fruit set (Table 2; Figure 4). Even then, the levels of fruit set obtained across the orchard were lower than those obtained in the five hand-pollinated trees (initial fruit set: 57.8 ± 2.0; final fruit set: 36.9 ± 1.8%) (Figure 2), indicating that pollination services could still be increased. Table 2. Output of LM model averaging relating initial (A) and final (B) fruit set to Osmia cornuta and Apis mellifera visitation. Estimated coefficients and their p-values are provided. Adjusted R2 of the best-fitted model (containing only O. cornuta visitation as predictor variable) are provided. (A) Initial Fruit Set. Estimate SE Adjusted SE z Value Pr(>|z|) (Intercept) 0.5545 0.028 0.031 17.62 <2e−16 O. cornuta 0.0099 0.003 0.003 3.08 0.002 A. mellifera −0.0004 0.001 0.001 0.39 0.689 Adjusted R2 : 0.44 (B) Final Fruit Set Insects 2020, 11, x Estimate SE Adjusted SE z Value Pr(>|z|) 7 of 13 (Intercept) 0.489 0.043 0.046 10.63 <2e−16 O. cornuta were lower than 0.009 0.003those obtained 0.003in the five hand-pollinated 2.70 trees (initial 0.007 A. mellifera fruit set: 57.80.001 −0.001 ± 2.0; final fruit0.001 set: 36.9 ± 1.8%) 1.17(Figure 2), indicating 0.241 that pollination services could Adjusted R2 : still 0.36 be increased. Figure Figure 4. Relationship 4. Relationship between between Osmia Osmia cornuta cornuta visitation visitation and and initial initial (A)(A) andand final(B) final (B)fruit fruitset set(% (%ofof flowers flowers setting setting fruit fruit in April in April andand June, June, respectively). respectively). Each Each dot dot representsa atree. represents tree.Gray Graybands bandsrepresent represent 95%95% confidence confidence intervals. intervals. 4. Discussion Table 2. Output of LM model averaging relating initial (A) and final (B) fruit set to Osmia cornuta and Apis Our first mellifera visitation. objective Estimated was toand coefficients assess theirwhether p-values are O. cornuta theprovided. population Adjusted hadbest-fitted R2 of the an impact model on only (containing almond production. O. cornuta visitationMost O. cornuta of thevariable) as predictor observed were recorded within 30 m are provided. from the nesting stations. Although Osmia females are able to locate their nest from (A) Initial Fruit Set. distances as far as 500–1800 m [28,29], populations established in orchards concentrate most of their foraging within 50 m ofSE Estimate Adjusted the nesting SE sites [23,30]. zThe Value negativePr(>|z|) relationship (Intercept) between fruit set and0.5545 distance from 0.028 0.031 closely paralleled the nesting stations 17.62 the<2e−16 distribution of O. cornuta across the orchard. The contribution of O. cornuta to pollination 0.002 O. cornuta 0.0099 0.003 0.003 3.08 service was A. mellifera further confirmed by-0.0004 0.001 the analysis identifying O.0.001 0.39 (but not 0.689 cornuta visitation A. mellifera visitation) as a significant predictor of initialRand Adjusted final fruit set. This result is remarkable 2: 0.44 given that A. mellifera visitation (B) wasFinal 10 times Fruit higher Set than O. cornuta visitation. Apis mellifera is consistently reported Estimate as the SE dominant Adjusted SE species pollinator in commercial z Value orchards Pr(>|z|) (Intercept) 0.489 0.043 0.046 10.63 <2e−16 O. cornuta 0.009 0.003 0.003 2.70 0.007 A. mellifera -0.001 0.001 0.001 1.17 0.241 Adjusted R : 0.36 2 Insects 2021, 12, 56 7 of 11 (e.g., [11,12,48]). However, its contribution to pollination service is strongly limited by its low per-visit pollination effectiveness on fruit tree flowers [20–25]. Our results are limited to a single orchard and a single year. However, they are in line with a previous study that found a significant impact of O. cornuta visitation on final seed-set in a pear orchard in which A. mellifera was 7 times more abundant [23]. These results are also in agreement with studies showing yield increases in orchards pollinated with other Osmia species [11,37,38]. Our study also shows that initial fruit set (commonly used as a proxy of pollination service) is a good predictor of final fruit set in almonds. Our second objective was to establish whether the O. cornuta population could be increased on site, and to identify the main factors limiting population growth. We released 198 females of which 182 emerged out of their cocoons (8.1% winter mortality). This winter mortality is similar to values obtained in other O. cornuta populations managed for orchard pollination (Table 3). Of the 182 females that emerged, 99 (54.4%) established in the nesting stations. Failure to establish can be caused by lack of vigor of emerging females [49], predation [50–52], and dispersal of pre-nesting females [53], but attributing a relative weight to each of these three factors is not easy. At any rate, the percent establishment obtained in our study was similar to values obtained with other O. cornuta populations released in orchards (Table 3). The 99 females that established in the nesting stations produced 906 cells (an average of 9.15 cells/female). This fecundity is again close to fecundity in other O. cornuta populations released in orchards (Table 3). Of these 906 cells, 253 contained female progeny (2.6 M/F sex ratio). This is considerably higher than that the sex ratio of the population released (1.8), but similar to sex ratios obtained from other O. cornuta populations released in orchards (Table 3). Developmental mortality (5.3%) and parasitism (1.8%) were also similar to mortality levels obtained in previous studies (Table 3). Ultimately, the female population recovered was 1.28 times larger than the female population released, providing evidence that Osmia populations can be sustained in orchard environments ([26] and references therein, [54]). Table 3. Population parameters of Osmia cornuta populations released in orchards. All means followed by standard error (SE). This Study Other Studies 7 Winter mortality 1 8.1% 5.5 ± 0.7 (n = 8) Female establishment 2 54.4% 50.7 ± 5.1 (n = 14) Fecundity 3 9.2 9.8 ± 0.9 (n = 10) Progeny sex ratio 4 2.6 2.5 ± 0.18 (n = 12) Developmental mortality 5 5.3% 8.5 ± 0.9 (n = 15) Parasitism 6 1.8% 1.6 ± 0.4 (n = 17) 1Individuals that did not emerge out of their cocoons; 2 Proportion of emerged females that established at the nesting stations provided; 3 Number of eggs laid; 4 Males/females; 5 Proportion of progeny that died from unknown causes; 6 Proportion of progeny that died from cleptoparasitism; 7 [21,23,53,55–58]. The results of Table 3 allow us to identify two important bottlenecks in the dynamics of O. cornuta populations managed for orchard pollination. The first one occurs in the establishment phase, during which the effective female population is reduced by ca. 50%. Previous studies have shown that Osmia establishment can be enhanced by releasing populations within their natal nests, rather than as loose cocoons [33,55,59–61]. This result is probably mediated by olfactory nest cues that enhance the tendency of females to re-nest in their natal nesting site (philopatry; [61]). In relation to this, an olfactory attractant similar to that developed for Osmia lignaria [54] could potentially enhance establishment of O. cornuta populations released in orchards. The second important bottleneck is the production of a male-biased sex ratio. In O. cornuta, the sex ratio of populations trap-nested in semi-natural areas is 1.5 ± 0.06 (n = 4 populations) [62,63], a figure that closely matches the theoretically optimal sex ratio based on male and female body weights (1.7) [42]. However, progeny sex ratios obtained from populations released in orchards are consistently higher (ca. 2.5; Table 3). In other words, in populations nesting in orchards, a considerable fraction of the Insects 2021, 12, 56 8 of 11 parental investment is devoted to the production of surplus males. The causes underlying these differences in sexual allocation are unclear, but a greater proportion of female cells in managed populations could be obtained by increasing the diameter [56,64,65], or the length [55,65] of the nesting cavities offered in orchard operations. A more balanced sex ratio (closer to 1.7) would increase the reproductive and pollinating potential of populations recovered from orchards. 5. Conclusions Our study demonstrates that even a small population of a highly effective pollinator may have a significant impact on crop pollination service and fruit set. Our results are encouraging not only for the use of Osmia spp. as managed pollinators but also for the implementation of measures to protect wild pollinator communities in orchard environ- ments. In addition to Osmia spp., various species of Andrena, Eucera and Bombus are highly effective fruit tree pollinators [5,22,24,25,36]. In the current scenario of pollinator declines, agri-environmental measures to enhance wild populations of these valuable pollinators could have important economic returns in terms of enhanced pollination service and fruit yields. Supplementary Materials: The following are available online at https://www.mdpi.com/2075-445 0/12/1/56/s1, dataset. Table S1: Fruit set vs. Distance; Table S2: Osmia vs. fset row1. Author Contributions: J.B., conceived the research. J.B., S.O.-C., and N.V. conducted the study. S.O.-C. analyzed the data. J.B., S.O.-C., N.V., and F.S. wrote the manuscript. J.B. secured funding. All authors have read and agreed to the published version of the manuscript. Funding: This study was supported by the DGICYT (project AGR. 91-0988-CO3) and the Spanish MEC (FPI grant to NV). Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data are available in the Supplementary Materials. Acknowledgments: We are grateful to J. Calzadilla and M. A. Escolano for their help with field work and nest analysis. Conflicts of Interest: The authors declare no conflict of interest. Appendix A Table A1. Output of Linear Mixed Model (LMM) relating initial fruit (A) and final fruit set (B) to distance from the Osmia cornuta nesting stations considering a linear fit. R2 m and R2 c are the marginal and conditional R2 of the best-fitted models. (A) Initial Fruit Set t-Value p-Value (Intercept) 32.03 <0.0001 Distance −4.73 <0.0001 R2 m: 0.24; R2 c: 0.66 (B) Final Fruit Set t-Value p-Value (Intercept) 29.22 <0.0001 Distance −4.31 0.0001 R2 m: 0.19; R2 c: 0.59 (B) Final Fruit Set R2m: 0.24; R2c: 0.66 t-Value p-Value (B) Final Fruit Set (Intercept) 29.22 <0.0001 t-Value p-Value Distance −4.31 0.0001 (Intercept) 29.22 <0.0001 Insects 2021, 12, 56 R2m: 0.19; R2c: 0.59 9 of 11 Distance −4.31 0.0001 R m: 0.19; R2c: 0.59 2 Figure A1. Structure of the study orchard and location of the Osmia cornuta nesting stations. Figure A1. Structure of the study orchard orchard and and location location of the Osmia of the Osmia cornuta cornuta nesting stations. nesting stations. Figure A2. Initial (A) and final (B) fruit fruit set set (% (% of of flowers flowerssetting settingfruit fruitininApril Apriland June, respectively) and June, respectively) at increasing distances from the Osmia cornuta nesting at increasing distances from the Osmia cornuta nesting shelters. Each dot represents a branch. Broken Figure A2. shelters. EachInitial dot (A) and final represents (B) fruitBroken a branch. set (% of flowers setting lines fruit in set April lines indicate mean fruit set in hand-pollinated treesindicate (5 trees;mean fruit 4 branches in tree). The gray bands per and June, respectively) hand-pollinated trees (5at increasing trees; 4 distances branches per from tree). the The Osmia gray cornuta bands nesting represent represent shelters. 95% dot Each confidence representsintervals. a branch. Broken lines indicate mean fruit set in 95% confidence intervals. hand-pollinated trees (5 trees; 4 branches per tree). The gray bands represent References 95% confidence intervals. References 1. Klein, A.-M.; Vaissière, B.E.; Cane, J.H.; Steffan-Dewenter, I.; Cunningham, S.A.; Kremen, C.; Tscharntke, T. Importance of 1. pollinators in Vaissière, Klein, A.-M.; changing landscapes B.E.; Cane, for world J.H.; crops. Proc. R.I.; Steffan-Dewenter, Soc. B 2007, 274, S.A.; Cunningham, 303–313. [CrossRef] Kremen, [PubMed] C.; Tscharntke, References 2. Winfree, R.; Williams, T. Importance N.M.;in of pollinators Gaines, changing H.; landscapes Ascher, J.S.;for Kremen, C. Wild world crops. beeR.pollinators Proc. Soc. B 2007,provide the majority of crop visitation 274, 303–313. 1. across Klein, land-use A.-M.; Vaissière, gradientsB.E.; Cane, in New J.H.; and Jersey Steffan-Dewenter, Pennsylvania, USA. I.; Cunningham, J. Appl. Ecol.S.A.; 2008,Kremen, C.; Tscharntke, 45, 793–802. [CrossRef] 3. T. Importance Garibaldi, L.A.;ofSteffan-Dewenter, pollinators in changing landscapes I.; Winfree, R.; Aizen,for M.A.; worldBommarco, crops. Proc.R.;R. Cunningham, Soc. B 2007, 274, 303–313. S.A.; Kremen, C.; Carvalheiro, L.G.; Harder, L.D.; Afik, O.; et al. Wild Pollinators enhance fruit set of crops regardless of honey bee abundance. Science 2013, 339, 1608–1611. [CrossRef] [PubMed] 4. Hevia, V.; Bosch, J.; Azcárate, F.M.; Fernández, E.; Rodrigo, A.; Barril-Graells, H.; González, J.A. Bee diversity and abundance in a livestock drove road and its impact on pollination and seed set in adjacent sunflower fields. Agric. Ecosyst. Environ. 2016, 232, 336–344. [CrossRef] 5. Park, M.G.; Raguso, R.A.; Losey, J.E.; Danforth, B.N. Per-visit pollinator performance and regional importance of wild Bombus and Andrena (Melandrena) compared to the managed honey bee in New York apple orchards. Apidologie 2016, 47, 145–160. [CrossRef] 6. Klein, A.-M.; Brittain, C.; Hendrix, S.D.; Thorp, R.; Williams, N.; Kremen, C. Wild pollination services to California almond rely on semi-natural habitat. J. Appl. Ecol. 2012, 49, 723–732. [CrossRef] 7. Woodcock, B.A.; Bullock, J.M.; Shore, R.F.; Heard, M.S.; Pereira, M.G.; Redhead, J.; Ridding, L.; Dean, H.; Sleep, D.; Henrys, P.; et al. Country-specific effects of neonicotinoid pesticides on honey bees and wild bees. Science 2017, 356, 1393–1395. [CrossRef] [PubMed] 8. Hass, A.L.; Kormann, U.G.; Tscharntke, T.; Clough, Y.; Bosem Baillod, A.; Sirami, C.; Fahrig, L.; Martin, J.-L.; Baudry, J.; Bertrand, C.; et al. Landscape configurational heterogeneity by small-scale agriculture, not crop diversity, maintains pollinators and plant reproduction in Western Europe. Proc. R. Soc. B 2018, 285, 20172242. [CrossRef] [PubMed] Insects 2021, 12, 56 10 of 11 9. Sirami, C.; Gross, N.; Baillod, A.B.; Bertrand, C.; Carrié, R.; Hass, A.; Henckel, L.; Miguet, P.; Vuillot, C.; Alignier, A.; et al. Increasing crop heterogeneity enhances multitrophic diversity across agricultural regions. Proc. Natl. Acad. Sci. USA 2019, 116, 16442–16447. [CrossRef] [PubMed] 10. Kennedy, C.M.; Lonsdorf, E.; Neel, M.C.; Williams, N.M.; Ricketts, T.H.; Winfree, R.; Bommarco, R.; Brittain, C.; Burley, A.L.; Cariveau, D.; et al. A global quantitative synthesis of local and landscape effects on wild bee pollinators in agroecosystems. Ecol. Lett. 2013, 16, 584–599. [CrossRef] [PubMed] 11. Pitts-Singer, T.L.; Artz, D.R.; Peterson, S.S.; Boyle, N.K.; Wardell, G.I. Examination of a managed pollinator strategy for almond production using Apis mellifera (Hymenoptera: Apidae) and Osmia lignaria (Hymenoptera: Megachilidae). Environ. Entomol. 2018, 47, 364–377. [CrossRef] [PubMed] 12. Roquer-Beni, L.; Alins, G.; Arnan, X.; Boreux, V.; García, D.; Hambäck, P.A.; Happe, A.-K.; Klein, A.-M.; Miñarro, M.; Mody, K.; et al. Management-dependent effects of pollinator functional diversity on apple pollination services: A response-effect trait approach. (Unpublished; Manuscript in Preparation). 13. McGregor, S.E. Insect Pollination of Cultivated Crop Plants; Agricultural Research Service: Washington, DC, USA, 1976; p. 411. 14. Free, J.B. Insect Pollination of Crops, 2nd ed.; Academic Press: London, UK, 1993. 15. Delaplane, K.S.; Mayer, D.R. Crop Pollination by Bees; CABI: Wallingford, UK, 2000. 16. Vicens, N.; Bosch, J. Weather-dependent pollinator activity in an apple orchard, with special reference to Osmia cornuta and Apis mellifera (Hymenoptera: Megachilidae and Apidae). Environ. Entomol. 2000, 29, 413–420. [CrossRef] 17. Visscher, P.K.; Seeley, T.D. Foraging strategy of honeybee colonies in temperate deciduous forest. Ecology 1982, 63, 1790–1801. [CrossRef] 18. Beekman, M.; Ratnieks, F.L.W. Long-range foraging by the honey-bee, Apis mellifera L. Funct. Ecol. 2000, 14, 490–496. [CrossRef] 19. Hagler, J.R.; Mueller, S.; Teuber, L.R.; Machtley, S.A.; Van Deynze, A. Foraging range of honey bees, Apis mellifera, in alfalfa seed production fields. J. Insect Sci. 2011, 11, 1–12. [CrossRef] [PubMed] 20. Bosch, J.; Blas, M. Foraging behaviour and pollinating efficiency of Osmia cornuta and Apis mellifera on almond (Hymenoptera, Megachilidae and Apidae). Appl. Entomol. Zool. 1994, 29, 1–9. [CrossRef] 21. Vicens, N.; Bosch, J. Pollinating efficacy of Osmia cornuta and Apis mellifera (Hymenoptera: Megachilidae, Apidae) on ‘Red Delicious’ apple. Environ. Entomol. 2000, 29, 235–240. [CrossRef] 22. Thomson, J.D.; Goodell, K. Pollen removal and deposition by honeybee and bumblebee visitors to apple and almond flowers. J. Appl. Ecol. 2001, 38, 1032–1044. [CrossRef] 23. Monzón, V.H.; Bosch, J.; Retana, J. Foraging behavior and pollinating effectiveness of Osmia cornuta (Hymenoptera: Megachilidae) and Apis mellifera (Hymenoptera: Apidae) on “Comice” pear. Apidologie 2004, 35, 575–585. [CrossRef] 24. Eeraerts, M.; Vanderhaegen, R.; Smagghe, G.; Meeus, I. Pollination efficiency and foraging behaviour of honey bees and non-Apis bees to sweet cherry. Agric. For. Entomol. 2020, 22, 75–82. [CrossRef] 25. Roquer-Beni, L.; Arnan, X.; Rodrigo, A.; Bosch, J. What Makes a Good Pollinator? Relationship between Pollinator Traits and Pollination Effectiveness in Apple Flowers. (Unpublished; Manuscript in Preparation). 26. Bosch, J.; Kemp, W.P. Developing and establishing bee species as crop pollinators: The example of Osmia spp. (Hymenoptera: Megachilidae) and fruit trees. Bull. Entomol. Res. 2002, 92, 3–16. [PubMed] 27. Sédivy, C.; Dorn, S. Towards a sustainable management of bees of the subgenus Osmia (Megachilidae; Osmia) as fruit tree pollinators. Apidologie 2014, 45, 88–105. [CrossRef] 28. Vicens, N.; Bosch, J. Nest site orientation and relocation of populations of the orchard pollinator Osmia cornuta (Hymenoptera: Megachilidae). Environ. Entomol. 29, 69–75. [CrossRef] 29. Guédot, C.; Bosch, J.; Kemp, W.P. Effect of three-dimension and color contrast on nest localization performance of two solitary bees (Hymenoptera: Megachilidae). J. Kansas Entomol. Soc. 2007, 80, 90–104. [CrossRef] 30. Biddinger, D.; Joshi, N.; Rajotte, E.; Halbrendt, N. An immunomarking method to determine the foraging patterns of Osmia cornifrons and resulting fruit set in a cherry orchard. Apidologie 2013, 44, 738–749. [CrossRef] 31. Torchio, P.F. Use of Osmia lignaria Say (Hymenoptera: Apoidea: Megachilidae) as a pollinator in an apple and prune orchard. J. Kansas Entomol. Soc. 1976, 49, 475–482. 32. Torchio, P.F. Field experiments with Osmia lignaria propinqua Cresson as a pollinator in almond orchards: II, 1976 studies (Hymenoptera: Megachilidae). J. Kansas Entomol. Soc. 1981, 54, 824–836. 33. Maeta, Y. Comparative studies on the biology of the bees of the genus Osmia in Japan, with special reference to their management for pollination of crops (Hymenoptera, Megachilidae). Bull. Tohoku Natl. Agric. Exp. Stn. 1978, 57, 1–221. (In Japanese) 34. Márquez, J.; Bosch, J.; Vicens, N. Pollens collected by wild and managed populations of the potential orchard pollinator Osmia cornuta (Latr.) (Hym., Megachilidae). J. Appl. Entomol. 1994, 117, 353–359. [CrossRef] 35. Maccagnani, B.; Ladurner, E.; Santi, F.; Burgio, G. Osmia cornuta (Hymenoptera, Megachilidae) as a pollinator of pear (Pyrus communis): Fruit- and seed-set. Apidologie 2003, 34, 207–216. [CrossRef] 36. Garratt, M.P.D.; Breeze, T.D.; Boreux, V.; Fountain, M.T.; McKerchar, M.; Webber, S.M.; Coston, D.J.; Jenner, N.; Dean, R.; Westbury, D.B.; et al. Apple pollination: Demand depends on variety and supply depends on pollinator identity. PLoS ONE 2016, 11, e0153889. [CrossRef] [PubMed] 37. Bosch, J.; Kemp, W.P.; Trostle, G.E. Bee population returns and cherry yields in an orchard pollinated with Osmia lignaria (Hymenoptera: Megachilidae). J. Econ. Entomol. 2006, 99, 408–413. [CrossRef] Insects 2021, 12, 56 11 of 11 38. Sheffield, C.S. Pollination, seed set and fruit quality in apple: Studies with Osmia lignaria (Hymenoptera: Megachilidae) in the Annapolis Valley, Nova Scotia, Canada. J. Pollinat. Ecol. 2014, 12. [CrossRef] 39. Velthuis, H.H.W.; van Doorn, A. A century of advances in bumblebee domestication and the economic and environmental aspects of its commercialization for pollination. Apidologie 2006, 37, 421–451. [CrossRef] 40. van der Steen, J.J.M. Indoor rearing of the solitary bee Osmia rufa L. Proc. Exp. Appl. Entomol. NEV Amst. 1997, 8, 81–84. 41. Ladurner, E.; Maccagnani, B.; Tesoriero, D.; Nepi, M.; Felicioli, A. Laboratory rearing of Osmia cornuta Latreille (Hymenoptera Megachilidae) on artificial diet. Boll. Dell’ist. Entomol. Univ. Bologna 1999, 53, 133–146. 42. Bosch, J.; Vicens, N. Body size as an estimator of production costs in a solitary bee. Ecol. Entomol. 2002, 27, 129–137. [CrossRef] 43. Kester, D.E.; Griggs, W.H. Fruit setting in the almond: The effect of cross-pollinating various percentages of flowers. Proc. Am. Soc. Hortic. Sci. 1959, 74, 206–213. 44. Burnham, K.P.; Anderson, D.R. Model Selection and Inference: A Practical Information-Theoretic Approach, 1st ed.; Springer: New York, NY, USA, 2002. [CrossRef] 45. Barton, K. ‘MuMIn’: Multi-Model Inference. R Package Version 1.43.17. Available online: https://CRAN.R-project.org/package= MuMIn (accessed on 11 January 2021). 46. Pinheiro, J.; Bates, D.; DebRoy, S.; Sarkar, D. ‘nlme’: Linear and Nonlinear Mixed Effects Models. R Package Version 3.1-151. Available online: https://CRAN.R-project.org/package=nlme (accessed on 11 January 2021). 47. R Core Team. R: A Language and Environment for Statistical Computing, Version 4.0.1; R Foundation for Statistical Computing: Vienna, Austria, 2020. 48. Melin, A.; Rouget, M.; Midgley, J.J.; Donadson, J.S. Pollination ecosystem services in South African agricultural systems. S. Afr. J. Sci. 2014, 110, 1–9. [CrossRef] 49. Sgolastra, F.; Arnan, X.; Pitts-Singer, T.L.; Maini, S.; Kemp, W.P.; Bosch, J. Pre-wintering conditions and post-winter performance in a solitary bee: Does diapause impose an energetic cost on reproductive success? Ecol. Entomol. 2016, 41, 201–210. [CrossRef] 50. Felicioli, A. Le osmie. In Api e Impollinazione; Pinzauti, M., Ed.; Giunta Regionale Toscana: Firenze, Italy, 2000; pp. 159–188. 51. Krunić, M.; Stanisaviljević, L.Ž. The Biology of European Orchard bee Osmia Cornuta (Latr.) (Hymenoptera: Megachilidae); Faculty of Biology University of Belgrade: Belgrade, Serbia, 2006. 52. Bosch, J.; Kemp, W.P. How to Manage the Blue Orchard Bee: As an Orchard Pollinator; Sustainable Agriculture Network: Beltsville, MD, USA, 2001. 53. Bosch, J.; Vicens, N. Relationship between body size, provisioning rate, longevity and reproductive success in females of the solitary bee Osmia cornuta. Behav. Ecol. Sociobiol. 2006, 60, 26–33. [CrossRef] 54. Boyle, N.K.; Pitts-Singer, T.L. The effect of nest box distribution on sustainable propagation of Osmia lignaria (Hymenoptera: Megachilidae) in commercial tart cherry orchards. J. Insect Sci. 2016, 17, 41. [CrossRef] 55. Bosch, J. Osmia cornuta Latr. (Hymenoptera, Megachilidae) as a potential pollinator in almond orchards: Releasing methods and nest-hole length. J. Appl. Entomol. 1994, 117, 151–157. [CrossRef] 56. Bosch, J. Improvement of field management of Osmia cornuta Latr. for almond pollination (Hymenoptera, Megachilidae). Apidologie 1994, 25, 71–83. [CrossRef] 57. Bosch, J. Parasitism in wild and managed populations of the almond pollinator Osmia cornuta Latr. (Hymenoptera, Megachilidae). J. Apicult. Res. 1992, 31, 77–82. [CrossRef] 58. Bosch, J. Comparison of nesting materials for the orchard pollinator Osmia cornuta Latr. (Hymenoptera, Megachilidae). Entomol. Gen. 1995, 19, 285–289. [CrossRef] 59. Torchio, P.F. Field experiments with the pollinator species, Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae) in apple orchards: III, 1977 studies. J. Kansas Entomol. Soc. 1984, 57, 517–521. 60. Torchio, P.F. Field experiments with the pollinator species, Osmia lignaria propinqua Cresson in apple orchards: V, 1979–1980, methods of introducing bees, nesting success, seed counts, fruit yields (Hymenoptera: Megachilidae). J. Kansas Entomol. Soc. 1985, 58, 448–464. 61. Pitts-Singer, T.L. Olfactory response of megachilid bees, Osmia lignaria, Megachile rotundata, and M. pugnata, to individual cues from old nest cavities. Environ. Entomol. 2007, 36, 402–408. [CrossRef] 62. Vicens, N.; Bosch, J.; Blas, M. Análisis de los nidos de algunas Osmia (Hymenoptera, Megachilidae) nidificantes en cavidades preestablecidas. Orsis 1993, 8, 41–52. 63. Bosch, J.; Vicens, N. Sex allocation in the solitary bee Osmia cornuta: Do females behave in agreement with Fisher’s theory? Behav. Ecol. Sociobiol. 2005, 59, 124–132. [CrossRef] 64. Tepedino, V.J.; Torchio, P.F. The influence of nest-hole selection on sex ratio and progeny size in Osmia lignaria propinqua (Hymenoptera: Megachilidae). Ann. Entomol. Soc. Am. 1989, 82, 355–360. [CrossRef] 65. Seidelmann, K.; Bienasch, A.; Pröhl, F. The impact of nest tube dimensions on reproduction parameters in a cavity nesting solitary bee, Osmia bicornis (Hymenoptera: Megachili-dae). Apidologie 2016, 47, 114–122. [CrossRef]