While Gleiser et al.’s recent analysis of yield variability across all crops in the FAO data found that increasing pollinator dependence was positively associated with yield variability, their inclusion of annual plants prevents meaningful inference for perennials. However, they observed greater interannual variability in woody than herbaceous plants, which is consistent with alternate bearing in long-lived perennials. Another difference is that we restricted our analysis to plant species known to be alternate bearing at the individual level. To the best of our knowledge, ours is the first study to use FAO data to assess patterns of yield variability specifically in alternate-bearing plants. Second, the horticultural literature has focused extensively on the genetic and hormonal bases of alternate bearing in an effort to achieve more stable yields. By contrast, genetic studies in wild-masting trees are scarce but could be valuable for understanding drivers of masting in the wild . Finally, the existing variability in farm practices and the potential to manipulate them through space and time offer novel opportunities to explore the roles of resource availability, habitat structure, phenology, planting blueberries in pots pollination dynamics and genetics in synchronous perennial plant reproduction.

We observed a national signal of alternate bearing in some insect-pollinated crops , despite a stronger tendency toward alternate bearing in wind-pollinated crops. Further work could explore whether crop alternate bearing emerged from a similar pathway across pollination syndromes or as a convergent trait. In wind-pollinated crops, breeding and cultivation may have served to shift the reproductive pattern from stochastic masting toward a relatively more deterministic biennial pattern of reproduction. For insect-pollinated crops, the alternate bearing could be a symptom of agricultural intensification if plants in their native habitat rarely ‘overinvested’ resources in reproduction and instead produced a steady, low number of seeds. Monocultures and high-density planting, for example, can make crops more susceptible to pest outbreaks and more exposed to environmental disasters, which would induce synchrony if there were an endogenous resource-driven mechanism involved in alternate bearing. Coffee illustrates this well; the plant is native to the shaded understory of Ethiopian rainforest but is now often cultivated in sun monocultures that have been shown to exaggerate alternate bearing and resource tradeoffs . Data on the reproductive patterns of wild/ancestral lineages of contemporary perennial crops are scarce but would shed valuable insight on the basis of contemporary alternate bearing. Unexpectedly, some crop series appeared to display longer-term cycles . Such a pattern could arise at farm scales as a result of pruning and subsequent recovery, but we would not anticipate these farmer practices to be synchronized at a national scale. Long-term climatic cycles, such as El Niño Southern Oscillation events, could also play a role, particularly in rainfed systems that we may expect to exhibit stronger variability than their irrigated counterparts.

ENSO phases have been shown to induce synchrony in masting systems, but knowledge of ENSO effects on crop plants is largely limited to annual crops. As a primary source of climate variation in Brazil and Iran, ENSO could be a cause of periodic yield in Brazilian tangerine and Iranian apricot . If future work confirms a biological basis of long cycles in some crops at a national scale, it would be profitable to understand the extent to which these longer-term crop fluctuations reflect exogenous forcing versus endogenous feedbacks . In conclusion, we have found that perennial crops frequently exhibit alternate bearing even at a national scale and that this is especially pronounced in wind-pollinated crops. This pattern is remarkable, given the general assumption that management practices have come to outweigh the ecological drivers that would synchronize country-wide production. Our results suggest that historic yield data present a thus far underused resource for further analyses on the mechanisms of reproductive synchrony across time, space and taxa. Future work could explore the intraspecific and intraregional differences in synchrony and the degree of overlap and divergence between patterns in natural and managed systems. We encourage strengthened collaborations between theoretical ecologists, applied ecologists and horticulturalists for the mutual benefits of achieving an enhanced understanding of the mechanisms of synchronous interannual variability and promoting stable yields and farmer livelihoods.The honey bee is the most prevalent single species of floral visitor worldwide both in agricultural fields and natural plant communities . In natural habitats where the honey bee is present, it accounts for nearly 13% of total floral visits and visits 37% of all plants present, on average, as documented in plant-pollinator network studies, which report the identity and frequency of pollinators that visit flowers of each plant species in a given study area .

While plant-pollinator network studies usually gauge pollinator importance solely by interaction frequency due to the strong correlation between plant pollinator interaction frequency and overall reproductive impact , there are cases where visitation frequency does not reflect importance, due to the low pollinator efficiency of some frequent floral visitors . The importance of a pollinator to plant reproduction is often defined as the product of interaction frequency and per-interaction effect . Despite the ubiquity of honey bee presence in natural systems worldwide, their relative performance as pollinators is largely unknown because of the difficulty of assessing pollinator efficiencies over a wide range of species. Our goal here is to assess whether visitation frequency adequately reflects average honey bee importance as a pollinator by determining whether the pollination efficiencies of honey bees on various species differ systematically from those of other floral visitors. We also assess the importance of honey bees as pollinators of species where both the per-visit efficiencies and visitation frequencies of various floral visitors were measured. To that purpose, we conducted a meta-analysis of published studies that report the efficiency of honey bees as pollinators relative to other floral visitors for both domesticated and undomesticated plants. Pollinator efficiency is the average result of a single visit by a particular species of floral visitor to a flower of a particular species . Currency of pollinator efficiency varies among studies as either the number of seeds set, the probability of fruit set, or the number of pollen grains deposited on the stigma resulting from a single visit. Additionally, for the subset of studies which report visitation data for honey bees and other pollinators, we analyzed the relative pollination importance where importance is the product of the relative pollination efficiencies of each pollinator type studied and that pollinator’s relative visitation frequency. Because pollinator efficiencies of honey bees are often measured in agricultural systems, we also assess whether the relative efficiency and importance of honey bees as pollinators differs between domesticated and undomesticated plant species The pollination efficiency of honey bees is of interest for several reasons. First, the honey bee might be expected to be less efficient compared to native pollinators in regions of the world where it has been introduced, though specialist and native pollinators don’t necessarily have higher pollinator efficiencies . Second, worldwide habitat fragmentation and climate change threaten to decrease pollinator diversity and abundance. As resource availability decreases with habitat fragmentation more sharply for specialist pollinators than generalist pollinators, specialists may experience greater population declines than generalists . Furthermore, in South America for example,populations of non-native honey bee and bumblebee species increase in abundance with habitat fragmentation and disturbance relative to native species . Therefore, evaluating the efficiency and importance of honey bees as pollinators to a variety of plant species, raspberries in pots could provide crucial information about how well honey bees might fill the void as pollinator diversity decreases.Data for efficiency analysis consist of single-visit pollination efficiency trials for single plant species, in which researchers investigated the efficiency of both honey bees and at least one other floral visitor. We used two approaches to compile our data set. First, we performed a literature search using the ISI Web of Science database with the search term [pollinat*] in combination with one of the following terms: [efficiency], [effectiveness], [“pollen deposition], [seed set], [fruit set], or [“pollination biology of”], from October 2014 to August 2016. Second, we examined the literature cited sections of each of the studies found through the first approach for additional studies that were not captured in the initial literature search.

The most recent study was used when more than one study was found for the same plant species. Selected papers experimentally compared pollinator efficiency of two or more pollinators using at least one of the following metrics: average pollen deposition, seed set, or fruit set resulting from a single visit by an individual pollinator species. If necessary, seed set was calculated as the number of seeds per fruit times percent fruit set. When multiple metrics were available from the same study, we chose seed set when reported, and fruit set if seed set was not reported, as seed set is the most accurate metric of fitness . Pollen deposition was used only when no other efficiency metric was available. In a small number of cases, we used Image-J to extract data from figures when raw data were not available. For every plant species in each study, we calculated the average single-visit pollination efficiency of non-honey bee pollinators by taking the mean efficiency metric of all non-honey bee pollinators studied. Single-visit pollination efficiency of the top non-honey bee pollinator was taken at the lowest taxanomic class provided by study . When there was a single non-honey bee pollinator studied, the efficiency measure for average and top non-honey bee pollinator was the same. Next, we calculated the relative single-visit pollination efficiency of honey bees by dividing honey bee efficiency by the average pollination efficiency of all non-honey bee pollinators or by the efficiency of the top non-honey bee pollinator. We then used one-sample t-tests to examine whether the relative efficiency of honey bees differed from that of the average or top non-honey bee pollinator, where the null hypothesis was honey bee relative pollination efficiency equals one. Additionally, for the subset of studies which report visitation rate data for honey bees and other pollinators, pollinator importance was calculated for each pollinator as [ pollination efficiency * visitation rate ]. Relative pollination importance of honey bees was then compared against the importance of the average or top non-honey bee pollinator as above. Plants were categorized as domesticated or undomesticated based on whether the species has been cultivated and grown by humans for a specific use . A plant species that has been cultivated, but was studied in its native habitat where it is a naturally occurring species was considered undomesticated. We ran four one-way anova tests, to determine whether there was a difference between domesticated and undomesticated plants in terms of relative honey bee efficiency and importance, compared to the average or top non-honey bee pollinator. All statistical tests were conducted on JMP version 11 software.Among the wide diversity of plant species studied, honey bees did not differ in efficiency from the average floral visitor, though they were somewhat less efficient than the most efficient non-honey bee species. In addition, there was no difference between efficiency of honey bees relative to the average floral visitor on domesticated versus undomesticated species. Therefore our total dataset of 35 plant species provides a broad sample with which to compare the efficiency of honey bees to those of other common floral visitors to these widely varied plant species. Our findings that honey bees perform as well as the average non-honey bee floral visitor imply that, while the relative efficiency of honey bees varies widely across species, their community-wide effect as pollinators may be reasonably approximated by their visitation frequency. For many of the efficiency studies analyzed, per-visit pollination success was compared between honey bees and the one or two other most frequent flower visiting species. This occurs because frequent visitors are often deemed, a priori, the most significant pollinators and because of the difficulty of obtaining enough data to accurately estimate the average single-visit efficiency of infrequent floral visitors. By contrast, some efficiency studies lump non-honey bee pollinators to taxonomic groupings such as genus, order, or functional type rather than a single pollinating species, therefore the gap between honey bee relative efficiency and the top pollinator may be, in some cases, underestimated. Honey bees are introduced to many areas of the world where they are currently abundant pollinators in natural ecosystems . In fact, in our dataset of n=20 undomesticated plant species in which the efficiency of honey bees was assessed relative to other floral visitors, only one study was performed within the native range of honeybees while all plant species studied were in their native ranges.