See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/328978780

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry

(Ericales: Ericaceae) Agroecosystem

Article  in  The Open Agriculture Journal · November 2018

DOI: 10.1515/opag-2018-0044

CITATIONS

0READS

4

3 authors:

Some of the authors of this publication are also working on these related projects:

Research on developing strategies for maximizing the health and size of the population of pollinators that provide buzz pollination View project

Sustainable Agricultural Intensification and Sustainable Agricultural Diversification Pathways for Brazil View project

Eric Venturini

The Xerces Society for Invertebrate Conservation

15 PUBLICATIONS   32 CITATIONS   

SEE PROFILE

Francis Drummond

University of Maine

283 PUBLICATIONS   2,097 CITATIONS   

SEE PROFILE

Aaron Kinyu Hoshide

University of Maine

39 PUBLICATIONS   101 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Eric Venturini on 16 November 2018.

The user has requested enhancement of the downloaded file.

Open Agriculture. 2018; 3: 393–403

evidence and economic justification that weeds must be controlled prior to planting and represents one of the first studies to empirically test organic strategies for wildflower establishment in an agricultural context.

Keywords: native bees, sustainable, agroecosystem, economics, Maine

1 IntroductionEfforts to conserve wild pollinators are ongoing in Europe (Osborne et al. 1991; Dicks et al. 2015; Dicks et al. 2016), and gaining momentum in North America (The White House 2015). Not-for-profit organizations (Xerces 2016) promote Pollination Reservoirs (PRs), which can boost wild bee diversity and abundance (Wood et al. 2015), and increase crop pollination services (Blaauw and Isaacs 2014; Pywell et al. 2015; Venturini et al. 2017a). These practices in the U.S. are incentivized by the United States Department of Agriculture’s Natural Resources Conservation Service (USDA NRCS) and recognized in the consumer market by Xerces’ certification program, “Bee Better Certified.” Pollinator dependent growers have the added incentive of increasing crop pollination services.

Pollination reservoirs are areas of pollen and nectar rich flowers that increase forage for pollinators, providing a reservoir of wild pollinators for crop pollination. PRs are increasingly common, however attempts to establish them too often fail (Ahern and Boughton 1994; J. Norment, personal communication). In England, grower-training initiatives have helped to increase establishment success (McCracken et al. 2015). Weeds are the most common cause of establishment failure, outcompeting slow growing perennial wildflowers and decreasing stand life (Ahern et al. 1992; Aldrich 2002). Organically approved weed control options are few and can delay planting by an entire year. As efforts to create pollinator habitat grow

https://doi.org/10.1515/opag-2018-0044received March 11, 2018; accepted September 12, 2018

Abstract: Pollination reservoirs are pollen and nectar rich wildflower plantings intended to enhance pollination services in pollinator-dependent crops. Despite government assistance, plantings often fail to establish. Our focal crop, wild blueberries, is a unique cropping-system native to the U.S.A. It is never planted or cultivated, and typically exists in isolated fields within a mostly coniferous forest matrix. Our study takes place in Maine, U.S.A., where growers could economically benefit by switching reliance from rented honey bees to native bee pollination. Lowbush blueberry growers support wild bee enhancement efforts, but the low pH (4.0-5.0) of this agroecosystem presents unique challenges to wildflower establishment. We sought to identify methods that Organic certified growers can use to successfully establish pollination reservoirs in this system. We tested the effects of nurse crops and mowing on the success of a custom wildflower mixture over four years. Success was considered in terms of longevity, sown species diversity, aboveground biomass, and the number and weight of inflorescences. The authors present an economic analysis of cost versus projected planting longevity. In the fourth year of establishment, sown plant diversity significantly decreased, Solidago spp. weeds became dominant, and treatments were not a strong determinant of planting success. The economic analysis suggests that the high cost of pollination reservoir establishment may be a barrier to grower adoption. This study provides

Research Article

E. M. Venturini, F. A. Drummond, A. K. Hoshide

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry (Ericales: Ericaceae) Agroecosystem

*Corresponding author: E. M. Venturini, School of Biology and Ecology, University of Maine, 5722 Deering Hall, Orono, ME 04469 (at time of study); The Xerces Society for Invertebrate Conservation, 1423 Broadway, Bangor, ME 04401 (current), E-mail: Eric.Venturini@xerces.orgF. A. Drummond, School of Biology and Ecology, University of Maine, 5722 Deering Hall, Orono, ME 04469; University of Maine Cooperative Extension, 495 College Avenue, Orono, ME 04469A.K. Hoshide, School of Economics, University of Maine, 206 Winslow Hall, Orono, ME 04469

Open Access. © 2018 E. M. Venturini et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 License. Unauthenticated

Download Date | 11/16/18 3:45 AM

394   E. M. Venturini, et al.

(The White House 2015), the development of effective and economic research-based strategies for establishing PRs is becoming increasingly important.

Most peer-reviewed wildflower establishment literature is focused on U.S. roadside plantings (Ahern et al. 1992; Ahern and Boughton 1994), but also beautification and landscaping (Weaner 1996; Love et al. 2016). The United States Department of Transportation in many states is replacing grassy swards with perennial wildflower plantings (Quarles 2003; Markwardt 2004). The Maine Department of Transportation recently conducted pollinator surveys to inform pollinator management strategies. Wildflower verges benefit a wide variety of wildlife and require less maintenance than grassy verges once established (Ahern et al. 1992), eventually resulting in cost savings (Bretzel et al. 2009). Ecological landscape design can also include wildflower meadows (Weaner 1996). The establishment success of PRs is largely driven by microsite scale factors (Frances et al. 2010). Although roadside planting and landscape design literature can provide a framework, these approaches must be adapted for the unique goals and site-specific challenges of agricultural applications in pollinator-dependent agroecosystems.

Our study took place in the lowbush blueberry

(Vaccinium angustifolium Aiton) agroecosystem in Maine, USA (Figure 1). This cropping system relies upon both wild and managed pollinators, without which blueberries would not set fruit (Free 1970). The state of Maine produces 97-99% of United States lowbush blueberries and contains over 24,000 ha of the crop (Strik and Yarborough 2005; Bell et al. 2009). Lowbush blueberries represent a significant source of income in Maine, contributing an estimated $250 million to the state each year (Smith, 2004). Managed biennially, the crop is harvested and pruned in alternate years. Post-harvest fields are either mowed or burned and virtually no aboveground biomass is left after pruning. Plants regrow shoots from underground rhizomes in the prune year. Weed control depends on scale, but relies upon biennial pruning, herbicides, and maintenance of a pH between 4.0 and 5.0 (Yarborough 2009). Organic lowbush blueberry growers do not use herbicides for weed management, relying only on the prune cycle, wood chip mulches, a low pH (as managed by sulfur application), and hand-pulling and cutting weeds (Drummond et al. 2009). Such low pH soils are not conducive to the establishment of wildflowers and present a challenge to PR establishment that is unique to lowbush blueberry. Maine lowbush blueberry growers are interested in pollinator enhancement strategies, typically

Figure 1: Photo of wild blueberry field in Maine, USA. This Illustrates a wild blueberry field and the surrounding landscape typical of the region. The field is surrounded by a hard edge of forest and except for the crop itself, very few flowering plants are present in the landscape to support wild pollinators throughout the season

UnauthenticatedDownload Date | 11/16/18 3:45 AM

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry  395

consider pollinators in their management decisions, and 94% consider native bees as either an important, or a somewhat important component of their crop pollination (Hanes et al. 2013).

The objectives of our study were to measure the effects of two nurse crops (sheep fescue (Festuca ovina L.), common oats (Avena sativa L.)) and a mowing regime on floral density, aboveground plant biomass, and plant species diversity of PRs. We hypothesized that (1) nurse crop treatments would increase establishment success in terms of floral density, sown species richness, and sown plant biomass; and (2) that the mow treatment would increase sown species diversity. We implemented a randomized complete block design to test the success of organically approved methods to establish a mixture of annual and perennial wildflowers in the margins of a lowbush blueberry field in Maine. We consider the effectiveness of our pre-planting site preparation strategy pooled across all blocks and treatments. Finally, we consider the costs of PR establishment as part of a representative lowbush blueberry enterprise budget.

2 Materials and MethodsOur study took place at the University of Maine’s lowbush blueberry research facility, Blueberry Hill Farm in Jonesboro, Washington Co., ME, USA (Figure 2). Maine has a temperate climate, with an average temperature of 10.3 C (2005-2015 ten year average) and average rainfall of 106.7 cm (NOAA National Centers for Environmental Information, retrieved on June 27, 2018 from https://www.ncdc.noaa.gov/cag). The soil is Colton gravelly sandy loam with a starting pH of 4.7. Before use in this study, the plot was managed as a lowbush blueberry field, which relies upon sulfur treatments to lower the pH below 5.0 to reduce grass weed pressure (Yarborough 2009). We applied lime at a recommended rate of 7846 kg/ha to raise the pH to 6.0. Plots were briefly stale seed bedded for weed control by shallow tillage, once on 18 May and again on 31 May 2012. A list of possible wildflower species was compiled based upon the literature (Stubbs et al. 1992; Loose et al. 2005; Tuell et al. 2008; Mader et al. 2011). The list was passed through a series of filters to develop a final wildflower mix. Filters included lowbush blueberry grower feedback. For a list of sown species and seeding rates see Table 1.

Seeds were approved for Organic use by the Maine Organic Farmers and Gardener’s Association and were sown, unscarified, on 8 July 2012 in accordance with the methods of Mader et al. (2011), Blaauw and Isaacs (2014) and Ahern et al. (1992). Plots were irrigated to receive 2.5

cm of rain per week. If rainfall was 2.5 cm or more in a week, no additional irrigation was applied.

Each of four replicates (statistical blocks) contained five treatment plots (4x8 m; Figure 3). Blocks were buffered by 1.0 m strips. Plots within blocks were adjacent to one another. All treatments were carried out in full compliance with organic standards. Treatment one was an unsown rototilled control. Remaining treatments were seeded to a mix of wildflowers (Table 1). Additionally, treatment two was mowed monthly during the first growing season and again once annually in October; treatment three was sown with a low density of F. ovina; treatment four included both mowing and F. ovina; and treatment five was subjected to no additional manipulation. Each treatment plot was split into two subplots. One subplot in each plot was sown with A. sativa. We cut ‘mow’ treatments with a gasoline powered line-trimmer each fall, cutting vegetation to 15.0-20.0 cm.

In September 2013, we randomly selected 0.5 m2 from a 1.0 m quadrat in the center of each subplot to assess plant biomass. All selected above-ground plant material in each subplot was cut, bagged by species, and dried for two weeks in a drying room (40-45oC) at the University

Figure 2: Map of site location. Blueberry Hill Farm is the University of Maine’s lowbush blueberry research farm and the site of this study. Its location is indicated on the map with a black star

UnauthenticatedDownload Date | 11/16/18 3:45 AM

396   E. M. Venturini, et al.

of Maine Analytical Laboratory and Maine Soil Testing Service in Orono, ME. All plant material was identified, labeled as sown or unsown (i.e. weedy volunteers), and weighed. Opened, unopened, and past inflorescences (i.e. cluster of florets) or single flowers were identified to species and recorded. In September 2015, we used the second 0.5 m2 from the same 1.0 m quadrat in the center of each subplot and estimated plant species richness, flower density and biomass of flowers. Richness was measured in small plots within each treatment, and should be considered relative to other treatments, not as an absolute richness for the plot as a whole.

In 2013 and 2015, a nested Poisson regression analysis (A. sativa nested within establishment treatment) was used to test the establishment techniques with respect to sown plant species richness. In 2013 a nested split-plot ANOVA was used to assess establishment techniques on total sown plant inflorescence density and dry weight (wgt) and non-sown (weed) plant inflorescence density and dry wgt. This was performed in JMP version 12 (SAS Institute Inc., Cary, NC). In 2015, nested split-plot ANOVA was used to assess establishment treatments and oat over-seeding on non-sown plant inflorescence density and dry wgt. Poisson regressions for 2015 were used for the analysis of sown inflorescence densities due to the prevalence of plots (controls) with no sown plant species in the mix (0 counts in plots).

Economic analysis of PR establishment and the

consequences of planting longevity were assessed by standard economic measures (Kay 2011) based upon a PR enterprise budget developed by Hoshide and Venturini (Venturini et al. 2017a) but included additional establishment costs (unpublished data, Dr. Cathy Neal, University of New Hampshire). This annual PR enterprise budget specified both variable and fixed costs. Variable costs include wildflower seed, labor and fuel. Fixed costs such as equipment depreciation were also calculated. Wildflower seed, site preparation labor, and lime (variable costs) spent during the initial establishment year were annualized over the stand life of the PR.

Ethical approval: The conducted research is not related to either human or animal use.

3 Results

3.1 Wildflower Species Richness

In September 2013, the second year after plot establishment, we cut, dried, and weighed a total of 7.12 kg of plant material, counted 747 open inflorescences and a total of 10,928 open, unopened, and dehisced flowers. Throughout the summer and early fall an average of 527 inflorescences/m2 bloomed and 340 g of plant dry matter/m2 was produced. Establishment technique significantly

Table 1: List of sown species, habit, and seeding rates for wildflower PR treatment. Except A. sativa and F. ovina, values shown are supplied by Diane Wilson at Applewood Seed Company. Standard Errors for pure live seed values (no. live seeds per unit area) are not available from this source. In the third column, ‘A’ denotes an annual and ‘P’ denotes a perennial

Species Common Name Habit No. live seeds/m2 No. live seeds/ha

Agastache foeniculum Lavender Hyssop P 56.9 93,240

Coreopsis lanceolata Lance-Leaved Coreopsis P 62.3 101,981

Coreopsis tinctoria Plains Coreopsis A 102.8 168,349

Desmodium canadense Canada Tick Trefoil P 19.6 32,051

Echinacea purpurea Purple Coneflower P 50.5 82,718

Eupatorium perfoliatum Common Boneset P 38.0 62,160

Gaillardia pulchella Indian Blanket A 82.5 135,003

Helianthus annuus Sunflower A 29.6 48,562

Monarda fistulosa Bergamot P 43.2 70,707

Symphyotrichum novae-angliae New-England Aster P 48.7 76,642

Avena sativaa Common Oats A - 12.20 kgb

Festuca ovinac Sheep Fescue P - 11.20 kgb

Total Wildflower: 534.4 874,412aAvena sativa included only in split-plot treatmentbShown as kg/ha and not included in live seed totals. CFestuca ovina included only in treatments 3 and 4.

UnauthenticatedDownload Date | 11/16/18 3:45 AM

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry  397

influenced the species richness of sown species (c2(4) = 19.19,

P < 0.001). Mowing had the greatest effect on sown species richness (Figure 4). However, it was only significantly different from unsown control plots. All treatments except wildflower seed alone had a significantly greater richness of sown species than the control. Avena sativa did not influence sown species richness (c2

(5) = 5.29, P = 0.381). In 2015, there were no differences in sown wildflower species richness among the treatments (Figure 4, c2

(4) = 6.699, P = 0.153). Sowing A. sativa as a nurse crop also did not affect the species richness of sown wildflowers (c2

(1) = 0.279, P = 0.597) and there was no species richness interaction between main plot treatments and the oat nurse crop split-plot treatments (c2

(4) = 1.756, P = 0.781).

3.2 Treatment effects on inflorescences and biomass

Coreopsis lanceolata L., Coreopsis tinctoria Nutt., Monarda fistulosa L., and Symphyotrichum novae-angliae (L.) G.L. Nesom, were the four most common plants in 2013 plot samples. There were no significant treatment effects on C. lanceolata dry matter wgt, blooming inflorescences, or total inflorescences. Coreopsis tinctoria finished flowering by the time of our sample, explaining why there was no treatment or split-plot effects on the number of open flowers. However, both C. tinctoria dry matter wgt (F(4,12)

= 4.76, P = 0.016) and total (open, closed, and dehisced) flower heads (F(4,12) = 4.43, P = 0.020) were affected by treatments. Means comparison (Tukey HSD) revealed that when the wildflower mix was seeded alone, C. tinctoria produced s ignificantly more flowers (443.5 flowers/m2) and plant material (31.5 g/m2) than any other treatments except when seeded with F. ovina (94.8 flowers/m2 and 5.9 g/m2, respectively). Inflorescence density and dry wgts of M. fistulosa were also not affected by establishment technique treatment or oat seeding. However, when wildflowers were sown alone and not mowed, M. fistulosa produced more dry matter (LS mean = 70.91 g/m2, F(4,12) = 2.44, P = 0.104). Neither treatments nor split-plot (A. sativa) effects influenced S. novae-angliae at the a= 0.05 level, however the interaction term treatment x split-plot significantly influenced the number of open flowers produced at the a= 0.10 level (F(4,12) = 2.45, P = 0.091), meaning that we can be 90% certain that our findings are not due to chance. Although not highly significant, treatments involving F. ovina and mowing consistently resulted in greater numbers of open S. novae-angliae flowers than wildflower seed alone.

In 2015, in rank order, the most common plants by inflorescence density (dehisced, in bloom, and unopened) were Desmodium canadense (L.) DC, M. fistulosa, and C. lanceolata. Twenty-six percent of the total floral resource (inflorescences/m2) in all plots was comprised of non-sown weedy species, mostly Solidago spp. (goldenrod

Figure 3: Late summer photo of pollination reservoir in Maine. Photograph of Block 1 study plots taken August 30th, 2013

UnauthenticatedDownload Date | 11/16/18 3:45 AM

398   E. M. Venturini, et al.

species) and sown wildflower richness decreased almost 70% from 2013. Total wildflower inflorescence density was only marginally influenced by establishment techniques (F(4,12) = 2.324, P = 0.082), and unaffected by the sowing of A. sativa (F(1,15) = 0.052, P = 0.821). Evidence at the a = 0.10 suggests that flower density was lower in control plots compared to mowing (Treatment 2) and mowing + F. ovina (Treatment 4). This was also the case when dry wgt/m2 of inflorescences was analyzed (F(4,12) = 1.879, P = 0.143). Avena sativa had no effect on floral density (F(1,15) = 0.013, P = 0.908). The response variables, wildflower density and wildflower dry wgt, were not significantly influenced by treatments or A. sativa (P > 0.05). Avena sativa did not significantly reduce Solidago spp. inflorescence density or Solidago spp. inflorescence dry wgt (square root transformed, P > 0.05). There was a trend (P > 0.05) toward higher Solidago spp. inflorescence densities and inflorescence dry wgts in unsown control plots. The exotic weed, Euthamia graminifolia (L.) Nutt. (lance-leaved goldenrod) had significantly higher inflorescence wgt and inflorescence density in A. sativa plots (F(1,15) = 3.971, P = 0.056; F(1,15) = 4.013, P = 0.051). However, there was no effect of planting treatment on total weed inflorescence density or total weed inflorescence dry wgt. We found evidence that A. sativa increased both total weed inflorescence density and total inflorescence dry wgt (F(1,15) = 6.014, P = 0.021; F(1,15) = 6.252, P = 0.012; weed flower density and dry wgt, respectively). We pooled the two most common weeds in terms of inflorescence density and dry wgt (Euthamia and Solidago spp., both commonly called goldenrods), and the effects of A. sativa. Avena sativa nurse crops increased inflorescence dry wgt (square root transformed: F(1,15) = 8.643, P = 0.010) and density (square root transformed: F(1,15) = 7.282, P = 0.017) of Euthamia and Solidago spp. in 2015. Treatment had no effect on inflorescence density or dry weight of Euthamia and Solidago spp. (P > 0.05).

We investigated individual sown plant species responses to establishment techniques and found no significant effects on D. canadense (square root transformed, P > 0.05) or M. fistulosa flower density or dry wgt (square root transformed, P > 0.05). We did find significant effects of establishment techniques on C. lanceolata floral density (F(4,12) = 3.637, P = 0.017) and flower dry wgt (F(4,15) = 3.542, P = 0.019). The mowing + fescue treatment had significantly higher densities and dry wgt of C. lanceolata flowers than the other treatments (LS means Student’s T, t=2.052, P = 0.050), although dry wgt of C. lanceolata in the fescue + mow treatment was only significantly different than mowing alone, fescue alone,

and the control (Figure 5). The mowing treatment and the mowing plus A. sativa treatment were not significantly different from one another, but the mowing treatment was different from the other three treatments, while the mowing plus A. sativa was not significantly different.

Figure 4: The richness of sown wildflower species under establish-ment treatments, 2013 and 2015. Significant differences are denoted by letters. When letters are different, differences are significant

Figure 5: 2015 Coreopsis lanceolata flower density and dry weight per m2. Significant differences are denoted by letters. When letters are different, differences are significant.

UnauthenticatedDownload Date | 11/16/18 3:45 AM

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry  399

3.3 Cost Analysis

We present costs and time to pay back of PRs under two scenarios, 1) the baseline site preparation used to establish plantings across all treatment plots, and 2) the same baseline site preparation plus more intensive weed control efforts. Upfront costs of PR installation may be less when growers undertake a full-season of weed control prior to planting (Table 2). If existing plant material is not fully decomposed by repeated tillage prior to planting, the seed bed must be raked of debris prior to seeding. We calculated manual raking cost, which represents a significant labor cost, and is responsible for the overall lower total variable costs of the full-season weed control scenario. If growers undertake one season of weed control (5 additional tillage events during the growing season) prior to planting, and can increase PR stand life to 8 years, PR costs/year will decrease to $974/ha/year. Considering variable costs alone, without considering land value, taxes, and equipment depreciation, a PR with an eight-year stand life will cost $590/ha/year. If weed control strategies are not implemented, we may expect a lesser stand life of 4 years and a total variable cost of $1,205/ha/year.

Desmodium canadense had the highest inflorescence densities in 2015. It was also the most expensive to include ($255.66/ha) due to a relatively high seeding rate (1.01 kg/ha) combined with a high price ($189.60/kg). Eupatorium perfoliatum L., while more expensive per kilogram of seed

($864.21/kg), is seeded at a much lower rate (0.07 kg/ha) making it more affordable ($108.37/ha). Reducing the seeding rate of D. canadense by half would reduce upfront wildflower seed cost by 16.6% from $1151/ha to $967/ha.

4 DiscussionAlthough useful guidelines for establishing wildflowers do exist (Ahern et al. 1992; Ahern and Boughton 1994), to our knowledge, no published studies have explored the factors that influence their establishment success in organic agriculture. This study builds upon the work of Ahern et al. (1992), which tested establishment strategies for wildflower meadows in Massachusetts’ roadsides. Our site preparation strategy was based upon their findings that wildflower seeds establish best without fertilizers, and with a fine soil texture that optimizes seed to soil contact. Like Ahern et al. (1992) we created a fine seed bed through tillage. Where Ahern et al. (1992) used post-emergent herbicides to control weeds, we used organically-approved weed control methods, annual mowing and stale seed bedding during site preparation. Despite differences in weed management strategies, our findings support those of Ahern et al. (1992), that competition from weeds is one of the major obstacles to the establishment of long-lived, successful PRs.

Mowing above the height of the growing wildflowers in the first year of PR establishment is a recommended

Table 2: Pollination reservoir (PR) costs. Costs are presented with and without a full-season of weed control. The full season of weed control scenario is annualized over 4, 6, and 8 years. Depreciation assumes 1% of farm budget

Itemized PR Costs ------ Upfront Cost------ ---Annualized over PR stand life---

No prepa Prepb 4-yearb 6-yearb 8-yearb

Variable Costs ($/ha) Seed $1,151 $1,151 $288 $192 $144 Vermiculitec $754 $754 $189 $126 $94 Lime $1,668 $1,668 $417 $278 $209 Labor $1,206 $984 $249 $167 $126 Fuel & Oil $42 $62 $16 $11 $9 Maintenance & Upkeep $8 $8 $8 $8 $8Total Variable Costs $4,829 $4,627 $1,167 $782 $590Fixed Costs ($/ha) Depreciation $175 $175 $175 $175 $175 Land $124 $124 $124 $124 $124 Taxes & Insurance $85 $85 $85 $85 $85Total Fixed Costs $384 $384 $384 $384 $384Total Costs ($/ha) $5,213 $5,011 $1,551 $1,166 $974a No pre-planting weed control. Labor costs higher from raking plant residue out of seed bed prior to planting. Plant residues decompose without raking in the stale-seed bedding scenario. b Full-season of weed control prior to fall seeding using repeated tillage.c Used as a seed bulking agent. Cost would be greatly reduced if growers used sand instead.

UnauthenticatedDownload Date | 11/16/18 3:45 AM

400   E. M. Venturini, et al.

practice (Norcini and Aldrich 2004; Vaughan et al. 2015). We found no evidence to support the effectiveness of this practice for this wildflower mix in our system. However, studies in other systems found that repeated mowing in the first, and even second year after planting results in significantly greater sown wildflower diversity (Cauwer et al. 2005; Williams et al. 2007). The relatively few species in our wildflower mix, in addition to the presence of some annuals, likely influenced our findings. The slow growth of perennial wildflowers is much more likely to benefit from repeated mowing. Of the ten flowering plants in our wildflower mix, only seven were perennials, one of which (C. lanceolata) is considered a short-lived perennial.

Although mowing in the first year was not a significant factor affecting any of our response variables, C. lanceolata inflorescence density and dry weight were significantly higher and the number of blooming S. novae-angliae inflorescences were marginally higher when plots were mowed in their first year and planted to F. ovina. All plots in our study were also mowed once per year late in the fall. Studies in other systems find evidence that weekly mowing decreases forb overwinter seedling mortality by 26% (Williams et al. 2007). Mowing twice per year and removing cuttings, in contrast to leaving cuttings on the ground, slowed the takeover of sown wildflower plantings by weeds (Cauwer et al. 2005). We mowed only once per year and allowed cut materials to stay on the ground. We did not remove cuttings in this study which may have negated some of the positive effects of mowing.

Our findings suggest a practical longevity of a PR in lowbush blueberry to be four years under the low intensity management that we implemented. If more intensive site preparation strategies are employed to reduce weed competition, PR longevity can be extended (C. Neal, personal communication). Love et al. (2016) reported the successful restoration of a weedy site using a grass-first strategy. Through careful selection of grass (Poaceae) species used, sown grasses may be able to outcompete weedy undesirable grasses (Cox and Anderson, 2004) prior to sowing wildflowers. However, such a strategy may not result in the artificially high density of floral resources that growers expect from a PR. In warm climates, soil solarization, a passive solar process where soil is heated by laying sheets of plastic over the seedbed, can be effective and economical (Stapleton 2000). Although uncontrolled, a California study found that temperatures under solarization plastic were consistently 38 – 41°C from August to October and reported subsequent successful establishment of a wildflower meadow (Stapleton and Jett, 2006). Solarization is recommended in fact sheets and guides (Lee-Mäder et al. 2013), although very little

empirical research has reported its effectiveness in north temperate agroecosystems. In north temperate zones, soil solarization may not reliably increase soil temperatures high enough to kill weed seeds and reduce the weed seed bank. A single Maine study found that soil solarization decreased the number of emerging seedlings by 80% over non-solarized controls, although they did not measure the weed seedbank (Gallandt et al., In Press). Whether this practice will increase the longevity of a wildflower planting requires further research.

Pollination reservoirs can take 3-4 years before causing measurable increases in pollination services (Blaauw and Isaacs 2014; Venturini et al. 2017a). To be more economically viable, their effective life needs to be extended over what we documented here. As in other PR studies Blaauw and Isaacs 2014; Carvell et al. 2006; Venturini et al. 2017a) we broadcast seeded a wildflower mix. Extending PR stand life closer to a decade using organic management will require more intensive site preparation and maintenance (e.g., stale seed bedding, fall seeding, mowing in year following establishment) but also possibly annual hand-rogueing of invasive weeds (C. Neal, personal communication). This is based on an ongoing long-term PR establishment trial at UNH where the inability to control annual weeds during establishment can reduce PR stand life to as little as two years. Our economic analysis of establishment costs and expected stand life demonstrate that investing in a full season of weed control is relatively inexpensive. Without a full season of stale seed bedding, which decomposes plant debris, growers may need to rake the seed bed by hand prior to sowing seeds. Hand-raking represents a significant labor cost. On large scales at least, a full season of stale-seed bedding which avoids hand raking, is less expensive. However, we do not consider the additional year of opportunity cost required to prepare the site for an entire season. If a full season of stale seed bedding, or a combination of stale seed bedding and soil solarization, can extend the longevity of PRs, then an up-front investment in weed control is economically justified. Assuming an eight-year stand life under the full-season of weed control scenario, total fixed and variable costs are $974/ha/year. This means that PRs must produce over their life, on average, ten honey bee colony’s worth of wild bee pollination services. Previous work suggests that for each unit of PR, wild bees and pollination services can be increased on anywhere from 10 to 100 units of cropland (Blaauw and Isaacs 2014; Dicks et al. 2015; Pywell et al. 2015; Venturini et al. 2017b). If we conservatively assume a one to ten ratio of PR to crop, PRs must provide the wild bee equivalent of 1 honey bee hive per hectare of crop field to break even

UnauthenticatedDownload Date | 11/16/18 3:45 AM

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry  401

sativa was sown as a nurse crop, the inflorescences of Solidago and Euthamia spp. were more abundant than in subplots without A. sativa, suggesting that A. sativa may have increased weed densities. In our plots, weeds (mostly Solidago spp.) comprised one quarter of the inflorescences (dehisced, in bloom, and unopened) in the fourth year, and corresponded to a decrease in sown species diversity. However, Solidago spp. inflorescences are smaller than those of most of our sown species; a measure of aboveground plant biomass in 2015 may have shown Solidago spp. to be less dominant than they appear in our results.

We compared treatments and considered overall pollination reservoir success in terms of longevity, sown species diversity, floral density, aboveground plant biomass, and cost. The small size of the study and a high variance between statistical blocks limited our findings. However, our results suggest that A. sativa is not an effective nurse crop for wildflower establishment in Maine lowbush blueberry soils, and that more intensive weed management strategies that significantly decrease weed competition should be carried out prior to sowing seed to increase the economic viability of PRs in cropping systems. This study highlights the great need for economical organically approved weed control strategies that may increase PR establishment success rates in north temperate agroecosystems.

Disclosure of Conflicts of Interest: The authors declare no conflicts of interest.

Acknowledgements: We thank our team of student technicians, Audrey Maddox, Sarah Watts, and Margaret McCullough. Cathy Neal, Brett Blaauw, and Rufus Isaacs shared valuable expertise. Diane Wilson of Applewood Seed Company provided guidance and supplied wildflower seeds. Technicians Josh Stubbs, Chris McManus, and the Farm Manager, Jeff Brann, of the University of Maine’s Blueberry Hill Farm provided labor, equipment, and land. Alison Dibble and Lois Stack provided invaluable edits on previous versions of this manuscript. This project would not have been possible without the support of the University of Maine, School of Biology and Ecology. We thank the United States Department of Agriculture National Institute of Food and Agriculture – Specialty Crops Research Initiative Grant No. 2011-51181-30673. This is Maine Agricultural and Forestry Experiment Station publication number 3629.

with costs. Optimal ratios in practice will vary with local and landscape scale factors, crop pollinator dependency, and spatial arrangement of PRs (Sardiñas and Kremen 2015; Scheper et al. 2015). Also, our costs estimates are inclusive. Some growers may not consider their own labor, or fixed costs (e.g., equipment depreciation), when making wild bee habitat management decisions.

As in other PR studies (Carvell et al. 2006; Blaauw and Isaacs 2014) we broadcast seeded a wildflower mix. If wildflower species were instead sown in blocks containing fewer species or at lower seeding rates for more expensive species, the more competitive sown species may not have dominated the planting so quickly and PR seed costs could be reduced. Wildflower seed comprises a significant proportion of PR costs in Maine wild blueberry farms (Venturini et al. 2017a) so reducing such costs in the future could increase grower adoption. Although it flowered at very low levels until 2015, Desmodium canadense had a higher inflorescence density than any other species in that year, suggesting that we may have been able to seed this species at a lower rate. Considering its expense, decreasing D. canadense seeding rates by half would decrease total PR costs by 16.6%. To increase grower adoption of PRs for pollination security, farm consultants (e.g., USDA NRCS certified Technical Service Providers), should closely consider the cost of seed when assisting growers with PRs.

We suggest that future research test the longevity of wildflower species sown alone or in lower diversity blocks of species that complement each other in terms of competitiveness, time to maturity, and lifespan. Two new PRs have been seeded at our study site in the fall of 2016 to test this approach. Even with more intensive pre-planting weed control, PRs will still require eventual reseeding. The inclusion of hedgerows in growers’ pollinator management plans, have also shown effectiveness for increasing pollination services (Morandin and Kremen 2013; M’Gonigle et al. 2015), and may provide pollinators with refuge during periods of PR reestablishment.

Four years after the PR was sown, the establishment techniques were not strong determinants of sown inflorescence density or floral biomass, except for C. lanceolata that appeared to be positively influenced by both mowing and seeding sheep fescue, F. ovina. Monarda fistulosa was positively influenced by seeding oat (A. sativa). One of the major weeds in our study was goldenrod, comprised of Euthamia spp. and Solidago spp. Although they are heavily visited by pollinators and are an important bee forage in Maine from mid-summer to late fall (Stubbs et al. 1992), lowbush blueberry growers consider both genera to be weeds. Although we did not measure aboveground biomass in 2015, when A.

UnauthenticatedDownload Date | 11/16/18 3:45 AM

402   E. M. Venturini, et al.

ReferencesAhern J., Niedner C.A., Barker, A. Roadside wildflower meadows:

summary of benefits and guidelines to successful establishment and management. In Transportation research record, No. 1334, planning, administration, and environment. Maintenance of pavement, lane markings, and roadsides. Transportation Research Board, NRC. Washington, D.C., 1992, 46-53

Ahern J., Boughton J. Wildflower meadows as sustainable landscapes, In: R.H. Platt, R.A. Rowntree, and P.C. Muick (Eds.), The ecological city: preserving and restoring urban diversity (1994, University of Massachusetts Press), Amherst, MA, 1994, 172-187

Aldrich J.H., Factors and benefits in the establishment of modest sized wildflower plantings: A review. 2002. Native Plants J, 3, 1, 67-86

Bell D.J., Rowland L.J., Smagula J., Drummond F., Recent advances in the biology and genetics of lowbush blueberry. Technical Bulletin 203. Maine Agricultural and Forest Experiment Station, 2009, The University of Maine, Orono, ME, 1-28

Blaauw B.R., Isaacs R., Flower plantings increase wild bee abundance and the pollination services provided to a pollination-dependent crop., 2014, J. Appl. Ecol, 51, 4, 890–898

Bretzel F., Pessarossa B., Carrai C., Malorgio F., Wildflower plantings to reduce the management costs of urban gardens and roadsides, 2009, Acta Hort, 813, 33, 263-270

Carvell C.W., Meek R., Pywell R.F., Goulson D., Nowakowski M., Comparing the efficacy of agri-environment schemes to enhance bumble bee abundance and diversity on arable field margins, 2006, J. Appl. Ecol, 44, 1, 29-40

Cauwer, B.D., Reheul D., D’hooghe K., Nijs I., Milbau A., Evolution of the vegetation of mown field margins over their first 3 years, 2005, Agric. Eco. Environ, 109, 87-86

Cox R.D., Anderson V.J., Increasing native diversity of cheatgrass-dominated rangeland through assisted succession, 2004, Rang. Eco. Manag., 57, 2, 203-210

Dicks L.V., Baude M., Roberts S.P.M., Phillips J., Green M., Carvell C., How much flower-rich habitat is enough for wild pollinators? Answering a key policy question with incomplete knowledge, 2015, Ecol. Entomol, 40, S1, 22-35

Dicks L.V., Viana B., Bonmarco R., Brosi B., del Coro Arizmendi M., Cunningham S.A., et al., Ten policies for pollinators, 2016, Science, 354, 6315, 975-976

Drummond F., Smagula J., Annis S., and Yarborough D., Organic wild blueberry production, 2009, University of Maine Agricul. and Forestry Exper. Stn. Tech. Bull., 852, 1-43

Frances A.L., Reinhardt Adams C., Norcini J.G., Importance of seed and microsite limitation: native wildflower establishment in non-native pasture, 2010, Restor. Ecol., 18, 6, 944-953

Free J.B., Insect pollination of crops, Academic Press, New York, NY, 1970

Gallandt E.R., Birthisel S.K., Brown B.J., McCollough M., Pickoff M. Organic farming and sustainable weed control. In: Korres N.E. (ed.), Weed control: Sustainability, hazards, and risks in cropping systems worldwide, CRC Press. In press.

Hanes S.P., Collum K.K., Hoshide A.K., Asare E., Grower perceptions of native pollinators and pollination strategies in the lowbush blueberry industry, 2013, Renew. Agr. Food Syst., 30, 2, 124-131

Kay R.D. Farm management: Planning, control, and implementation, 7th ed., McGraw-Hill, Inc., New York, NY, 2011

Lee-Mäder E., Borders B., Minnerath A., Establishing pollinator meadows from seed, The Xerces Society for Invertebrate Conservation, Portland OR, 2013

Loose J.L., Drummond F.A., Stubbs C.S., Woods S.A. Conservation and management of native bees in cranberry, 2005, Technical Bulletin 191, The University of Maine Cooperative Extension, Orono, ME, 1-27

Love S.L., Hutchinson P.J.S., Price W.J., Managing weeds during wildflower meadow establishment in the arid Intermountain West: Efficacy of a grass-first strategy for sites with heavy annual weed pressure, 2016, NPJ, 17, 3, 216-229

Mader E., Shepherd M., Vaughn M., Black S.H., LeBuhn G., Attracting native pollinators: Protecting North America’s bees and butterflies. A Xerces Society Guide. Story Publishing, North Adams, MA, 2011.

Markwardt, D., Texas roadside wildflowers, 2004, NPJ, 6, 1, 69-71McCracken M.E., Woodcock B.A., Lobley M., Pywell R.F., Saratsi E.,

Swetnam R.D., et al., Social and ecological drivers of success in agri-environmental schemes: The roles of farmers and environmental context, 2015, J. Appl. Ecol, 52, 696-705

M’Gonigle L.K., Ponisio L.C., Cutler K., Kremen C., Habitat restoration promotes pollinator persistence and colonization in intensively managed agriculture, 2015, Ecol. Appl., 25, 6, 1557-1565

Morandin L.A., Kremen C., Hedgerow restoration promotes pollinator populations and exports native bees to adjacent fields, 2015, Ecol. Appl., 23, 4, 829-839

Norcini J.G., Aldrich J.H., Establishment of native wildflower plantings by seed, ENH 968, Environmental Horticulture Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. North Florida Research and Education Center, Quincy, FL, 2004

Osborne J.L., Williams I.H., Corbet S., Bees, pollination, and habitat change in the European community, 1991, Bee World, 72, 3, 99-116

Pywell R.F., Heard M.S., Woodcock B.A., Hinsley S., Ridding L., Nowakowski M., et al., Wildlife-friendly farming increases crop yield: evidence for ecological intensification, 2015, Proc. Biol. Sci., 282, 1816, 20151740

Quarles W., Native plants and integrated roadside vegetation management, 2003, IPM J. XXV, ¾, 1-9

SAS Institute, JMP user guide, release 7, SAS Institute, Cary, NC, 2009

Sardiñas H.S., Kremen C., Pollination services from field-scale agricultural diversification may be context-dependent, Agric. Ecosyt. Environ, 2015, 207, 17-25

Scheper J., Bonmarco R., Holzschuh A., Potts S.G., Riedinger V., Roberts S.P.M., et al., Local landscape-level floral resources explain effects of wildflower strips on wild bees across four European countries. 2015, J Appl. Ecol., 52, 5, 1165-1175

Smith S.N., Maine agriculture: a natural resource based industry constantly adapting to change. Maine State Planning Office, Augusta, ME, 2004

Stapleton J.J. Soil solarization in various agricultural production systems. 2000, Crop Prot., 19, 8, 837-841

UnauthenticatedDownload Date | 11/16/18 3:45 AM

Organic Establishment of Pollination Reservoirs in the Lowbush Blueberry  403

Stapleton J.J., Jett S., A large-scale demonstration of solar inactivation of invasive weed propagules for revegetation of California native wildflower communities, In: The proceedings of California Conference on Biological Control V, Riverside, CA, 2006

Strik B.C., Yarborough D., Blueberry production trends in North America, 1992 to 2003, and predictions for growth, 2005, HortTech., 15, 2, 391-398

Stubbs C.S., Jacobson H.A., Osgood E.A., Drummond F.A., Alternative forage plants for native (wild) bees associated with lowbush blueberry, Vaccinium spp., in Maine, Technical Bulletin 148, University of Maine Cooperative Extension, Orono, ME, 1992

Tuell J.K., Fiedler A.K., Landis D.A., Isaacs R., Visitation by wild and managed bees (Hymenoptera: Apoidea) to Eastern U.S. native plants for use in conservation programs, 2008, Environ. Entomol, 37, 3, 707-713

Vaughan M., Lee-Mäder E., Cruz J.K., Goldenetz-Dollar J., Borders B., Conservation cover (327) for pollinators: Florida installation guide. The United States Department of Agriculture, Natural Resources Conservation Service, and The Xerces Society for Invertebrate Conservation. Portland, OR, 2015

Venturini E.M., Drummond F.A., Hoshide A.K., Dibble A.C., Stack L.B., Pollination reservoirs in lowbush blueberry (Ericales: Ericaceae), 2017a, J Econ. Entomol., 110, 2, 333-346

Venturini E.M., Drummond F.A., Hoshide A.K., Dibble A.C., Stack L.B., Pollination reservoirs for wild bee habitat enhancement in cropping systems: A review, 2017b, Agroecol. Sust. Food. 41, 2, 101-142

Weaner L., Seeded meadow ten years after planting, 1996, Land. Des. 9, 1, 1-5

White House, The, National strategy to promote the health of honey bees and other pollinators. Pollinator Health Task Force. Washington, D.C., 2015

Williams D.W., Jackson L.L., Smith D.D., Effects of frequent mowing on survival and persistence of forbs seeded into a species-poor grassland, 2007, Restor. Ecol., 15, 1, 24-33

Wood, T.J., Holland J.M., Hughes W.O.H., Goulson D., Targeted agri-environmental schemes significantly improve the population size of common farmland bumblebee species, 2015, Mol. Ecol., 24, 1668-1680

Xerces, Bring back the pollinators, annual report. The Xerces Society for Invertebrate Conservation. Portland, OR, 2016

Yarborough D., Wild blueberry culture in Maine, University of Maine Cooperative Extension Fact Sheet no. 2088, University of Maine, Orono, ME, 2009

UnauthenticatedDownload Date | 11/16/18 3:45 AM

View publication statsView publication stats