The Impacts of Two Protein Supplements on Commercial Honey Bee

(Apis mellifera L.) Colonies

Marianne Lamontagne-Droleta, Olivier Samson-Roberta, Pierre

Giovenazzob and Valérie Fourniera*

aCentre de recherche et innovation en sciences végétales (CRIV) and Département de

Phytologie, Université Laval, Québec, Canada; bDépartement de Biologie, Université

Laval, Québec, Canada

Full correspondence details:

Marianne Lamontagne-Drolet: CRIV, Pavillon Envirotron, Laval University, 2480,

Hochelaga Boulevard, Québec, Qc, Canada, G1V 0A6, marianneld@hotmail.com.

Olivier Samson-Robert: CRIV, Pavillon Envirotron, Laval University, 2480, Hochelaga

Boulevard, Québec, Qc, Canada, G1V 0A6, olivier.samson-robert.1@ulaval.ca

Pierre Giovenazzo: Department of biology, Pavillon Vachon, Laval University, Québec,

Qc, Canada, G0A 1S6, pierre.giovenazzo@bio.ulaval.ca

*Valerie Fournier : CRIV, Pavillon Envirotron, Laval University, 2480, Hochelaga

Boulevard, Québec, Qc, Canada, G1V 0A6, valerie.fournier@fsaa.ulaval.ca

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

The Impacts of Two Protein Supplements on Commercial Honey Bee

(Apis mellifera L.) Colonies

Honey bees (Apis mellifera L.) are pollinators of major importance for crop

production. In recent years, colony management has become more difficult due to

multiple problems such as pesticide exposure, exotic parasites, pathogens and

nutritional deficiencies. The latter has incited beekeepers to provide protein

supplements to their colonies to make up for the lack of pollen resources in the

environment. However, their efficiency varies depending on their composition

and the surrounding landscape. In this field study, we provided two different

protein supplements (Global Patties and Ultra Bee) to colonies with either

limited or unlimited access to natural pollen to assess their impacts on various

colony and individual bee parameters. We used 50 colonies distributed among

three sites in the Montérégie area, in Quebec, Canada. We found that

supplemented colonies limited in pollen collection were able to raise the same

amount of brood than control colonies. Nurse bees in supplemented colonies also

had a higher protein content compared to control bees. However, bees from

supplemented colonies displayed shorter lifespan, which casts a doubt on the

suitability of these products for honey bee nutrition. The supplement containing

natural pollen, Global Patties, was the most consumed and the most beneficial

of the two for the colonies. Finally, colonies from the apiary surrounded by the

highest proportion of cultivated land in a 5-km radius performed better toward

the end of the season, which could be due to the presence of nutritionally

interesting plants specific to the agricultural landscape at that time of the year.

Keywords: Apis mellifera; nutritional stress; pollen substitute; brood; pollen

consumption; protein content

Introduction

Over the last decade, commercial honey bee (Apis mellifera L.) colonies have suffered

high mortality rates in North America and Europe (Intergovernmental Science-Policy

Platform on Biodiversity and Ecosystem Services, 2016; Kulhanek et al., 2017;

Oldroyd, 2007), ranging from 29 to 35% from 2007 to 2009, and from 16 to 29% from

2010 to 2015 in Canada (Ministère de l'Agriculture des Pêcheries et de l'Alimentation

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

du Québec, 2016). Multiple factors appear to act synergistically to cause this problem,

including pesticide exposure, exotic parasites and pathogens, management practices,

climate changes and lack of floral resources (Klein, Cabirol, Devaud, Barron, &

Lihoreau, 2017).

Abundant and diversified floral resources are crucial for honey bees to acquire

their essential nutrients through consumption of pollen and nectar (Wright, Nicolson, &

Shafir, 2018). In the current beekeeping industry, pollen nutrition is of special concern

for multiple reasons. Firstly, pollen is the only source of proteins, lipids, vitamins and

minerals for the honey bee; nectar being mainly a source of carbohydrates. Secondly,

pollen storage within the hive in the form of bee bread is limited and can deplete rapidly

during periods of weather unfavorable to foraging (Schmickl and Crailsheim, 2002).

Pollen must therefore be available throughout the brood producing season, in adequate

quantity. Finally, as different flowers produce pollen of distinctive nutritional content,

flower diversity is fundamental to insure an appropriate and complete bee diet (Di

Pasquale et al., 2013; Roulston and Cane, 2000; Roulston, Cane, & Buchmann, 2000).

However, in intensive agricultural environments or when colonies are used for

commercial pollination, these requirements are not always met (Danner, Keller, Hartel,

& Steffan-Dewenter, 2017; Di Pasquale et al., 2016; Donkersley, Rhodes, Pickup,

Jones, & Wilson, 2014; Girard, Chagnon, & Fournier, 2012; M. D. Smart, Pettis, Euliss,

& Spivak, 2016). Poor pollen nutrition negatively affects individual honey bee health as

well as colony performances (Brodschneider and Crailsheim, 2010; Di Pasquale, et al.,

2016; Di Pasquale, et al., 2013; Scofield and Mattila, 2015), rendering bees less

resistant to other stressors such as pesticides (Schmehl, Teal, Frazier, & Grozinger,

2014; Tosi, Nieh, Sgolastra, Cabbri, & Medrzycki, 2017) and pathogens (Alaux,

Ducloz, Crauser, & Le Conte, 2010; Di Pasquale, et al., 2013).

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

To counter this problem, beekeepers provide protein supplements to their

colonies during periods of poor floral abundance or diversity (Brodschneider and

Crailsheim, 2010) (see Table 1). These supplements are made from ingredients rich in

proteins, such as soy or yeast, and may also contain natural pollen (Eccles, Kempers,

Gonzalez, Thurston, & Borges, 2016). It is generally recognised that feeding such

artificial diets is beneficial to colonies facing poor foraging conditions or environments

(Brodschneider and Crailsheim, 2010). However, the benefit of providing these

products is not clear and vary depending on their composition, especially when

compared to pollen (Alqarni, 2006; De Jong, Da Silva, Kevan, & Atkinson, 2009;

DeGrandi-Hoffman et al., 2016; Mortensen et al., 2018; Peng, D'Antuono, & Manning,

2012). Among other things, a minimal protein content, usually of 20-30%, is required

for supplements to be functional (Li, Xu, Wang, Feng, & Yang, 2012; Li, Xu, Wang,

Yang, & Yang, 2014; Morais et al., 2013). According to this criterion, frequently cited

commercial products such as Bee-Pro, MegaBee and Feed-Bee contain enough

protein (De Jong, et al., 2009; DeGrandi-Hoffman, Chen, Huang, & Huang, 2010).

However, Global Patties, Ultra Bee and other unnamed products used in a study by

DeGrandi-Hoffman et al. (2008) do not reach 20% of protein. The choice of ingredients

can also affect the performance of the product as well. Adding pollen to the diet, for

example, can stimulate consumption of the supplement as well as its positive impacts on

honey bee health (Alqarni, 2006; Manning, Rutkay, Eaton, & Dell, 2007). On the

contrary, soy-based diets tend to be less palatable and overall less beneficial to the

honey bees, bees consuming it display either lower brood production, lower honey

harvested, lower body protein content, lower adult populations and higher Nosema

infestation levels (De Jong, et al., 2009; DeGrandi-Hoffman, et al., 2016; Fleming,

Schmehl, & Ellis, 2015; A. Saffari, P. G. Kevan, & J. Atkinson, 2010a; A. Saffari, P. G.

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

Kevan, & J. L. Atkinson, 2010b). Even the surface area of the applied product can be an

important factor to consider, as a larger surface reportedly increases the quantity of

supplement consumed and its positive impacts on the colony (Avni, Dag, & Shafir,

2009). However, despite the surprisingly wide range of homemade and commercial

supplements available, basic knowledge as to the essential nutritional requirements of

the honey bees are still lacking, especially for lipids, vitamins and minerals (Wright, et

al., 2018). Thus, an ideal protein supplement formulation, which could effectively

replace pollen, has yet to be found (Wright, et al., 2018).

[Table 1 near here]

Efficacy of different protein supplements has been the subject of relatively few

scientific studies, especially under field conditions. Also, comparison between studies is

difficult considering the variability of supplements, control treatments, feeding periods,

and pollen traps used to simulate dearth. Furthermore, the landscape has a significant

impact on pollen and nectar sources having a direct effect on the nutritional status of the

colony (Donkersley, et al., 2014). Moreover, to our knowledge, the impact of

supplemental feeding on commercial colonies has not been investigated yet. As stated

by Mattila and Otis (2006), commercially managed colonies accumulate stress during

the season, and the impact of supplementation would most probably be different on

these colonies than on low-density and relatively undisturbed research colonies.

Therefore, the objectives of this study are to 1) compare the health and strength (sealed

brood surface, collected pollen weight, foraging effort, protein content of adult honey

bees, lifespan and Nosema and Varroa infestation) of commercial honey bee colonies

supplemented or not with a protein supplement, 2) compare the consumption and impact

on honey bee health of two commercial protein supplements and 3) evaluate the impact

of surrounding landscape on the nutritional status of colonies. Our hypotheses were that

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

1) supplemented colonies would be stronger and healthier, 2) pollen-enriched protein

supplement would be more consumed and have a higher positive impact on colony

health and 3) colonies based in a more suitable environment (notably less cultivated

area) would perform better regardless of the treatment.

Materials and Methods

Sites and Colony Management

The study was conducted from May to September 2016 in Montérégie, known as the

most intensive agricultural region in the province of Québec, Canada, located south-east

of the island of Montreal. Fifty colonies of similar strength (around 5 frames of bees

and 4 frames of brood for each colony) and genetic origin were provided by the

commercial beekeeper Les Ruchers Gauvin Inc. (Saint-Hyacinthe, QC, Canada). Each

colony consisted of a single brood chamber Langstroth hive. The colonies were

randomly placed on three sites in the vicinity of the towns of Saint-Hyacinthe and La

Présentation (Site 1 – 16 colonies: 45°38'43.0"N, 72°55'26.3"W, Site 2 – 16 colonies:

45°42'20.7"N, 73°05'30.3"W, Site 3 – 18 colonies: 45°39'38.8"N, 72°51'42.9"W).

Distances between the three sites were 14.5 km (Site 1 – Site 2), 5.1 km (Site 1 – Site

3), and 18.5 km (Site 2 – Site 3), respecting the minimal distance of 4 km between

apiaries recommended in Quebec (Centre de référence en agriculture et agroalimentaire

du Québec, 2011). The hives were managed in accordance with local professional

beekeeping practices. Varroa mite treatments were applied early August and early

September, according to the treatment calendar recommended by the Québec’s Ministry

of Agriculture, Fisheries, and Food (Ministère de l'Agriculture des Pêcheries et de

l'Alimentation du Québec, 2014). Nosema spp. was also monitored during the

experiment.

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

In 2017, the study was repeated and modified to measure the lifespan of honey

bees depending on their experimental group (see below for details). Fifteen colonies of

similar strength were used, all placed in the same apiary in St-Hyacinthe (45°38'43.0"N,

72°55'26.3"W). Their management was the same as in 2016.

Experimental Design and Treatments

In 2016, protein supplementation and access to pollen were manipulated in the 50

experimental colonies. Colonies were either not supplemented in protein

(“unsupplemented”), supplemented with a commercial soy-based protein patty

containing 15% pollen (distributed by Global Patties, Airdrie, AB, Canada) or

supplemented with a commercial plant-based (containing soy but not as a main source

of proteins) protein patty containing no pollen (Ultra Bee, distributed by Mann Lake

Limited, Hackensack, MN, USA). Global Patties are made of sugar, soy, yeast, a mix

of fats, vitamins and minerals provided by Latshaw Apiaries

(http://www.latshawapiaries.com/), and sterilized (electron-beam processed) wildflower

pollen from China. Ultra Bee patties recipe is a trade secret, but ingredients include

plant protein, high fructose corn syrup, sugar, canola oil, soybean oil, palm oil,

lemongrass oil, spearmint extract, probiotics, wheat flour and various vitamins and

minerals additives. The protein content of the tested patties are found in Table 1. The

products tested are commercially available for beekeepers in Québec. The patties were

offered ad libitum to the colonies, from May 3rd to August 3rd. Pollen resources are

usually limited during the spring in Quebec, which unfortunately coincides with a high

demand of strong colonies by the industry for pollination services and colony

multiplication. Thus, beekeepers will feed a protein supplement to their colonies to

increase brood production. Pollen availability usually increases in the summer, but it

can still be limited in intensive agricultural environments. Access to pollen was either

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

unrestricted or restricted by the addition of pollen traps under the hives. This was done

to assess the pollen harvest of colonies and mimic pollen restrictive conditions. Pollen

traps were not installed on unsupplemented colonies to avoid malnutrition and death,

which was not acceptable for the owners of the beekeeping operation. Pollen traps were

installed from May 3rd-5th to August 3rd, at which point they were activated every other

week and then completely removed by mid-September. All the possible combinations of

supplementation and access to pollen resulted in five treatments: 1) Control (no

supplement, no pollen traps); 2) pollen-enriched supplement (PES); 3) pollen-enriched

supplement + pollen traps (PES + traps); 4) pollen-free supplement (PFS); 5) pollen-

free supplement + pollen traps (PFS + traps). Ten colonies were attributed to each

treatment, for a total of 50 colonies. As stated previously, colonies were placed on three

sites and each site included at least two repetitions of each treatment. Treatments within

sites were randomly assigned to the hives.

In 2017, only protein supplementation (no supplement, pollen-enriched

supplement or pollen-free supplement) was manipulated in the experimental colonies.

No pollen traps were installed. Five colonies were randomly assigned to each treatment,

hence 15 colonies in total. As before, supplements were given ad libitum, but only for

four weeks starting on the 27th of April. After four weeks, brood frames were taken out

of the colonies to perform a longevity test in laboratory.

Supplement Consumption

In 2016 and 2017, protein supplement consumption was measured every week of the

feeding period during the experiment. In 2016, to compare the consumption of the

supplements, two patties of about 500 g each (weighed to measure their exact mass)

were placed on the top of the brood chamber frames of treated colony. Every week, new

patties were provided, and leftovers were weighed again.

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

Pollen Weight

The amount of pollen collected by the colonies was measured by weighing the content

of the pollen traps every week (or every other week in August and September).

Sealed Brood Surface

The sealed brood surface was evaluated every three weeks from May to September

2016. Each frame of the brood chamber containing sealed brood was photographed on

both sides without bees. Then, as described by Delaplane, van der Steen, & Guzman-

Novoa (2013), a grid was applied on the photographs using the software ImageJ. Each

square division of the grid was estimated as 0, 25, 50, 75 or 100% filled with sealed

brood. The total per frame was converted in square centimeters and added up for the

hive.

Foraging Effort

Foraging effort was also evaluated every three weeks from May to September 2016. The

entrance of each hive was filmed for 1 minute to count the number of honey bees

returning to the hive with and without corbicular pollen loads. Longer observation

periods would have added activity variability due to time of day. Ratio of bees bringing

back pollen to the colony was then calculated. The recordings were always made before

opening the hives and between 10 AM and 14 PM to maximize activity and minimize

differences due to the time of the day (Delaplane, et al., 2013).

Protein Content of Adult Honey Bees

Measuring the protein content of honey bees is a well-known method to compare the

quality of different protein diets (De Jong, et al., 2009; Li, et al., 2012). Five nurse bees

per hive were collected every four weeks of the experiment to measure total protein

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

content. Only nurse bees are used to minimize potential protein differences due to age.

Nurse bees were identified by observing a brood frame and picking up bees that tended

to the larvae. If no larva were found in the hive, no bees were picked up and a missing

value was recorded. Collected bees were placed in a tube and immediately chilled on

ice. Within 24 hours, they were stored at -80°C until further analysis. Upon analysis,

five frozen workers were crushed to fine powder using liquid nitrogen, and 200 mg of

bee powder was mixed with a PBS buffer solution containing 1% (v/v) PMSF protease

inhibitor at the ratio of 1:3. The mixture was then centrifuged at 4°C, 20 000 xg for 10

minutes and the resulting supernatant was used for the Bradford protein assay

(Bradford, 1976). Samples were done in triplicate. Standard curves were prepared using

bovine serum albumin (BSA) and protein absorbance was measured at 595 nm.

Bee Lifespan

Lifespan is also a commonly evaluated parameter to test the suitability of different diets

in honey bees (Alqarni, 2006; Hocherl, Siede, Illies, Gatschenberger, & Tautz, 2012;

Ihle, Baker, & Amdam, 2014; Li, et al., 2014; Manning, et al., 2007; van der Steen,

2007). Longevity tests were carried in 2016 and 2017. In 2016, 1 colony per treatment

(5 treatments) were used for the test and 90 newly emerged bees per hive were

randomly picked and placed in cages (30 bees per cages), for a total of 3 replicates and

90 bees per treatment. In 2017, 15 colonies were used (5 colonies/treatment, 3

treatments) for the test and 60 newly emerged bees per hive were randomly picked and

placed in cages (30 bees per cages), for a total of 10 replicates and 300 bees per

treatment. In 2016 the bees were picked in late July, and in 2017 the test was carried in

late May after four weeks of colony supplementation, ensuring that newly emerged bees

were fed supplemented (or unsupplemented) diet as larvae. To obtain newly emerged

bees, we first removed adult bees from a brood frame, and then put the brood frame in a

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

mesh net. The next day, bees present in the net (newly emerged bees) were picked for

the test. In both years, the plexiglass cages (13.5 x 12.5 x 17.5 cm) containing the bees

were kept in an incubator at 30°C, 75% humidity, in total darkness (Williams et al.,

2013). Bees were fed with 50% (weight/volume) sucrose solution through two 1.5 ml

Eppendorf tubes per cage. Feeding tubes were changed daily. Dead bees were counted

and removed approximately every 24 hours, until no more bees remained. The trial

lasted for 50 days in 2016 and 45 days in 2017.

Nosema and Varroa Mite Infestation

Nosema spp. and Varroa destructor were both monitored during the season. Nosema

infestation levels were calculated at the end of July 2016 using the spore count method

described in Pernal and Clay (2015). Varroa mite levels were monitored on May 3rd and

July 27th 2016 using the 75% ethanol wash method (Dietemann et al., 2013).

Landscape Analyses

To assess the impact of environment on the colonies, landscape was characterized

within a radius of 5 km around the apiaries. QGIS software (2.18.14) was used to

calculate the ratios of cultivated land and specific crop (La Financière agricole du

Québec, 2016), wooden area (Ministère des Ressources naturelles et de la Faune du

Québec, 2000) and urban landscape (DMTI Spatial, 2012).

Statistical Analyses

Since measurement were taken across time on the same hives, the variables supplement

consumption, pollen weight, sealed brood surface, foraging effort, protein content of the

honey bees and varroa mite infestation were analysed using repeated measures ANOVA

models. However, for Nosema spp. infestation, we used a two-way ANOVA model

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

since measurements were taken at one specific time point. In the repeated measures

ANOVA models, the factors sites and treatments are between-hives sources of

variation, while time is a within-hive source of variation. The best correlation structure

between observations taken on the same hive through time was selected based on the

Akaike Information Criterion (AIC). Response variables were transformed when

necessary to meet the assumption of normality, but the least square means and their

corresponding confidence intervals were back-transformed at their original scale to

make comparison with other studies easier. Following a significant effect in any

ANOVA table, post-hoc multiple comparisons were performed using the Tukey-Kramer

method in order to control the type I error rate. Lifespan was computed using Kaplan-

Meier curves of honey bee survival, and a log-rank test was performed to assess for

significant differences between curves. Data was right censored if a honey bee died of

non-natural cause (crushed or drowned), indicating it could have lived longer than the

value entered. All statistical analyses were performed using the R software (R Core

Team, 2016) at the significance level of α=5 %.

Results

Supplement Consumption

Daily supplement consumption ranged from 0.51 to 270 g per colony. ANOVA

revealed significant site x time (F22, 363 = 7.5714, p < 0.0001) and site x treatment (F6, 22 =

4.4267, p = 0.0044) interactions. For the site x time interaction, multiple comparisons

indicated that nine weeks out of 12 showed a significant difference between sites (see

figure 1a). Up until the end of June, daily consumption was generally highest at Site 3,

lowest at Site 1 and Site 2 displayed an intermediate level of consumption. However,

after that, consumption at Site 2 decreased, while at Site 1 and especially Site 3

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

consumption kept rising. During the last week of supplementation, Site 3 colonies

consumed, on average, 158 g [95% CI 126.8–196.1 g] of supplement daily, which was

57 g more than colonies at Site 1, and almost 100 g more than colonies at Site 2. As for

the site x treatment interaction, as shown in figure 2, the pollen-enriched supplements

were generally more consumed than the pollen-free supplements across sites.

Pollen Weight

Daily collection of pollen by the traps ranged from 0 to 180 g per colony. There was no

difference between treatments. A significant site x time interaction (F30, 223 = 2.4892, p =

0.0001) was detected. However, following multiple comparisons, we found no

significant difference between sites except for the last two weeks of the experiment, late

August and early September (see figure 1b). At that time, the pollen traps at Site 3

weighed more than those at Site 1 and 2, which contained less than 10 g of pollen

pellets on average.

Sealed Brood Surface

Model revealed a significant site x time interaction (F10, 210 = 3.9585, p = 0.0001) for

sealed brood area. There was no difference between treatments. As shown in figure 1c,

multiple comparisons indicated differences between sites on July 6th, where Site 2

displayed on average less brood (1961 cm2, 95% CI 1359–2564 cm2 per colony) than

the other sites (3628 cm2, 95% CI 3028–4228 cm2 per colony for Site 1 and 3544 cm2,

95% CI 2973–4114 cm2 per colony for Site 3), and on September 6th, where Site 3

displayed on average more brood (1962 cm2, 95% CI 1392–2533 cm2 per colony) than

the other sites (1069 cm2, 95% CI 469–1669 cm2 per colony for Site 1 and 970 cm2,

95% CI 368–1573 cm2 per colony for Site 2).

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

Foraging Effort

Foraging effort, evaluated by the proportion of foragers returning to the hive with

corbicular pollen, ranged from 0 to 72%. A significant effect of treatment (F4, 28 =

7.5641, p = 0.0003) and a site x time interaction (F10, 208 = 4.3091, p < 0.0001) were

found. As shown in figure 3, multiple comparisons confirmed that foraging effort was

significantly higher for the PFS + traps treatment, which was around the double the

value observed for the control and “without traps” treatments (PFS and PES). As for the

site x time interaction, significant differences between sites were found at 3 sampling

dates out of 6, but no clear tendency could be observed.

Protein Content of Adult Honey Bees

Total protein content of the nurse bees ranged from 26.2 to 47.8 mg of protein per gram

of honey bee. There was a significant effect of treatment (F4, 28 = 4.490, p = 0.0063) and

a site x time interaction (F6, 126 = 3.108, p = 0.0071). As shown in figure 4, the colonies

receiving no supplement (without traps) produced bees with a significantly lower

content in protein compared to other treatments, except PFS (without trap) which

displayed an intermediate value. As for the site x time interaction, significant

differences between sites were found for 2 sampling dates out of 4. On May 31th, bees

from Site 1 contained significantly more protein on than bees from Site 3, and on July

27th, significantly more than bees on both Site 2 and 3.

Bee Lifespan

As shown in figure 5, in both 2016 and 2017, the log-rank test showed that honey bees

from colonies fed no supplement lived longer than honey bees from colonies fed pollen-

enriched supplements (2016: p = 5.2e-12, 2017: p = 0.0091), which lived longer than

honey bees from colonies fed pollen-free supplement patties (2016: p = 0.0049, 2017: p

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

= 0.0018).

Nosema and Varroa Mite Infestation

Nosema infestation levels were significantly different between sites (F2, 28 = 5.236,

p = 0.0117), colonies at Site 3 being less infected (658 405 spores/bee, 95% CI

368 658–1 175 880 spores/bee) than colonies of other sites (Site 1: 1 592 995

spores/bee, 95% CI 851 460–2 980 332 spores/bee; Site 2: 2 463 544 spores/bee, 95%

CI 1 316 771–4 609 041 spores/bee). There was no difference between treatments. For

varroa mite infestation there was a significant difference between treatments (F4, 28 =

5.3827, p = 0.0024) and significant time effect (F1, 39 = 171.3488, p < 0.0001), but no

interaction between these factors. As expected, infestation was higher at the end of July

than in early May, rates ranging from 0 to 1.5% in May compared to 0.3 to 10.6% in

July. As for treatments, as shown in figure 6, the PES + traps treatment had significantly

higher infestation rates compared to other treatments (p = 0.05), except for PES

(without traps) which displayed an intermediate value.

Landscape Surrounding the Apiaries

In 2016, in a 5 km radius around the sites, cultivated land occupied 58%, 61% and 70%

of Site 1, 2 and 3 respectively (see figure 7 for details). Corn and soy were the main

crops in all three sites. Wooded area occupied 4%, 31% and 13% of Site 1, 2 and 3

respectively. Urban area occupied 29% of Site 1, but only 2% of Site 3 and there was no

urban area in Site 2.

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

Discussion

Supplemented vs. Unsupplemented Colonies

The first purpose of this study was to compare the strength and health of commercial

honey bee colonies provided with a protein supplement or not. Our hypothesis, that

supplemented colonies would be stronger and healthier compared to unsupplemented

colonies, was partially confirmed. Indeed, our results indicate regarding brood surface

and protein content hint that supplements may be beneficial to honey bees, but longevity

is impaired when feeding them, which indicates that the underlying mechanics is more

complex than we think. Supplements may thus represent a temporary solution to avoid

dwindling of the colonies in unfavorable foraging conditions but cannot possibly be

envisioned as a long-term solution to the lack of pollen abundance and diversity.

Brood surface did not significantly differ between any of the treatments. Supplemented

colonies that were limited in pollen (with pollen traps) reared similar amounts of brood

compared to unrestricted colonies. This suggest that during periods of floral resources

scarcity, protein supplements provide sufficient nutrients to the colonies to maintain

brood rearing. Other studies placing colonies in actual pollen shortage conditions

showed that supplemented colonies reared more brood than control colonies (although

this was not true for all the supplements tested in these studies) (Avni, et al., 2009;

DeGrandi-Hoffman, et al., 2008; Morais, et al., 2013; Saffari, et al., 2010b).

Supplemented colonies not restricted in pollen also reared similar amounts of

brood compared to control colonies. It could be because bees preferred to feed on the

collected pollen and ignored the protein patties, a phenomenon observed for some

supplements (Keller, Fluri, & Imdorf, 2005; Mattila and Otis, 2006). However, we

observed the bees actively feeding on the patties and found no sign that they were not

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

consumed (patty fragments at the entrance of the hive for example), which discards this

possibility. Therefore, floral resources available during the experiment were sufficient

in quality and quantity for optimal brood rearing, which we did not expect. This is

supported by others, who reported no benefit in terms of brood rearing when feeding

supplements to colonies that were in good foraging conditions (Avni, et al., 2009;

DeGrandi-Hoffman, et al., 2008; Mattila and Otis, 2006). In our case, despite being in

an intensive agricultural region, foragers may have found satisfying pollen resources in

trees and weeds (although we did not perform palynological analyses to confirm this),

which can constitute a surprisingly high proportion of the honey bee diet at critical

periods in the season (Girard, et al., 2012; Requier et al., 2015). In addition, during

spring and summer of 2016 the weather was generally warm and there was little rainfall

(good foraging conditions), allowing a lot of time for foraging activities and possibly

preventing stored pollen shortage.

Supplementation generally led to a higher protein content in honey bees

compared to control bees, regardless of the presence of pollen traps. While the

difference may seem small, we believe it is still biologically relevant. In the very

challenging context of modern apiculture, everything we can gain from better

apicultural practices can help. Some factors may not seem to impact the colonies

notably, but when combined together or in interaction with others can make a great

difference in colony performances. Expectedly, colonies without pollen traps contained

more protein since they were able to consume both real pollen and protein supplements.

However, supplemented bees that were restricted in pollen also had a higher protein

content compared to control. This was not expected, since both supplements had a

protein content just below 20%, which is reportedly not optimal (Herbert, Shimanuki, &

Caron, 1977). De Jong, et al. (2009) also found that young bees fed exclusively with

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

Feed-Bee showed higher haemolymph protein contents than bees fed exclusively with

pollen. However, the authors suspected it was because the bees consumed freshly

collected pollen and not actual bee bread, the latter reportedly leading to higher protein

contents in the haemolymph of bees consuming it (Cremonez, De Jong, & Bitondi,

1998). Other studies reported an equivalent effect of pollen and protein supplements on

the protein content of honey bees (DeGrandi-Hoffman, et al., 2010; DeGrandi-Hoffman,

et al., 2016; Li, et al., 2012; Morais, et al., 2013; van der Steen, 2007), while some

noted that pollen was better than supplements in that regard (Amro, Omar, & Al-

Ghamdi, 2016; van der Steen, 2007). In the end, the results probably depend more on

the nutrients present in these food sources than on their origin. For the honey bee,

higher protein content is considered a sign of good health, since proteins are needed to

synthetize the various tissues during growth (Haydak, 1970), as well as for immunity

purposes (Alaux, et al., 2010) and the workings of antioxidant enzymes which are

believed to slow down aging (Li, et al., 2014).

Surprisingly, in both 2016 and 2017, bees reared in a supplemented colony had

significantly shorter lifespan than control bees. This was unexpected because only

colonies that had no pollen traps were used for this experiment. Supplemented colonies

had access to pollen as well as supplements and likely consumed more nutrients (our

previous results confirmed their higher protein content). This contradicts the findings of

Li, et al. (2014) who observed that bees fed with diets richer in protein had increased

body protein and longevity. As previously stated, they found protein consumption was

implicated in preventing oxidative stress, resulting in enhanced longevity. However,

other factors are likely to influence longevity as well, since van der Steen (2007)

observed that honey bees with the highest protein content did not live the longest. In our

case, it is possible that nutrients other than proteins, such as lipids and vitamins, present

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

in pollen but absent or rare in the supplements tested promoted longevity. Indeed,

although the foraging effort did not differ between the colonies used for this test,

supplemented colonies may have preferentially stored the pollen and consumed the

supplements first. A reduction of pollen consumption as low as 5-10% can negatively

affect vitellogenin expression and longevity (Di Pasquale, et al., 2016). Alternatively,

the lower digestibility of the protein in the patties may have been the cause of the

reduced longevity observed in our study. Researcher Ihle, et al. (2014) noted that bees

weighed less and lived longer when fed a diet high in carbohydrates and low in protein.

They stated that a high protein ingestion may cause a deleterious effect on the bees.

Indeed, proteins are demanding to digest and metabolize, which may explain reduced

longevity. Also, it is possible that the supplements contained ingredients harmful to bee

longevity. Finally, another possibility is that the results observed in our laboratory

longevity tests may not reflect what really happens in the hive. As observed by

Schmehl, et al. (2014), when not exposed to pesticides, bees fed with protein had

reduced lifespan compared to bees fed with sucrose only. However, when exposed to

the pesticide chlorpyrifos, the bees fed protein lived longer. This may be part of the

explanation of our results. This also explains why, despite shorter longevity in the

laboratory trials, we found no difference in brood surface in the field. In the end, as we

cannot verify if the longevity would have differed within the hive, this result still

dampers other beneficial effects of the supplements observed in this study and

highlights the present lack of knowledge in bee nutrition to formulate a supplement

equivalent to pollen. In addition, our results emphasize the importance of measuring

multiple colony parameters to assess the suitability of a diet.

Pollen-Enriched vs. Pollen-Free Supplement

The second goal of this study was to compare the consumption and impact on honey bee

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

health of two commercial protein supplements. Our hypothesis that the pollen-enriched

supplement (Global Patties) would be more consumed than the pollen-free

supplement (Ultra Bee) was confirmed. Indeed, the pollen-enriched supplement was

generally more consumed than the pollen-free supplement, which was expected since

phagostimulant components are present in pollen (Hopkins, Jevans, & Boch, 1969;

Schmidt and Hanna, 2006). Our results are in accordance with the work of Alqarni

(2006), who reported better consumption of artificial diets enriched with pollen.

The pollen-enriched supplement also appeared to be more suitable to honey bees

than the pollen-free supplement. We observed no significant difference between the two

supplements in terms of brood rearing, but lifespan was better when using the PES.

Also, compared to the control, the PES allowed bees to attain a higher protein content in

colonies with and without pollen traps, as opposed to PFS which was significant only

for colonies with pollen traps. Finally, although it did not influence weight of trap-

collected pollen, foraging effort was significantly higher for PFS + traps colonies.

Colonies can react to pollen trapping by increasing their foraging effort (Keller, et al.,

2005), which would explain why we also observed a higher value for the PES + traps

treatment (although it was not significantly different from the “no traps” treatments). If

the environment around the colonies was pollen-limited, this increased foraging effort

may have not resulted in a higher weight in the pollen traps, explaining our results.

Foraging efforts are also expected to increase with colony size (Keller, et al., 2005), but

we found no significant difference between treatments in that regard. Moreover,

foragers can balance colony nutrients by adapting their foraging behaviour (Wright, et

al., 2018). When restricted in pollen, PFS patties may have triggered increased foraging

effort because they were not consumed enough to meet the nutritional requirements of

the colonies or because they lacked specific essential nutrients, which suggest the

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

inadequacy of this diet. The positive impacts noted in colonies fed with the PES may be

simply due to its greater consumption compared to PFS. Avni, et al. (2009) noted that a

greater consumption of the same supplement led to a better brood production and tended

toward higher honey yields as well. The superiority of the pollen-enriched supplement

could also be caused by the presence of pollen in the product. Alqarni (2006) observed

better hypopharyngeal glands development and longevity when feeding bees

supplements with pollen compared to a “traditional” supplement. Manning, et al. (2007)

also found that adding pollen to diets based on soy flour significantly increased

longevity. Finally, Schmehl, et al. (2014) noted that bees fed with sucrose and pollen

and exposed to the pesticide chlorpyrifos lived longer than the bees in the same situation

but fed with a sucrose and soy protein patty. The hives in this study, being placed in an

agricultural environment where chlorpyrifos was likely present, this may in part explain

our results.

Regarding diseases and parasites, Nosema infection did not significantly differ

between treatments, but varroa mite infestation did, as well as trough time. Varroa

infestation was higher in July than in May, which was expected since varroa populations

are positively correlated with brood surface, which was also higher at this point in the

season. However, the difference between treatments for varroa infestation is surprising,

PES + traps colonies being more infected than all the other treatments, except for PES

(without traps) which displayed an intermediate value. What caused this higher

infestation rate is unknown, as we possess little knowledge concerning the impact of

nutrition on this major parasite. Requier, Odoux, Henry, & Bretagnolle (2017)

observed that varroa mite loads were correlated with the decline in pollen harvest of

their colonies. However, they could not confirm if these higher infestations were linked

directly to poorer nutrition or to the low honey bee populations at that time. The effect

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

of nutrition on varroa infestation could be limited according to Alaux, Dantec,

Parrinello, & Le Conte (2011), since pollen feeding is not able to reverse the negative

impacts of varroa on bee metabolism and immune functions. In this study the PES +

traps treatment performed well in all the other measured variable, making it hard to

pinpoint what could have caused the higher mite loads.

The Impact of Sites and Landscape

Overall colony performance was better at Site 3 compared to colonies at Site 1 and 2,

especially toward the end of the season. Indeed, brood surfaces measured on September

6th were greater at Site 3 and weight of trap-collected pollen was also greater for this site

at this time of the year. This increase in brood rearing could be due to the rise of

supplement consumption that occurred at this site from July onwards, or to higher floral

availability and/or quality at this time and location. Finally, Site 3 colonies displayed

the lowest Nosema infection rates. Site 1 also seemed mildly favorable as nurse bees

displayed higher protein contents compared to the other sites at two sample dates out of

four.

Site 3 was characterized by the highest proportion of cultivated land: 70%

compared to 58% for Site 1 and 61% for Site 2. This goes against our third hypothesis,

which was that colonies based in an environment with less cultivated area would

perform better regardless of the treatment. This is surprising, since some authors

reported a negative impact of intensively cultivated areas on honey bee colony nutrition

(Donkersley, et al., 2014; M. Smart, Otto, Cornman, & Iwanowicz, 2018). In addition,

the main crop cultivated around the apiaries was corn and its pollen is known to be

nutritionally deficient for the honey bee (Hocherl, et al., 2012). Contrastingly, other

authors observed a positive impact of cultivated land on honey bee nutrition. In one

study, colonies gained weight over the season when placed in agricultural areas of high

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

and moderate intensity, while colonies placed in non-agricultural area gained

significantly less weight and faced starvation, leading to colony loss in some cases

(Alburaki et al., 2017). Brood production was also greater in agricultural areas of

moderate and low intensity when compared to non-agricultural area. Another study

found a positive correlation between food reserves in the hive and cropland, while the

correlation was negative with forest and grassland (Sponsler and Johnson, 2015). In

both studies the authors emphasized that weeds present in cropland are an important

part of the honey bees diet, which was also confirmed by Requier, et al. (2015) and M.

D. Smart, et al. (2016). Goldenrods (Solidago spp. L.) are especially important food

sources late in the season, yet they tend to be abundant in unmanaged areas or

conservation strips present in agricultural areas (Sponsler and Johnson, 2015). This

could explain why in our study Site 3, which had the greatest proportion of cultivated

land, displayed better performances late in the season. Indeed, we observed many

goldenrods in the vicinity of this site. The various weeds present in agricultural

landscape could have also provided a good diversity of pollen to the colonies, which can

reportedly increase tolerance to pathogens (Di Pasquale, et al., 2016). Some crops, such

as soybeans, can also represent an important source of pollen when floral resources are

limited (Gill and O'Neal, 2015; Sponsler and Johnson, 2015). Site 3 also had the largest

proportion of pastures which, if managed not too intensively, can provide abundant

floral resources. As for Site 1, it was characterized by a relatively important urban area,

mostly residential. Urban landscape, especially suburban gardens, can present high

floral diversity and promote nectar and pollen intake by bees (Kaluza, Wallace, Heard,

Klein, & Leonhardt, 2016). This type of environment can allow honey bees to acquire

all essential amino acids to build up their body proteins. This could make quite a

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

difference in late July, when pollen from agricultural areas mainly originates from corn

and does not promote honey bee health (Di Pasquale, et al., 2016).

Conclusion

In conclusion, the protein supplements tested in this study promoted brood production

in pollen-limited condition. However, the shorter lifespan of the bees consuming them

suggests that the supplements tested are nutritionally deficient. These products are

useful as a temporary solution to avoid dwindling of the colonies in unfavorable

foraging conditions, but beekeepers should show caution when using them as a long-

term solution to the lack of pollen abundance and diversity. In this study, we

demonstrate the importance of measuring multiple colony and individual bee parameters

to test the suitability of a diet, as honeybee nutrition evaluation is complex to measure.

Our findings support a growing body of literature that emphasize the importance of

longer-term nutritional studies to clarify the nutritional requirements and specifics of

artificial feeding of honey bee colonies in order to promote honey bee health, as the

environment they face is becoming more challenging.

Finally, we believe it would be important to repeat similar study over several years to

examine how control/unsupplemented colonies perform in the long run, across

landscape and weather variations. Indeed, when formulating our hypotheses, we thought

the chosen environments were poor in terms of floral resources and that the

unsupplemented colonies would dwindle at certain critical times during the season.

However, except for the protein content of nurse bees, unsupplemented colonies

performed similarly to supplemented ones, which may be due to sufficient floral

resources and exceptionally good weather during the year of the study. Different results

might be obtained during a year with less favorable weather, as it was the case for

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

Mattila and Otis (2006). Moreover, the addition of a control without supplements but

with pollen traps, which we could not include in our study, should be used to confirm

the negative impacts of pollen shortage on colonies.

Acknowledgments

We are grateful to Les Ruchers Gauvin Inc. who offered access to their hives for the duration of

the study; Dominique Michaud and Marie-Claire Goulet for their precious help with protein

analyses; all the field and lab assistants (Lucie Alexandre, Thais Andro, Aurélie Boilard,

Audrey Boivin, Tristan Cloutier, Andréa Duclos, Guillaume Guengard, Clémence Landreau and

Damien Le Botlan) for their diligent help with data collection; and Gaétan Daigle and Awa

Diop for their guidance with statistical analyses. We are also thankful to the beekeeping team at

the Centre de recherche en sciences animals de Deschambault (CRSAD), who kindly lent us the

cages and the pollen traps and helped with Nosema spp. monitoring.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by The Natural Sciences and Engineering Research Council of Canada

(NSERC), Discovery Grant and Engage Grant Programs to V.F. and NSERC and FQRNT

scholarships to M. L.-D.

References

Alaux, C., Dantec, C., Parrinello, H., & Le Conte, Y. (2011). Nutrigenomics in honey bees: digital gene expression analysis of pollen's nutritive effects on healthy and varroa-parasitized bees. Bmc Genomics, 12, p 496. doi:10.1186/1471-2164-12-496.

Alaux, C., Ducloz, F., Crauser, D., & Le Conte, Y. (2010). Diet effects on honeybee immunocompetence. Biology Letters, 6(4), pp. 562-565. doi:10.1098/rsbl.2009.0986.

Alburaki, M., Steckel, S. J., Williams, M. T., Skinner, J. A., Tarpy, D. R., Meikle, W. G., . . . Stewart, S. D. (2017). Agricultural Landscape and Pesticide Effects on Honey Bee (Hymenoptera: Apidae) Biological Traits. Journal of Economic Entomology, 110(3), pp. 835-847. doi:10.1093/jee/tox111.

Alqarni, A. S. (2006). Influence of some protein diets on the longevity and some physiological conditions of honeybee Apis mellifera L. workers. Journal of Biological Sciences, 6(4), pp. 734-737. doi:10.3923/jbs.2006.734.737.

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600601602603604605606607608609610611612613

Amro, A., Omar, M., & Al-Ghamdi, A. (2016). Influence of different proteinaceous diets on consumption, brood rearing, and honey bee quality parameters under isolation conditions. Turkish Journal of Veterinary & Animal Sciences, 40(4), pp. 468-475. doi:10.3906/vet-1507-28.

Avni, D., Dag, A., & Shafir, S. (2009). The effect of surface area of pollen patties fed to honey bee (Apis mellifera) colonies on their consumption, brood production and honey yields. Journal of Apicultural Research, 48(1), pp. 23-28. doi:10.3896/ibra.1.48.1.06.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry, 72(1), pp. 248-254. doi:10.1016/0003-2697(76)90527-3.

Brodschneider, R., & Crailsheim, K. (2010). Nutrition and health in honey bees. Apidologie, 41(3), pp. 278-294. doi:10.1051/apido/2010012.

Centre de référence en agriculture et agroalimentaire du Québec. (2011). Gestion optimale du rucher (2 ed.) Québec, Canada.

Cremonez, T. M., De Jong, D., & Bitondi, M. M. G. (1998). Quantification of hemolymph proteins as a fast method for testing protein diets for honey bees (Hymenoptera: Apidae). Journal of Economic Entomology, 91(6), pp. 1284-1289. doi:10.1093/jee/91.6.1284.

Danner, N., Keller, A., Hartel, S., & Steffan-Dewenter, I. (2017). Honey bee foraging ecology: Season but not landscape diversity shapes the amount and diversity of collected pollen. Plos One, 12(8), p 14. doi:10.1371/journal.pone.0183716.

De Jong, D., Da Silva, E. J., Kevan, P. G., & Atkinson, J. L. (2009). Pollen substitutes increase honey bee haemolymph protein levels as much as or more than does pollen. Journal of Apicultural Research, 48(1), pp. 34-37. doi:10.3896/IBRA.1.48.1.08.

DeGrandi-Hoffman, G., Chen, Y. P., Huang, E., & Huang, M. H. (2010). The effect of diet on protein concentration, hypopharyngeal gland development and virus load in worker honey bees (Apis mellifera L.). Journal of Insect Physiology, 56(9), pp. 1184-1191. doi:10.1016/j.jinsphys.2010.03.017.

DeGrandi-Hoffman, G., Chen, Y. P., Rivera, R., Carroll, M., Chambers, M., Hidalgo, G., & de Jong, E. W. (2016). Honey bee colonies provided with natural forage have lower pathogen loads and higher overwinter survival than those fed protein supplements. Apidologie, 47(2), pp. 186-196. doi:10.1007/s13592-015-0386-6.

DeGrandi-Hoffman, G., Wardell, G., Ahumada-Segura, F., Rinderer, T., Danka, R., & Pettis, J. (2008). Comparisons of pollen substitute diets for honey bees: consumption rates by colonies and effects on brood and adult populations. Journal of Apicultural Research, 47(4), pp. 265-270. doi:10.3896/ibra.47.4.06.

Delaplane, K. S., van der Steen, J., & Guzman-Novoa, E. (2013). Standard methods for estimating strength parameters of Apis mellifera colonies. [Review]. Journal of Apicultural Research, 52(1), p 12. doi:10.3896/ibra.1.52.1.03.

Di Pasquale, G., Alaux, C., Le Conte, Y., Odoux, J., Pioz, M., Vaissiere, B. E., . . . Decourtye, A. (2016). Variations in the availability of pollen resources affect honey bee health. Plos One, 11(9), pp. 1-15. doi:10.1371/journal.pone.0162818.

Di Pasquale, G., Salignon, M., Le Conte, Y., Belzunces, L. P., Decourtye, A., Kretzschmar, A., . . . Alaux, C. (2013). Influence of pollen nutrition on honey bee health: Do pollen quality and diversity matter? Plos One, 8(8), p 13. doi:10.1371/journal.pone.0072016.

614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661

Dietemann, V., Nazzi, F., Martin, S. J., Anderson, D. L., Locke, B., Delaplane, K. S., . . . Ellis, J. D. (2013). Standard methods for varroa research. [Review]. Journal of Apicultural Research, 52(1), p 54. doi:10.3896/ibra.1.52.1.09.

DMTI Spatial. (2012). CanMap RouteLogistics - v2012.3. Markham, OntarioDonkersley, P., Rhodes, G., Pickup, R. W., Jones, K. C., & Wilson, K. (2014).

Honeybee nutrition is linked to landscape composition. Ecology and Evolution, 4(21), pp. 4195-4206. doi:10.1002/ece3.1293.

Eccles, L., Kempers, M., Gonzalez, R. M., Thurston, D., & Borges, D. (2016). Canadian best management practices for honey bee health: industry analysis and harmonization. Canada.

Fleming, J. C., Schmehl, D. R., & Ellis, J. D. (2015). Characterizing the impact of commercial pollen substitute diets on the level of Nosema spp. in honey bees (Apis mellifera L.). Plos One, 10(7), p 14. doi:10.1371/journal.pone.0132014.

Gill, K. A., & O'Neal, M. E. (2015). Survey of Soybean Insect Pollinators: Community Identification and Sampling Method Analysis. Environmental Entomology, 44(3), pp. 488-498. doi:10.1093/ee/nvv001.

Girard, M., Chagnon, M., & Fournier, V. (2012). Pollen diversity collected by honey bees in the vicinity of Vaccinium spp. crops and its importance for colony development.. Botany-Botanique, 90(7), pp. 545-555. doi:10.1139/b2012-049.

Haydak, M. H. (1970). Honey bee nutrition. [Review]. Annual Review of Entomology, 15, pp. 143-156. doi:10.1146/annurev.en.15.010170.001043.

Herbert, E. W., Shimanuki, H., & Caron, D. (1977). Optimum protein levels required by honey bees (Hymenoptera, Apidae) to initiate and maintain brood rearing. Apidologie, 8(2), pp. 141-146. doi:10.1051/apido:19770204.

Hocherl, N., Siede, R., Illies, I., Gatschenberger, H., & Tautz, J. (2012). Evaluation of the nutritive value of maize for honey bees. Journal of Insect Physiology, 58(2), pp. 278-285. doi:10.1016/j.jinsphys.2011.12.001.

Hopkins, C. Y., Jevans, A. W., & Boch, R. (1969). Occurrence of octadeca-trans-2,cis-9,cis-12-trienoic acid in pollen attractive to the honey bee. Canadian journal of biochemistry, 47(4), p 433. doi:10.1139/o69-067.

Ihle, K. E., Baker, N. A., & Amdam, G. V. (2014). Insulin-like peptide response to nutritional input in honey bee workers. Journal of Insect Physiology, 69, pp. 49-55. doi:10.1016/j.jinsphys.2014.05.026.

Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. (2016). Summary for policymakers of the assessment report of the IPBES on pollinators, pollination and food production. Bonn, Germany

Kaluza, B. F., Wallace, H., Heard, T. A., Klein, A. M., & Leonhardt, S. D. (2016). Urban gardens promote bee foraging over natural habitats and plantations. Ecology and Evolution, 6(5), pp. 1304-1316. doi:10.1002/ece3.1941.

Keller, I., Fluri, P., & Imdorf, A. (2005). Pollen nutrition and colony development in honey bees - Part II. Bee World, 86(2), pp. 27-34.

Klein, S., Cabirol, A., Devaud, J. M., Barron, A. B., & Lihoreau, M. (2017). Why bees are so vulnerable to environmental stressors. [Review]. Trends in Ecology & Evolution, 32(4), pp. 268-278. doi:10.1016/j.tree.2016.12.009.

Kulhanek, K., Steinhauer, N., Rennich, K., Caron, D. M., Sagili, R. R., Pettis, J. S., . . . vanEngelsdorp, D. (2017). A national survey of managed honey bee 2015-2016 annual colony losses in the USA. Journal of Apicultural Research, 56(4), pp. 328-340. doi:10.1080/00218839.2017.1344496.

La Financière agricole du Québec. (2016). Base de données des parcelles et productions agricoles déclarées. Québec, Canada

662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711

Li, C. C., Xu, B. H., Wang, Y. X., Feng, Q. Q., & Yang, W. R. (2012). Effects of dietary crude protein levels on development, antioxidant status, and total midgut protease activity of honey bee (Apis mellifera ligustica). Apidologie, 43(5), pp. 576-586. doi:10.1007/s13592-012-0126-0.

Li, C. C., Xu, B. H., Wang, Y. X., Yang, Z. B., & Yang, W. R. (2014). Protein content in larval diet affects adult longevity and antioxidant gene expression in honey bee workers. Entomologia Experimentalis Et Applicata, 151(1), pp. 19-26. doi:10.1111/eea.12167.

Manning, R., Rutkay, A., Eaton, L., & Dell, B. (2007). Lipid-enhanced pollen and lipid-reduced flour diets and their effect on the longevity of honey bees (Apis mellifera L.). Australian Journal of Entomology, 46, pp. 251-257. doi:10.1111/j.1440-6055.2007.00598.x.

Mattila, H. R., & Otis, G. W. (2006). Influence of pollen diet in spring on development of honey bee (Hymenoptera : Apidae) colonies. Journal of Economic Entomology, 99(3), pp. 604-613.

Ministère de l'Agriculture des Pêcheries et de l'Alimentation du Québec. (2014). Calendrier de contrôle de la varroase - Comment utiliser le calendrier. Québec, Canada.

Ministère de l'Agriculture des Pêcheries et de l'Alimentation du Québec. (2016). Enquête sur la mortalité hivernale des colonies d'abeilles au Québec en 2015 (Bulletin zoosanitaire). Québec, Canada.

Ministère des Ressources naturelles et de la Faune du Québec. (2000). Base de données topographiques du Québec, 1:20 000, feuilles 21L/02-200-202, 21L/02-200-201 version 1.0. Québec.

Morais, M. M., Turcatto, A. P., Pereira, R. A., Francoy, T. M., Guidugli-Lazzarini, K. R., Goncalves, L. S., . . . De Jong, D. (2013). Protein levels and colony development of Africanized and European honey bees fed natural and artificial diets. Genetics and Molecular Research, 12(4), pp. 6915-6922. doi:10.4238/2013.December.19.10.

Mortensen, A. N., Jack, C. J., Bustamante, T. A., Schmehl, D. R., & Ellis, J. D. (2018). Effects of Supplemental Pollen Feeding on Honey Bee (Hymenoptera: Apidae) Colony Strength and Nosema spp. Infection. Journal of Economic Entomology doi:10.1093/jee/toy341

Oldroyd, B. P. (2007). What's Killing American Honey Bees? PLoS Biology, 5(6), p e168. doi:10.1371/journal.pbio.0050168.

Peng, Y., D'Antuono, M., & Manning, R. (2012). Effects of pollen and artificial diets on the hypopharyngeal glands of newly hatched bees (Apis mellifera L.). Journal of Apicultural Research, 51(1), pp. 53-62. doi:10.3896/ibra.1.51.1.07.

Pernal, S. F., & Clay, H. (2015). Maladie et organismes nuisibles de l’abeille domestique (3ème édition ed.) Beaverlodge, AB, Canada: Association canadienne des professionnels de l’apiculture (ACPA).

R Core Team. (2016). R: A language and environment for statistical computing. Retrieved from https://www.R-project.org/.

Requier, F., Odoux, J.-F., Henry, M., & Bretagnolle, V. (2017). The carry-over effects of pollen shortage decrease the survival of honeybee colonies in farmlands. Journal of Applied Ecology, 54(4), pp. 1161-1170. doi:10.1111/1365-2664.12836.

Requier, F., Odoux, J.-F., Tamic, T., Moreau, N., Henry, M., Decourtye, A., & Bretagnolle, V. (2015). Honey bee diet in intensive farmland habitats reveals an

712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760

unexpectedly high flower richness and a major role of weeds. Ecological Applications, 25(4), pp. 881-890. doi:10.1890/14-1011.1.

Roulston, T. H., & Cane, J. H. (2000). Pollen nutritional content and digestibility for animals. [Review]. Plant Systematics and Evolution, 222(1-4), pp. 187-209. doi:10.1007/bf00984102.

Roulston, T. H., Cane, J. H., & Buchmann, S. L. (2000). What governs protein content of pollen: Pollinator preferences, pollen-pistil interactions, or phylogeny? [Review]. Ecological Monographs, 70(4), pp. 617-643. doi:10.1890/0012-9615(2000)070[0617:wgpcop]2.0.co;2.

Saffari, A., Kevan, P. G., & Atkinson, J. (2010a). Consumption of three dry pollen substitutes in commercial apiaries. Journal of Apicultural Science, 54(1), pp. 13-20.

Saffari, A., Kevan, P. G., & Atkinson, J. L. (2010b). Palatability and consumption of patty-formulated pollen and pollen substitutes and their effects on honeybee colony performance. Journal of Apicultural Science, 54(2), pp. 63-71.

Schmehl, D. R., Teal, P. E. A., Frazier, J. L., & Grozinger, C. M. (2014). Genomic analysis of the interaction between pesticide exposure and nutrition in honey bees (Apis mellifera). Journal of Insect Physiology, 71, pp. 177-190. doi:10.1016/j.jinsphys.2014.10.002.

Schmickl, T., & Crailsheim, K. (2002). How honeybees (Apis mellifera L.) change their broodcare behaviour in response to non-foraging conditions and poor pollen conditions. Behavioral Ecology and Sociobiology, 51(5), pp. 415-425. doi:10.1007/s00265-002-0457-3.

Schmidt, J. O., & Hanna, A. (2006). Chemical nature of phagostimulants in pollen attractive to honeybees. Journal of Insect Behavior, 19(4), pp. 521-532. doi:10.1007/s10905-006-9039-y.

Scofield, H. N., & Mattila, H. R. (2015). Honey bee workers that are pollen stressed as larvae become poor foragers and waggle dancers as adults. Plos One, 10(4), p 19. doi:10.1371/journal.pone.0121731.

Smart, M., Otto, C., Cornman, R., & Iwanowicz, D. (2018). Using Colony Monitoring Devices to Evaluate the Impacts of Land Use and Nutritional Value of Forage on Honey Bee Health. Agriculture-Basel, 8(1), p 14. doi:10.3390/agriculture8010002.

Smart, M. D., Pettis, J. S., Euliss, N., & Spivak, M. S. (2016). Land use in the Northern Great Plains region of the US influences the survival and productivity of honey bee colonies. Agriculture Ecosystems & Environment, 230, pp. 139-149. doi:10.1016/j.agee.2016.05.030.

Sponsler, D. B., & Johnson, R. M. (2015). Honey bee success predicted by landscape composition in Ohio, USA. Peerj, 3, p 16. doi:10.7717/peerj.838.

Tosi, S., Nieh, J. C., Sgolastra, F., Cabbri, R., & Medrzycki, P. (2017). Neonicotinoid pesticides and nutritional stress synergistically reduce survival in honey bees. Proceedings of the Royal Society B-Biological Sciences, 284(1869), p 9. doi:10.1098/rspb.2017.1711.

van der Steen, J. (2007). Effect of a home-made pollen substitute on honey bee colony development. Journal of Apicultural Research, 46(2), pp. 114-119.

Williams, G. R., Alaux, C., Costa, C., Csaki, T., Doublet, V., Eisenhardt, D., . . . Brodschneider, R. (2013). Standard methods for maintaining adult Apis mellifera in cages under in vitro laboratory conditions. [Review]. Journal of Apicultural Research, 52(1), p 35. doi:10.3896/ibra.1.52.1.04.

761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809

Wright, G. A., Nicolson, S. W., & Shafir, S. (2018). Nutritional Physiology and Ecology of Honey Bees. Annual Review of Entomology, 63, pp. 327-344. doi:10.1146/annurev-ento-020117-043423.

810811812813

Table 1. Principal characteristics of frequently used commercial protein supplements.

Product DistributorCrude protein

content Protein source(s) PollenGlobal Patties Global Patties 17% (patties) Soy / yeast Yes

Ultra Bee Mann Lake Ltd. 18% (patties) Plant protein products (no soy) NoBee-Pro Mann Lake Ltd. 29,9% (powder)a Soy No

Feed-Bee Grain Process Enterprises Ltd. 36,4% (powder)a Plant protein products (no soy) NoMegaBee MegaBee 40% (powder) Plant products (no soy) No

Hearty Bee Purina Animal Nutrition LLC. 56% (powder) Unknown Unknowna : De Jong, et al., 2009.

814

815

816

Figure 1. Least square means of weight of supplement consumed per day per hive (a),

weight of pollen collected by the pollen traps per day per hive (b), and sealed brood

surface per hive (c) ± 95% CI, for each site in 2016 over time (n = 10). For each date,

different letters indicate significant differences (Tukey-Kramer, p < 0.05). No letters

were indicated if there was no significant difference. Least square means for the main

effect of site, over all treatments, are presented.

817

818

819

820

821

822

823

Figure 2. Least square means of weight of supplement consumed per day per hive ±

95% CI for each treatment and site in 2016 (n = 10). For each site, different letters

indicate significant differences (Tukey-Kramer, p < 0.05). Means and confidence

intervals were back-transformed to the original scale of the data.

824

825

826

827

828

829

Figure 3. Mean proportion of foragers bringing back pollen to the hive ± 95% CI for

each treatment in 2016 (n = 10). Different letters indicate significant differences

(Tukey-Kramer, p < 0.05). Means and confidence intervals were back-transformed to

the original scale of the data. Since there was no interaction between treatments and

time or sites, least square means for the main effect of treatments, over all sites and

times, are presented.

830

831

832

833

834

835

836

Figure 4. Least square means of protein content of nurse honey bees (in mg of protein

per g of honey bees) ± 95% CI, for each treatment in 2016 (n = 10). Different letters

indicate significant differences (Tukey-Kramer, p < 0.05). Since there was no

interaction between the treatments and time or sites, least square means for the main

effect of treatments, over all sites and times, are presented.

837

838

839

840

841

842

843

Figure 5. Kaplan-Meier curves for honey bee survival in (left) 2016 and (right) 2017. In

both 2016 and 2017, the honey bees from unsupplemented colonies lived longer than

honey bees from colonies fed the pollen-enriched supplement (PES) (2016: p = 5.2e-12,

2017: p = 0.0091), which lived longer than honey bees from colonies fed the pollen-free

supplement (PFS) (2016: p = 0.0049, 2017: p = 0.0018). In 2016, about 90 honey bees

were used for each treatment, and 300 were used for each treatment in 2017.

844

845

846

847

848

849

850

851

852

Figure 6. Mean varroa mite infestation ± 95% CI for each treatment in 2016 (n = 10).

Different letters indicate significant differences (Tukey-Kramer, p < 0.05). Means and

confidence intervals were back-transformed to the original scale of the data. Since there

was no interaction between treatments and time or sites, least square means for the main

effect of treatments, over all sites and times, are presented.

853

854

855

856

857

858

859

860

Figure 7. Landscape structure within a 5 km radius around each site (apiary). On the

map, orange surfaces represent cultivated land, green surfaces are wooded areas and

grey surfaces, urban areas. Cultures present in 2016 are detailed in the diagrams on the

left, which represent the proportion of the zone occupied by each land use. Cultivated

land occupied 58%, 61% and 70% of Site 1, 2 and 3 respectively (Site 1: 22% corn,

17% soy, 7% cereals, 6% pasture, 5% other crops. Site 2: 32% corn, 16% soy, 3%

cereals, 3% pasture, 7% other crops. Site 3: 31% corn, 18% soy, 10% cereals, 8%

pasture, 3% other crops). Wooded area occupied 4%, 31% and 13% of Site 1, 2 and 3

respectively. Urban area occupied 29%, 0% and 2% of Site 1, 2 and 3 respectively.

861

862

863

864

865

866

867

868

869

870

871