NEST, SOIL AND POLLEN PREFERENCES OF MASON BEES (HYMENOPTERA:
MEGACHILIDAE: Osmia) IN SOUTHWEST MICHIGAN FARMS
By
Mario Simon Pinilla Gallego

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Entomology - Master of Science
2017

ABSTRACT
NEST, SOIL AND POLLEN PREFERENCES OF MASON BEES (HYMENOPTERA:
MEGACHILIDAE: Osmia) IN SOUTHWEST MICHIGAN FARMS
By
Mario Simon Pinilla Gallego
Bees (Hymenoptera: Apoidea) are the most important pollinators due to their active collection of
pollen and frequent visits to flowers. Although honey bees (Apis mellifera) are still the most
economically important commercial crop pollinators, other bee species are being developed,
including the mason bees (Hymenoptera: Megachilidae: Osmia), which are efficient pollinators
of tree-fruit crops. However, we still need to improve the management of these bees and
understand how they use resources in their environment. I studied the nesting preferences, soil
use and pollen preferences of two common mason bees in Michigan: Osmia lignaria and O.
cornifrons. Several nesting shelters and nesting substrates were offered to bees, revealing that the
combination of wooden shelters with natural reeds and an attractant resulted in the most nesting.
Analysis of soil brought back to the nest showed that most nest partitions are loam or sandy
loam, with the two species showing different soil preferences. In the brood provision, even when
the nests were in the center of a large blueberry field, O. lignaria did not use this crop as a
dominant pollen resource. Results presented here will help to improve management of these bees
by offering more suitable nesting materials and shelters, as well as highlighting that O. lignaria
is likely not very suitable for blueberry pollination. The soil texture analysis provides new
insights into the soil preferences and requirements of mason bees.

ACKNOWLEDGEMENTS

I want to specially thank Margaret Hartman and Nora Straquadine for their invaluable help with
the field work, processing and analysis of samples. Without their help this project could not have
been possible. To Rufus Isaacs for being a great mentor and guiding me during the development
of this project and my masters degree. To the committee members, James Crum, Matthew
Grieshop and Larry Gut, for their valuable contributions to this project. To Jason Gibbs for the
original idea and initial work that gave shape and guided this project. To Theresa Pitts-Singer at
Utah State University for providing the bee attractant used in this experiment. To all the people
and lab members and colleagues that in several ways contributed to this project, helping me with
lab and field work. To Heather Lenartson-Kluge at the Department of Entomology for her help
and assistance regarding administrative and school processes. To Jon Dahl at the Soil and Plant
Nutrient Laboratory for allowing me to use his lab for the soil texture analysis, and to Randall
Schaetzl at the Department of Geography for allowing me to use the laser diffractometer for the
particle size analysis. To the Fulbright Program, Ecopetrol, Michigan State University, and the
Department of Entomology for funding my graduate studies and research.

	

iii	

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... vi
LIST OF FIGURES ...................................................................................................................... vii
CHAPTER 1. LITERATURE REVIEW: NESTING, POLLEN AND SOIL PREFERENCES OF
MASON BEES (HYMENOPTERA: MEGACHILIDAE: Osmia) ................................................ 1
INTRODUCTION ...................................................................................................................... 1
Alternative pollinators ............................................................................................................ 2
Life cycle of mason bees......................................................................................................... 3
Mason bees (Osmia spp.) as crop pollinators ......................................................................... 5
Nesting preferences of Osmia ................................................................................................. 7
Pollen used by Osmia............................................................................................................ 11
Vaccinium pollination by Osmia........................................................................................... 12
Soil use by Osmia ................................................................................................................. 15
SUMMARY .............................................................................................................................. 16
CHAPTER 2. IMPROVING MASS NESTING OF TWO Osmia (HYMENOPTERA:
MEGACHILIDAE) SPECIES BY NEST DESIGN AND AN ATTRACTANT SPRAY ........... 18
INTRODUCTION .................................................................................................................... 18
METHODS ............................................................................................................................... 20
Shelters and nesting substrate study ..................................................................................... 22
Attractant to increase Osmia nesting .................................................................................... 25
Data analysis ......................................................................................................................... 27
RESULTS ................................................................................................................................. 28
Shelter/substrate experiment ................................................................................................. 28
Attractant spray experiment .................................................................................................. 34
DISCUSSION ........................................................................................................................... 37
CHAPTER 3. SOIL USE BY O. lignaria and O. cornifrons (HYMENOPTERA:
MEGACHILIDAE) IN AN AGRICULTURAL ENVIRONMENT ............................................ 42
INTRODUCTION .................................................................................................................... 42
METHODS ............................................................................................................................... 45
Soil sampling ........................................................................................................................ 47
Particle size analysis ............................................................................................................. 49
Partition color analysis.......................................................................................................... 50
Data analysis ......................................................................................................................... 50
RESULTS ................................................................................................................................. 51
DISCUSSION ........................................................................................................................... 62
CHAPTER 4. POLLEN USE BY Osmia lignaria (HYMENOPTERA: MEGACHILIDAE) IN
HIGHBUSH BLUEBERRY ......................................................................................................... 67
INTRODUCTION .................................................................................................................... 67

	

iv	

METHODS ............................................................................................................................... 70
Study site............................................................................................................................... 70
Experimental setup................................................................................................................ 72
Pollen samples from nests..................................................................................................... 74
Pollen sampling from plants ................................................................................................. 74
Pollen processing .................................................................................................................. 75
Data analysis ......................................................................................................................... 75
RESULTS ................................................................................................................................. 76
DISCUSSION ........................................................................................................................... 82
CHAPTER 5. FUTURE DIRECTIONS ....................................................................................... 88
APPENDIX ................................................................................................................................... 92
LITERATURE CITED ................................................................................................................. 94

	

v	

LIST OF TABLES
	
Table 3.1. Soil texture (range and average ± SE) of complete nests soil samples collected by O.
cornifrons and O. lignaria. Means with the same letters within the same column are not
significantly different (p > 0.05). .................................................................................................. 54
Table 3.2. Soil texture (range and average ± SE) of samples from individual cell partitions
collected by O. cornifrons and O. lignaria ................................................................................... 55
Table 3.3. Color and texture of the different partition’s color identified in O. cornifrons and O.
lignaria nests................................................................................................................................. 61
Table 4.1. Blooming plants present in the blueberry field during the period Osmia lignaria was
active (May 10 – June 17, 2016), and their presence in the field interior (IN) or exterior (OUT).
....................................................................................................................................................... 76
Table 4.2. Mean percentage (± SE) of total pollen in O. lignaria brood cells nesting inside (IN)
or at the edge (OUT) of blueberry fields in west Michigan. Within each location, means with the
same letter are not significantly different, p > 0.05. ..................................................................... 81
Table A. List of voucher specimens ............................................................................................. 93

	

vi	

LIST OF FIGURES

Figure 1.1. Osmia life cycle, presenting the main development stages of the bee. From Mader et
al. (2010). ........................................................................................................................................ 5
Figure 2.1. A) Wooden shelter with nesting tubes at Clarksville Research Center showing the
basic layout of the shelters. B) Nesting substrates used in the shelter/substrate experiment,
including the emergence box (upper left to right: 7 mm, 8 mm, 6 mm, 5 mm cardboard tubes,
natural reeds; middle center: emergence box with exit hole; bottom center: wood block). C)
Three bundles of 7 mm cardboard tubes used in the attractant spray experiment. ....................... 22
Figure 2.2. Shelter locations at Trevor Nichols Research Center at Fennville, MI. Magenta pins
correspond to locations where O. cornifrons was released, yellow pins correspond to locations
where O. lignaria was released. .................................................................................................... 24
Figure 2.3. Shelter locations at Clarksville Research Center, Clarksville, MI. Magenta pins
correspond to locations where O. cornifrons was released, yellow pins correspond to locations
where O. lignaria was released. .................................................................................................... 26
Figure 2.4. Total number of nests obtained in locations where the target species was released or
not for the shelter/substrate experiment at Fennville, MI. A) O. cornifrons. B) O. lignaria. Bars
with the same letter are not significantly different, p > 0.05. ....................................................... 29
Figure 2.5. Percentage of cavities of each substrate used by O. cornifrons in each shelter type, A)
wooden, B) blue, and C) white shelters. Natural reeds and 5 mm cardboard tubes are the
substrates with the higher proportion of use. Bars within a graph with the same letter are not
significantly different, p > 0.05. ................................................................................................... 30
Figure 2.6. Percentage of cavities in each shelters type used by O. cornifrons for the different
substrates. A) 5 mm tubes, B) 6 mm tubes, C) 7 mm tubes, D) 8 mm tubes, E) wood block, F)
natural reeds. Bars within a graph with the same letter are not significantly different, p > 0.05. 31
Figure 2.7. Percentage of cavities of each substrate used by O. lignaria. Natural reeds is the
substrate with the higher proportion of use. Bars with the same letter are not significantly
different, p > 0.05. ........................................................................................................................ 32
Figure 2.8. Number of cells build per nest in the most used substrates for A) O. cornifrons, and
B) O. lignaria. Bars with the same letter are not significantly different, p > 0.05. ...................... 33
Figure 2.9. Number of females produced per nest in the most used substrates for A) O.
cornifrons, and B) O. lignaria. Bars with the same letter are not significantly different, p > 0.05.
....................................................................................................................................................... 34

	

vii	

Figure 2.10. Total number of nest obtained in locations where the target species was released or
not in the attractant spray experiments at Clarksville, MI. A) O. cornifrons. B) O. lignaria. No
significant difference was found between treatments. .................................................................. 35
Figure 2.11. Number of nest build per shelter when the nesting materials are sprayed or not with
decanoic acid. A) O. cornifrons and B) O. lignaria. Bars with the same letter are not significantly
different, p > 0.05. ........................................................................................................................ 36
Figure 2.12. A) Number of cells per nest and B) number of females per nest build by O.
cornifrons when the nesting materials are sprayed or not with decanoic acid. Bars with the same
letter are not significantly different, p > 0.05. .............................................................................. 37
Figure 3.1. Maps of the Trevor Nichols Research Center near Fennville, MI, showing: A) surface
soil (0-25 cm) map generated with the USDA Soil Survey online database in relation to the
locations of shelters for Osmia bees. B) measured soil texture of surface samples (0-10 cm) taken
from the research center. The estimated area available for bee foraging depicted in the figure was
based on a 250 m foraging range from each shelter location. Alternative locations correspond to
Osmia shelters in the stations installed for other experiments, from where nests were taken to
increase the number of soil samples from bees’ nests. ................................................................. 46
Figure 3.2. Wooden shelters used at the research station, including the emergence box (bottom)
and bamboo stems (top) offered to Osmia bees for nesting.......................................................... 47
Figure 3.3. Texture of soil samples from A) the whole nest and B) from individual partitions for
O. cornifrons (red) and O. lignaria (blue). The clay content in O. cornifrons samples was
significantly higher than in O. lignaria samples (p < 0.05). ......................................................... 52
Figure 3.4. Biplot of the principal component analysis for the soil texture of whole nest samples.
PC1 (sand) and PC2 (clay) are shown. Each sample is represented by one dot and the color label
correspond to the species, O. cornifrons (red) and O. lignaria (blue). ......................................... 54
Figure 3.5. Texture of soil samples from individual partitions of the nest of A) O. cornifrons and
B) O. lignaria. Dots of the same color correspond to partitions taken from the same nest.......... 56
Figure 3.6. Biplot of the principal component analysis for the soil texture from individual
partitions for A) O. cornifrons and B) O. lignaria. PC1 (sand) and PC2 (clay) are shown. Each
sample is represented by one dot and the color label (cell) corresponds to the order of the
partitions in the nest, with 0 being the most inner partition.......................................................... 57
Figure 3.7. Mean ± SE of sand, silt and clay content of soil partitions for nests of A) O.
cornifrons and B) O. lignaria. ...................................................................................................... 58
Figure 3.8. Mean ± SE percentage of clay, silt and sand in individual partitions from O.
cornifrons nests from different zones in the research station. Silt content in partitions from zone
4 showed a trend towards higher particle size (p = 0.07). ............................................................ 58

	

viii	

Figure 3.9. Percentage of clay (mean ± SE) in O. cornifrons partitions at the different zones in
TNRC. Partitions showed significant differences in clay content (p < 0.05) only in zone 2. ...... 59
Figure 3.10. Different color patterns observed on the cell partitions of O. cornifrons and O.
lignaria. Patterns can vary from spotted (A-C) to spiral or concentric (E-H). Soil of different
colors suggest that they were collected at different location or at different depths in the same
location.......................................................................................................................................... 60
Figure 3.11. Percentage of partition of each color identified in A) O. cornifrons and B) O.
lignaria nests................................................................................................................................. 61
Figure 4.1. Blueberry Valley farm in Pullman Michigan outlined in red, and the locations (IN or
OUT) where Osmia bee shelters were located. ............................................................................. 71
Figure 4.2. Nest shelter located at A) the OUT location at the edge of the field and B) at the IN
location in the blueberry field. ...................................................................................................... 73
Figure 4.3. Percentage of blueberry pollen in O. lignaria nests located inside (IN) or at the edge
(OUT) of a blueberry farm in West Michigan. ............................................................................. 79
Figure 4.4. Mean (± SE) percentage of the most common pollen found in the brood cells of O.
lignaria nesting inside (IN) or at the edge (OUT) of blueberry fields in west Michigan. The *
represent statistical differences between positions (IN and OUT) (p < 0.05). ............................. 80
Figure 4.5. Percentage of blueberry pollen in Osmia lignaria cell provisions collected at the
interior (IN) and exterior (OUT) of the field as a function of their position in the tubes, equating
to their timing of completion of the cells. ..................................................................................... 81

	

ix	

CHAPTER 1. LITERATURE REVIEW: NESTING, POLLEN AND SOIL
PREFERENCES OF MASON BEES (HYMENOPTERA: MEGACHILIDAE: Osmia)

INTRODUCTION
Seventy-five percent of crops around the world depend, to some extent, on animal pollination,
and it is required by many crops to achieve economic yields (Klein et al. 2007). Bees
(Hymenoptera: Apoidea) are the most important pollinators due to their active collection of
pollen and frequent visits between conspecific flowers (Gallai et al. 2009, Garibaldi et al. 2011,
Nogué et al. 2016), contributing approximately $14.6 billion per year in the United States, of
which $3.07 billion is provided by wild bees (Morse and Calderone 2003, Losey and Vaughn
2006, Klein et al. 2007). European honey bees (Apis mellifera L.) are the most commonly used
commercial pollinator, however, in the last few decades their populations have declined
(VanEngelsdorp et al. 2008, Jaffé et al. 2010, Potts et al. 2010, VanEngelsdorp and Meixner
2010). Decreasing honey bee populations have sparked an interest in studying alternative
pollinators, which can be more efficient per visit than honey bees for many crops (Garibaldi et al.
2013).
Bees in the genus Osmia (Megachilidae) are an important group of alternative pollinators.
These solitary, cavity-nesting bees emerge early in the spring, and some species have a strong
preference for tree-fruit crops, which makes them good pollinators of these crops (Bosch and
Kemp 2001, Mader et al. 2010). In the United States, Osmia lignaria Say and O. cornifrons
(Radoszkowski) are the two species with the highest potential to be used in commercial orchards
(Mader et al. 2010, Sedivy and Dorn 2014). Several strategies have been developed to rear and
release Osmia in crops, but there are still issues related to how to improve their retention and

	

1	

reproductive success in commercial orchards (Biddinger et al. 2008, Sedivy and Dorn 2014).
Variables such as the number and location of nesting sites, number of bee release sites, type of
nesting materials, effect of landscape structure and use of chemical nesting attractants still need
to be assessed (Artz et al. 2013, 2014). Furthermore, the few studies that have considered Osmia
species as potential pollinators of non-Rosaceae plants such as blueberry (Vaccinium spp.), found
that some species can be effective depositing pollen on the stigma (Stubbs et al. 1997, Dogterom
1999, West and McCutcheon 2009, Bushmann and Drummond 2015), so the potential of Osmia
bees as pollinators of other crops should be assessed in the field.

Alternative pollinators
There are over 20,000 described bee species worldwide (Michener 2007), and the majority
collect pollen as the primary source of protein for their larvae. In many cases, wild bees can be
better pollinators than honey bees due to specialized behavioral and morphological adaptations to
specific flowers (Garibaldi et al. 2015). Although honey bee hives are commonly installed in
many crops during bloom to ensure pollination, wild bees can greatly contribute to pollination
(Garibaldi et al. 2013) and in some cases they can be the dominant pollinators, providing enough
pollination for small crop fields and making supplementary pollination unnecessary (Ratti et al.
2008, Winfree et al. 2008, Isaacs and Kirk 2010). Furthermore, pollinator richness can increase
yield and reduce the variance in yield between fields and across years due to complementary
pollination among species (Morris et al. 2010, Blitzer et al. 2016, Garibaldi et al. 2016). Mason
bees within the genus Osmia are one potential source of alternative pollinators.

	

2	

Life cycle of mason bees
Osmia spp. are univoltine (one generation per year), solitary, cavity-nesting bees (Levin 1966),
present mainly in the norther hemisphere (Michener 2007). In Michigan, adults emerge early in
the spring, around April and May, with males emerging approximately 3-6 days earlier than
female. Upon emergence, females mate and start locating nesting locations (Torchio 1989, 1990)
(Fig 1.1). Usually, mason bee populations are biased towards males, with 1.5 - 2 males per
females (Torchio and Tepedino 1980). In the wild, mason bees nest mainly in the soil or in
existing cavities made by wood-boring insects, but they can also nest in fabricated structures
such as wood blocks with drilled holes, cardboard straws, hollow reeds or bamboo (Levin 1966,
Bosch and Kemp 2002, Cane et al. 2007, Mader et al. 2010).
Females of O. lignaria and O. cornifrons tend to return to the locations where they
emerged, attracted by volatiles present in old cocoons (Torchio 1984, Pitts-Singer 2007). Once
they select a nesting cavity, it is marked with free fatty acids, hydrocarbons and wax esters that
allow bees to identify the cavity and warn other females that it is in use (Raw 1992, Guedot et al.
2006). This behavior is especially useful for species that tend to nest in close proximity to other
individuals, allowing them to distinguish individual nests (Fauria and Campan 1998).
Nests of O. lignaria and O. cornifrons usually consist of straight cavities in wooden
substrates. Females clean the selected cavity and build a partition at the back of the nest (Torchio
1989). Some species use mud to make the partitions between cells (Bosch and Kemp 2000),
while other species in the genus use leaves or a mix of plant materials and mud when partitioning
cells (Torchio 1989, Cane et al. 2007, Sampson et al. 2009). Once the first partition is built,
females start foraging for pollen and nectar, with 15 to 40 foraging trips required to fill one
brood cell (Levin 1966, Torchio 1989, Mader et al. 2010). Although some species can fly as far a

	

3	

1200 m (Guedot et al. 2009), their effective foraging range is estimated to be between 100-200 m
(Bosch and Kemp 1999), and most of the foraging activity occurs within 30 to 50 m of the nest
when abundant resources are available (Monzón et al. 2004, Matsumoto et al. 2009, Biddinger et
al. 2013).
Females then lay a single egg directly on the pollen ball and start collecting soil to build a
partition (Torchio 1989, Bosch and Kemp 2000). Four to 12 cells are typically built in each
cavity, with the highest probability of female eggs in the innermost cells, and male eggs in the
cells closest to the entrance of the nest. With good conditions females can build between 1 to 6
nests within one month (Medler 1967, Torchio 1982a, 1984, Rust 1998).
Osmia spp. larval development is temperature dependent (Bosch and Kemp 2003, Kemp
and Bosch 2005), and for O. lignaria, development from larva to pupa can take approximately 97
days in field conditions (Bosch and Kemp 2000). After reaching the final instar, larvae spin a
cocoons and pupate in it (Torchio 1989) (Fig 1.1), with female cocoons being larger than male
cocoons (Torchio and Tepedino 1980). Bees finish their development by August, after which
they are required to go through a winter diapause to successfully emerge the next spring. If bees
are deprived of this cold period, they will consume their fat tissues and starve (Sgolastra et al.
2011, 2012). Overwintering adults will remain in their cocoons until next spring, with
temperature increase triggering emergence (White et al. 2009) (Fig 1.1).

	

4	

Figure 1.1. Osmia life cycle, presenting the main development stages of the bee. From Mader et
al. (2010).

Mason bees (Osmia spp.) as crop pollinators
Mason bees are common pollinators of early spring plants, including crops such as apple, cherry,
pear and plum (Bosch and Kemp 1999, Ladurner et al. 2004, Monzón et al. 2004, Matsumoto et
al. 2009, Mader et al. 2010, Sheffield 2014). For this group of rosaceous plants, mason bees are
considered more efficient pollinators than honey bees because they mainly forage for pollen,
constantly make contact with the reproductive parts of the flowers, and can fly at lower
temperatures than honey bees (Bosch and Kemp 2001, Mader et al. 2010, Matsumoto and
Maejima 2010). Importantly, they tend to gather pollen from at least two cultivars during
foraging trips, promoting cross pollination (Matsumoto et al. 2009).

	

5	

Osmia species that have been studied as alternative commercial pollinators include O.
lignaria in the USA (Bosch and Kemp 2000, 2002), O. cornifrons in Japan (Matsumoto and
Maejima 2010) and Osmia cornuta (Latreille) in Europe (Bosch 1994a). In Japan, O. cornifrons
has been used as an orchard pollinator since the 1940s, and was introduced to the United States
in 1977 by the USDA Bee Lab in Beltsville, Maryland (Batra 1978). Osmia lignaria is native to
the USA, with two subspecies; O. lignaria propinqua Cresson and O. lignaria lignaria Say,
which are found in western and eastern regions of the continent, respectively (Rust 1974). In
some areas, Osmia species are commercially available and their emergence time can be
manipulated with temperature to coincide with bloom of different crops such as almond (Bosch
and Kemp 2000, Artz et al. 2014) and blueberry (West and McCutcheon 2009).
Even though Osmia bees are recognized as good pollinators of tree crops, their
populations in orchards can be at low levels (Vicens and Bosch 2000a, Gardner and Ascher
2006, Schlueter and Stewart 2015). If an effective pollinator is not present in adequate numbers,
increasing the density of pollinators can help to achieve a high pollination rate. One example is
the case of honey bees pollinating blueberries, in which the fruit-set increases when the number
of visits per flower increases (Dedej and Delaplane 2003). In the case of stem nesting bees, the
number of bees in the field can be increased by providing suitable nesting options and/or by
releasing individuals in the field. Some studies report a substantial increase in Osmia populations
in crops after releasing an initial group of females and providing adequate nesting materials
(Bosch and Kemp 1999, Bosch et al. 2006, Krunic and Stanisavljevic 2006).
Given that mason bees need to go through a period of cold temperature to be able to
emerge in the spring (Sgolastra et al. 2011), several studies have addressed the effects of
temperature, pre-wintering and winter duration on mason bee development, survival and

	

6	

emergence time (Bosch and Kemp 2003, 2004, Kemp and Bosch 2005, Sheffield, Westby,
Kevan, et al. 2008, Radmacher 2011, Sgolastra et al. 2011, 2012). This knowledge allows for
control of bee emergence to match with crop bloom. The recommended density for O. lignaria is
620 and 740 females/ha in apples and almond orchards, respectively. For apples, bees can be
released when 30% of flowers are open (Vicens and Bosch, 2000a), while for almonds it is
recommended to release bees when 5-10% of flowers are in bloom (Artz et al. 2013).
Recently, in order to improve the use of these bees in crops, the effect of pesticides on
mason bees has been studied in more detail. Several authors report that Osmia species can be
more sensitive than honey bees to some insecticides, and that fungicides, even without a lethal
effect on bees, can disrupt the nest recognition abilities (Ladurner et al. 2005, 2008, Biddinger,
Robertson, et al. 2013, Artz and Pitts-Singer 2015). Assessment of nesting substrates/shelters
have been running for many years (Torchio and Tepedino 1980, Torchio 1982a, 1982b, 1984,
Artz et al. 2014), but there are still many questions about their management that need to be
answered to achieve sustainable and reliable use of these bees in agriculture.

Nesting preferences of Osmia
In order to develop effective strategies for managing mason bees as commercial pollinators, it is
essential to understand their nesting preferences given that a lack of nesting sites can limit the
size of wild populations (Yoon et al. 2015). Studies describing mason bee nesting date back to
1925 (Fabre 1925) with later reviews of their biology (Levin 1966, Raw 1972). Species such as
O. lignaria and O. cornifrons also have a natural tendency for building their nests in close
proximity to other conspecific bees (Pitts-Singer 2007). This gregarious behavior facilitates

	

7	

maintaining a large number of nests and individuals in a single place that can be used for crop
pollination (Mader et al. 2010).
The reported preferred nest dimensions for O. lignaria are 7.5 mm diameter holes, that
are 152 mm deep (Mader et al. 2010). However, Torchio (1984) reports that dimensions seem to
be less important than the attractiveness of the nesting material for this species. Osmia cornifrons
prefer nests 5.8-6.0 mm in diameter, and 120-140 mm deep. Inter-hole distance is also important
and is suggested to be 20 mm, facilitating orientation and nest location (Mader et al. 2010). The
sex ratio usually varies between 1.5 to 2.0 males per female, and is partly determined by the
depth and diameter of the nest cavity, with deeper tunnels favoring the production of females and
narrower tunnels favoring males. Bees commonly build 5 to 13 cells per nest, with a decrease in
the number of cells per nest later in the flight period. Nests built at this time also tend to contain
a higher proportion of males (Torchio and Tepedino 1980, Mader et al. 2010). Female cells are
bigger than male cells and contain larger pollen provisions, resulting in sexual dimorphism with
females being bigger than males (Torchio and Tepedino 1980).
For management of large numbers of Osmia bees, nest shelter design should also be taken
into account. Nests need protection from direct sunlight and rain with shelters, that can also serve
as a visual landmark for bees (Mader et al. 2010). Shelters should not be moved after bees have
started nesting; this will result in females abandoning their nests (Vicens and Bosch 2000c).
Shelters that are more than 0.9 m depth are too dark and less attractive to bees (Maccagnani et al.
2007). Painting the shelters in colors that contrast with the landscape, or creating contrasting
patterns as visual clues for bees can help females to easily identify their nest location (Fauria and
Campan 1998). Torchio (1982a, 1984) recommends placement of the shelters facing east or

	

8	

south to encourage flight activity early in the day. However, the direction of the nest does not
affect the number of females nesting in a location (Yoon et al. 2015).
Dispersal away from the nest site is also a factor that can limit the effective use of mason
bees as pollinators (Bosch and Kemp 2002). Osmia bicornis (L.) dispersal distance is very low
with most bees dispersing less than 100 m (Gruber et al. 2011). Large females are less likely to
exhibit pre-nesting dispersal (Steffan-Dewenter and Schiele 2004, Bosch and Vicens 2006), as
well as bees forced to chew out of their natal nests (10-15% dispersal) (Bosch and Kemp 1999).
If loose cocoons are used, they should be released in an emergence box consisting of a dark
container with a single exit hole (Mader et al. 2010), but even in this case dispersal rates can be
as high as 48% for O. cornuta (Bosch 1994), and can increase up to 81% for the same species
when releasing emerged adults (Bosch 1994b). Bosse et al. (2014) report that performing
multiple releases can increase the number of O. bicornis females nesting in a location, likely
because females from the second release are attracted to the location where females from the first
release are nesting because of their gregarious nesting behavior. Dispersal rates can also be
reduced when bees are released in separate small groups, or by increasing attraction to nesting
material (Bosch 1994a). Rough handling of pre-emergent adults does not affect the establishment
of O. lignaria females when they emerge from their natal nest (Stanley et al. 2011), but the effect
of rough handling on bee dispersal rate still needs to be assessed when managing loose cocoons.
Cavities that have been previously occupied by bees can be more attractive to nesting
bees than new adjacent holes, however, the risk of diseases and parasites is higher in the old
nests (Torchio 1984). Females may prefer to nest close to the place where they emerged because
of olfactory cues from the old nests. Pitts-Singer (2007) found that O. lignaria females show a
positive olfactory response to cocoons in a Y-tube bioassay, suggesting that the bees could use

	

9	

them as an olfactory cue that attracts them to nest in a specific place. Cues used by bees to select
potential nesting places include visual orientation cues, chemical cues and physical
characteristics of the nesting material (Pitts-Singer 2007).
Enhanced bee retention and nesting distribution in commercial orchards are important
factors when developing a management system for alternative bee pollinators. Variables such as
the number and location of nesting sites, number of bee release sites, type of nesting materials,
effect of landscape structure and use of chemical nesting attractants should be assessed (Artz et
al. 2013, 2014). Osmia lignaria and O. bicornis prefer nest sites near orchard perimeters and
along bee-ways (roads where honey bee hives are placed during bloom) (Matsumoto and
Maejima 2010, Gruber et al. 2011, McKinney and Park 2012). The amount of fallow land
surrounding apple orchards in Germany was the most important parameter for the establishment
of O. bicornis in the orchards (Gruber et al. 2011), presumably because this provided pollen
resources when the orchard was not in bloom.
As distance to the release point increases, the number of nests started by O. bicornis
decreases (Gruber et al. 2011). Both O. lignaria (Artz et al. 2013) and O. cornuta (Bosch 1994a)
increase the number of nests built when more small shelters with fewer stems are used instead of
few large shelters with many stems. This indicates that it is easier for bees to find shelters that
are in high density (Artz et al. 2013).
Once O. lignaria females have already established their nest, they show little hesitation
when entering the nest, and if they accidentally enter the wrong nest, they quickly exit the cavity
and scan over to locate the correct one. This suggests that O. lignaria uses olfactory cues to
recognize their nests (Guedot et al. 2006). When exiting the nest, females drag the tip of the
abdomen along the entire tube, marking not only the entrance but the complete tube with free

	

10	

fatty acids, long chain hydrocarbons and wax esters (Guedot et al. 2006). The amount of
compounds that females deposit varies, and can also vary between the entrance and middle
sections of the tube, indicating that the nest marking chemistry is unique for each female. As can
be expected, females are sensitive to the replacement of the entrance and middle section of the
nest, making them hesitate and increasing recognition attempts, or even abandoning the nest
(Guedot et al. 2006). This behavior indicates that chemical labeling of the nest is essential for
nest recognition and its role on bee retention should be studied in more detail.

Pollen used by Osmia
Bees collect pollen and nectar to feed their larvae. Osmia lignaria requires 15 to 40 foraging
trips to gather the necessary pollen and nectar to provision one cell (Mader et al. 2010) and O.
cornifrons females can provision a cell in about 30 minutes when weather conditions are good
and pollen sources abundant (McKinney and Park 2012)
Osmia spp. show a strong preference for Rosaceae trees, and during the crop bloom
period they mainly collect pollen from these plants (Mader et al. 2010). In a cherry and peach
orchard in Utah, the percentage of crop pollen (Prunus spp.) in O. lignaria loads varied between
68 and 98% (Bosch et al. 2006). In apple orchards in Canada, O. lignaria pollen loads contained
72-93% apple pollen (Sheffield et al. 2008). Monzón et al. (2004) examined O. cornuta larvae
fecal samples, and found that overall, 94.4% of the pollen was pear pollen, and 75.5% of the
samples contained exclusively pear pollen. Vicens and Bosch (2000a) reported that 94% of the
cells made by O. cornuta in apple orchards contain 100% apple pollen. This highlights the strong
preferences they can have for tree-orchard crops if their abundance is high enough.

	

11	

Mason bees also show strong floral preferences when they are in natural habitats.
Radmacher and Strohm (2010) found that even when there were 107 plant species blooming
around the nesting site, O. bicornis collect almost exclusively oak and maple pollen, and 39% of
the cell pollen consists of a single (>95%) pollen type. Felicioli et al. (2004) report that the
pollen of anemophilous plants like Salix spp., Acer spp., and Quercus spp. can be very attractive
for O. cornuta and O. bicornis.
Furthermore, some plants can be even more attractive than the actual crop for Osmia
bees. In Virginia and North Carolina apple orchards, Kraemer et al. (2014) found that the most
abundant pollen in O. lignaria cells was redbud (Cercis canadensis, Fabaceae), ranging from 45
to 82%, and only in the farm without redbud trees they found that the orchard pollen reach 53%.
Other pollen types reported in Osmia spp. loads include pollen from the families Fabaceae
(Lupinus polyphyllus) Asteracea, Ranunculaceae, Papaveraceae, and other genera like Erica,
Quercus, Brassica, Carya, Liquidambar (Monzón et al., 2004; Radmacher and Strohm, 2010;
Sheffield et al., 2008). The fact that some plants can be more attractive that the actual crop for
Osmia bees make important to consider the alternative floral resources present in the crop area,
given that floral resources planted aiming to support bees could distract them from visiting the
crop of interest

Vaccinium pollination by Osmia
Blueberries (Vaccinium spp.) are an economically important crop that requires insect pollination.
The flowers have poricidal anthers that release pollen by mechanical stimulus (McGregor 1976,
Abrol 2015). Bumblebees are able to “buzz-pollinate”, in which they vibrate their thoracic
muscles when visiting the flowers, so the pollen is released (Buchmann and Hurley 1978). Other

	

12	

wild bees such Habropoda laboriosa (Fab) are recognized as good blueberry pollinators (Cane
1997, Sampson and Cane 2000), but to provide high abundance of bees, honey bee hives are
placed close to the fields during bloom to ensure good pollination levels (Scherm et al. 2001,
Dedej and Delaplane 2003). However, many other native bees visit blueberry flowers, and Osmia
species can be observed visiting the flowers (Dogterom 1999, Sampson and Cane 2000, Isaacs et
al. 2009, Cutler et al. 2015). Dogterom (1999) found O. lignaria propinqua in 55% of 80
blueberry sites in British Columbia, Canada, but with low abundance. Bushmann and Drummond
(2015) report nine Osmia species visiting blueberry field in Maine, and Tuell et al. (2009) report
13 Osmia species visiting blueberry flowers in southwest Michigan.
Many authors report Osmia species as potential pollinators of blueberries. Stubbs and
Drummond (1997) report that Osmia atriventris Cresson deposits more pollen on lowbush
blueberry (Vaccinium angustifolium) flowers per visit than Bombus impatiens Cresson and A.
mellifera in a cage experiment in Maine. In Poland, Osmia bicornis provide good pollination
levels to V. corymbosum under isolation (Wilkaniee and Giejdasz, 2003), and O. cornifrons
provide the same pollination level as bumblebees in a cage study with V. corymbosum in West
Virginia (West and McCutcheon 2009). Osmia ribifloris Cockerell is another potential pollinator
of blueberries, because their visits led to a high fruit set in a single visit experiment in a
screenhouse with V. ashei for Tifblue and Climax cultivars. However, the same species does not
pollinate the Premier cultivar efficiently, due to the slipper-like shape of the flowers. Despite
these promising results, fruits pollinated by O. ribifloris contained fewer seeds than fruits
pollinated by honey bees (Sampson and Cane 2000, Sampson et al. 2004, 2013). However, most
of these experiments were conducted in controlled situations, where bees have blueberry flowers
as their only pollen and nectar source, and may not be a good representation of what happens in

	

13	

open field settings. Another difficulty that Osmia can face when used for blueberry pollination is
that Vaccinium spp. have a very small opening and a deep corolla, which can make it difficult for
this non-buzz-pollinating bee to access the anthers and release the pollen. Cultivars with wider
corollas and longer stigmas have greater probabilities of being pollinated by Osmia (Felicioli et
al. 2004).
Osmia lignaria propinqua is one of the few Osmia species commercially available for
crop pollination, and is primarily used for pollinating orchard crops. However, in V. corymbosum
var. Bluecrop, O. lignaria deposit a similar number of pollen grains on the stigmas as honey
bees, and eighty-six percent of O. lignaria females captured at the nest entrances had blueberry
pollen on their body (Dogterom 1999). It is still unclear what percentage of the pollen carried by
bees belonged to blueberry. On the other hand, Sampson et al. (2009) do not recommend O.
lignaria as a pollinator of blueberry because even if O. lignaria can produce viable progeny
when confined on blueberry, this bee has a strong preference for tree crops.
In some regions nesting activity of natural population of O. lignaria occurs between early
May to mid-June, coinciding with blueberry bloom (Dogterom 1999). In Michigan, O. lignaria
lignaria fly between April and May, so their activity period does not completely coincide with
blueberry bloom, which happens during May (Garcia-Salazar 2002). However, loose cocoons
can be stored in cold temperatures safely for two months longer than the natural emergence
period (Mader et al. 2010), and the emergence time can be altered to coincide with blueberry
bloom. The efficiency of this bee for pollinating blueberries in the field remain to be tested in
Michigan farms.

	

14	

Soil use by Osmia
Most bees in the genus Osmia use leaf material or a mix of leaves and soil to build the partitions
between cells (Cane et al. 2007); however, O. lignaria and O. cornifrons exclusively use soil as
material for the partitions and nest plugs once the nest is complete (Torchio 1989, Mader et al.
2010). Therefore, soil is an essential resource for these bees (Krunic and Stanisavljevic 2006),
yet there is relatively little known about the composition of the soil that they collect or the
ecology of their soil collection.
Osmia lignaria females congregate at soil collecting sites, where the soil is wet and they
have been reported to fly up to 54 m away from the nesting location to collect soil particles
(Torchio 1989). Sometimes females can excavate tunnels in inclined soil surfaces to reach a soil
layer of a particular moisture content, and several individuals can use the same excavated hole
(Torchio 1989). An O. lignaria female gathers soil by scraping with her mandibles and forelegs,
then shapes it to create a round pellet 2-3 mm in diameter using her forelegs, mandibles and the
ventral surface of the posterior tip of the abdomen. Then she holds the pellet between her
mandibles and flies back to the nest (Levin 1966). A similar behavior has been reported for O.
bicornis (Raw, 1972 as O. rufa).
The average time spent collecting a soil pellet is about a minute (Levin 1966), and 8 to 12
soil collecting trips are necessary in order to build one cell partition (Torchio 1989). The soil is
deposited as a foundation that circumscribes the inner wall of the tube. Additional soil is added
until a solid wall is formed across the nest. More soil is then added to the outer perimeter of the
partition which is then polished (Torchio 1989). Females of O. californica Cresson can spend on
average 4 minutes depositing each soil ball into the nest, while O. lignaria females spend an
average of 2 minutes (Levin 1966). The total time required by O. cornifrons to complete a

	

15	

partition is 20 minutes (McKinney and Park 2012). When females finish the last cell, they build a
soil plug at the nest entrance that for O. lignaria can be 4-8 mm thick. Usually the plug is
protruding slightly from the nest entrance (Torchio 1989).
Few studies have addressed the soil preferences of mason bees and ground nesting bees,
which is important because it is an indispensable resource for these bees. Cane (1991) studied the
composition of soil used by ground-nesting bees for their subterranean nests, finding that the
percentage of sand is one of the most important factors for nesting site selection for groundnesting bees. Bosch (1994) tried to study O. cornuta soil preferences by offering soil of different
textures in dispensers, but bees preferred to collect moist soil from the ground. There is still
much to learn about mason bee soil preferences if we want to establish them as commercial
pollinators in orchards, and questions related to soil selection plasticity, preferred soil texture,
and variability among individuals, populations, and species regarding soil preferences still need
to be addressed.

SUMMARY
Pollination is an essential input for many crops, and farmers of large commercial scale crops
have depended on honey bees to ensure good levels of pollination. To reduce the pressure and
dependence on honey bees, it is necessary to find alternative pollinators that can be managed and
used in commercial orchards. Mason bees have the potential to be managed in tree-fruit crops,
where they exhibit good performance and are considered better pollinators than honey bees.
However, in order to achieve sustainable management of these bees, improve female retention,
reproduction success and pollination delivery, we need to increase our understanding of several
aspects of their nesting preferences, including the interaction between nest substrates and shelter

	

16	

type, effect of shelter density, release location density, factors inducing a high dispersal rate of
females, soil use and preferences, and explore whether they can be used in other non-Rosaceae
fruit crops.

To address some of these questions, the objectives of this research thesis are to:
1) Elucidate the nesting material and shelters preferences of O. lignaria and O. cornifrons in
and agricultural environment in Michigan.
2) Determine whether the attractant spray developed from Osmia cocoons is an effective
way to increase nesting by O. lignaria and O. cornifrons.
3) Determine the effect of nest location on blueberry pollen collection by O. lignaria.
4) Characterize the texture of the soil used by O. lignaria in an agricultural environment.

	

17	

CHAPTER 2. IMPROVING MASS NESTING OF TWO Osmia (HYMENOPTERA:
MEGACHILIDAE) SPECIES BY NEST DESIGN AND AN ATTRACTANT SPRAY

INTRODUCTION
Many plants rely on animals to transfer pollen from anthers to the stigmas, and bees
(Hymenoptera: Apoidea) are the most effective and common pollinators of angiosperm plants,
including a large proportion of crops around the world (Klein et al. 2007, Gallai et al. 2009,
Abrol 2015). The European honey bee (Apis mellifera L.) is the most commonly used
commercial pollinator, however, in the last few decades their populations have declined in some
parts of the world (VanEngelsdorp et al. 2008, Jaffé et al. 2010, Potts et al. 2010, VanEngelsdorp
and Meixner 2010). This decrease in honey bee populations has sparked increased interest in
studying alternative pollinators (Canto-Aguilar and Parra-Tabla 2000, Rader et al. 2009,
Christmann and Aw-Hassan 2012, Holzschuh et al. 2012, Sedivy and Dorn 2014), with some
studies showing that many other bee species are more efficient pollinators than honey bees in
many crops, due to their behavioral and morphological adaptations to some kinds of flowers and
their floral preferences (Javorek et al. 2002, Krug et al. 2010, Sheffield 2014, Junqueira and
Augusto 2016).
Mason bees, Osmia (Megachilidae), are an important group of crop pollinators. Some
species show high pollination efficiency for rosaceous crops, and a small number of Osmia
species have been explored as commercial pollinators of apple, cherry, pear and plum (Batra
1978, Bosch and Kemp 1999, Ladurner et al. 2004, Monzón et al. 2004, Matsumoto et al. 2009,
Sheffield 2014). In North America, Osmia lignaria Say (blue orchard bee) and O. cornifrons
Radoszkowski (Japanese hornfaced bee) are the two species with the highest potential for use in

	

18	

commercial orchards. Osmia lignaria is native to the USA, with two subspecies; O. lignaria
propinqua, which is found in the west part of the country, and O. lignaria lignaria, which is
present in the east (Rust 1974). Osmia cornifrons was introduced to the United States in the
1970s from Japan to be used as an apple pollinator (Batra 1978), and is now widely distributed in
the eastern United States (Biddinger et al. 2009). Both are univoltine, solitary, cavity-nesting
bees (Bosch and Kemp 2000, McKinney and Park 2012). In the wild, females nest in existing
cavities made by wood-boring insects, but they can also nest in man-made structures such as
wood blocks with drilled holes, cardboard straws, hollow reeds or bamboo (James and PittsSinger 2008, Sampson et al. 2009). Additionally, they have a natural tendency for building their
nests in close proximity to other conspecific bees (Pitts-Singer 2007), facilitating their rearing.
Several strategies have been developed to rear and release mason bees in crops, but there
are still issues related to how to improve the retention and reproductive success of these bees in
commercial orchards (Sedivy and Dorn 2014). Especially considering that mass released, Osmia
bees show high dispersal rates that can be as high as 50% (Torchio 1982b, Bosch 1994a). When
bees are released as emerged adults the dispersal rates can reach 81% (Bosch 1994b). One way
to reduce dispersal of young Osmia bees is to increase attractiveness of the nesting materials
offered. This can be achieved by selecting the preferred nesting material and dimensions.
Nesting preference studies for mason bees have shown that they preferred wood substrates over
artificial substrates such as cardboard tubes, and preferred diameters between 6-9 mm for
different species (Torchio 1982b, 1984, Rust 1998, Krunic and Stanisavljevic 2006, Biddinger et
al. 2008, Sampson et al. 2009). Another way to increase attractiveness of substrates is by using
chemical attractants that induce bees to nest in a specific location. Early studies recognized that
several Osmia species prefer cavities from where they emerged over new cavities (Torchio

	

19	

1984). Later studies showed that O. lignaria have an antennal response to fatty acids (decanoic
acid, dodecanoic acid, and tetradecanoic acid) emitted from old cocoons (Pitts-Singer 2007) and
are attracted to old nests by these volatiles (Pitts-Singer et al. 2016). Based on these results, a
commercial Osmia attractant was developed from fatty acids isolated from O. lignaria propinqua
nest-trapped in Utah. According to the patent, this attractant is expected to increase the number
of nests of several Osmia species (e.g. O. bicornis, O. bucephala, O. cornifrons, O. aglaia, O.
ribifloris, O. bruneri, and O. cornuta) (Pitts-Singer et al. 2016).
Despite existing knowledge of Osmia nesting preferences and management, there are still
improvements needed for mason bees management in the field. These gaps include the effect of
combinations of different shelters and nesting substrates, and the efficacy of the commercial
attractant in the field for O. lignaria and other Osmia species. Variables such as the number and
location of nesting sites, number of bee release sites, and effect of landscape are areas where
more information is needed to improve the use of these bees as crop pollinators (Artz et al. 2013,
2014). Therefore, the aims of this study were to 1) elucidate the nesting substrate and shelter
materials preferred by O. lignaria and O. cornifrons and 2) determine whether an attractant spray
derived from Osmia cocoons is an effective way to increase the number of nests filled by Osmia
female in a Michigan farm landscape setting.

METHODS
The experiments were conducted at two research centers of Michigan State University (MSU)
during 2016. Both have experimental plots of tree fruit crops, including apple, cherry and plum;
annual crops such as corn, and berries including blueberry and raspberry. Both centers also have
small patches of woods and grassy areas, providing resources for Osmia bees.

	

20	

For all experiments, boxes of 45 x 32 x 35 cm (length x width x height) with one open
face were used as shelters. Inside each shelter, the nesting substrates were attached to a stake at
the upper section so they were not in contact with any of the shelter walls. The open face of the
box was covered with chicken wire (2.5 cm mesh size) to exclude birds and rodents (Fig. 2.1AB). Shelters were attached to two metal posts and held approximately 1.5 m above the ground
(Fig. 2.1A), with the open side facing east or southeast for exposure to sunlight early in the day
(Torchio 1982a, 1982b, 1984). When more than one box was placed in a location they were at
least 5 m away from each other. Each location was at least 60 m away from the others.
Osmia cocoons for release at the experiment locations were obtained from trap nests
collected from the Horticultural Demonstration Garden at MSU (East Lansing, MI) in September
2015. Nests were opened and cocoons were extracted and sorted by sex based on size, since
female cocoons are bigger (» 11 mm long) than male cocoons (» 7 m long) (Mader et al. 2010).
Cocoons were stored in an environmental chamber at 4ºC from November 2015, until their
release. A sub-sample of 10 cocoons was taken from each batch once each week during February
and March (2016), and emerged by holding them at laboratory conditions (» 25ºC) to determine
the proportion of each species and when they were ready to emerge quickly. For the experiments,
bees were released as loose cocoons in cardboard emergence boxes (9.3 x 9.3 x 4.3 cm) with one
10 mm hole on one of the walls to allow bees to exit the box after emerging from the cocoons.

	

21	

Figure 2.1. A) Wooden shelter with nesting tubes at Clarksville Research Center showing the
basic layout of the shelters. B) Nesting substrates used in the shelter/substrate experiment,
including the emergence box (upper left to right: 7 mm, 8 mm, 6 mm, 5 mm cardboard tubes,
natural reeds; middle center: emergence box with exit hole; bottom center: wood block). C)
Three bundles of 7 mm cardboard tubes used in the attractant spray experiment.

Shelters and nesting substrate study
This experiment was conducted at Trevor Nichols Research Center (TNRC), in Fennville, MI
(USA) (42°35'38.26"N, 86° 9'18.03"W), between April-June 2016. For this experiment the
variables evaluated were the shelter type and nesting substrates, both were tested together using a
nested design. In this design, the main factor was the shelter type, with three levels
corresponding to three shelter types: blue and white plastic totes (ULINE, Pleasant Prairie, WI)
	

22	

and pine plywood boxes. The nesting substrates evaluated within each shelter type were a wood
block with 48 holes (8 mm in diameter), natural reeds (Phragmites sp.) (7-10 mm inner
diameter) and cardboard tubes with internal diameters of 5, 6, 7, or 8 mm. All the substrates were
approximately 15 cm long. Wood blocks were closed at one end by a piece of cardboard, reeds
were closed at one end by the natural node of the plant, cardboard tubes were closed at one end,
and a paper straw open at both ends was inserted in each cardboard tube to facilitate removal
from nests. One wood block, 40 natural reed tubes and 50 cardboard tubes each of 5, 6, 7 and 8
mm inner diameter were placed in each shelter. The relative placement of the substrates in the
shelter was randomized, except for the wood block that was always placed in the lower section
(Fig. 2.1B). The response variables were the proportion of the holes in each substrate used for
nesting and the number of cells and female offspring per nest.
Ten locations were selected at TNRC for placement of Osmia nest shelters (Fig. 2.2). On
April 19, 2016, one shelter of each of the three types was placed in a random order at each
location. One emergence box was placed in the central shelter. In five locations the emergence
boxes contained 60 female and 88 male cocoons that were at least 67% O. cornifrons (with the
remaining being O. lignaria). In the remaining five locations the emergence boxes contained 55
female and 69 male cocoons that were at least 73% O. lignaria (with the remaining being O.
cornifrons). These numbers were calculated to have at least 40 females of the desired species
emerging at each location.

	

23	

Figure 2.2. Shelter locations at Trevor Nichols Research Center at Fennville, MI. Magenta pins
correspond to locations where O. cornifrons was released, yellow pins correspond to locations
where O. lignaria was released.

Shelters were checked weekly to count the number of nest cavities that had been filled by
nesting Osmia bees. Those with mud caps at the entrance of the cavity were considered
completed. When the number of completed nests stopped increasing, in early June, shelters were
brought back to the MSU Entomology Research Center (Lansing, MI) and kept at outside

	

24	

ambient temperature until September 2016, when bees have finished their development into
adults. Nests were then opened and the number of nests filled per treatment, number of cells per
nest tube, and the number of females per nest tube were determined. One cocoon per nest tube
was opened and the bee inside was placed in a jar with cyanide to kill the bee, and then pinned in
order to identify the species that built each nest using the key by Arduser (2009).

Attractant to increase Osmia nesting
The attractant spray experiment was carried out at the Clarksville Research Center (CRC) in
Clarksville, MI (USA) (42°52'17.92"N, 85°15'29.29"W), between April-June 2016. For this
experiment, I evaluated whether application of an attractant spray to the nesting substrate
affected the number of cavities filled by Osmia bees, and the number of cells and females
produced per nest tube. The attractant spray consisted of decanoic acid (CH3(CH2)8COOH)
provided by Dr. Theresa Pitts-Singer of the USDA-ARS in Logan, Utah. To prepare the
attractant, 5.8 g of decanoic acid was dissolved in 100 ml of ethyl acetate. The solution was
transferred to a spray bottle that delivered 0.22 ml per spray, corresponding to 50 cocoons
equivalents (CE) per application.
For this experiment, I used pine plywood shelters built by the author and 7 mm cardboard
tubes as the nesting substrate. Each shelter contained three bunches of 50 tubes each (150 tubes
per shelter) (Fig. 2.1C). Eight locations were selected at CRC (Fig. 2.3), and two shelters were
placed at each location on April 26, 2016. Tubes in one shelter per pair were treated with the
attractant, while tubes in the second shelter were left untreated. Treated tubes were sprayed with
the attractant 24 hours before being placed in the field, and were kept apart from the untreated

	

25	

tubes to avoid cross contamination. Each bundle of tubes was sprayed once, for a total of 1 CE
per tube and 150 CE per shelter.

Figure 2.3. Shelter locations at Clarksville Research Center, Clarksville, MI. Magenta pins
correspond to locations where O. cornifrons was released, yellow pins correspond to locations
where O. lignaria was released.

Shelters were checked weekly to count the number of nest tubes filled during that week in
each shelter. When the number of completed tubes stopped increasing around early June, boxes
were brought back to the MSU Entomology Research Center (Lansing, MI) and stored until bees
completed development and nesting activity quantified using the same methods as described
above in the shelter/substrate experiment.

	

26	

Data analysis
Shapiro–Wilk tests and diagnostic graphs (qq plots and residual plots) were used to assess data
normality, and a Bartlett test was used to assess the homogeneity of variance among treatments
(Shanks and Hutton 1976). Non-normal data were square root transformed prior to analysis. A
generalized linear mixed model (GLMM) with beta distribution and logit link function in PROC
GLIMMIX in SAS version 9.4 (SAS Institute 2016) was used to determine the effect of the
shelter and substrate as well as the interaction between the two factors. If any of the factors was
significant (p < 0.05), Tukey’s test was used to determine the significance of differences among
treatments. An analysis of variance (ANOVA) was used to determine whether shelter order at the
location (left, center or right) affected nesting, using R version 3.1.1 (R Development Core
Team, Vienna, Austria). For the analysis of the number of cells and females per nest, only the
data from treatments where more than four nests were obtained were used. For O. cornifrons this
included data from the following treatments (shelter/substrate): blue-5 mm, blue-6 mm, bluenatural reeds, wood-5 mm, wood-6 mm, wood-7 mm, wood-natural reeds, white-natural reeds.
For O. lignaria this included data from the following treatments: blue-natural reeds, blue-wood
block, white-natural reeds, wood-natural reeds.
To test the effect of the attractant, I only included data from locations where bees nested.
This resulted in four replicates for O. cornifrons and three replicates for O. lignaria. To compare
the number of nests obtained and the number of cells and females per nest in each treatment I
used the non-parametric Mann-Whitney test, given that the data did not respond to
transformations. For O. lignaria, given that there were only three replicates, I used the Kruskal–
Wallis test to compare the number of nests obtained, and it was not possible to perform a

	

27	

statistical test on the number of cells per nest because there was only one nest in all of the nospray shelters.

RESULTS
Shelter/substrate experiment
Seventy one percent of the nests obtained in this experiment belonged to O. cornifrons, and 29%
to O. lignaria. The sex ratio for both species was 1.2 males per female. The order of the shelters
at the location (left, center or right) did not affect the number of nests obtained, for O. cornifrons
(F(2, 12) = 1.215, p = 0.331) or O. lignaria (F(2, 12) = 0.752, p = 0.492). More O. cornifrons nests
(7.2 fold) were obtained in the location where this species was released (F(1, 3) = 10.76, p =
0.046), but the number of O. lignaria nests was not different between locations regardless
whether it was released there or not (F(1, 3) = 3.901, p = 0.143) (Fig. 2.4).

	

28	

Figure 2.4. Total number of nests obtained in locations where the target species was released or
not for the shelter/substrate experiment at Fennville, MI. A) O. cornifrons. B) O. lignaria. Bars
with the same letter are not significantly different, p > 0.05.
	
For O. cornifrons, there was a significant effect of the shelter type (F(2, 15) = 5.36, p =
0.0175), nest tube substrate (F(5, 15) = 9.12, p = 0.0004), and the interaction between the two
factors (F(2, 15) = 4.87, p = 0.0235) on the percentage of cavities used. When splitting groups by
shelter type, in the wooded and blue shelters, natural reeds were the most used substrate
(wooden-reeds: 20 ± 7 %, blue-reeds: 10 ± 3 %) (mean ± SE) (all p < 0.05), followed by 5 mm
(wooden-5 mm: 4 ± 2 %, blue-5 mm: 3 ± 2 %) and 6 mm tubes (wooden-6 mm: 4 ± 2 %, blue-6
mm: 4 ± 2 %) (Fig. 2.5 A and B). In white shelters, all substrates showed a low use percentage
close to 1% (all p > 0.05) (Fig. 2.5B).

	

29	

Figure 2.5. Percentage of cavities of each substrate used by O. cornifrons in each shelter type, A)
wooden, B) blue, and C) white shelters. Natural reeds and 5 mm cardboard tubes are the
substrates with the higher proportion of use. Bars within a graph with the same letter are not
significantly different, p > 0.05.

	

30	

When splitting groups by substrate, O. cornifrons used 8.6% more reeds when they were
in wooden shelters (reeds-wood: 13.4 ± 3.8 %) than when they are in blue shelters (reeds-blue:
4.8 ± 1.6 %) (all p < 0.05) (Fig. 2.6 F). The percentage of cavities used for the rest of the
substrates did not change among the shelter types (all p > 0.05) (Fig. 2.6 A-E).

Figure 2.6. Percentage of cavities in each shelters type used by O. cornifrons for the different
substrates. A) 5 mm tubes, B) 6 mm tubes, C) 7 mm tubes, D) 8 mm tubes, E) wood block, F)
natural reeds. Bars within a graph with the same letter are not significantly different, p > 0.05.
	

31	

For O. lignaria, there was no significant effect of the shelter type (F(2, 8) = 0.57, p =
0.587), a significant effect of the substrate materials (F(3, 8) = 1.94, p = 0.202) and no significant
interaction between the two factors (F(1, 8) = 0.63, p = 0.452). Natural reeds were the most used
substrate (8.5 ± 1.1 %) (Tukey test, p > 0.05), with few individuals nesting in the other options
available (Fig. 2.7).

Figure 2.7. Percentage of cavities of each substrate used by O. lignaria. Natural reeds is the
substrate with the higher proportion of use. Bars with the same letter are not significantly
different, p > 0.05.

For the number of cells per nest for O. cornifrons, there was no significant effect of the
shelter type (F(2, 183) = 0.698, p = 0.498), a significant effect of substrate (F(3, 183) = 5.964, p =
0.0006) and no significant effect of the interaction between the two factors (F(2, 183) = 0.559, p =
0.572). Among the substrates include in the analysis, 7 mm cardboard tubes and natural reeds
	

32	

had higher number of cells per nest (6.0 ± 1.7 individuals/nest and 6.0 ± 0.2 individuals/nest,
respectively) than the other sizes (all p < 0.05) (Fig. 2.8A). Shelter type did not significantly
affect the number of females produced per nest for O. cornifrons (F(2, 183) = 0.485, p = 0.616),
there was a significant effect of substrate (F(3, 183) = 8.096, p < 0.0001) but no significant
interaction between the two factors (F(2, 183) = 1.735, p = 0.179). On average, one female more
per nest was produced in the 7 mm tubes (2.0 ± 0.1 females/nest) and reeds (3.3 ± 0.4
females/nest) than in the 5 mm (1.3 ± 0.6 females/nest) and 6 mm tubes (1.8 ± 0.5 females/nest)
(Fig. 2.9A).

Figure 2.8. Number of cells build per nest in the most used substrates for A) O. cornifrons, and
B) O. lignaria. Bars with the same letter are not significantly different, p > 0.05.

Osmia lignaria had a similar number of cells per nest in both natural reeds (7.4 ± 0.2
cells/nest) and wood blocks (8.6 ± 1.0 cells/nest) (Fig. 2.8B), with no significant effect of the
shelter type (F(2, 65) = 0.533, p = 0.589), or substrate material (F(1, 65) = 0.429, p = 0.514). I was
unable to include the interaction in the model since almost all nesting occurred in the natural

	

33	

reeds. Regarding the number of females produced per nest for O. lignaria, there was no
significant effect of the shelter (F(2, 65) = 0.187, p = 0.829), no significant effect of substrate (F(1,
65)

= 1.148, p = 0.287) and was not possible to include the interaction in the model since almost

all nests where obtained in the natural reeds, obtaining 3.1 ± 0.3 females/nest for natural reeds
and 2.0 ± 0.8 females/nest for wooden blocks (Fig. 2.9B).

Figure 2.9. Number of females produced per nest in the most used substrates for A) O.
cornifrons, and B) O. lignaria. Bars with the same letter are not significantly different, p > 0.05.

Attractant spray experiment
In this experiment, 86% of the nests obtained belonged to O. cornifrons, and 14% to O. lignaria,
similar to the experiment reported above. I found on average 17 more O. cornifrons nests in
locations where I released this species, but in this case the differences where not significant (X2 1
= 1, p = 0.317) (Fig. 2.10A). For O. lignaria the total number of nests was similar in locations

	

34	

where it was released or not (X2 1 = 1, p = 0.317) (Fig. 2.10B). The sex ratio for O. cornifrons
was 1.7 males per female, while for O. lignaria it was 1.0 male per female.

Figure 2.10. Total number of nest obtained in locations where the target species was released or
not in the attractant spray experiments at Clarksville, MI. A) O. cornifrons. B) O. lignaria. No
significant difference was found between treatments.

For O. cornifrons, I obtained six-fold more nests in the shelters treated with the attractant
lure (15.0 ± 4.9 nests/shelter) than in the untreated shelters (2.0 ± 1.1 nests/shelter), (W = 2.0, p
= 0.035) (Fig. 2.11A). There were no significant differences between treatments in the number of
cells built per nest (spray: 7.1 ± 0.2 cells/nest; no spray: 8.4 ± 0.7 cells/nest) (W = 172, p =
0.193) or the number of females produced per nest (spray: 2.2 ± 0.2 females/nest; no spray: 2.1 ±
0.4 females/nest) (W = 259.5, p = 0.83) (Fig. 2.12).

	

35	

Figure 2.11. Number of nest build per shelter when the nesting materials are sprayed or not with
decanoic acid. A) O. cornifrons and B) O. lignaria. Bars with the same letter are not significantly
different, p > 0.05.

For O. lignaria there were no significant differences in the number of nests obtained
between treatments, although there was a small increase in the number of nests per shelter when
the attractant was used (spray: 3.3 ± 2.8 nest/shelter; no spray: 0.1 ± 0.0 nest/shelter) (X2 1 = 1.5,
p = 0.221) (Fig. 2.11B). It was not possible to compare the number of cells and females per nest
given that only one nest in the no spray treatment was obtained.

	

36	

Figure 2.12. A) Number of cells per nest and B) number of females per nest build by O.
cornifrons when the nesting materials are sprayed or not with decanoic acid. Bars with the same
letter are not significantly different, p > 0.05.

DISCUSSION
Different nesting shelters and substrates can be more attractive than others for Osmia bees, and
can influence the number of cells made and the number of females produced per nest. For O.
cornifrons, when given a choice among different materials, the shelters and substrates with the
highest use percentage were blue and wooden shelters with natural reeds and 5-6 mm cardboard
tubes. In the case of O. lignaria, bees nested almost exclusively in natural reeds, regardless of the
shelter type.
Earlier studies found that the percentage of cavities utilized is between 10-20 percent
when bees are mass released in apple orchards (Torchio 1982a, 1982b, Yoon et al. 2015). This is
similar to what I found in the wooden shelter/natural reeds treatment. Torchio (1982b) found that
O. lignaria nest more in wooden materials than in milk cartons, which also coincides with
observations in this experiment. Bosse et al. (2014) also report that O. bicornis (L.) prefer to nest

	

37	

in reeds rather than in paper tubes. In this study I also found the typical male-biased sex ratio that
is usually reported for Osmia species (Torchio and Tepedino 1980), as well as the typical range
for the number of cells per nest (1-10 cells/nest) (Bosch and Kemp 2002, Mader et al. 2010).
Initial studies into Osmia nesting biology focused on how attractive different nesting
materials are for bees (Torchio 1982a, 1982b, 1984, Bosch 1994b), followed by studies
regarding the influence of different length and diameter of the cavities on bees preferences (Rust
1998). More recently, studies have addressed the effect of shelter color (Artz et al. 2014) as well
as the number of shelters and cavity density for improved bee retention and reproductive success
(Artz et al. 2013). However, few studies have addressed the interaction between shelter
color/material and substrate type. In that sense, this study provides a new perspective for the
strategies that can be used to manage Osmia bees in orchards. My research supports the
conclusion that wooden substrates are the most attractive nesting materials for mason bees
(Mader et al. 2010, MacIvor 2016). However, cardboard tubes are easier to process to obtain
cocoons, it is possible to reuse them several times, and they are less expensive than wood
materials or natural reeds, so they could be more suitable for mass productions of mason bees.
Other studies have also shown a sufficient nesting success when using 7 mm diameter cardboard
tubes (Artz et al. 2013, 2014). Another point to consider is that, because wooded blocks were in
the base of the box, they had a higher chance of remaining wet after rain, possibly decreasing the
attractiveness of this substrate, something that was also observed by Artz et al. (2013).
Some authors report that the color and patterns of shelters are important because it can be
used by bees as a visual cue (Fauria and Campan 1998, Biddinger et al. 2008, Mader et al. 2010).
Artz et al. (2014) found more O. lignaria nests and more females per nest in blue plastic boxes
than in orange or yellow boxes. I also found differences in the rate of substrate use for the

	

38	

different shelter types for O. cornifrons, particularly for natural reeds, for which wooden shelters
yield a higher use proportion. It is possible that the combination of wood materials makes this
arrangement more attractive to bees because it resembles their natural nesting environment. In
the case of O. lignaria, I did not find a significant effect of the shelter type, but given the
relatively small percentage of O. lignaria nests obtained, it is possible that the results observed
could change if there was a larger number of females nesting in the trap nests. Other factors that
can affect Osmia nesting success include the cavity density per shelter, as O. lignaria has been
shown to prefer lower densities of tubes per box (Artz et al. 2013, 2014), the high dispersion rate
when bees are mass released, the distance of the nests from the release site, and the response to
weather conditions (Torchio 1982a). Additionally, interactions between Osmia species such as
competition for nesting sites remains an under-studied aspect of their biology.
In this study, the number of cells or females per nest of O. cornifrons did not vary among
shelter types, however, bees showed a strong preference for 7 mm cardboard tubes and natural
reeds, with higher numbers of completed cells and a greater investment in females per nest in
these nest resources. Artz et al. (2014) found that more O. lignaria females are produced per nest
tube in blue plastic shelters than in orange or yellow shelters. However, in this case I did not find
differences among shelter type, perhaps due to the low numbers of nests of O. lignaria obtained.
Several authors report that nest cavities with greater diameters can promote more females per
nest (Levin 1966), because for many Osmia species, females are bigger than males, so they are
mainly produced in cavities of greater diameter. There is an upper limit to this, however, when
the tube becomes too wide to be accepted (» 12 mm) (Bosch and Kemp 2001). Given that in this
experiment, Osmia bees mainly nested in natural reeds, few other substrates could be included in
the analysis, but a similar pattern was observed for O. cornifrons, in which the substrates with

	

39	

the wider diameter (reeds 8-10 mm) had more cells and more females per nest. The type of
shelter can also affect the number of cells per nest for O. lignaria (Torchio 1982a, 1982b).
Several authors have found evidence that Osmia species rely on volatiles for nest location
and recognition. Some studies showed that O. lignaria females use chemical cues produced by
the Dufour's gland to mark and recognize their nests (Guedot et al. 2006, Pitts-Singer et al.
2012). Osmia lignaria is also more attracted to old previously used cavities than new ones
(Torchio, 1984a), apparently driven by volatiles emitted by the old cocoons (Pitts-Singer 2007).
The nesting-site selection behavior indicates that females tend to come back to the location
where they emerge, and then scout several cavities, entering and exiting them, until a cavity is
selected, with old cavities showing a higher probability of been selected (Levin 1966, Torchio
1984). Although not clear yet, this behavior suggests that females use the volatiles from the old
cocoons as a local cue to come back to the location where they originally emerged, and then
select a cavity from the options available. This suggests that the decanoic acid is not functioning
as an attractant, but rather as an arrestant and identifier of nest tube locations, to increase the
acceptance of holes for nesting by females. There is no direct evidence of this being a long-range
attractant that would attract bees from downwind to orient towards a nesting location.
In this study I found that decanoic acid can increase the number of O. cornifrons females
nesting in a location, as reported by the patent (Pitts-Singer et al. 2016), making this
semiochemical a good option to increase retention of this species when mass released. The fact
that decanoic acid did not affect the number of cells or the number of females per nest indicates
that this molecule does not influence the investment that females make in their nests. In the case
of O. lignaria, it is possible that bees did not nest much in any of the treatments because the
nesting substrate offered was cardboard tubes, which based on the shelter/substrate experiment

	

40	

described above, is not a preferred substrate. It is also possible that as this molecule was
identified in cocoons of the western subspecies (O. lignaria propinqua), it could be less
attractive for Michigan populations that belong to the eastern subspecies (O. lignaria lignaria).
Although Pitts-Singer et al. (2016) report that the attractant is efficient for several Osmia species,
further evaluations are needed to determine the range of species with which this attractant will be
effective. It is also important to test whether this or future lures can also attract bee natural
enemies, as Glasser and Farzan (2016) showed that Melittobia acasta (Hymenoptera:
Eulophidae), an Osmia parasitoid, can be attracted to O. lignaria propinqua cocoons.
My results highlight the potential of the two Osmia species to be used as alternative
pollinators in Michigan fruit crops. The higher success rate of O. cornifrons and its better
response to the attractant spray suggests that this bee is easier to manage than O. lignaria.
However, the use of non-native species for pollination should be carefully considered, given that
in many cases they transmit diseases and parasites (Goulson and Hughes 2015), compete with
native species for resources (Thomson 2004, 2006), and can pollinate invasive species (Hanley
and Goulson 2003). Further, many studies have shown that O. lignaria, a native species, is a
promising species and it is possible to manage them in commercial settings. This reinforces the
importance of continuing to explore strategies to improve the retention and reproductive success
of mason bees in agricultural landscapes. The knowledge about preferred trap nest for bees can
be also used to increase populations of Osmia spp. in natural areas with conservation goals, or to
increase natural populations that support pollination service (Stubbs et al. 1997).

	

41	

CHAPTER 3. SOIL USE BY O. lignaria and O. cornifrons (HYMENOPTERA:
MEGACHILIDAE) IN AN AGRICULTURAL ENVIRONMENT

INTRODUCTION
Most bee species (Hymenoptera: Apoidea) build their nests underground, excavating tunnels and
building cells where they deposit pollen and nectar to feed their larvae (Roubik 1989, Michener
2007). However, many bee species in the families Apidae and Megachilidae nest above-ground
(Linsley 1958, Krombein 1967). Nests can be built in cavities between rocks (e.g. Euglossini),
preexisting cavities in wood (e.g. Megachilini, Osmiini), excavating tunnels in decaying wood
(e.g. Xylocopini), in abandoned mammal burrows (e.g. Bombini), or on trees (e.g. Apini,
Meliponini) (Linsley 1958, Roubik 1989, Torchio 1989, Nates-Parra and Gonzalez 2000, NatesParra 2001, Gonzalez et al. 2004, 2009, Marchi and Melo 2010). Studies on the nesting
preferences of solitary bees have been relatively scarce (Levin 1966, Torchio 1982a, 1982b,
Cane 1991, Wuellner 1999, Cane et al. 2007). In recent years more attention has been given to
this topic, due to increasing interest in having alternative pollinators for crops, including studies
of ground nesting and stem nesting bees (Klein et al. 2007, Christmann and Aw-Hassan 2012,
Sedivy and Dorn 2014).
Bees in the genus Osmia (Megachilidae) are univoltine (one generation per year),
solitary, and cavity-nesting (Levin 1966). In the wild, mason bees nest mainly in soil or existing
cavities made by wood-boring insects, but they can also nest in man-made structures like wood
blocks with drilled holes, cardboard straws, hollow reeds or bamboo (Levin 1966, Bosch and
Kemp 2002, Cane et al. 2007). Most species use a mix of plant materials and mud to build the
partitions between cells (e.g. Osmia kenoyeri Cockerell, Osmia unca Michener), but some

	

42	

species exclusively use soil to build the partitions (e.g. Osmia lignaria Say and Osmia cornuta
Latreille) (Linsley 1958, Cane et al. 2007).
For the species that use only soil for partitions, it is common to find females congregating
at mud collecting sites where the soil is wet, with several females taking soil from the same
excavated hole (Torchio 1989). Females gather mud by scraping with the mandibles and
forelegs, then shaping it to create a round pellet 2-3 mm in diameter using the forelegs,
mandibles, and the ventral surface of the posterior tip of the abdomen. She then holds the pellet
between her mandibles and flies back to the nest (Levin 1966, Raw 1972). The soil is deposited
as a foundation that circumscribes the inner wall of the nest. Additional soil is added in
concentric circles around the foundation until a solid wall is formed across the nest (Torchio
1989). Females can spend 2 - 4 minutes depositing each mud ball into the nest (Levin 1966). A
total of 8 to 12 mud collecting trips are necessary in order to have enough soil for one cell
partition, requiring about 20 min to build each partition (Torchio 1989, McKinney and Park
2012).
Some of the species that collect exclusively soil as a material for cell partitions are
particularly important for agriculture. Osmia cornifrons (Radoszkowski) has been used as an
managed orchard pollinator in Japan since the 1940s, and was introduced into United States in
1977 for crop pollination (Batra 1978). Osmia lignaria (native to the USA), Osmia bicornis (L.)
and O. cornuta (Latreille) (native to Europe) have been proposed as alternative commercial
pollinators for rosaceous tree crops, given their strong preferences for this group of plants and
their high efficiency at pollinating them (Bosch and Kemp 2000, 2002, Vicens and Bosch 2000b,
Ladurner et al. 2004, Gruber et al. 2011, Sheffield 2014).

	

43	

Most studies on the nesting preferences of stem nesting bees, including Osmia, have
focused on the kind of preferred cavities and shelters, effect of cavy density and release density
(Sampson et al. 2009, Artz et al. 2013, 2014, Yoon et al. 2015). Even though soil is an essential
and limiting resource for many bee species (both ground nesting bees and the ones that use it for
cell partitions) (Cane 1991, Krunic and Stanisavljevic 2006), few studies have investigated what
parameters of the soil (texture, structure, moisture level, organic matter content, etc.) are
important for bees when selecting a nesting site or soil collecting sites. Soil texture, defined as
the percentage of clay (< 2 µm particles), silt (3-50 µm particles) and sand (51-2000 µm
particles) in a soil sample (Foth 1990), is one of several factors that can influence nest site
selection of ground nesting bees and wasps (Krombein 1967, Cane 1991, Scott 2016), so it is
possible that it will also play a role in nest construction by Osmia species.
Both O. lignaria and O. cornifrons are present in Michigan, where they are effective
pollinators of important fruit crops such as apples, cherries and pears (Bosch and Kemp 1999,
Ladurner et al. 2004, Monzón et al. 2004, Matsumoto et al. 2009, Sheffield 2014). Determining
the type of soil these bees use in agricultural landscapes should provide avenues for increasing
their role as alternative pollinators. Understanding the soil that these bees prefer may also
improve our ability to maximize their propagation, by ensuring that soil is not a limiting factor.
Osmia bees are able to make choices over the plants they visit for pollen (Monzón et al. 2004,
Bosch et al. 2006, Sheffield et al. 2008), and the substrates they use for nesting (Torchio 1982a,
1984, Artz et al. 2013, Chapter 2), so I expect that they can also select soil from the options
available to them. However, it is not known how selective they may be in soil selection and
whether they only select from the immediate soil type or will fly further for their preferred type.
Given this, the objectives of this study were to 1) describe the texture of the soil used by O.

	

44	

lignaria and O. cornifrons in an orchard setting and 2) explore whether the soil used by these
Osmia bees is similar to or distinct from the characteristics of the soil surrounding the nest.

METHODS
The experiment was conducted at Trevor Nichols Research Center (TNRC), in Fennville, MI
(USA) (42°35'38.26"N, 86° 9'18.03"W), between May-June 2016. The station has experimental
plots of tree fruit crops, including apple, cherry and plum; annual crops such as corn, and berries
including blueberry and raspberry, as well as patches of woods and grassy areas. There are small
streams and side ditches that can provide wet soil for bees. A surface (0-25 cm) soil texture map
of the station and surrounding areas was generated using the USDA Soil Survey website
(https://websoilsurvey.sc.egov.usda.gov/App/HomePage.htm). According to this map (Fig.
3.1A), and the pesticide application plan of the station, 12 locations were selected that provided a
variety of soil texture conditions and low risk of pesticide exposure. At each location, one bee
shelter was installed. Nesting shelters consisted of pine plywood boxes (26 x 28 x 28 cm) with an
open face. Inside each shelter, I placed 50 bamboo stems, approximately 15 cm long with an
inside diameter of 7-10 mm. Shelters were held 1.5 m above ground by a metal post, and the
open face of the shelter was covered with chicken wire (2.5 cm mesh size) to exclude birds and
rodents (Fig. 3.2).

	

45	

Figure 3.1. Maps of the Trevor Nichols Research Center near Fennville, MI, showing: A) surface
soil (0-25 cm) map generated with the USDA Soil Survey online database in relation to the
locations of shelters for Osmia bees. B) measured soil texture of surface samples (0-10 cm) taken
from the research center. The estimated area available for bee foraging depicted in the figure was
based on a 250 m foraging range from each shelter location. Alternative locations correspond to
Osmia shelters in the stations installed for other experiments, from where nests were taken to
increase the number of soil samples from bees’ nests.

At each shelter, I placed an emergence box containing 50 female and 60 male loose
Osmia cocoons. Female cocoons were 73% O. lignaria and 27% O. cornifrons, while male
cocoons were 88% O. lignaria and 12% O. cornifrons. The emergence box consisted of a

	

46	

cardboard box of 9.3 x 9.3 x 4.3 cm, with a hole on one face to allow bees to leave the box after
emergence. Cocoons were gathered from trap nests located at Michigan State University (MSU)
campus in East Lansing, MI, in September 2015. Nests were opened and cocoons were extracted
and sorted by sex based on size, since female cocoons are bigger (» 11 mm) than male cocoons
(» 7 m) (Torchio and Tepedino 1980). Cocoons were stored in an environmental chamber at 4ºC
from November 2015 until their release. A sub-sample of 50 cocoons was taken from each batch
and emerged in the lab by holding them at laboratory conditions (» 25ºC) to determine the
proportion of each species.

Figure 3.2. Wooden shelters used at the research station, including the emergence box (bottom)
and bamboo stems (top) offered to Osmia bees for nesting.

Soil sampling
To confirm the soil texture reported by the Soil Survey, soil samples were collected at the
research stations and surrounding areas. The foraging range of O. lignaria and O. cornifrons is

	

47	

between 30 to 150 m from the nest (Bosch and Kemp 1999, Monzón et al. 2004, Matsumoto et
al. 2009, Biddinger, Joshi, et al. 2013), and they have been reported to collect soil from locations
54 m away from their nest (Torchio 1989), so an area of 250 m radius was established around
each nesting shelter, to represent the area available for bees to forage (Fig. 3.1). The research
station site was sampled using a total of 62 soil samples taken in and around the station,
including locations in each of the areas with different texture reported by the Soil Survey. At
each of the locations selected for soil samples, five subsamples from the upper 10 cm layer were
collected in a 2 m radius from the selected point. The subsamples were then mixed to complete
approximately 500 g of soil per sampling point.
To obtain soil samples from bee nests, bamboo tubes with a mud plug from each shelter
where bees nested were brought back from the field after the nesting period was complete
(approximately June 10, 2016). Nests were kept at the MSU Entomology Research Center
(Lansing, MI) at outside ambient temperature until September 2016, when bees have finished
their development into adults. The bamboo tubes were then opened and all soil partitions present
in the nest were collected from 24 O. cornifrons and 22 O. lignaria nests. Partitions of the same
nest were mixed and homogenized. Thirteen of the O. cornifrons nests and four of the O.
lignaria nests came from alternative nest shelters placed in the station for other experiments, but
that were close to the shelters used in this experiment (zone 2 and 3) (Fig. 3.1B). Additional
tubes were opened to collect individual partitions from eight O. cornifrons nests (total 67
partitions) and six O. lignaria nests (total 44 partitions). In this case, each partition was kept
apart from each other, and the order in which the partitions were in the nest was noted. To
confirm the species that built each nest, one cocoon per nest was opened and the bee inside was

	

48	

placed in a jar with cyanide to kill the bee, which was then pinned and identified using the key
by Arduser (2009).

Particle size analysis
A particle size (texture) test was use to analyze soil samples. Soil samples from TNRC were
analyzed using the hydrometer method at the Soil and Plant Nutrient Laboratory at the
Department of Plant, Soil and Microbial Sciences of MSU. Samples were processed following
the standard laboratory protocols. Briefly, 100 g of soil per sample was oven dried for 12 h, then
100 ml of 5 % sodium metaphosphate (NaO3P) solution were added to disperse particles, that
were then placed in a shaker overnight. The percentage of sand, silt and clay was then
determined using the standard hydrometer methodology (Foth 1990).
Texture analysis of soil samples from Osmia nest and individual partitions was performed
using the laser diffractometry technique at the Department of Geography, Environmental and
Spatial Sciences of MSU. This technique allows texture analyzes of samples as small as 0.5 g,
and the results are directly comparable to the results obtained by the hydrometer method (Miller
and Schaetzl 2012). Samples were gently ground and dispersed by adding 2 ml of 5 % sodium
hexametaphosphate ((NaPO3)13Na2O) solution and shaken for 10 min. Samples were then run
using a Malvern Mastersizer 2000E laser particle size analyzer with a Hydro 2000MU pump
accessory (Malvern Instruments Ltd., Worcestershire, UK), following the laboratory protocols.
This provided the particle size analysis of the soil samples in terms of percentage of clay, silt and
sand.

	

49	

Partition color analysis
Soil partitions from 40 O. cornifrons nests and 20 O. lignaria nest were sorted by color. Pictures
of partitions with distinctive color patterns were taken using a Leica S8AP0 microscope with a
MC120 HD camera attached. Partitions of the same color were ground and mixed together.
Moistened samples were compared to a Munsell soil color chart to determine the color (hue,
value and chroma). A subsample from each color was run for texture analysis using the
diffractometry technique described previously.

Data analysis
A principal component analysis (PCA) with R version 3.1.1 (R Development Core Team,
Vienna, Austria) was used to explore patterns in the composition of the nest partitions for the
two species of Osmia, in relation to sand, silt and clay content. A generalized linear mixed model
(GLMM) with beta distribution and logit link function was used. Nest tube was included as a
random effect to take account of partitions from the same tube being more similar given than
they were collected by the same females. For the GLMM, I used PROC GLIMMIX in SAS
version 9.4 (SAS Institute 2016) to compare between species, and among cells and soil zones.
For soil samples from the whole nest, each texture component (sand, silt and clay) was
individually analyzed for differences between the two Osmia species, among the four soil zones,
and the interaction effect. Soil from individual partitions was analyzed separately for each
species, to determine whether the partition’s position (as a ordinal variable) in the nest or the soil
zone affected the soil composition, and whether there was an interaction between these factors. If
these factors were significant, a Tukey test was used to separate the means (p < 0.05).

	

50	

RESULTS
The soil samples taken in and around the research station showed different soil textures than the
map generated with the online Soil Survey database (Fig. 3.1B). Only three of the six textures
reported by the soil survey were actually present (loam, loamy sand and sandy loam), but two
additional textures were found (sandy clay loam and clay loam). Based on the measured soil
texture of the site, locations of the Osmia shelters were classified into four zones, depending on
the soil texture near the nest location (Fig. 3.1B). In zone 2, which had mainly sandy clay loam
texture, there were not enough O. lignaria nests for individual partition analysis, so only the
whole nest analysis was conducted. Consequently, the individual partition analysis includes
samples only from three zones, where zone 1 was mainly sandy loam, zone 3 was mainly sandy
loam and loamy sand, and zone 4 that was mainly loamy sand.

	

51	

Figure 3.3. Texture of soil samples from A) the whole nest and B) from individual partitions for
O. cornifrons (red) and O. lignaria (blue). The clay content in O. cornifrons samples was
significantly higher than in O. lignaria samples (p < 0.05).

	

52	

When analyzing soil samples from complete nests, most of the samples of both species
were classified as loam or sandy loam (Fig. 3.3A). The PCA shows a clear separation between
the two species (Fig 3.4), with the analysis highlighting variation between the species based on
only sand in the X axis, and clay in the Y axis. Silt composition was the least variable component
of the soil and was not included in the PCA plot. These differences were mainly represented by
an average 5.8% higher clay content in O. cornifrons samples (F(1, 38) = 12.77, p = 0.001) (Table
3.1), without an effect of the location of the nests (F(3, 38) = 1.94, p = 0.139) or an interaction
between factors (F(3, 38) = 1.04, p = 0.385). Silt and sand content in samples of both species were
similar (all p > 0.05). Clay content in O. cornifrons samples had a slightly wider range (6.1 20.9 %) than O. lignaria (2.1 – 15.2 %), with the opposite for the silt and sand content, for which
O. lignaria showed a wider range than O. cornifrons (Table 3.1).

	

53	

Figure 3.4. Biplot of the principal component analysis for the soil texture of whole nest samples.
PC1 (sand) and PC2 (clay) are shown. Each sample is represented by one dot and the color label
correspond to the species, O. cornifrons (red) and O. lignaria (blue).
	
Table 3.1. Soil texture (range and average ± SE) of complete nests soil samples collected by O.
cornifrons and O. lignaria. Means with the same letters within the same column are not significantly
different (p > 0.05).
Range

O.
cornifrons
O.
lignaria

	

Average

n

Clay (%)

Silt (%)

Sand (%)

Clay (%)

Silt (%)

Sand (%)

24

6.1-20.9

23.5-45.1

36.7-69.9

14.1±0.9 a

34.1±1.2 a

51.9±1.7 a

22

2.1-15.2

15.5-49.3

40.5-82.5

8.3±0.8 b

37.1±2.1 a

54.5±2.1 a

54	

Table 3.2. Soil texture (range and average ± SE) of samples from individual cell partitions collected
by O. cornifrons and O. lignaria
Range

O.
cornifrons
O.
lignaria

Average

n

Clay (%)

Silt (%)

Sand (%)

Clay (%)

Silt (%)

Sand (%)

67

2.1-34.2

20.3-76.3

3.6-77.7

12.4 ± 0.9

39.4 ± 1.4

48.2 ± 2.1

44

1.5-27.8

19.6-86.5

0.0-78.8

8.9 ± 0.8

46.4 ± 2.9

44.7 ± 2.0

When analyzing individual partitions of both species, the range of all particle components
was wider than the whole nest samples (Table 3.2). Similar to the whole nest samples, the clay
content range was wider for O. cornifrons, but silt and sand content ranges were wider for O.
lignaria (Table 3.2). Due to this high variation, partitions were categorized as loamy sand, sandy
loam, loam, silt loam or silty clay loam depending on the specific sample (Fig. 3.3B), although
most of the samples were in the range of sandy loam and loam. It was also observed for both
species that when considering partitions from the same nest, sometimes all the partitions were
grouped and had a common texture, while in other nests each partition had different textures,
varying mainly in sand content (Fig. 3.5).

	

55	

Figure 3.5. Texture of soil samples from individual partitions of the nest of A) O. cornifrons and
B) O. lignaria. Dots of the same color correspond to partitions taken from the same nest.
	
The PCA suggests similar soil composition for partitions in different positions in the
bamboo nests (Fig. 3.6), and for O. cornifrons, neither the zone of the nest or the position of the
partition in the nest (Fig. 3.7) had a significant effect on any of the texture components (clay:

	

56	

zone F(3, 54) = 0.86, p = 0.46, cell position F(1, 54) = 0.09, p = 0.76; silt: zone F(3, 54) = 2.49, p =
0.070, cell position F(3, 54) = 0.23, p = 0.63; sand: zone F(3, 54) = 2.01, p = 0.12, cell position F(1, 54)
= 0.05, p = 0.82). However, there was a marginally significant effect of the zone of the nest on
the silt content of the partitions (F(3, 54) = 2.49, p = 0.0704), with partitions of nests from zone 4
(mainly loamy sand) showing a 10 % lower silt content than those in zone 1 (mainly loam) (Fig.
3.8). Also, there was a significant interaction between factors (F(3, 54) = 3.23, p = 0.029) for the
clay content of the partitions. This significant interaction without a significant effect of the main
factors was due to finding differences in the clay content between partitions only in zone 2 (T(54)
= -3.21, p = 0.0023) (Fig. 3.9). For O. lignaria, there were no effects of the positions of the
partition in the nest, the zone of the nest or an interaction between factors on any of the texture
components (all p > 0.05).

Figure 3.6. Biplot of the principal component analysis for the soil texture from individual
partitions for A) O. cornifrons and B) O. lignaria. PC1 (sand) and PC2 (clay) are shown. Each
sample is represented by one dot and the color label (cell) corresponds to the order of the
partitions in the nest, with 0 being the most inner partition.
	

57	

Figure 3.7. Mean ± SE of sand, silt and clay content of soil partitions for nests of A) O.
cornifrons and B) O. lignaria.

Figure 3.8. Mean ± SE percentage of clay, silt and sand in individual partitions from O.
cornifrons nests from different zones in the research station. Silt content in partitions from zone
4 showed a trend towards higher particle size (p = 0.07).
	

58	

Figure 3.9. Percentage of clay (mean ± SE) in O. cornifrons partitions at the different zones in
TNRC. Partitions showed significant differences in clay content (p < 0.05) only in zone 2.
	
Partitions from nests were sorted in six different colors, plus a “mix” category, when a
partition showed more than one color (Fig. 3.10). Some of the colors, even when classified as
different colors at first, had the same hue, value and chroma values (Color 1 and 4), but they had
different texture. In other cases, they showed different hue, value and chroma values (Color 3
and 5), but they showed the same texture category (Table 3.3). Additionally, mixed color
partitions, which account for approximately 8 % of the partitions for both species (Fig 3.11),
presented a highly variable pattern of colors, from spirals and circular patterns that follow the
construction configuration of the partition, to spotted color patterns (Fig. 3.10).

	
	

	

59	

Figure 3.10. Different color patterns observed on the cell partitions of O. cornifrons and O.
lignaria. Patterns can vary from spotted (A-C) to spiral or concentric (E-H). Soil of different
colors suggest that they were collected at different location or at different depths in the same
location.
	

60	

Table 3.3. Color and texture of the different partition’s color identified in O. cornifrons and
O. lignaria nests.
Color

Texture

hue

Value

Chroma

Clay (%)

Silt (%)

Sand (%)

Category

Color 1

2.5y

4

4

13.5

33.6

52.8

Sandy Loam

Color 2

2.5y

3

3

8.6

37.4

54.1

Sandy Loam

Color 3

10yr

4

6

12.4

38.2

49.4

Loam

Color 4

2.5y

3

3

9.4

38.6

52.0

Sandy Loam

Color 5

2.5y

4

4

13.9

36.6

49.5

Loam

Color 6

2.5y

3

2

10.7

40.4

48.9

Loam

Mix

2.5y

4

4

11.4

40.2

48.4

Loam

Figure 3.11. Percentage of partition of each color identified in A) O. cornifrons and B) O.
lignaria nests
	

	

61	

DISCUSSION
This study revealed variability with some species-specific differences in composition of soil use
by a native and a non-native Osmia bee species. To the best of my knowledge, this is the first
report of the texture of the soil collected by O. cornifrons and O. lignaria, and it sheds light on
the soil preferences and ecological adaptations of these species. It also has to be stated that there
could be Japanese literature regarding O. cornifrons that is not accessible to me. The few reports
of soil texture preferred by mason bees indicate that they prefer soil that is soft and damp with a
high clay content (Rau and Mo 1937, Medler 1967, Mader et al. 2010). However, these reports
are usually based on anecdotal observations and there are no descriptions of the particle size
composition. Although other soil factors not measured in this experiment, such as organic matter
content or moisture content may influence soil selection by mason bees, soil texture is probably
one of the most important variables driving Osmia soil selection, given that partitions and
entrance plugs must be resistant enough to prevent parasites and predators to enter the nest, and
soil texture has been found to be an important factor for bees nest site selection by several
ground-nesting bee species (Cane 1991, Potts and Willmer 1997).
Using a unique application of laser diffractometry employed in the geological sciences, I
found wide variation in the soil texture that bees bring back to the nest, indicating that they are
flexible in the type of soil used to make nest partitions. The sand content in partitions ranged
from 0.0 to 78%, but the clay content was never higher than 34%, with an average between 810%, so most of the samples can be classified as loam or sandy loam. Although these results
contradict what was previously thought about mason bee soil preferences, it has to be considered
that the soil in the area where bees were nesting was in the sandy spectrum, with a few locations
presenting clayey texture. Given this, my results only indicate that O. cornifrons and O. lignaria

	

62	

are able to use sandy soil to build the nest partitions. Experiments where bees have access to soil
of several textures are needed to explore their soil preferences. Cage choice experiments where
soil of different textures is offered to females to evaluate their preferences and constancy of soil
use would give a better idea of their soil preferences. Cage experiments where only one type of
soil is offered could help to elucidate the level of plasticity exhibited by these bees at the time of
using soil of a specific texture. This type of experiment also would allow a more detailed
manipulation of the soil offered.
In this study, I found that O. lignaria collect soil with a clay content 5.8% lower than O.
cornifrons. Given than O. cornifrons is an introduced species from Japan, it may have evolved in
places with sandier soil types than the ancestral home of O. lignaria. But the fact that it has been
successful at establishing in North America (Biddinger et al. 2009), suggests that this species has
sufficient plasticity to use different soils than the ones found in their native range. On the other
hand, O. lignaria is a species that is widely distributed in North America (Rust 1974), also
indicating that it has to be able to use a wide range of soil textures, given that the available soil
resources change between geographical areas. One possible explanation for the average low clay
content in the soil collected by bees is that soils with high clay content shrink when dry
(Schoeneberger et al. 2012), so if bees build partitions with wet clayey soil, it could crack when
dry, opening nests to parasitoids and predators. This soil may also be more challenging to collect
and/or mold for mason bees, a question that could be answered through manipulative
experiments. The fact that soil of relatively large particle sizes (sandy loam and loam) is used by
these bees, suggests that females use a sticky substance to help keep soil particles together. If
present, this substance is probably secreted by the salivary glands and present in bee’ saliva,
given that many other bees secrete substances to waterproof the nest or keep together building

	

63	

materials, and even nectar could also be used as a sticky substance (Michener 2007). This mix of
large particles in the mud and a substance that sticks them together could make partitions harder
than if they were made with mud of a smaller particle size. Harder partitions could more
efficiently prevent the entrance of parasites or natural enemies to the nest.
For both species of Osmia studied in this research, the range of soil textures used
increased when considering individual partitions rather than the whole nest samples. This
indicates that even when the general soil collected by both species is loam or sandy loam, they
can collect soil from areas with a wide range in sand content. Interestingly, for both species, I
found that partitions from the same nest can all have a common texture or a large variation in
texture. This seems to indicate that some females collect soil from locations with different
textures to build each partition, while other females collect soil of similar textures (from the
same or different locations) to make all the partitions in the nest. This observation seems to
indicate that within a population there are differences in the preferences and consistency of soil
selection between individuals. However, a larger sample size is needed to determine whether this
variability is common and to evaluate the role of available soil on the level of variation
individual bees show. A hypothesis that could be tested is that if more preferred soil is offered,
the level of variation among individuals will decrease.
Despite the variation among individual partitions, the percentage of the different texture
components was relatively constant among zones and along the nesting period. The similar soils
collected in different zones could be due to several factors. First, the soil in and around the
research station turned out to have much less texture variability than expected based on the Soil
Survey, with most of the soil in the sandy spectrum. So is not surprising that most of the samples
collected by bees were sandy loam or loam. Second, it is possible that bees are flying more than

	

64	

250 m to collect their preferred soil for the nest. The foraging range (for pollen and nectar) of
these bees is between 50-150 m (Bosch and Kemp 1999, Dogterom 1999), but individual flights
can extend beyond 600 m (Zurbuchen et al. 2010), making it possible that bees collected soil
outside of our study area. Future studies should also include a larger sample size regarding the
number of nests from where samples are taken, and a larger study area. Having a detailed and
accurate soil map of the area would also help to determine the degree of correlation between the
type of soil available and the soil that females bring back to their nests.
The effect of the soil in the area also needs further investigation, given that I observed a
weak effect of the soil zone on soil texture that bees collected. Changes in clay content over time
observed in zone 2 could be due to changes in the availability of wet soil in that region, given
that soils of different textures vary in their water retention capacity. If a water source is not
constant across this site (i.e. rain, rivers, lakes, crop irrigation), soils that retain water for longer
periods (silty loam or clayey soils) would be available to be used by bees for longer periods than
soil with lower water-holding capacity (dry, sandy soil). This highlights the importance of taking
the hydrology and climate of a zone into account when considering mason bees as pollinators,
since a lack of water sources could make it difficult for bees to use soil, even if adequate textures
are in the area.
To build one partition, Osmia bees need to make 8-12 mud collecting trips (Torchio
1989), meaning that they can potentially collect soil from multiple locations for constructing
each partition. The differences in soil color that I observed in single partitions may be due to
differences in organic matter content or the presence of some elements that alter the soil color
(Schoeneberger et al. 2012). The presence of areas of different colors in an individual partition
suggests that bees collected soil from more than one location to build one single partition, or that

	

65	

they dig into different layers of soil in the same place as they collect sequential soil loads. The
relationship between color and texture of the samples is not clear, as two samples of the same
color can have different textures, so it is also possible that partitions of a uniform color are made
of soil collected at different locations. This makes it difficult to establish correlations between
the soil in an area and the soil found in individual partitions or nests, given that partitions are
created by a mix of soil from different locations, and in the same way, soil samples from the
whole nest are a mix of several partitions. On the other hand, if females coming back to the nest
with mud balls were captured, these samples would be too small to run a texture analysis. This
makes it necessary to develop a different approach that could give information on where bees are
collecting soil, such as tracking bees to the mud collecting sites or analyzing the soil from
locations where females have been observed collecting soil.
Despite these remaining gaps in our understanding of soil use by Osmia bees, data from
the present study provide an important first step for understanding the soil use and preferences of
these insects, showing that they use soil from a broader range of textures than previously
thought. I also found that Osmia species and individuals vary in the type of soil collected, with
females even bringing mud from different locations to build one single partition. Future studies
of soil collected by Osmia should also take into account variables such as organic matter content,
moisture content and bulk density. Future studies could also address differences in soil
preferences between different populations and subspecies of O. lignaria, given that phenological
differences have been already shown between the subspecies (Pitts-Singer et al. 2014), and this
information will be relevant when planning to use these bees as orchard pollinators in regions
with different soils.

	

66	

CHAPTER 4. POLLEN USE BY Osmia lignaria (HYMENOPTERA: MEGACHILIDAE)
IN HIGHBUSH BLUEBERRY

INTRODUCTION
Bees in the genus Osmia (Megachilidae) are an important group of pollinators with potential for
pollinating fruit crops (Bosch and Kemp 1999, Monzón et al. 2004, Matsumoto et al. 2009).
These univoltine, solitary, cavity-nesting bees are commonly known as mason bees because
some species use mud to make the partition between cells (Bosch and Kemp 2000). Osmia
lignaria Say is a species native to North America that emerges early in the spring and has a
strong preference for plants in the family Rosaceae, including crops such as apple, cherry, pear
and plum (Bosch and Kemp 1999, Ladurner et al. 2004, Monzón et al. 2004, Matsumoto et al.
2009, Mader et al. 2010, Sheffield 2014). They are considered more efficient pollinators than
honey bees for this group of plants because they mainly forage for pollen, constantly make
contact with the reproductive parts of the flowers, and tend to gather pollen from at least two
cultivars during the foraging trips, promoting cross pollination (Bosch and Kemp 2001,
Matsumoto et al. 2009, Matsumoto and Maejima 2010).
In the United States, O. lignaria is commercially available in some areas (Bosch and
Kemp 2000, 2002, Mader et al. 2010), and its emergence can be manipulated by controlling
temperature and keeping dormant cocoons in cold storage until the target crop bloom or by
breaking the dormancy period by increasing temperature (Mader et al. 2010). Due to this, O.
lignaria has been successfully used as a pollinator of other crops whose bloom period do not
match with the natural flying period of this bee, such as almonds (Artz et al. 2013, 2014).

	

67	

Highbush blueberry (Vaccinium corymbosum) is an economically important crop that
requires insect pollination due to its poricidal anthers, dense pollen, and floral morphology that
promotes cross-pollination. Additionally, some varieties are not self-compatible or show a
significant increase in fruit quality when cross-pollinated (McGregor 1976, Ritzinger and Lyrene
1999). Bumblebees (Bombus spp.) are recognized as some of the best pollinators of this plant
given their ability to “buzz-pollinate” the flowers (Stubbs and Drummond 1997, Brevis 2001,
Mackenzie 2009, Scott 2016). Osmia bees can also visit and pollinate blueberries (Dogterom
1999, Sampson and Cane 2000, Cutler et al. 2015). Osmia lignaria has been found in 55% of 80
blueberry sites sampled in British Columbia, Canada, (Dogterom, 1999), and nine Osmia species
have been reported to visit blueberries in Maine, but with low relative abundance (Bushmann
and Drummond, 2015). In southwest Michigan, Osmia species can be observed in blueberry
fields during the pre-bloom and bloom periods (Tuell et al. 2009), suggesting that these bees
have potential for pollination of blueberries.
Several species of Osmia have been tested in controlled experiments (greenhouse or
cage) for their ability to pollinate blueberry, and positive results have been obtained for O.
lignaria, O. atriventris Cresson, O. cornifrons Radoszkowski and O. ribifloris Cockerell
regarding transferring pollen for different Vaccinium species, in some cases achieving the same
level of pollination as bumblebees (Stubbs and Drummond 1997, Dogterom 1999, Sampson and
Cane 2000, Wilkaniee and Giejdasz 2003, Sampson et al. 2004, West and McCutcheon 2009).
Dogterom (1999) reports that in blueberry fields in British Columbia, 86% of O. lignaria females
captured at the nest entrances had blueberry pollen on their body. However, other reports show
that fruit of rabbiteye blueberry (V. ashei) pollinated by O. ribifloris in southeastern United
States contained fewer seeds than fruits pollinated by honey bees (Sampson and Cane 2000).

	

68	

Sampson et al. (2009) did not recommend O. lignaria as a pollinator of blueberry, because even
if O. lignaria can use its pollen and produce viable progeny when confined on blueberries, this
bee has a strong preference for tree crops. This is supported by many other reports that indicate
that most of the pollen found in Osmia nests located in agricultural environments belong to treecrops, especially rosaceous trees (Vicens and Bosch 2000a, Monzón et al. 2004, Bosch et al.
2006, Sheffield, Westby, Smith, et al. 2008). Also, even if the foraging range is usually between
50-150 m from the nest (Bosch and Kemp 1999, Dogterom 1999), the possible flight range is
close to 600 m (Zurbuchen et al. 2010), highlighting the importance of considering the location
of the nest in the field.
There is growing interest in the use of alternative pollinators among growers who are
experiencing rising costs for honey bee rentals, and increased variability in spring weather.
Given this, it is important to explore pollen use by O. lignaria in field experiments in large areas
of blueberry production to determine whether this bee can be used as an alternative pollinator of
this crop, or whether the promising results described above are relevant only to those specific
environments. To address this question, the objectives of this study were 1) to determine whether
blueberry is a main pollen source for O. lignaria in a commercial field in Michigan through the
bloom season, 2) to determine whether the location of the nest in the field influences the pollen
source used by O. lignaria, and 3) to determine the plant species collected by Osmia in this
setting.

	

69	

METHODS
Study site
This experiment was carried out at the Blueberry Valley farm in Pullman, Michigan (USA)
(42°28'40.42"N 86° 1'38.50"W). The farm has an area of approximately 79 ha, with 63 ha
dedicated to highbush blueberry (V. corymbosum). The varieties grown and the percentage of
area they occupy in the field were Jersey (33.1%), Bluecrop (24.8%), Nelson (14.2%), Elliott (14
%), Rubel (4.4 %), Northland (3.6 %), Blue Jay (3%), Aurora (2.3 %) and Blue Gold (0.6 %).
The farm uses conventional management practices, including insecticide and fungicide
applications. During the bloom period insecticide applications are reduced and only
methoxyfenozide was applied during this study, but fungicide applications continued, including
chlorothalonil, captan and metconazole. Periodic mowing approximately every two weeks in and
around the field was performed by the farmers. However, weeds and wild plants grow in the
drainage channels around the edges of the field. The field is mainly surrounded by deciduous
forest (Figure 4.1).

	

70	

Figure 4.1. Blueberry Valley farm in Pullman Michigan outlined in red, and the locations (IN or
OUT) where Osmia bee shelters were located.

A total of 16 locations in the field were selected for Osmia nests. Eight locations were
positioned at the east or west border of the farm (OUT location), and the remaining eight
locations were inside the farm (IN location) (Fig. 4.1). The inside locations were at least 200 m
away from the border of the field, given that the foraging rang of O. lignaria is estimated to be
between 50-150 m (Bosch and Kemp 1999, Dogterom 1999, Mader et al. 2010). Each location
was a minimum of 100 m away from each other. When logistically feasible, I selected locations
as far as possible (>50 m) from commercial honey bee hives placed within the field.
	

71	

Experimental setup
Nesting shelters consisted of pine plywood boxes (26 x 28 x 28 cm) with an open face. Inside
each box, I placed 25 bamboo stems and 25 natural reeds approximately 15 cm long with an
inside diameter of 7-10 mm (Fig. 4.2). Boxes were held 1.5 m above ground by a metal post, and
the open face of the box was covered with chicken wire to exclude birds and rodents. Shelters at
the border of the field were placed facing into the field (east or west), and shelters in the field
were placed with the opening parallel to the orientation of the plant rows (Fig. 4.2B). Shelters
remained in the field between May 10 and June 17, 2016.
An emergence box containing 40 female and 60 male loose cocoons (at least 73% O.
lignaria, with the remaining proportion being O. cornifrons) were placed in each shelter. The
emergence box consisted of a cardboard box of 9.3 x 9.3 x 4.3 cm, with a hole on one face to
allow bees to go out of the box after emergence. Cocoons were gathered from trap nests located
at Michigan State University (MSU) campus in East Lansing, MI, in September 2015. Nests
were opened and cocoons were extracted and sorted by sex based on size since female cocoons
are bigger (» 11 mm) than male cocoons (» 6 m) (Torchio and Tepedino 1980). Cocoons were
stored in an environmental chamber at 4ºC from November 2015 until their release. A subsample of 50 cocoons was taken from each batch and emerged in the lab by holding them at
laboratory conditions (» 25ºC) to determine the proportion of each species.

	

72	

Figure 4.2. Nest shelter located at A) the OUT location at the edge of the field and B) at the IN
location in the blueberry field.

	

73	

Pollen samples from nests
Shelters were inspected at least once per week during the study period to record the number of
completed nests and to observe nesting bees. I observed each shelter for 15 min, identifying the
species actively nesting there. On June 8, 2016, when blueberry bloom was over, nests with a
mud plug at the entrance were brought back to the laboratory and kept in cold storage (-23ºC)
until processing. All nests were opened and pollen from individual brood cells was collected. I
also recorded the total number of nests in each location and the number of cells per nest.

Pollen sampling from plants
Once per week after placement of the boxes and until the nest were collected, I sampled plants
blooming around each shelter. Two transects of 100 x 3 m were established at each location,
where I recorded presence of flowering plants. At the OUT locations, transects were done in
north and south direction. At the IN locations, transects followed the orientation of the blueberry
rows. The first time each plant was observed, a reference specimen was collected and pressed for
further identification using Petrides (1972), Newcomb (1977), Peterson and McKenny (1996).
Additionally, several flower buds were collected from each plant for pollen extraction. If flower
buds were not available, open flowers were collected. Flowers and flower buds were stored in
70% ethanol until processing. Pollen was extracted by macerating the anthers into a container
with ethanol, to later centrifuge the solution for 5 min at 3000 rpm. This pollen was processed by
acetolysis (see below) and used as a reference library for identifying pollen collected from bee
nests.

	

74	

Pollen processing
All pollen samples were processed by acetolysis, following the protocol of Louveaux et al.
(1978). Briefly, pollen was treated with an acetic anhydride acid and sulfuric acid solution (9:1
ratio), and then subjected to washes of 70% ethanol and then distilled water to dissolve tissues
and remove lipids and debris. This protocol makes the diagnostic characteristics of the pollen
grains more visible (Jones 2014). Acetolyzed samples from each brood cell were set up on
microscope slides. For each slide, I counted and identified all pollen grains to the lowest
taxonomic level possible in 15 fields of view at a magnification of 400x. Each field of view had
a diameter of 450 µm. Pollen identification was done with the assistance of the reference pollen
library from herbaceous and woody plants collected at the field site.

Data analysis
Only the data from O. lignaria nesting in shelters were included in the analysis. The proportion
of each pollen type in each brood cell was calculated based on the number of grains. ShapiroWilk tests and diagnostic graphs (qq plots and residual plots) were used to assess the normality,
and Bartlett’s test was used to assess variance homogeneity between the locations (Shanks and
Hutton 1976). In order to meet the normality and equal variance assumptions, I used an arcsine
transformation on the data for the parametric analysis.
To determine which pollen type was more abundant in each location, the proportion of
each pollen type was compared within each group by an analysis of variance (ANOVA) and a
Tukey’s test (p<0.05). To evaluate the effect of location (IN or OUT) on the percentage of each
pollen species, and to compare the proportion of blueberry pollen between cells within each
location group, as well as the number of nests and cells per nest obtained between locations, I

	

75	

used the non-parametric Kruskal-Wallis Rank Sum Test (p<0.05). All data analyses were
conducted in R version 3.1.1 (R Development Core Team, Vienna, Austria) using the
“Agricolae” package.

RESULTS
Fifty different blooming plant species were recorded during the sampling (Table 4.1). A total of
47 species were identified at the OUT locations, with 7.9 ± 0.3 species/location (mean ± SE). At
the IN locations, I observed 37 species and 6.5 ± 0.3 species/location. Sixty-eight percent of the
species were common to both location. The number of O. lignaria nests built at IN and OUT
locations (4.5 ± 0.4 and 4.2 ± 0.3 nest/location respectively) was similar, with no significant
differences (X2 1 = 0.228, p = 0.63), as well as the number of cells build per nest (IN: 5.5 ± 0.2
cells/nest; OUT: 5.6 ± 0.1 cells/nest) (X2 1 = 0.125, p = 0.72).

Table 4.1. Blooming plants present in the blueberry field during the period Osmia lignaria was active
(May 10 – June 17, 2016), and their presence in the field interior (IN) or exterior (OUT).
	

	

Family

Species

Common name

Location

Asteraceae

Taraxacum officinale

Common dandelion

Both

Hieracium pratense

Field hawkweed

Both

Senecio vulgaris

Common groundsel

IN

Erigeron annuus

Daisy fleabane

Both

Senecio jacobaea

Tansy ragwort

OUT

Achillea millefolium

Common yarrow

Both

76	

Table 4.1 (cont’d)
Asteraceae

Chrysanthemum leucanthemum Oxeye daisy

OUT

Hieracium aurantiacum

Orange hawkweed

Both

Hypochaeris radicata

Cat's ear

Both

Boraginaceae

Borago officinalis

Borage

IN

Brassicaceae

Barbarea vulgaris

Winter cress

Both

Brassica rapa

Field mustard

Both

Nasturtium officinale

Watercress

OUT

Rorippa sp.

Yellowcresses

Both

Sisymbrium sp.

Hedge mustard

IN

Thlaspi arvense

Pennycress

Both

Lonicera tatarica

Tartarian honeysuckle

OUT

Viburnum recognitum

Arrowwood

Both

Stellaria media

Common chickweed

Both

Silene alba

White campion

Both

Cornus florida

Flowering dogwood

OUT

Cornus racemosa

Gray dogwood

Both

Caprifoliaceae

Caryophyllaceae

Cornaceae

	

Cyperaceae

Carex sp.

Both

Ericaseae

Vaccinium corymbosum

Highbush blueberry

Both

Fabaceae

Trifolium repens

White clover

Both

Trifolium pratense

Red clover

Both

Medicago lupulina

Black medick

Both

Vicia cracca

Cow vetch

OUT

77	

Table 4.1 (cont’d)
Geraniaceae

Erodium cicutarium

Redstem filaree

OUT

Iridaceae

Iris versicolor

Blue flag iris

Both

Sisyrinchium montanum

Blue eyed grass

Both

Lamiaceae

Lamium amplexicaule

Henbit

Both

Oxalidaceae

Oxalis stricta

Common wood sorrel

Both

Portulacaceae

Claytonia virginica

Spring beauty

Both

Ranunculaceae

Ranunculus ophioglossifolius

Spearwort

Both

Ranunculus hispidus

Swamp buttercup

OUT

Fragaria virginiana

Wild strawberry

Both

Potentilla simplex

Oldfield cinquefoil

Both

Prunus serotine

Black cherry

OUT

Rubus sp.

Bramble

Both

Rubus allegheniensis

Raspberry

OUT

Rubiaceae

Galium asprellum

Rough bedstraw

Both

Salicaceae

Salix pentandra

Laurel willow

Both

Scrophulariaceae

Nuttallanthus canadensis

Blue toadflax

Both

Solanaceae

Solanum dulcamara

Deadly nightshade

OUT

Violaceae

Viola kitaibeliana

Field pansy

Both

Viola sororia

Common blue violet

Both

Viola lanceolata

Lance leaved violet

Both

Rosaceae

	

Unidentified bush 1 (P17)

OUT

Unidentified tree 2 (P21)

OUT

78	

There were 23 nests included in the pollen analysis, nine nests (with 83 cells total) from
the IN location, and 14 nests (with 105 cells total) from the OUT locations. Overall, 81.8% of
cells contained blueberry pollen, but only in 2.3% of the cells the percentage of blueberry pollen
was higher than 50%, and only in 1.5% the contribution was higher than 90% (Fig. 4.3). The
percentage of blueberry pollen in cells was not different between locations (IN: 4.9 ± 1.0 %;
OUT: 5.5 ± 1.1 %) (X2 1 = 0. 450, p = 0.502) (Fig. 4.3). Nine other pollen types were identified in
cell provisions (Fig. 4.4), including black cherry (Prunus serotina, Rosaceae) and white clover
(Trifolium repens, Fabaceae). At the OUT locations, percentage of pollen from white clover and
Rubus spp. were 6.4 and 3.4 % higher than at the IN locations respectively (all p < 0.05), while
black cherry showed a 12.4% highest abundance at the IN locations (X2 1 = 6.96, p = 0.008) (Fig.
4.4). The percentage of blueberry pollen did not change among cells of the same nest regardless
of the locations, and therefore the timing of their construction (all p > 0.05) (Fig. 4.5).

Figure 4.3. Percentage of blueberry pollen in O. lignaria nests located inside (IN) or at the edge
(OUT) of a blueberry farm in West Michigan.

	

79	

Figure 4.4. Mean (± SE) percentage of the most common pollen found in the brood cells of O.
lignaria nesting inside (IN) or at the edge (OUT) of blueberry fields in west Michigan. The *
represent statistical differences between positions (IN and OUT) (p < 0.05).
	
In both locations the most abundant pollen belonged to black cherry and white clover,
together accounting for 84% of the pollen provision at the IN locations and 79% at the OUT
locations. The percentage of blueberry pollen was significantly lower than these two plants in
both location (all p < 0.05) (Table. 4.2). Other pollen types present in the brood cells in
percentages lower than 0.1% include Rubus allegheniensis (Rocaceae), Silene alba
(Caryophyllaceae) and an unidentified pollen type.

	

80	

Figure 4.5. Percentage of blueberry pollen in Osmia lignaria cell provisions collected at the
interior (IN) and exterior (OUT) of the field as a function of their position in the tubes, equating
to their timing of completion of the cells.

Table 4.2. Mean percentage (± SE) of total pollen in O. lignaria brood cells nesting inside (IN) or at
the edge (OUT) of blueberry fields in west Michigan. Within each location, means with the same
letter are not significantly different, p > 0.05.

	

Prunus

Trifolium

Salix

Vaccinium

serotina

repens

pentandra

corymbosum

IN

54.1 ± 1.7 a

30.7 ± 1.9 b

6.8 ± 3.7 c

6.3 ± 4.0 c

OUT

41.7 ± 0.7 a

37.1 ± 2.2 a

10.6 ± 3.2 a

4.7 ± 3.5 a

81	

DISCUSSION
In this study I found that O. lignaria nest cells contain pollen from blueberry plants regardless of
whether they are nesting inside or at the edge of the field. However, blueberry is not one of the
main pollen sources used by these bees, representing a low proportion of total collected pollen
(Fig. 4.4). Additionally, the proportion of blueberry pollen did not change among cells in the
nest, suggesting that the use of blueberry pollen was relatively consistent over time (Fig. 4.5),
independent of the volume of crop bloom open in the field.
Osmia species, including O. lignaria, are able to gather blueberry pollen despite of its
bell-shaped corolla and poricidal anthers that can restrict the access to many other pollinators
(Stubbs and Drummond 1997, Dogterom 1999, Wilkaniee and Giejdasz 2003, West and
McCutcheon 2009). I also found a few cells that contained more than 90% blueberry pollen,
which further indicates that access to the flowers is not a limitation for these bees. However, O.
lignaria, as well as other Osmia species, show strong floral preferences, mainly for plants of the
family Rosaceae (Williams and Tepedino 2003, Sampson et al. 2009, Mader et al. 2010). When
bees are released during the bloom period of tree crops such as apples, cherries or pears, most of
the pollen found in their nests (up to 95%) corresponds to pollen from the main crop (Vicens and
Bosch 2000b, Bosch et al. 2006, Sheffield, Westby, Smith, et al. 2008, Kraemer et al. 2014).
Even when released in orchards, however, not-target plants can be more attractive than the target
crop. Kraemer et al. (2014), reported that in some apple orchards the most abundant pollen in O.
lignaria brood cells was Eastern redbud (Cercis canadensis: Fabaceae), and even in natural
habitats mason bees also show strong floral preferences (Radmacher and Strohm 2010). In this
case I found that the most abundant pollen in cell provisions belonged to back cherry and white

	

82	

clover, both plants in families that have been reported to be highly attractive to Osmia (Rosaceae
and Fabaceae).
In general, O. lignaria tend to collect resources that are close to their nests, but this can
be affected by plant density and plant species (Williams and Tepedino 2003). Black cherry is a
dominant native species of tree in the woods of this region and is commonly found in woodlands
that were cleared to create fields for blueberry fields. Black cherry was present only at the edge
of the fields, so females nesting in the field had to fly at least 200 m to collect this pollen. The
foraging range of O. lingaria is between 50-150 m (Bosch and Kemp 1999, Dogterom 1999), but
they can fly up to 600 m (Zurbuchen et al. 2010), so it is perhaps not surprising that bees left the
field to collect pollen from plants that are highly attractive to them. Additionally, blueberry
pollen has a low protein content (13.9%) (Somerville 2001), suggesting that it is not sufficiently
nutritious for Osmia to use this as a dominant pollen to feed their larvae, a fact that has been also
observed in honey bees (Topitzhofer 2014).
Whether Osmia bees are effective pollinators may have more to do with their interaction
with the flowers than whether their pollen provisions contain a high proportion of blueberry
pollen. The pollination efficiency of some bees on blueberries is variety-dependent. For example,
O. ribifloris effectively pollinates Tifblue and Climax rabbiteye blueberry, but not Premier,
while A. mellifera is an effective pollinator of Premier and Climax, but not Tifblue (Sampson
and Cane 2000). Visitation rates of O. ribifloris decrease for varieties with narrow corollas
(Courcelles et al. 2013), while varieties with long pistils are more efficiently pollinated because
pollen transfer is made easier for the bees (Sampson et al. 2013). It is possible that the small
peaks observed in blueberry pollen collected during the nesting period of O. lignaria (Fig 4.5),
correspond to the blooming period of different blueberry cultivars with differences in their floral

	

83	

morphology, that could make easier for bees to harvest its pollen. Therefore, it would be
necessary to evaluate whether pollination efficiency and flower preferences change for different
varieties of highbush blueberry, through direct observations of the bees in multi-variety
plantings.
The presence of honey bees in high densities is also a factor that can negatively affect
visitation rate of wild bees (Lindström et al. 2016). Bumblebee visits to wild plants increase as
distance from honey bees hives increase (Thomson 2006), and they can also shift between plant
species when honey bee density increases (Walther-Hellwig et al. 2006). Osmia bees can be
affected by a high density of honey bees, decreasing visitation rate to blueberries and reducing
the number of cells built per nest by O. bicornis in a cage experiment (Hudewenz and Klein
2015). In our study site, honeybee hives were placed into the blueberry field to ensure good
pollination levels, and it is possible that the high rate of honey bees visiting blueberry flowers
could have negatively influenced visitation rate by O. lignaria, making them more likely to visit
other plants.
Another farm practice that can affect bees is the applications of pesticides (Johnson
2015). At the Blueberry Valley site, they restrict the applications of insecticides during bloom
only to pesticides with high safety for bees, but there are applications of fungicides during that
period. No lethal effects of fungicides have been reported for O. lignaria, but after applications
of iprodione, propiconazole, benomyl, and captan, bees can remain inactive for a few hours
before resuming activity (Ladurner et al. 2008). Artz and Pitts-Singer (2015) found that
iprodione, pyraclostrobin and the adjuvant N-90 disrupt the nest recognition abilities of O.
lignaria females. Though the fungicides (chlorothalonil, captan and metconazole) and insecticide
(methoxyfenozide) applied in our experimental site did not seem to discourage bees from nesting

	

84	

at the IN locations, they could have caused a repellent effect on O. lignaria, making blueberry
flowers less attractive to bees. This could be another possible explanation for the variation
observed on the percentage of blueberry pollen in different cells, due to that application of
pesticides could induce bees to not visit blueberry flowers for a short period. However, it is
difficult to estimate the exact day a specific cell was build, so it is difficult to establish a
relationship between blueberry pollen use and pesticides applications. Furthermore, this was not
measured directly, and further studies of pollinator-pesticide interactions in this system would be
important for better understanding the importance of these potential interactions.
In Michigan, O. lignaria fly between April and May, so their nesting period does not
completely coincide with blueberry bloom, which typically happens during May (Garcia-Salazar
2002). The possibility of adjusting the timing of O. lignaria emergence is one of the most
appealing characteristics of this species, because it allows for the ability to release a large
number of individuals to cover the bloom period of the crop. However, releasing bees out of their
normal flight period also potentially exposes them to plants different from those they would
normally encounter, resulting in females bringing atypical pollen back to the nest. Pollen of
different plant species differ in their nutritional value and digestibility (Roulston and Cane 2000),
and they can also differentially influence larval growth and development in several bee species,
including O. lignaria (Levin and Haydak 1957, Tasei and Aupinel 2008). The use of non-host
species can have negative effects on larvae health, as in the case of some populations of O.
cornuta, that when reared with Ranunculus acris (Ranunculaceae) pollen, a non-host species for
this bee, had higher larval mortality (Haider et al. 2013). This negative effect of non-host pollen
can be due to physiological limitations related to pollen digestion, including lack of essential
nutrients or the presence of toxins in the pollen (Williams 2003, Sedivy et al. 2008). This

	

85	

highlights the importance of considering several factors when releasing O. lignaria in a time
period different to their natural emergence time and in locations where the species does not
naturally occur (Pitts-Singer et al. 2014).
Plant diversity in and around farms has been proposed and evaluated as a method to
increase pollinator diversity in agroecosystems (Potts, Biesmeijer, et al. 2010). By providing
alternative pollen and nectar sources, bees can establish before the main crop blooms, and
continue providing resources after it is over, to allow bees to have the highest reproductive
success rate (Stubbs et al. 1992, Sedivy and Dorn 2014, Werling et al. 2014). However, coblooming alternative foraging plants should not be more attractive to bees than the crop of
interest, otherwise the objective of pollinating the crop may not be accomplished (Bosch and
Kemp 2002, Mader et al. 2010). Considering that O. lignaria preferred to forage on alternative
plants in this system, this species may not be the best option to use as an alternative pollinator for
blueberry, and there should be future exploration for additional bee species that may have greater
fidelity to blueberry.
With the results obtained in this experiment I can conclude that in field conditions O.
lignaria does not visit blueberries as a main source of pollen, and therefore their contribution to
pollination of this crop is probably low. The fact that O. lignaria does not have a high level of
visitation to blueberries, even when there is a large amount of pollen available during bloom,
suggest that different factors including floral preferences, competition with honey bees, and
application of fungicides during bloom may limit its ability to provide high levels of pollination.
Even though O. lignaria does not appear to be a promising pollinator of blueberry, there
are many other Osmia species that have been observed visiting blueberry fields, including O.
atriventris, O. pumila and O. bucephala (Stubbs et al. 1992, Stubbs and Drummond 1997,

	

86	

Sampson et al. 2009b, Tuell et al. 2009; J. Gibbs unpublished). Although their relative
abundance can be low (Ratti 2006, Bushmann and Drummond 2015, Scott et al. 2016), it could
be increased by providing suitable nesting substrates (Torchio 1982a, Mader et al. 2010, MacIvor
2016). Several studies show that increasing diversity and abundance of bees in agricultural areas
can increase yield stability, resilience, and decrease dependence and pressure on honey bees
(Garibaldi et al. 2013, 2016, Senapathi et al. 2015, Blitzer et al. 2016). Given this, practices that
help to establish and increase wild pollinator populations around farms can have a positive effect
on yield, as shown by Gibbs et al. (2016), who found that in British Columbia blueberry farms,
wild bee abundance is related to the proportion of semi-natural habitat around the field, and the
pollination service increased with abundance of wild bumble bees.
In this study I found that blueberry pollen can be found in O. lignaria nest provisions, but
at low levels in comparison with other pollen types. Additionally, the proportion of blueberry
pollen does not change during the nesting period or depending on the location of the nest (in the
field or at the border of the field). These finding are important given the growing interest of
using alternative pollinators for crops, and in this case show that O. lignaria would not be the
best option to pollinate blueberries. Future studies that aim to use Osmia as blueberry pollinators
could focus on other Osmia species that naturally visit blueberries and how to increase their
visitation rate. Plants with high pollen contribution identified in this study could be used as
pollen sources for O. lignaria in non-agricultural setting, like urban environments, where they
can be important pollen sources for several bee species, including Osmia.

	

87	

CHAPTER 5. FUTURE DIRECTIONS
Mason bees (Osmia spp.) are a promising group of alternative pollinators for rosaceous fruitcrops. Although reports about mason bees biology and behavior started early in 1925, and
several studies about their nesting preferences and potential for crop pollinations were published
in the 1980s, only in recent years has there been a concerted effort to have these pollinators
readily available for commercial orchards. Indeed there are now commercial suppliers of these
bees, supplying the residential and farm markets for Osmia bees. However, the extensive use of
mason bees in orchards requires us to have sufficient knowledge of their biology and behavior to
be able to ensure an adequate supply of bees to provide for crop pollinations. While the results
obtained in the present thesis contributes to the body of knowledge that will help to improve
management of Osmia bees as alternative pollinators, there are still issues and questions. Next I
will present future directions and further questions generated from my results that still need to be
addressed.
In the shelter/substrate experiments I found that a higher percentage of cavities are used
by O. cornifrons than O. lignaria, and a larger number of female offspring is produced when
using wooden shelters and reeds. So this combination should be tested in commercial settings to
establish the degree of success in the environments that bees are being used for commercial
pollination. However, this combinations of shelter and substrate materials is more expensive than
plastic shelter and cardboard tubes, so there is also a need for economic analyses of whether the
investment in more expensive materials is balanced by the increase in pollination service or
number of offspring produced that could be use in the next growing season.
The attractant spray tested in this research proved to be effective at increasing the number
of cavities used by O. cornifrons, so further research could focus on testing different

	

88	

concentrations and the degradation time of the active ingredient, so it can be established whether
and when reapplication is needed. Additionally, the attractant should be tested in a broader range
of species. Another point that should be addressed is that even when the attractant was originally
developed from O. lignaria cocoons, in this experiment I observed a low response of this species
to the attractant. These results could be influenced by the high dispersal rate that is commonly
reported for O. lignaria, so a larger number of females should be used in future experiments to
be able identify possible effects of the attractant on this species. This also highlights the
importance of developing rearing and releasing methodologies that could decrease dispersal of
Osmia bees in the field.
Research that focuses on the ecology of species that have the potential to be used as
commercial pollinators is also an important topic. Given that O. cornifrons is a successful
introduced species that has spread widely through the eastern US, possible negative effects on
native species due to competition for floral or nesting resources should be conducted before
introducing this species in areas where it is still not present. Studies on the effects of trapping of
wild populations of Osmia with the aim of extracting individuals to be used to supplement
artificial rearing, or for research purposes, would be important to establish the degree in which
that practice can affect natural populations.
Regarding soil used by Osmia, this is the first study that has addressed detailed particle
size analysis of the soil collected by bees. I found that bees can used a wide range of soil
textures, unlike the information provided in previous publications, and that there are differences
in the texture of the soil collected by both species. I also found variation in the consistency of
individual females Osmia bees for collecting soil, with different soil types being found within a
single partition, among partitions, and among nests. This generates a significant set of new

	

89	

questions regarding soil used by mason bees, including; what soil texture would females prefer if
a wide range of soil textures is offered? Are females able to use soil of least preferred texture if
other options are not available? Does the soil texture of the partitions influence its effectiveness
at preventing the entrance of parasites/predators into the nest? Do different populations or
subspecies prefer soil of different textures? What other parameters (like organic matter content,
bulk density, moisture content) are important for bees when selecting soil sources? How far can
females fly to collect soil of their preferred characteristics? These and several more questions are
of interest not only to better understand the ecology of mason bees, but also to be able to
determine whether a particular region offers soil suitable for bees or whether it is necessary to
provide additional soil sources of the adequate characteristics.
For the blueberry pollen use experiment, I found that even when nests are located in a big
blueberry field more than 200 m away from the border of the field, O. lignaria females prefer to
collect pollen from alternative plants that bloom in or around the field. These results support
other studies that suggest that this species is not a suitable pollinator for blueberries given that
they do not visit the target flowers if other floral options are available. These results are
important given that recently more attention has been paid to alternative pollinators, and growers
could mistakenly interpret that O. lignaria is an effective pollinator of blueberries in the field
However, O. cornifrons still remains to be tested in the field to determine whether they will
constantly visit blueberries. Additionally, other Osmia species have been reported to naturally
visit blueberry flowers, like O. atriventris, O. pumila and O. bucephala, so they can be asses for
their effectiveness and potential as blueberry pollinators. Studies about their nesting preferences
would be helpful even if the species are not suitable for mass rearing, given that nesting
substrates can be provided around fields to increase natural occurring populations of these bees

	

90	

Even if substantial progress has been made on mason bee rearing and management for
orchard pollination in recent years, is clear that we still have to solve several questions that will
allow us to have a sustainable and reliable system to rear and manage Osmia bees for crop
pollination, reducing dependence and pressure on honey bees. In that sense, the present work
makes a substantial contribution. However, we would also have to implement management
strategies that are compatible with the use these pollinators, and recognize that it will not be
possible to have commercial pollinators for every crop, and that in many cases integrated
management of the field is necessary to ensure a long-term pollination service that does not only
rely on human inputs.

	

91	

APPENDIX

	

92	

APPENDIX A
RECORD OF DEPOSITION OF VOUCHER SPECIMENS
The specimens listed below have been deposited in the named museum as samples of those
species or other taxa, which were used in this research. Voucher recognition labels bearing the
voucher number have been attached or included in fluid preserved specimens.

Voucher Number: 2017-06
Author and Title of thesis: Mario Simon Pinilla Gallego. “Nest, soil and pollen preferences of
mason bees (Hymenoptera: Megachilidae: Osmia) in Southwest Michigan farms”.

Museum where deposited: Albert J. Cook Arthropod Research Collection, Michigan State
University (MSU)
Specimens:

Table A. List of voucher specimens

	

Family

Genus-Species

Life Stage

Sex

Quantity

Preservation

Hymenoptera

Osmia lignaria

Adult

Male

10

Pinned

Hymenoptera

Osmia lignaria

Adult

Female

10

Pinned

Hymenoptera

Osmia cornifrons

Adult

Male

10

Pinned

Hymenoptera

Osmia cornifrons

Adult

Female

10

Pinned

93	

LITERATURE CITED
	
	

	

94	

LITERATURE CITED

Abrol, D. 2015. Pollination Biology, Vol. 1 Pest and pollinators of fruit crops. Springer, New
York.
Arduser, M. 2009. Key to Osmia females known from eastern North America (east of 100th
meridian). Online key website.
Artz, D. R., M. J. Allan, G. I. Wardell, and T. L. Pitts-Singer. 2013. Nesting site density and
distribution affect Osmia lignaria (Hymenoptera: Megachilidae) reproductive success and
almond yield in a commercial orchard. Insect Conserv. Divers. 6: 715–724.
Artz, D. R., M. J. Allan, G. I. Wardell, and T. L. Pitts-Singer. 2014. Influence of nest box
color and release sites on Osmia lignaria (Hymenoptera: Megachilidae) reproductive
success in a commercial almond orchard. J. Econ. Entomol. 107: 2045–2054.
Artz, D. R., and T. L. Pitts-Singer. 2015. Effects of fungicide and adjuvant sprays on nesting
behavior in two managed solitary bees, Osmia lignaria and Megachile rotundata. PLoS
One. 10: e0135688.
Batra, S. 1978. Osmia cornifrons and Pithitis smaragdula, two Asian bees introduce into the
United States for crop pollination. Proc. IVth Int. Symp. Pollinat. 1: 307–312.
Biddinger, D. J., J. Frazier, E. Rajotte, L. R. Donovall, and T. Leslie. 2009. Solitary bees as
alternative pollinators of Pennsylvania fruit crops. Penn Fruit News. 89: 84–94.
Biddinger, D. J., N. K. Joshi, E. G. Rajotte, N. O. Halbrendt, C. Pulig, K. J. Naithani, and
M. Vaughan. 2013. An immunomarking method to determine the foraging patterns of
Osmia cornifrons and resulting fruit set in a cherry orchard. Apidologie. 44: 738–749.
Biddinger, D. J., H. Ngugi, J. Frazier, M. Frazier, E. Rajotte, T. Leslie, and L. R. Donovall.
2008. Development of the mason bee, Osmia cornifrons, as an alternative pollinator to
honey bees and as a targeted delivery system for biocontrol agents in the management of
fire.
Biddinger, D. J., J. L. Robertson, C. A. Mullin, J. Frazier, S. A. Ashcraft, E. G. Rajotte, N.
K. Joshi, and M. Vaughn. 2013. Comparative toxicities and synergism of apple orchard
pesticides to Apis mellifera (L.) and Osmia cornifrons (Radoszkowski). PLoS One. 8:
e72587.
Blitzer, E. J., J. Gibbs, M. G. Park, and B. N. Danforth. 2016. Pollination services for apple
are dependent on diverse wild bee communities. Agric. Ecosyst. Environ. 221: 1–7.

	

95	

Bosch, J. 1994a. Improvement of field management of Osmia cornuta (Latreille) (Hymenoptera,
Megachilidae) to pollinate almond. Apidologie. 25: 71–83.
Bosch, J. 1994b. Osmia cornuta Latr. (Hym., Megachilidae) as a potential pollinator in almond
orchards. J. Appl. Ent. 117: 151–157.
Bosch, J., and W. P. Kemp. 1999. Exceptional cherry production in an orchard pollinated with
blue orchard bees. Bee World. 80: 163–173.
Bosch, J., and W. P. Kemp. 2000. Development and emergence of the orchard pollinator Osmia
lignaria (Hymenoptera: Megachilidae). Environ. Entomol. 29: 8–13.
Bosch, J., and W. P. Kemp. 2001. How to manage the blue orchard bee. Sustainable
Agriculture Network (SAN), Baltimore.
Bosch, J., and W. P. Kemp. 2002. Developing and establishing bee species as crop pollinators:
the example of Osmia spp. (Hymenoptera: Megachilidae) and fruit trees. Bull. Entomol.
Res. 92: 3–16.
Bosch, J., and W. P. Kemp. 2003. Effect of wintering duration and temperature on survival and
emergence time in males of the orchard pollinator Osmia lignaria (Hymenoptera :
Megachilidae). Environ. Entomol. 32: 711–716.
Bosch, J., and W. P. Kemp. 2004. Effect of pre-wintering and wintering temperature regimes
on weight loss, survival, and emergence time in the mason bee Osmia cornuta
(Hymenoptera: Megachilidae). Apidologie. 35: 452–461.
Bosch, J., W. P. Kemp, and G. E. Trostle. 2006. Bee population returns and cherry yields in an
orchard pollinated with Osmia lignaria (Hymenoptera: Megachilidae). J. Econ. Entomol.
99: 408–413.
Bosch, J., and N. Vicens. 2006. Relationship between body size, provisioning rate, longevity
and reproductive success in females of the solitary bee Osmia cornuta. Behav. Ecol.
Sociobiol. 60: 26–33.
Bosse, G., T. Jütte, and O. Klein. 2014. Experimental designs for field and semi-field studies
with solitary wild bees, pp. 226–229. In 12th Int. Symp. ICP-PR Bee Prot. Gr. Ghent.
Brevis, P. A. 2001. Factors controlling fruit set of rabbiteye blueberry (Vaccinium ashei Reade).
Buchmann, S., and J. P. Hurley. 1978. A biophysical model for buzz pollination in
angiosperms. J. Theor. Biol. 72: 639–657.
Bushmann, S. L., and F. A. Drummond. 2015. Abundance and diversity of wild bees
(Hymenoptera: Apoidea) found in lowbush blueberry growing regions of downeast Maine.
Environ. Entomol. 44: 975–989.

	

96	

Cane, J. H. 1991. Soils of ground-nesting bees (Hymenoptera: Apoidea): texture, moisture, cell
depth and climate. J. kansas Entomol. Soc. 64: 406–413.
Cane, J. H. 1997. Lifetime monetary value of individual pollinator: the bee Habropoda
laboriosa at rabbiteye blueberry (Vaccinium ashei Reade). Acta Hort. 446: 67–70.
Cane, J. H., T. Griswold, and F. D. Parker. 2007. Substrates and materials used for nesting by
North American Osmia bees (Hymenoptera: Apiformes: Megachilidae). Ann. Entomol. Soc.
Am. 100: 350–358.
Canto-Aguilar, M. A., and V. Parra-Tabla. 2000. Importance of conserving alternative
pollinators: assessing the pollination efficiency of the squash bee, Peponapis limitaris in
Cucurbita moschata (Cucurbitaceae). J. Insect Conserv. 4: 201–208.
Christmann, S., and A. A. Aw-Hassan. 2012. Farming with alternative pollinators (FAP)-An
overlooked win-win-strategy for climate change adaptation. Agric. Ecosyst. Environ. 161:
161–164.
Courcelles, D. M. M., L. Button, and E. Elle. 2013. Bee visit rates vary with floral morphology
among highbush blueberry cultivars ( Vaccinium corymbosum L.). J. Appl. Entomol. 137:
693–701.
Cutler, G. C., V. O. Nams, P. Craig, J. M. Sproule, and C. S. Sheffield. 2015. Wild bee
pollinator communities of lowbush blueberry fields: Spatial and temporal trends. Basic
Appl. Ecol. 16: 73–85.
Dedej, S., and K. S. Delaplane. 2003. Honey bee (Hymenoptera: Apidae) pollination of
rabbiteye blueberry Vaccinium ashei var. “Climax” is pollinator density-dependent. J. Econ.
Entomol. 96: 1215–20.
Dogterom, M. H. 1999. Pollination by four species of bees on highbush blueberry. Funct. Ecol.
Fabre, J. H. 1925. The Mason-Bees. Garden city publishing company, Garden city.
Fauria, K., and R. Campan. 1998. Do solitary bees Osmia cornuta Latr. and Osmia lignaria
Cresson use proximal visual cues to localize their nest? J. Insect Behav. 11: 649–669.
Felicioli, A., M. Krunic, and M. Pinzauti. 2004. Rearing and using Osmia bees for crop
pollination: a help from a molecular approach, pp. 161–174. In Freitas, B.M., Pereira, J.O.
(eds.), Solitary Bees Conserv. Rearing Manag. Pollinat. Imprensa Universitaria, Fortaleza.
Foth, H. D. 1990. Fundamentals of soil science, 8th ed. Wiley.
Gallai, N., J.-M. Salles, J. Settele, and B. E. Vaissière. 2009. Economic valuation of the
vulnerability of world agriculture confronted with pollinator decline. Ecol. Econ. 68: 810–
821.

	

97	

Garcia-Salazar, C. 2002. Crop timeline for blueberries in Michigan and Indiana.
Gardner, K. E., and J. S. Ascher. 2006. Notes on the native bee pollinators in New York apple
orchards. J. New York Entomol. Soc. 114: 86–91.
Garibaldi, L. A., I. Bartomeus, R. Bommarco, A. M. Klein, S. A. Cunningham, M. Aizen,
V. Boreux, M. P. D. Garratt, L. G. Carvalheiro, C. Kremen, C. L. Morales, C.
Schüepp, N. P. Chacoff, B. M. Freitas, V. Gagic, A. Holzschuh, B. K. Klatt, K. M.
Krewenka, S. Krishnan, M. M. Mayfield, I. Motzke, M. Otieno, J. Petersen, S. G.
Potts, T. H. Ricketts, M. Rundlöf, A. Sciligo, P. A. Sinu, I. Steffan-Dewenter, H. Taki,
T. Tscharntke, C. H. Vergara, B. F. Viana, and M. Woyciechowski. 2015. Trait
matching of flower visitors and crops predicts fruit set better than trait diversity. J. Appl.
Ecol. 52: 1436–1444.
Garibaldi, L. A., L. G. Carvalheiro, B. E. Vaissiere, B. Gemmill-Herren, J. Hipolito, B. M.
Freitas, H. T. Ngo, N. Azzu, A. Saez, J. Astrom, J. An, B. Blochtein, D. Buchori, F. J.
C. Garcia, F. Oliveira da Silva, K. Devkota, M. d. F. Ribeiro, L. Freitas, M. C.
Gaglianone, M. Goss, M. Irshad, M. Kasina, A. J. S. P. Filho, L. H. P. Kiill, P.
Kwapong, G. N. Parra, C. Pires, V. Pires, R. S. Rawal, A. Rizali, A. M. Saraiva, R.
Veldtman, B. F. Viana, S. Witter, and H. Zhang. 2016. Mutually beneficial pollinator
diversity and crop yield outcomes in small and large farms. Science (80-. ). 351: 388–391.
Garibaldi, L. A., I. Steffan-Dewenter, C. Kremen, J. M. Morales, R. Bommarco, S. A.
Cunningham, L. G. Carvalheiro, N. P. Chacoff, J. H. Dudenhöffer, S. S. Greenleaf, A.
Holzschuh, R. Isaacs, K. M. Krewenka, Y. Mandelik, M. M. Mayfield, L. Morandin, S.
G. Potts, T. H. Ricketts, H. Szentgyörgyi, B. F. Viana, C. Westphal, R. Winfree, and A.
M. Klein. 2011. Stability of pollination services decreases with isolation from natural areas
despite honey bee visits. Ecol. Lett. 14: 1062–1072.
Garibaldi, L. A., I. Steffan-Dewenter, R. Winfree, M. Aizen, S. A. Cunningham, C.
Kremen, L. G. Carvalheiro, L. Harder, O. Afik, I. Bartomeus, F. Benjamin, V.
Boreux, D. P. Cariveau, N. P. Chacoff, J. H. Dudenhöffer, B. M. Freitas, J. Ghazoul, S.
S. Greenleaf, J. Hipólito, A. Holzschuh, B. G. Howlett, R. Isaacs, S. K. Javorek, C. M.
Kennedy, and S. Krishnan. 2013a. Wild pollinators enhance fruit set of crops regardless
of honey bee abundance. Science.
Gibbs, J., E. Elle, K. Bobiwash, T. Haapalainen, and R. Isaacs. 2016. Contrasting pollinators
and pollination in native and non-native regions of highbush blueberry production. PLoS
One. 11: e0158937.
Glasser, S., and S. Farzan. 2016. Host-associated volatiles attract parasitoids of a native
solitary bee, Osmia lignaria Say (Hymenoptera, Megachilidae). J. Hymenopt. Res. 51: 249–
256.

	

98	

Gonzalez, V. H., M. M. Gonzalez, and Y. Cuellas. 2009. Notas biológicas y taxonómicas sobre
los abejorros del maracuyá del género Xylocopa (Hymenoptera: Apidae, Xylocopini) en
Colombia. Acta biológica Colomb. 14: 31–40.
Gonzalez, V. H., A. Mejia, and C. Rasmussen. 2004. Ecology and nesting behavior of Bombus
atratus Franklin in Andean highlands (Hymenoptera : Apidae). J. Hym. Res. 13: 234–242.
Goulson, D., and W. O. H. Hughes. 2015. Mitigating the anthropogenic spread of bee parasites
to protect wild pollinators. Biol. Conserv. 191: 10–19.
Gruber, B., K. Eckel, J. Everaars, and C. F. Dormann. 2011. On managing the red mason bee
(Osmia bicornis) in apple orchards. Apidologie. 42: 564–576.
Guedot, C., J. Bosch, and W. P. Kemp. 2009. Relationship between body size and homing
ability in the genus Osmia (Hymenoptera; Megachilidae). Ecol. Entomol. 34: 158–161.
Guedot, C., T. L. Pitts-Singer, J. S. Buckner, J. Bosch, and W. P. Kemp. 2006. Olfactory
cues and nest recognition in the solitary bee Osmia lignaria. Physiol. Entomol. 31: 110–
119.
Haider, M., S. Dorn, and A. Müller. 2013. Intra- and interpopulational variation in the ability
of a solitary bee species to develop on non-host pollen: implications for host range
expansion. Funct. Ecol. 27: 255–263.
Hanley, M. E., and D. Goulson. 2003. Introduced weeds pollinated by introduced bees: Cause
or effect? Weed Biol. Manag. 3: 204–212.
Holzschuh, A., J. H. Dudenhöffer, and T. Tscharntke. 2012. Landscapes with wild bee
habitats enhance pollination, fruit set and yield of sweet cherry. Biol. Conserv. 153: 101–
107.
Hudewenz, A., and A. M. Klein. 2015. Red mason bees cannot compete with honey bees for
floral resources in a cage experiment. Ecol. Evol. 1–8.
Isaacs, R., and A. K. Kirk. 2010. Pollination services provided to small and large highbush
blueberry fields by wild and managed bees. J. Appl. Ecol. 47: 841–849.
Isaacs, R., J. Tuell, A. K. Fiedler, M. Gardiner, and D. Landis. 2009. Maximizing arthropodmediated ecosystem services in agricultural landscapes: the role of native plants. Front Ecol
Env. 7: 193–203.
Jaffé, R., V. Dietemann, M. H. Allsopp, C. Costa, R. M. Crewe, R. Dall’olio, P. de la Rúa,
M. A. A. El-niweiri, I. Fries, N. Kezic, M. S. Meusel, R. J. Paxton, T. Shaibi, E. Stolle,
and R. F. A. Moritz. 2010. Estimating the density of honeybee colonies across their natural
range to fill the gap in pollinator decline censuses. Conserv. Biol. 24: 583–593.

	

99	

James, R., and T. L. Pitts-Singer. 2008. Bee pollination in agricultural ecosystems. Oxford
University Press, New York.
Javorek, S. K., K. E. Mackenzie, and S. P. Vander Kloet. 2002. Comparative pollination
effectiveness among bees (Hymenoptera: Apoidea) on lowbush blueberry (Ericaceae:
Vaccinium angustifolium). Ann. Entomol. Soc. Am. 95: 345–351.
Johnson, R. M. 2015. Honey bee toxicology. Annu. Rev. Entomol. 60: 415–434.
Jones, G. D. 2014. Pollen analyses for pollination research, acetolysis. J. Pollinat. Ecol. 13: 203–
217.
Junqueira, C. N., and S. C. Augusto. 2016. Bigger and sweeter passion fruits: effect of
pollinator enhancement on fruit production and quality. Apidologie.
Kemp, W. P., and J. Bosch. 2005. Effect of temperature on Osmia lignaria (Hymenoptera:
Megachilidae) prepupa-adult development, survival, and emergence. J. Econ. Entomol. 98:
1917–1923.
Klein, A. M., B. E. Vaissiere, J. H. Cane, I. Steffan-Dewenter, S. A. Cunningham, C.
Kremen, and T. Tscharntke. 2007. Importance of pollinators in changing landscapes for
world crops. Proc. R. Soc. B Biol. Sci. 274: 303–313.
Kraemer, M. E., F. D. Favi, and C. E. Niedziela. 2014. Nesting and pollen preference of
Osmia lignaria lignaria (Hymenoptera: Megachilidae) in Virginia and North Carolina
orchards. Environ. Entomol. 43: 932–941.
Krombein, K. V. 1967. Trap-nesting wasps and bees life histories, nests and associates,
Smithson. Press. Washington, D.C.
Krug, C., I. Alves-dos-Santos, and J. H. Cane. 2010. Visiting bees of Cucurbita flowers
(Cucurbitaceae) with emphasis on the presence of Peponapis fervens Smith (Eucerni Apidae) - Santa Catarina, Southern Brazil. Oecologia Aust. 14: 128–139.
Krunic, M., and L. Stanisavljevic. 2006. Augmentation of managed populations of Osmia
cornuta and O. rufa (Hymenoptera: Megachilidae) in Southeastern Europe. Eur. J. Entomol.
103: 695–697.
Ladurner, E., J. Bosch, W. P. Kemp, and S. Maini. 2005. Assessing delayed and acute
toxicity of five formulated fungicides to Osmia lignaria Say and Apis mellifera.
Apidologie. 36: 449–460.
Ladurner, E., J. Bosch, W. P. Kemp, and S. Maini. 2008. Foraging and nesting behavior of
Osmia lignaria (Hymenoptera: Megachilidae) in the presence of fungicides: cage studies. J.
Econ. Entomol. 101: 647–653.

	

100	

Ladurner, E., L. Recla, M. Wolf, R. Zelger, and G. Burgio. 2004. Osmia cornuta
(Hymenoptera Megachilidae) densities required for apple pollination: a cage study. J. Apic.
Res. 43: 118–122.
Levin, M. D. 1966. Biological notes on Osmia lignaria and Osmia californica (Hymenoptera:
Apoidea, Megachilidae). J. Kansas Entomol. Soc. 39: 524–535.
Levin, M. D., and M. H. Haydak. 1957. Comparative value of different pollens in the nutrition
of Osmia lignaria. Bee World. 38: 221–226.
Lindström, S. A. M., L. Herbertsson, M. Rundlöf, R. Bommarco, and H. G. Smith. 2016.
Experimental evidence that honeybees depress wild insect densities in a flowering crop.
Proc. R. Soc. B Biol. Sci. 283: 20161641.
Linsley, E. G. 1958. The ecology of solitary bees. Hilgardia. 27: 543–599.
Losey, J., and M. Vaughn. 2006. The economic value of ecological services provided by
insects. Bioscience. 56: 311–323.
Louveaux, J., A. Maurizio, and G. Vorwohl. 1978. Methods of melissopalynology. Bee world.
139–153.
Maccagnani, B., G. Burgio, L. Ž. Stanisavljević, and S. Maini. 2007. Osmia cornuta
management in pear orchards. Bull. Insectology. 60: 77–82.
MacIvor, J. S. 2016. Cavity-nest boxes for solitary bees: a century of design and research.
Apidologie.
Mackenzie, K. E. 2009. Pollination practices and the use of bees in Vaccinium crops. Acta Hort.
810: 527–538.
Mader, E., M. Spivak, and E. Evans. 2010. Managing alternative pollinators, 1th ed. NRAES,
Ithaca.
Marchi, P., and G. A. R. Melo. 2010. Biologia de nidificação de Xylocopa (Neoxylocopa)
frontalis (Olivier) (Hymenoptera, Apidae, Xylocopini) em Morretes, Paraná. Oecologia
Aust. 14: 210–231.
Matsumoto, S., A. Abe, and T. Maejima. 2009. Foraging behavior of Osmia cornifrons in an
apple orchard. Sci. Hortic. (Amsterdam). 121: 73–79.
Matsumoto, S., and T. Maejima. 2010. Several new aspects of the foraging behavior of Osmia
cornifrons in an apple orchard. Psyche A J. Entomol. 1–6.
McGregor, S. E. 1976. Insect pollination of cultivated crop plants. Agricultural Reseach
Service, Washington, D.C.

	

101	

McKinney, M. I., and Y.-L. Park. 2012. Nesting activity and behavior of Osmia cornifrons
(Hymenoptera: Megachilidae) elucidated using videography. Psyche A J. Entomol. 2012:
1–7.
Medler, J. T. 1967. Biology of Osmia in trap nests in Wisconsin (Hymenoptera: Megachilidae).
Ann. Entomol. Soc. Am. 60: 338–344.
Michener, C. 2007. The bees of the world, 2th ed. The Johns Hopkins University Press,
Baltimore.
Miller, B. a., and R. J. Schaetzl. 2012. Precision of soil particle size analysis using laser
diffractometry. Soil Sci. Soc. Am. J. 76: 1719–1727.
Monzón, V., J. Bosch, and J. Retana. 2004. Foraging behavior and pollinating effectiveness of
Osmia cornuta (Hymenoptera: Megachilidae) and Apis mellifera (Hymenoptera: Apidae) on
“Comice” pear. Apidologie. 35: 575–585.
Morris, W., D. Vazquez, and N. P. Chacoff. 2010. Benefit and cost curves for typical
pollination mutualisms. Ecology. 91: 1276–1285.
Morse, R. A., and N. W. Calderone. 2003. The value of honey bees as pollinators of U.S. crops
in 2000. Bee Cult. 128: 1–15.
Nates-Parra, G. 2001. Las abejas sin aguijón (Hymenoptera: Apidae: Meliponini) de Colombia.
Biota Colomb. 2: 233–248.
Nates-Parra, G., and V. H. Gonzalez. 2000. Notas solobre el nido de Eulaema polychroma
(Hymenoptera: Apidae: Euglossini). Actual Biol. 22: 83–90.
Newcomb, L. 1977. Wildflower guide. Little, Brown, New York.
Nogué, S., P. R. Long, A. E. Eycott, L. de Nascimento, J. M. Fernández-Palacios, G.
Petrokofsky, V. Vandvik, and K. J. Willis. 2016. Pollination service delivery for
European crops: challenges and opportunities. Ecol. Econ. 128: 1–7.
Peterson, R. T., and M. McKenny. 1996. Wildflowers Northeastern/ North-central North
America. Houghton Mifflin, New York.
Petrides, G. A. 1972. Trees and shrubs, 2th ed. Houghton Mifflin, New York.
Pitts-Singer, T. L. 2007. Olfactory response of megachilid bees, Osmia lignaria, Megachile
rotundata, and M. pugnata, to individual cues from old nest cavities. Environ. Entomol. 36:
402–408.

	

102	

Pitts-Singer, T. L., J. S. Buckner, T. P. Freeman, and C. Guedot. 2012. Structural
examination of the Dufour’s gland of the solitary bees Osmia lignaria and Megachile
rotundata (Hymenoptera: Megachilidae). Ann. Entomol. Soc. Am. 105: 103–110.
Pitts-Singer, T. L., J. H. Cane, and G. Trostle. 2014. Progeny of Osmia lignaria from distinct
regions differ in developmental phenology and survival under a common thermal regime. J.
Insect Physiol. 67: 9–19.
Pitts-Singer, T. L., W. P. Kemp, D. Moreland, S. Peterson, J. S. Buckner, and M. M.
Hagen. 2016. Bee attractants.
Potts, S. G., J. C. Biesmeijer, C. Kremen, P. Neumann, O. Schweiger, and W. E. Kunin.
2010. Global pollinator declines: trends, impacts and drivers. Trends Ecol. Evol. 25: 345–
353.
Potts, S. G., S. Roberts, R. Dean, G. Marris, M. Brown, R. Jones, P. Neumann, and J.
Settele. 2010. Declines of managed honey bees and beekeepers in Europe. J. Apic. Res. 49:
15.
Potts, S. G., and P. Willmer. 1997. Abiotic and biotic factors influencing nest-site selection by
Halictus rubicundus, a ground-nesting halictine bee. Ecol. Entomol. 22: 319–328.
Rader, R., B. G. Howlett, S. A. Cunningham, D. A. Westcott, L. E. Newstrom-Lloyd, M. K.
Walker, D. A. J. Teulon, and W. Edwards. 2009. Alternative pollinator taxa are equally
efficient but not as effective as the honeybee in a mass flowering crop. J. Appl. Ecol. 46:
1080–1087.
Radmacher, S. 2011. Temperature effects on the development in the solitary bee Osmia bicornis
(Hymenoptera, Megachilidae).
Radmacher, S., and E. Strohm. 2010. Factors affecting offspring body size in the solitary bee
Osmia bicornis (Hymenoptera, Megachilidae). Apidologie. 41: 169–177.
Ratti, C. M. 2006. Bee abundance and diversity in berry agriculture.
Ratti, C. M., H. A. Higo, T. Griswold, and M. L. Winston. 2008. Bumble bees influence berry
size in commercial Vaccinium spp. cultivation in British Columbia. Can. Entomol. 140:
348–363.
Rau, P., and K. Mo. 1937. The life-history of Osmia lignaria and O. cordata, with notes on O.
conjuncta. Annals of entomological society of America. 30: 324–344.
Raw, A. 1972. The biology of the solitary bee Osmia rufa (L.)(Megachilidae). Trans. R. ent.
Soc. L. 124: 213–229.

	

103	

Raw, A. 1992. Solitary bee (Hymenoptera: Megachilidae), restricted to identical resources from
nesting, recognized their own nests: an example of genetically determined personal scents?
Entomol. 2: 79–87.
Ritzinger, R., and P. M. Lyrene. 1999. Flower morphology in blueberry species and hybrids.
HortScience. 34: 130–131.
Roubik, D. 1989. Ecology and natural history of tropical bees. Cambridge University Press,
New York.
Roulston, T. H., and J. H. Cane. 2000. Pollen nutritional content and digestibility for animals.
Plant Syst. Evol. 222: 187–209.
Rust, R. 1974. The systematics and biology of the genus Osmia, subgenera Osmia, Chalcosmia,
and Cephalosmia (Hymenoptera: Megachilidae). Wasmann J. Biol. 32: 1–93.
Rust, R. 1998. The effects of cavity diameter and lenght on the nesting biology of Osmia
lignaria propinqua Cresson (Hymenoptera: Megachilidae). J. Hymenopt. Res. 7: 84–93.
Sampson, B. J., and J. H. Cane. 2000. Pollination efficiencies of three bee (Hymenoptera:
Apoidea) species visiting rabbiteye blueberry. J. Econ. Entomol. 93: 1726–31.
Sampson, B. J., J. H. Cane, G. T. Kirker, S. J. Stringer, and J. M. Spiers. 2009. Biology and
management potential for three orchard bee species (Hymenoptera: Megachilidae): Osmia
ribifloris Cockerell, O. lignaria (Say) and O. chalybea Smith with emphasis on the Former,
pp. 549–555. In Acta Hort.
Sampson, B. J., S. J. Stringer, J. H. Cane, and J. M. Spiers. 2004. Screenhouse evaluations of
a mason bee Osmia ribifloris (Hymenoptera: Megachilidae) as a pollinator for blueberries in
the southeastern United States. Small Fruits Rev. 3: 381–392.
Sampson, B. J., S. J. Stringer, and D. a. Marshall. 2013. Blueberry floral attributes and their
effect on the pollination efficiency of an oligolectic bee, Osmia ribifloris Cockerell
(Megachilidae: Apoidea). HortScience. 48: 1–7.
Scherm, H., D. S. Nesmith, D. L. Horton, and G. Krewer. 2001. A survey of horticultural and
pest management practices of the Georgia blueberry industry. Small Fruits Rev. 1: 17–28.
Schlueter, M. A., and N. G. Stewart. 2015. Native bee (Hymenoptera: Apoidea) abundance and
diversity in north Georgia apple orchards throughout the 2010 growing season (March to
October). Southeast. Nat. 14: 721–739.
Schoeneberger, P. J., D. A. Wysocki, and E. C. Benham. 2012. Field book for describing and
sampling soils, Agriculture. Natural Resources Conservation Service, National Soil Survey
Center, Lincoln, NE.

	

104	

Scott, Z. 2016. Native bee biversity and pollen foraging specificity in cultivated highbush
blueberry (Ericaceae: Vaccinium corymbosum L.) plantings in Rhode Island.
Scott, Z., H. S. Ginsberg, and S. R. Alm. 2016. Native bee diversity and pollen foraging
specificity in cultivated highbush Blueberry (Ericaceae: Vaccinium corymbosum ) in Rhode
Island. Environ. Entomol. nvw094.
Sedivy, C., and S. Dorn. 2014. Towards a sustainable management of bees of the subgenus
Osmia (Megachilidae; Osmia) as fruit tree pollinators. Apidologie. 45: 88–105.
Sedivy, C., C. J. Praz, A. Müller, A. Widmer, and S. Dorn. 2008. Patterns of host-plant
choice in bees of the genus Chelostoma: the constraint hypothesis of host-range evolution in
bees. Evolution (N. Y). 62: 2487–2507.
Senapathi, D., J. C. Biesmeijer, T. D. Breeze, D. Kleijn, S. G. Potts, and L. G. Carvalheiro.
2015. Pollinator conservation - the difference between managing for pollination services
and preserving pollinator diversity. Curr. Opin. Insect Sci. 12: 93–101.
Sgolastra, F., W. P. Kemp, J. S. Buckner, T. L. Pitts-Singer, S. Maini, and J. Bosch. 2011.
The long summer: Pre-wintering temperatures affect metabolic expenditure and winter
survival in a solitary bee. J. Insect Physiol. 57: 1651–1659.
Sgolastra, F., W. P. Kemp, S. Maini, and J. Bosch. 2012. Duration of prepupal summer
dormancy regulates synchronization of adult diapause with winter temperatures in bees of
the genus Osmia. J. Insect Physiol. 58: 924–933.
Shanks, A. M., and J. C. Hutton. 1976. Bartlett’s test. Biostatistics.
Sheffield, C. S. 2014. Pollination, seed set and fruit quality in apple: studies with Osmia lignaria
(Hymenoptera: Megachilidae) in the Annapolis valley, Nova Scotia, Canada. J. Pollinat.
Ecol. 12: 120–128.
Sheffield, C. S., S. Westby, P. G. Kevan, and R. F. Smith. 2008. Winter management options
for the orchard pollinator Osmia lignaria Say (Hymenoptera: Megachilidae) in Nova Scotia.
J. Entomol. Soc. Ontario. 139: 3–18.
Sheffield, C. S., S. M. Westby, R. F. Smith, and P. G. Kevan. 2008. Potential of bigleaf lupine
for building and sustaining Osmia lignaria populations for pollination of apple. Can.
Entomol. 140: 589–599.
Somerville, D. C. 2001. Nutritional value of bee collected pollens, Rural Ind. Res. Dev. Corp.
ANSW agriculture.
Stanley, C. A., T. L. Pitts-Singer, and J. Bosch. 2011. Influence of rough handling on Osmia
lignaria (Hymenoptera: Megachilidae) nest establishment in commercial orchards. J. Econ.
Entomol. 104: 750–752.

	

105	

Steffan-Dewenter, I., and S. Schiele. 2004. Nest-site fidelity, body weight and population size
of the red mason bee, Osmia rufa (Hymenoptera: Megachilidae), evaluated by markrecapture experiments. Entomol. Gen. 27: 123–131.
Stubbs, C., and F. A. Drummond. 1997. Blueberry and cranberry (Vaccinium spp.) pollination:
a comparison of managed and native bee foraging behavior. Acta Hort. 437: 341–344.
Stubbs, C., F. A. Drummond, and S. Allard. 1997. Bee conservation and increasing Osmia
spp. in Maine lowbush blueberry fields. Northeast. Nat. 4: 133–144.
Stubbs, C. S., H. A. Jacobson, E. A. Osgood, and F. A. Drummond. 1992. Alternative forage
plants for native (wild) bees associated with lowbush blueberry, Vaccinium spp., in Maine.
Maine Agric. Exp. Stn.
Tasei, J.-N., and P. Aupinel. 2008. Nutritive value of 15 single pollens and pollen mixes tested
on larvae produced by bumblebee workers (Bombus terrestris, Hymenoptera: Apidae).
Apidologie. 39: 397–409.
Thomson, D. 2004. Competitive Interactions between the Invasive European honey bee and
native bumble bees. Ecolo. 85: 458–470.
Thomson, D. 2006. Detecting the effects of introduced species: a case study of competition
between Apis and Bombus. OIKOS. 114: 407–418.
Topitzhofer, E. 2014. Effects of pollen collected by honey bees from pollination dependent
agricultural cropping systems on honey bee nutrition. Can. Entomol.
Torchio, P. F. 1982a. Field experiments with the pollinator species, Osmia lignaria propinqua
Cresson, in apple orchards: I, 1975 Studies (Hymenoptera: Megachilidae). J. Kansas
Entomol. Soc. 55: 136–144.
Torchio, P. F. 1982b. Field experiments with the pollinator species, Osmia lignaria propinqua
Cresson, in apple orchards: II, 1976 studies (Hymenoptera: Megachilidae). J. Kansas
Entomol. Soc. 55: 759–778.
Torchio, P. F. 1984. Field experiments with the pollinator species, Osmia lignaria propinqua
Cresson (Hymenoptera: Megachilidae) in apple Orchards: III, 1977 studies. J. Kansas
Entomol. Soc. 57: 517–521.
Torchio, P. F. 1989. In-nest biologies and development of immature stages of three Osmia
species (Hymenoptera: Megachilidae). Ann. Entomol. Soc. Am. 82: 599–615.
Torchio, P. F. 1990. Osmia ribifloris, a native bee species developed as a commercially
managed pollinator of highbush blueberry (Hymenoptera: Megachilidae). Source J. Kansas
Entomol. Soc. J. KANSAS Entomol. Soc. 63: 427–436.

	

106	

Torchio, P. F., and V. J. Tepedino. 1980. Sex ratio, body size and seasonality in a solitary bee,
Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae). Evolution (N. Y). 34:
993–1003.
Tuell, J. K., J. S. Ascher, and R. Isaacs. 2009. Wild bees (Hymenoptera: Apoidea: Anthophila)
of the Michigan highbush blueberry agroecosystem. Ann. Entomol. Soc. Am. 102: 275–
287.
VanEngelsdorp, D., J. Hayes, R. M. Underwood, and J. Pettis. 2008. A survey of honey bee
colony losses in the U.S., fall 2007 to spring 2008. PLoS One. 3: e4071.
VanEngelsdorp, D., and M. D. Meixner. 2010. A historical review of managed honey bee
populations in Europe and the United States and the factors that may affect them. J.
Invertebr. Pathol. 103: S80–S95.
Vicens, N., and J. Bosch. 2000a. Weather-dependent pollinator activity in an apple orchard,
with special reference to Osmia cornuta and Apis mellifera (Hymenoptera: Megachilidae
and Apidae). Environ. Entomol. 29: 413–420.
Vicens, N., and J. Bosch. 2000b. Pollinating efficacy of Osmia cornuta and Apis mellifera
(Hymenoptera: Megachilidae, Apidae) on “red delicious” apple. Popul. Ecol. 29: 235–240.
Vicens, N., and J. Bosch. 2000c. Nest site orientation and relocation of populations of the
orchard pollinator Osmia cornuta (Hymenoptera: Megachilidae). Environ. Entomol. 29: 69–
75.
Walther-Hellwig, K., G. Fokul, R. Frankl, R. Büchler, K. Ekschmitt, and V. Wolters. 2006.
Increased density of honeybee colonies affects foraging bumblebees. Apidologie. 37: 517–
532.
Werling, B. P., T. L. Dickson, R. Isaacs, H. R. Gaines, C. Gratton, K. L. Gross, H. Liere, C.
M. Malmstrom, T. D. Meehan, L. Ruan, B. Robertson, G. P. Robertson, T. M.
Schmidt, A. C. Schrotenboer, T. Teal, J. Wilson, and D. Landis. 2014. Perennial
grasslands enhance biodiversity and multiple ecosystem services in bioenergy landscapes.
Proc. Natl. Acad. Sci. U. S. A. 111: 1652–7.
West, T., and T. McCutcheon. 2009. Evaluating Osmia cornifrons as pollinators of highbush
blueberry. Int. J. fruit Sci. 9: 115–125.
White, J., Y. Son, and Y. Park. 2009. Temperature-dependent emergence of Osmia cornifrons
(Hymenoptera: Megachilidae) adults. J. Econ. Entomol. 102: 2026–2032.
Wilkaniee, Z., and K. Giejdasz. 2003. Effects of the pollination of highbush blueberry by the
solitary bee Osmia rufa L. (Apoidea: Megachilidae) under isolation. J. Apic. Sci. 47: 25–28.

	

107	

Williams, N. M. 2003. Use of novel pollen species by specialist and generalist solitary bees
(Hymenoptera: Megachilidae). Oecologia. 134: 228–237.
Williams, N. M., and V. J. Tepedino. 2003. Consistent mixing of near and distant resources in
foraging bouts by the solitary mason bee Osmia lignaria. Behav. Ecol. 14: 141–149.
Winfree, R., N. M. Williams, H. Gaines, J. S. Ascher, and C. Kremen. 2008. Wild bee
pollinators provide the majority of crop visitation across land-use gradients in New Jersey
and Pennsylvania, USA. J. Appl. Ecol. 45: 793–802.
Wuellner, C. T. 1999. Nest site preference and success in a gregarious, ground-nesting bee
Dieunomia triangulifera. Ecol. Entomol. 24: 471–479.
Yoon, H. J., K. Y. Lee, S. Y. Kim, Y. B. Lee, N. Kim, and B. R. Jin. 2015. Effects of location,
direction, altitude, and placement of trap nests on the rate of trap-nesting of Osmia solitary
bees. J. Asia. Pac. Entomol. 18: 695–700.
Zurbuchen, A., L. Landert, J. Klaiber, A. Müller, S. Hein, and S. Dorn. 2010. Maximum
foraging ranges in solitary bees: only few individuals have the capability to cover long
foraging distances. Biol. Conserv. 143: 669–676.

	

108