European bee-eaters (Merops apiaster) and apiculture: understanding their interactions and the usefulness of nonlethal techniques to prevent damage at apiaries

Abstract

Finding a balance between wildlife conservation and land use protection when conflicts arise requires applying measures based on scientific knowledge and applicability. This is because of the complexity of interactions and the social dimensions of human–wildlife relationships. One of the conflicts in the Mediterranean basin is between European bee-eaters Merops apiaster and beekeeping. The objective of this study was to evaluate the activity of bee-eaters in relation to honey bees Apis mellifera in Extremadura (W Spain), and the possible correlation with traits and viability of beehives. We also aimed to assess the effectiveness of damage prevention measures, to be able to predict whether they could assist in the future sustainability of apicultural management. A monitoring program was conducted in 58 apiaries to observe bee and bee-eater activity and abundance. Additionally, we measured the effect of environmental variables and the amount of resources in the hives, by means of honey, pollen, and brood. We found that bee-eater negatively correlated to honey bee activity and the more abundant it was the fewer resources (honey, pollen, and brood) and lower abundance of bees in the hives, mainly during their migratory season (August). However, there were no negative effects on the survival and viability of the hives. The preventive and deterrent measures offered promising results, especially the installation of shading meshes around the apiary that prevented the predation of bees in the immediate surroundings of the beehives. Further studies on the effectiveness of preventive techniques and their practical implementation through different mechanisms of economic compensation must be enhanced to reduce this social conflict.

Full text

ORIGINAL ARTICLE
European bee-eaters (Merops apiaster) and apiculture: understanding
their interactions and the usefulness of nonlethal techniques to prevent
damage at apiaries
Rubén Moreno-Opo
1,2
&Juan Carlos Núñez
3
&Manuel Pina
3
Received: 28 February 2018 /Revised: 3 July 2018 /Accepted: 21 August 2018
#Springer-Verlag GmbH Germany, part of Springer Nature 2018
Abstract
Finding a balance between wildlife conservation and land use protection when conflicts arise requires applying measures based
on scientific knowledge and applicability. This is because of the complexity of interactions and the social dimensions of human–
wildlife relationships. One of the conflicts in the Mediterranean basin is between European bee-eaters Merops apiaster and
beekeeping. The objective of this study was to evaluate the activity of bee-eaters in relation to honey bees Apis mellifera in
Extremadura (W Spain), and the possible correlation with traits and viability of beehives. We also aimed to assess the effective-
ness of damage prevention measures, to be able to predict whether they could assist in the future sustainability of apicultural
management. A monitoring program was conducted in 58 apiaries to observe bee and bee-eater activity and abundance.
Additionally, we measured the effect of environmental variables and the amount of resources in the hives, by means of honey,
pollen, and brood. We found that bee-eater negatively correlated to honey bee activity and the more abundant it was the fewer
resources (honey, pollen, and brood) and lower abundance of bees in the hives, mainly during their migratory season (August).
However, there were no negative effects on the survival and viability of the hives. The preventive and deterrent measures offered
promising results, especially the installation of shading meshes around the apiary that prevented the predation of bees in the
immediate surroundings of the beehives. Further studies on the effectiveness of preventive techniques and their practical
implementation through different mechanisms of economic compensation must be enhanced to reduce this social conflict.
Keywords Beekeeping .Extremadura .Honey bee .Human–wildlife conflict .Pollination .Preventive techniques
Introduction
The resolution of human–wildlife conflicts is a widespread
area of study in environmental management and research. It
aims to reconcile the conservation of wildlife species with
socioeconomic uses and human security (Woodroffe et al.
2005; Redpath et al. 2013). To achieve this compatibility, it
is necessary to implement solutions with a proven efficacy
that are truly applicable and ensure the favorable conservation
status of the populations under management (Conover 2002;
Caro 2007). Livestock plays a major role among the socioeco-
nomic scenarios in which most human–wildlife conflicts oc-
cur. In general, the most widely studied cases focus on the
predation of domestic or huntable ungulates, lagomorphs, or
poultry by carnivores and raptors (i.e., Conover 2002;
Woodroffe et al. 2005), as well as on health issues related to
epizootics shared between wild and domestic fauna (Gortázar
et al. Gortazar et al. 2015).
Another type of husbandry, beekeeping, has received less
attention with regard to human–wildlife conflict. There is an
increasing concern of beekeeper organizations about the pre-
dation by European bee-eater Merops apiaster and the alleged
negative effects on bees in the Mediterranean basin (Fry 1983;
Galeotti and Inglisa 2001; Farinós-Celdrán et al. 2016). The
European bee-eater is a gregarious bird species, especially
during the postbreeding and migratory periods, forming flocks
moving simultaneously (Fry and Fry 1992;FryandBoesman
2017). They feed mainly on medium and large-sized insects,
with Hymenoptera and Coleoptera the most represented
*Rubén Moreno-Opo
rmorenoopo@gmail.com
1
General Directorate on Biodiversity and Environmental Quality,
Ministry for the Ecological Transition, Pza. San Juan de la Cruz s/n,
E-28071 Madrid, Spain
2
Evolution and Conservation Biology Research Group, University
Complutense of Madrid, C/José Antonio Novais 12,
E-28040 Madrid, Spain
3
TRAGSATEC, C/Julián Camarillo 6B, 28037 Madrid, Spain
European Journal of Wildlife Research (2018) 64:55
https://doi.org/10.1007/s10344-018-1215-9

orders in the diet, along with Odonata and Orthoptera
(Galeotti and Inglisa 2001; Massa and Rizzo 2002;Research
Group in Conservation Biology-UNEX 2007;Farinós-
Celdrán et al. 2016). The intake for each bee-eater is estimated
to be 1333.39 ± 759.8 honey bees Apis mellifera/year in
Mediterranean areas (Farinós-Celdrán et al. 2016), which rep-
resent approx. 42% of its total diet (Research Group in
Conservation Biology-UNEX 2007).
Beekeeping provides resources such as honey, pollen, wax,
and royal jelly. It annually yields around1.79 M tons of honey
(22% in Europe) and 66,622 tons of beewax worldwide (Food
and Agriculture Organization 2018). Beekeeping employed
600,000 people in Europe in 2010 (Chauzat et al. 2013). It
also plays a key ecological role since honey bees contribute to
the pollination of angiosperm flora, which has become partic-
ularly important given the global decline of wild pollinating
insects (Potts et al. 2010). The main environmental threat fac-
ing beekeeping is the phenomenon of disappearance of bees in
different regions (Moritz and Erler 2016), caused by political
and socioeconomic changes in honey trade,toxic chemicals,
diseases and parasites, invasive alien species such as the Asian
hornet Vespa velutina, and issues related to a loss of genetic
diversity, overexploitation, or stress derived from inadequate
beekeeping practices (De la Rúa et al. 2009; Vanbergen 2013;
Doublet et al. 2015; Goulson et al. 2015; Moritz and Erler
2016). Moreover, the colony collapse disorder (CCD)—un-
derstood as the sudden loss of bee population in a hive, while
resources like honey, pollen, and/or brood are abundant—is
influenced by different factors (vanEngelsdorp et al. 2009;
Paxton 2010) but the role of the bee-eater generating this
phenomenon remains poorly known.
With the purpose of reducing or mitigating the potential
conflict between beekeeping and bee-eaters, we developed a
monitoring program with the dual purpose of quantifying the
way in which damage occurs, and proposing measures to pre-
vent or mitigate damage. In this sense, we installed three dif-
ferent preventive and/or deterrent measures in the nearby sur-
roundings or inside the apiaries with the aim of testing their
effectiveness in decreasing bee-eater abundance and, conse-
quently, improving the vigor of the hives. The present study is
based on data from the aforementioned program and had the
following objectives: (1) evaluate the correlation between bee-
eater and honey bee activity; (2) test whether beehive traits
(such as abundance of honey, pollen, brood, and bees) influ-
ence the bee-eater–honey bee correlation; and (3) assess the
effect of preventive measures of bee-eater damage that are
sustainable and applicable in beekeeping management. In par-
ticular, we hypothesize that (1) there is a link between the
abundance of bee-eaters and bees, (2) there are changes in
the vigor of the hives in terms of honey, pollen, brood, and
bee abundance levels related to the abundance of bee-eaters
feeding in the surroundings of the apiary, and (3) there is a
comparative increase in the vigor of the hives and a decrease
in bee-eater activity in apiaries in which preventive measures
are applied.
Methods
Study area and period
The study was carried out in the region of Extremadura
(Western Spain, Fig. 1). This region has a great importance
for apiculture with 471,350 beehives representing 19.2% of
the 2.4 M totals existing in Spain and 3.4% of the 13.8 M
totals in Europe (Chauzat et al. 2013). Extremadura is the
top producer in wax and pollen and the third largest producer
in honey in Spain with 6450 t annually (MAGRAMA
2013).There has been an increasing amount of concerns raised
by apicultural organizations about supposed damage from
bee-eaters. In the region, bee-eaters are widely distributed
with an estimated breeding population of 1.3 M individuals
(Carrascal and Palomino 2008). Additionally, a contingent of
pre-migratory and migratory birds from other Western
European regions (Yosef et al. 2006; Valera et al. 2011)also
stop over at apiaries in Extremadura during August and
September.
The apiary and the beehives within each apiary were con-
sidered as experimental units in our study. We selected eight
areas based on the following criteria: (1) high density of api-
aries; (2) higher number of complaints by beekeepers about
bee-eaters damage, based on official records of the regional
authority (Junta de Extremadura) in the 2010–2012 period; (3)
willingness of the beekeeper to collaborate in the study; and
(4) minimization of the length of the routes made by field
technicians for implementing the monitoring activities to re-
duce economic costs and emissions of greenhouse gases
(Fig. 1).
We selected 47 different apiaries made up of about 80
beehives each (Table 1). During the 2 years of study (2013
and 2015), a total of 58 apiaries from the 47 initially se-
lected were chosen for their monitoring, such that 11 api-
aries were sampled both in 2013 and in 2015. Within each
selected apiary and before the start of the study, a subset of
12 hives from the about 80 composing the apiary was se-
lected. These 12 hives were marked and identified with
permanent stickers, to quantify their characteristics (see
BFieldwork^and BVariables considered^). These hives
were randomly chosen, avoiding those with symptoms of
disease or poor honeycomb condition.
The present study was carried out in 2013 and 2015.
Monitoring programs for bee-eaters and honey bees
55 Page 2 of 11 Eur J Wildl Res (2018) 64:55

abundance and activity in apiaries were developed between
April and September. During late September, the state of the
hives was evaluated to determine their vigor and its relation-
ship with bee-eater effects.
Tested preventive measures
The preventive measures were placed in March before starting
the field work and removed in October after the bee-eaters left
the study areas. In addition to the apiaries with deterrent
systems, there were control apiaries without these systems
with which to compare results (Table 1). None of the control
apiaries disposed of any of the elements of the measures test-
ed, including drinking troughs. The three deterrent and pre-
ventive techniques were the following (Fig. 2):
1. Decoys of bee-eater’s predatory species along with
sounds (Bdecoys^). This consisted of the installation of
eagle owl Bubo bubo and buzzard Buteo sp. decoys with
autonomous mobility placed above the hives,
Fig. 1 aDistribution of the
European bee-eater Merops
apiaster in the circum-
Mediterranean region during the
breeding season (April to
September, source: IUCN, http://
maps.iucnredlist.org/map.html?
id=22683756). bRegion of
Extremadura (in box) within
Spain (in gray). cDelimitation of
the working areas selected to
assess the effect of the European
bee-eater in beekeeping in
Extremadura in 2013 and 2015.
The points in gray show the
locations of all the beekeeping
settlements in Extremadura in
2012. Within eight areas (striped
polygons), the apiaries selected
are shown as black points,
indicating the number of apiaries
evaluated in each area (47 in total)
Eur J Wildl Res (2018) 64:55 Page 3 of 11 55

accompanied by a device emitting repelling sounds of
predatory species and alarms of different birds
(BirdXpeller and Sentinel models) powered with conven-
tional (AA) and/or rechargeable batteries (22A-70A).
2. Perching structure triggering minor electric shocks
(Belectrified perch^). This consisted of the installation of
a device on which bee-eaters could perch, formed by par-
allel wires held up with tension posts, generating electric
shocks when the bird perched on the wires. This dis-
charge was considered high voltage (approx. 4000 V)
but for a short duration, so as to not physically damage
the bird and only to induce a negative experience associ-
ated with the hives. The structure was set up around and
above the hives.
3. Protection of the apiary with meshes, nets, and a water
supply (Bshading^). A shaded mesh was placed above all
the hives of the apiary, supported by poles and tensioners,
and at the lowest possible height above the hives. The
shading mesh was complemented with plastic side nets
to prevent the access of bee-eaters to the inmediate prox-
imity of the hives. Drinking troughs were placed among
the hives so that bees had a water supply as close as
possible to their hives.
We ensured that tested systems would not infringe on the
legal protection of the bee-eater at the European level (Birds
Directive 2009/147/EC). Namely, we avoided systems that
impacted the life prospects of the birds (live or dead captures,
shots, changes in natural habitat, etc.)
Fieldwork
We conducted fieldwork for monitoring the target species dur-
ing the season in which the bee-eater is present in Spain (April
to September). For the bee-eater census, an observer was lo-
cated between 100 m and 150 m from the hives, so that the
human presence would not interfere with the bee-eater activ-
ity. To determine the distance at which the bee-eaters were
detectable, previous training on the auditory capacity of the
observers was carried out, using bee-eaters sound recordings.
The maximum linear distance at which it was possible to
detect a bee-eater flock was 463 ± 19.9 m, with light breeze
conditions and open land without trees. This distance offers a
proxy of the radius with respect to the observation point, so
that we ensured that the counted bee-eaters were in the close
surroundings of the hives. Binoculars (8.5–10 × 42) were used
to recognize the activity of the bee-eaters at the greatest pos-
sible distance. The census effort occurred on a monthly basis
in each apiary except for August, in which we perfomed cen-
sus in two different days due to the predicted greater abun-
dance of bee-eaters around the apiaries, as indicated by the
beekeepers. In each apiary, 2 hours per day were invested,
distributed homogenously along different periods of the day.
During these 2 hours, bee-eaters seen or heard by the observer
were counted, according to their activity (totals,interacting
with hives,andattempting to capture bees,seeTable2).
To monitor honey bee activity, a bee counter was installed
on each census day in two different hives. The bees that en-
tered and left the beehive were counted such that an activity
index of bee movements was obtained by calculating a joint
mean of 1-hour movements in the two hives for each day in
the apiary. The counter model used was the ApiSCAN-Plus
(Lowland Electronics).
In addition to the direct observation of bee-eater predatory
activity and bee movements, the internal temperature of the
hive was used as an index related to the bee activity and
colony vigor (Seeley 2014). In March, digital electronic re-
corders were placed in two hives of each apiary recording data
continuously until their removal in September. These were
placed in top of the hives, between the upper edge of the
frames, and the cover of the hive. It was not possible to install
the recorders in the brood chamber, which could have provid-
ed more accurate data in the most sensitive area of the hive.
The recorders were programmed to obtain temperature re-
cords every 10 min.
An internal assessment of the 12 hives selected in each
apiary was carried out in September and October once the
bee-eaters left the study areas after their post-breeding migra-
tion. We tried to determine the amount of elements related to
the vigor of the hive (Seeley 2009; Odoux et al. 2014).
Specifically,fromeachhivewenoted:
Table 1 Total number of apiaries sampled and their distribution
between sedentary (present in the same location throughtout the study
period) or migratory (change of location and moving of apiaries to or
from Extremadura during the period of presence of the European bee-
eater Merops apiaster, between April to September)
Type of management Total apiaries
sampled
Type of preventive-deterrent
measure
Sedentary Migratory
Control 36 11 47
Decoys 3 3
Electrified perch 3 3
Shading 5 5
Tot al 47 11 58
Also shown is the type of preventive-deterrent technique applied (Decoys
= activation of predatory decoys + alarm sounds of birds of prey;
Electrified perches = installation of elevated wires to perch on with sup-
ply of minor electric discharges; and Shading = apiary protected by mesh
shade, side nets and water supply) that would limit and/or deter access to
apiaries by the European bee-eater,and apiaries without measures (con-
trol), in the eight study areas of Extremadura (WSpain) in 2013 and 2015
55 Page 4 of 11 Eur J Wildl Res (2018) 64:55

&Amount of honey and pollen, as the number of sides (un-
derstood as the two parts into which the frames or combs
that make up the hive are divided) filled with stored honey
and/or pollen in a proportion greater than ¼ of its area;
&Amount of brood, as the number of sides of each frame
with presence of bees brood (eggs or larvae);
&Amount of honey,pollen,and brood, as the sum of the
number of sides of the two previous variables;
&Amount of bees, as the number of sides with more than
half of their area filled by bees;
&Weig ht in kg. of the whole hive;
&Collapsed or unviable beehive (yes/no), recognizing the
presence of honeycombs collapsed by heat, the absence of
a queen bee in the hive through the observation of the
proportion of drones to workers, or visual signs of the
presence of diseases (varroasis, chalkbrood, beetles, etc.;
Higes et al. 2010; Sammataro and Yoder 2011).
Variables considered
A preliminary correlation analysis was made with the vari-
ables under study, selecting those with a higher correlation
coefficient and offering, therefore, a greater explanatory pow-
er in subsequent analyses. In particular, the variables related to
the vigor of the hives’amounts of honey–pollen,brood,and
bees showed high correlation coefficients among each other
(always higher than ρ=0.53).Theweight of the hive showed
a greater correlation with the amount of honey,pollen,and
brood (ρ= 0.53). For this reason, the variable amount of bees
Fig. 2 Preventive-deterrent
measures tested for evaluating
their effectiveness in reducing the
interaction of the European bee-
eater Merops apiaster in
beekeeping in Extremadura in
2013 and 2015. aDecoys of
predatory species of the bee-eater,
with sounds (decoys). bParallel
wires held up with tension posts
for bee-eaters to perch on,
generating electric shocks
(electrified perch). c,dProtection
oftheapiarywithmeshesand
nets, and water supplied
(shading)
Eur J Wildl Res (2018) 64:55 Page 5 of 11 55

Table 2 Variables measured during this study to analyze the interaction between European bee-eater Merops apiaster and honey bee Apis mellifera in Extremadura (W Spain), indicating the scope of
study of each variable, if the variable was considered as response (dependent) or explanatory (independent), and quantitative or qualitative in the analyses and their definition
Scope of the variable Variable Type Characteristics Definition
Bee-eater activity Tot al nu mb er o f
bee-eaters observed
Response and explanatory Quantitative Number of individuals detected (seen/heard) during the field censusin the surroundings of the apiary, independent of
their activity
Number of bee-eaters
interacting
Response and explanatory Quantitative Number of individuals observed pursuing and/or flying over the hives at low height (< 20 m approx.)
Number of bee capture
attempts by
bee-eaters
Response and explanatory Quantitative Number of attempts, succesful or not, of capture of bees by bee-eaters through pursuit in flight or stalking and
capture from perches
Vigor of the beehive Amount of honey, pollen
and brood
Response Quantitative Sum of the number of sides (as the two parts into which the frames or combs that make up the hive are divided) filled
with stored honey and/or pollen in a proportion greater than ¼ of its area, and the number of sides of each frame
with presence of bees brood (eggs or larvae)
Amount of bees Response Quantitative Number of sides filled by more than half of their area by bees
Collapsed or unviable
beehive
Response Qualitative Presence (yes/no) of honeycombs collapsed by heat, the absence of a queen bee in the hive through the observation
of the proportion of drones to workers or signs of presence of diseases (varroasis, chalkbrood, beetles)
Honey bee activity Honey bee activity Response and explanatory Quantitative Number of bees entering or departing the beehive per hour
Temperature in the
beehive
Mean temperature Response and explanatory Quantitative Mean temperatures (in °C) in two beehives of each apiary
Maximum temperature Response and explanatory Quantitative Maximum temperature (in °C) in two beehives of each apiary
Minimum temperature Response and explanatory Quantitative Minimum temperature (in °C) in two beehives of each apiary
Preventive-deterrent
measures
Presence of
preventive-deterrent
measure
Explanatory Qualitative Presence/absence of preventive-deterrent measures towards the bee-eater
Type of
preventive-deterrent
measure
Explanatory Qualitative With the options control (without preventive-deterrent measure), electrified perch (parallel wires hold up with
tension posts for bee-eaters to perch on, generating electric shocks when the bird perches on the wires), shading
(shaded mesh above the apiary with plastic side nets and drinking troughs), and decoys (eagle owl Bubo bubo and
buzzard Buteo sp. decoys with autonomous mobility, and repelling sounds of predatory species and alarms)
Weather Environmental
temperature
Explanatory Quantitative Environmental temperature in °C at the beginning of field census
Cloudiness Explanatory Qualitative BCloudy^(> 75% of the sky covered by clouds), Bfew clouds^(> 25 to 75% of the sky covered by clouds), Bsunny^
(<25%oftheskycoveredbyclouds)
Rain Explanatory Qualitative Presence/absence of rain during the field census
Time conditions Hour Explanatory Quantitative Starting hour of the field census
Month Explanatory Qualitative April to September
Period Explanatory Qualitative Regarding the life-cycle of the bee-eater: Bbreeding^(April to July) and Bmigration^(August and September)
Type of management Migration Explanatory Qualitative Movement of the apiary from its location, assigning to the categories Bsedentary^(present in the same location
throughout the study period) or Bmigratory^(change of location and moving to or from Extremadura between
April and September)
Type of beehive Explanatory Qualitative According to its model, assigning to the categories BLayens^or BLangstroth^
Number of hives in the
apiary
Explanatory Quantitative Number of beehives that make up the apiary
55 Page 6 of 11 Eur J Wildl Res (2018) 64:55

and amount of honey,pollen,and brood were selected to de-
velop the analyses related to the vigor of the hive, given also
the high correlation of the latter with the amount of honey and
pollen (ρ= 0.78) and with the amount of brood (ρ=0.56).
A total of 21 variables were selected, of which 10 were
responses and 18 were explanatory, with seven evaluated both
as responses and as explanatory (Table 2). The analyses were
carried out according to the hypotheses reported: thus, the
group of variables related to the activity of the bee-eaters
was analyzed with respect to bee activity, preventive-
deterrent measures, weather, time conditions, type of manage-
ment, and temperatures inside the hive. The variables related
to the vigor of the hives were related to the activity of the bee-
eaters and bees, the temperature in the hive and the
preventive-deterrent measures. The activity of the bees was
analyzed with respect to the activity of the bee-eaters, the
temperature in the hive, the preventive-deterrent measures,
the weather, and the time conditions, while the temperature
in the hive was evaluated in relation to the existence of
preventive-deterrent measures.
Statistical analyses
We first explored the quantitative values of the different var-
iables to check their adjustment to a normal distribution, in
order to apply parametric analysis. None of the variables, even
after logarithmic transformation, showed a normal distribution
based on Shapiro–Wilk and Kolmogorov–Smirnov tests, so
nonparametric analyses were carried out in all cases by pairs
of variables (response and explanatory).
For the analysis of quantitative response variables and
qualitative explanatory variables, a Mann–Whitney/
Wilcoxon test was used to compare means in the case that
the explanatory variable was divided into only two categories.
When the explanatory variable was assigned to more than two
categories, a comparison of means was made using a Kruskal–
Wallis test. When both variables (response and explanatory)
were quantitative, a nonparametric correlation analysis was
applied, showing the correlation coefficients (Spearman’sR)
and a ttest for comparison of samples. In the case of categor-
ical response variables, frequency homogeneity tests were
performed for each category, by means of a chi-square analy-
sis. The complete results (statistics, degrees of freedom, and
values of statistical significance) of the analyses that were
statistically significant with a confidence level greater than
95% (p< 0.05) are shown. All analyses were conducted with
the software Statistica 7.0 (StatSoft, Tulsa).
Results
We made 307 field visits (about 6 visits/apiary per year) for
censusing bee-eaters and obtaining bee activity data, from
April 9 to September 13, in 2013 and 2015. We also examined
696 beehives for the evaluation of their status and vigor, be-
tween September 19 and October 10.
European bee-eater predatory activity
During the fieldwork, a total of 6533 bee-eaters were recorded
(mean of 21.2 ± 41.4 per census), of which 3004 were ob-
served interacting with the apiaries, as well as 8456 registered
capture attempts of bees. The number of bee-eaters interacting
with apiaries was related to the activity of the bees: there was a
marginally significant negative relationship between numbers
of bee-eaters and the number of the bees going in and out of
the hive (R=−0.196, p= 0.061, Fig. 3). The same was true
for the number of times bee-eaters attempted to capture bees
(R=−0.201, p=0.018,Fig. 3).
The bee-eaters interacting with apiaries and the number of
attempts to capture bees were lower in the apiaries protected
with shading meshes compared to the controls (Z=1922,p=
0.018, and Z= 2372, p= 0.017, respectively; Fig. 4). There
were no differences in the number of bee-eaters at apiaries
with other measures with respect to the control apiaries (elec-
trified perch p = 0.423 and p= 0.715, decoys p = 0.521 and
p= 0.477, for bee-eaters interacting with apiaries and capture
attempts, respectively).
Rainy days increased the number of bee-eaters interacting
with apiaries (Z=16.89, p< 0.001) as did a higher environ-
mental temperature (R= 0.131, t= 2.214, p=0.027).
Regarding the time conditions, the interacting bee-eaters were
more abundant during the migratory period (Z=−2850, p=
0.004), especially during August (H
7,307
=77.48, p<0.001).
Similarly, in the middle hours of the day, a lower number of
bee-eaters interacted with apiaries although in a marginally
significant way (R= 0.113, t=1.759,p=0.071).
Honey bee activity
The bee-movement recorders placed for 1 hour in two hives of
each apiary recorded a total of 3,160,560 entries/departures of
bees from the hives (mean 4087.9 ± 4672.0 bees/h). Once the
negative correlation of the bee activity with the abundance of
bee-eaters was shown (Fig. 3), a significant effect of weather
variables was also indicated. Thus, higher values of tempera-
ture resulted in a higher bee activity rate, for the environmen-
tal temperature (R=0.197,t=3.178,p= 0.001), for the mean
temperature inside the hive (R=0.211; t=2.815; p=0.005),
and for the maximum and minimum temperatures recorded
(R=0.175, t=2.319, p=0.021 and R=0.248, t= 3.351,
p<0.000,respectively).
The activity of the bees was related to the type of
preventive-deterrent measure (H
3,276
=8.260,p= 0.049), with
the apiaries with decoys having a greater number of bees en-
tering and leaving the hives and the apiaries with shading
Eur J Wildl Res (2018) 64:55 Page 7 of 11 55

showing the least bee traffic with respect to the control apiar-
ies. Regarding the time conditions, the number of bees/h en-
tering and leaving the hive was proportionally higher in the
middle hours of the day (R=0.191, t=3.087, p=0.002) and
during June and July (H
4,277
=36.95, p<0.001).
State and vigor of the beehives
From the inspections carried out in 696 beehives, we observed
evidence of inviability due to diseases, melting of honeycomb
wax, or queen bee loss in 60 cases (8.6%). The stored honey,
pollen, and brood was negatively related to a greater number
of interacting bee-eaters (R=−0.113; t=−3.007; p=0.002,
Fig. 5). The amount of bees in the hive was also lower when a
higher number of bee-eaters were interacting (R=−0.212, t=
−5.689, p<0.001,Fig.5). Collapsed or unviable hives due to
the melting of honeycombs, loss of queens, or diseases were
not related to greater numbers of bee-eater (p= 0.987). On the
other hand, the higher activity of bees did not influence the
amount of honey, pollen, and brood (p=0.199).
The vigor of the hive was related to the presence of
preventive-deterrent measures. Thus, the stored honey, pollen,
and brood was higher in the apiaries where measures were
implemented to dissuade bee-eaters (Z=−2420, p=0.015).
Specifically, in the apiaries with decoys the beehives showed
a higher rate of frames filled with honey, pollen, and brood
(H
3,689
= 13.36, p= 0.003) and with bees (H
3,691
= 19.50,
p<0.001).
Temperatures inside the beehive
The mean temperature inside all hives was 31.2 ± 1.6 °C. The
maximum temperature recorded was 59.6 °C and the mini-
mum 6 °C. The mean temperature during the entire study
period varied according to the type of deterrent measure im-
plemented (H
3,183
= 11.08, p= 0.011), with apiaries with
shading showing values 3–6 °C lower than the rest. In this
sense, the shaded apiaries tempered the extreme temperatures
with respect to the control apiaries, the maximums being low-
er (Z=3.872, p< 0.001) and the minimum temperatures less
different from the mean (Z=2.227;p=0.025).
The temperatures inside the hive showed no significant
results regarding the vigor and possibilities of survival of the
Fig. 3 Number of European bee-
eaters Merops apiaster interacting
with apiaries (right Y-axis, empty
squares) and number of attempts
to capture honey bees Apis
mellifera (left Y-axis, black dots)
with respect to the number of
honey bees/h. This latter rate was
recorded in the counterslocated at
the entrance of two beehives in
each studied apiary. Linear
regressions are shown with
continuous (bee-eaters
interacting) and dashed (capture
attempts) lines
Fig. 4 Mean (± 95% CI) number of Europeanbee-eaters Merops apiaster
interacting with apiaries (empty squares) and the number of attempts to
capture honey bees Apis mellifera (black dots) in the surroundings of the
apiaries of types control (without preventive-deterrent measures) and
shading (with protecting meshes and nets, and water supply), observed
during bee-eater censuses
55 Page 8 of 11 Eur J Wildl Res (2018) 64:55

hive, except in the case of the the minimum temperatures
which were negatively related to the amount of honey, pollen,
and brood as well as the amount of bees (R=0.212,t=4.795,
p<0.001,andR=0.122,t=2.688,p=0.024).
Discussion
Our study is the first to explore in depth the alleged impacts of
bee-eaters on beekeeping and its viability. However, this must
be considered a first approach to an issue more geographically
widespread, given that bee-eater could have different impacts
in other regions (Galeotti and Inglisa 2001; Alfallah et al.
2010; Farinós-Celdrán et al. 2016). This study was carried
out in 2 years which limits the scope of more robust conclu-
sions due to the limited weather variability included. In addi-
tion, the sample size to test the effectiveness of the different
preventive measures would have to be augmented in future
complementary studies in relation to the number of control
apiaries, with the purpose of strengthening the robustness of
our results. Logistic and economic limitations prevented us to
increase this test of preventive measures on a wider scale.
Interaction between European bee-eaters and honey
bees
A negative relationship between the activities of the bee-eater
and the bees was indicated so this outcome is discussed in
terms of cause–effect implications. On the one hand, bee ac-
tivity could determine the presence of the bee-eater: a greater
foraging activity of the domestic honey bees would coincide
with a greater general entomological abundance and availabil-
ity according to weather conditions (Vicens and Bosch 2000),
such that bee-eaters would dilute their effect in the apiaries
and would predate to a greater extent on natural prey (Fry
1983). However, when honey bees are scarce, wild insects
are also scarce due to the influence of weather (Motten
1986; Lee and Barnard 2015). Subsequently, bee-eaters are
forced to feed at locations where preys congregate such as
apiaries (Cramp 1998). On the other hand, the bee-eater could
determine bee activity through the inhibition of their normal
behavior due to their repeated presence around the apiary. This
would entail a general reduction in the movements of entrance
to and departure from the hives, foraging activity, obtaining
water, the ventilation of the hive, and the breeding output
when detecting the presence of predators (Research Group
in Conservation Biology -UNEX 2007).
Another important finding is that the greatest interaction
occurs during bee-eater migration. After the breeding season,
the recruitment of juveniles and the arrival of breeding birds
from other latitudes increase the population of bee-eaters in
our study region. In addition, the abundance of wild insects
decreases in August with respect to spring (Herrera 1988;
Bosch et al. 1997).
Condition of the beehives
The condition of the hives correlated with the activity of the
bee-eaters, thus inferring a lower yield due to a greater pres-
ence of the bird. This does not mean these hives are no longer
viable, given that those actually detected as unviable in our
study (due to diseases, lack of a queen bee, etc.) did not cor-
relate with bee-eaters. Regarding temperature, it was remark-
able that the amount of honey, brood, and bees was negatively
affected by low temperatures reducing the potential yield of
the hive. In this sense, bees are favored by high temperatures,
Fig. 5 Linear regression between
the number of sides of frames of
the beehive with stored honey,
pollen, and brood (black dots,
continuous line) and the number
of sides of frames of the beehive
filled by more than half of their
area by honey bees Apis mellifera
(crosses,dashed line) with respect
to the number of European bee-
eaters Merops apiaster interacting
with the apiary
Eur J Wildl Res (2018) 64:55 Page 9 of 11 55

which increase bee activity, but not at extreme levels that
could have negative effects on their survival prospects
(Seeley 2009,2014).
According to our results, the use of honeycombs filled with
brood, honey, and bees as an indicator can be useful to deter-
mine the prospective survival and vigor of the hive.
Nevertheless, it is important taking into consideration that
several anthropic (hive management, number of honey or pol-
len harvests, use of pesticides, etc.) and natural circumstances
(type and abundance of flowering at regional or local scale)
could determine the amount of resources and bees in the hives
(Odoux et al. 2014).
Effectiveness of preventive measures
In the case of the bee-eaters interacting with apiaries, the
shading was the only measure that protected the immediate
surroundings of the hives and favored honey bees in three
ways. First, this measure allows bees to leave the hives safely
without being predated in the first few meters from the hive
until they passed the protection mesh (see Fig. 2) which less-
ened the mortality risk. Secondly, the water supply inside the
mesh reduces the need for long trips during the summer in
search of this key resource (Seeley 2009; Kühnholz and
Seeley 1997). In fact, in our study, we checked for a continued
use of the drinking troughs by a large number of bees, which
in any case, is a good beekeeping practice (Sammataro and
Avitabi le 1998). To really know the effect of water supply
in improving the condition of hives, with respect to the rest
of the elements composing of the shading measure, we
should have tested it in control apiaries as a covariate.
But we did not separate it from the other elements of the
shading measure and could not check its own effectiveness
in optimizing bee activity.Finally, the apiaries with shading
showed more temperate temperatures. The reduction of
very high temperatures would reduce the hazard of
overheating (Southwick and Heldmaier 1987;Seeley
2009,2014), as well as a less need for movements of bees
outside the hive to obtain water and to ventilate. Thus, the
temperature of the brood chamber, which must remain con-
tinuouslybetween32°Cand36°C(SeelyandHeinrich
Seeley and Heinrich 1981), would be lower.
Management implications
The present study has shown that the European bee-eater is a
factor to consider in the field of beekeeping management, and
especially during post-breeding migration when its population
increases. A greater abundance of bee-eaters was negatively
related to honey bee performance and to the amount of stored
honey, pollen, and brood in the hives. However, bee-eater
does not influence beehive survival and productivity, which
is the main concern of beekeepers. In fact, mortality was
detected in only 8.6% of beehives and most of them showed
diseases from months prior to the study, not attributable to the
bee-eater. Extremely high or low temperatures and diseases
could be important factors to consider in beekeeping to reduce
beehive losses.
The shading technique reduced the level of predation of
bees in the immediate surroundings of the apiary, attenuated
the extreme temperatures that disfavored apicultural produc-
tion, and could lower bee efforts in the search of water outside
the area protected with nets and mesh. The approximate cost
of this measure is 1200-1500€for an apiary of 80 hives, in
terms of equipment and staff for its installation.
It is recommended the implementation of shading roofs in
those apiaries that remain in the same location during the
period of bee-eater presence. In other apiaries, this measure
may not be viable due to the need for assembly and disman-
tling of the structure or the possible reluctance of the land-
owners where the apiaries are located, which often are not the
beekeepers themselves. In addition, there are other schemes
that can help in implementing these preventive and other com-
pensatory measures, such as payments to beekeepers via dif-
ferent mechanisms (Conover 2002; Nyhus et al. 2003;
Woodroffe et al. 2005). First, financing programs of agri-
environmental measures from the Common Agricultural
Policy of the European Union (mainly FEADER and
FEAGA) that are channeled through the National
Apicultural Program in Spain already consider these types of
preventive measures (MAGRAMA 2013), and thus should be
utilized more. Second, the promotion and enhancement of
insurance to mitigate wildlife damage in apiaries could pro-
vide compensatory payments. Thus, a specific insurance line
for apiculture affected by bee-eaters subsidized with public
resources has been recently enabled in Spain (MAGRAMA
2017).
Acknowledgements R. Higuero, P. Montoto, F. Aparicio, F. Guil, J. M.
Julián, A. Mogena, M. Abascal, D. Muñoz, R. Martínez, and L. M.
González helped in different phases of the fieldwork. This work was
carried out within the framework of the project Guidelines for the assess-
ment and prevention of damage from bee-eater to apiculture in
Extremadura, commissioned by the Spanish Ministry of Environment
and developed by TRAGSATEC. The Junta de Extremadura provided
essential help and authorized the implementation of the project (M. J.
Palacios, E. Jiménez, A. Civantos). We are obliged to the beekeepers
who provided access to their apiaries and thank them for their fruitful
collaboration.
References
Alfallah HM, Alfituri M, Hmuda M (2010) The impact of bee eater
Merops apiaster on the behavior of honey bee Apis mellifera L.
during foraging. Mansoura J Plant Protect Pathology 1:1023–1034
Bosch J, Retana J, Cerdá X (1997) Flowering phenology, floral traits and
pollinator composition in a herbaceous Mediterranean plant com-
munity. Oecologia 109:583–591
55 Page 10 of 11 Eur J Wildl Res (2018) 64:55

Caro T (2007) Behavior and conservation: a bridge too far? Trends Ecol
Evol 22:394–400
Carrascal LM, Palomino D (2008) Las aves comunes reproductoras en
España. Población en. SEO/BirdLife, Madrid, pp 2004–2006
Chauzat M-P, Cauquil L, Roy L, Franco S, Hendrikx P, Ribière-Chabert
M (2013) Demographics of the European Apicultural Industry.
PLoS One 8:e79018
Conover M (2002) Resolving human-wildlife conflicts: the science of
wildlife damage management. Lewis Publisher, Florida
Cramp S (ed) (1998) The complete birds of the Western Palearctic con
CD-Rom. Oxford University Press, Oxford
De la Rúa P, Jaffé R, Dall'Olio R, Muñoz I, Serrano J (2009) Biodiversity,
conservation and current threats to European honeybees. Apidologie
40:263–284
Doublet V, Labarussias M, Miranda JR, Moritz RF, Paxton RJ (2015)
Bees under stress: sublethal doses of a neonicotinoid pesticide and
pathogens interact to elevate honey bee mortality across the life
cycle. Envir Microbiol 17:969–983
Farinós-Celdrán P, Zapata VM, Martínez-López V, Robledano F (2016)
Consumption of honey bees by Merops apiaster Linnaeus, 1758
(Aves: Meropidae) in Mediterranean semiarid landscapes: a threat
to beekeeping? J Apicultural Res 55:193–201
Food and Agriculture Organization (2018) FAOSTAT; Livestock prima-
ry. 2016. FAO-United Nations. http://www.fao.org/faostat/en/#data/
QL
Fry CH (1983) Honeybee predation by bee-eaters, with economic con-
siderations. Bee World 64:65–78
Fry H, Boesman P (2017) European bee-eater (Merops apiaster). In: del
Hoyo J, Elliott A, Sargatal J, Christie DA, De Juana E (eds)
Handbook of the birds of the world alive. Lynx Edicions, Barcelona
Fry CH, Fry K (1992) Kingfishers, bee-eaters and rollers. A&C Black,
London
Galeotti P, Inglisa M (2001) Estimating predation impact on honey bees
Apis mellifera L. by European bee-eater Merops apiaster L. Revue
d’Ecologie (Terre Vie) 56:373–387
Gortazar C, Diez-Delgado I, Barasona JA, Vicente J, De La Fuente J,
Boadella M (2015) The wild side of disease control at the wildlife-
livestock-human interface: a review. Front Vet Sci 1:27
Goulson D, Nicholls E, Botías C, Rotheray EL (2015) Bee declines driv-
en by combined stress from parasites, pesticides, and lack of flowers.
Science 347:1255957
Herrera J (1988) Pollination relationships in southern Spanish
Mediterranean shrublands. J Ecol 76:274–287
Higes M, Martín-Hernández R, Martínez-Salvador A, Garrido-Bailón E,
González-Porto AV, Meana A, Del Nozal MJ, Bernal J (2010) A
preliminary study of the epidemiological factors related to honey
bee colony loss in Spain. Envir Microbiol Rep 2:243–250
Kühnholz S, Seeley TD (1997) The control of water collection in honey
bee colonies. Behav Ecol Sociobiol 41:407–422
Lee AT, Barnard P (2015) Spatial and temporal patterns of insect-order
activity in the fynbos, South Africa. J Entomol Zool Stud 3:95–102
MAGRAMA (2013) Programa nacional de medidas de ayuda a la
apicultura. España:2014–2016 http://www.mapama.gob.es/es/
ganaderia/temas/produccion-y-mercados-ganaderos/Plan_
Nacional_Ap%C3%ADcola_2014-2016_tcm7-311228.pdf
MAGRAMA (2017) Orden APM/617/2017, de 12 de junio, por la que se
definen las explotaciones, animales y producciones asegurables (…)
con el seguro de explotación de apicultura (…) Plan de Seguros
Agrarios Combinados. Boletín Oficial del Estado 153:54723–54737
Massa B, Rizzo MC (2002) Nesting and feeding habits of the European
bee-eater (Merops apiaster L.) in a colony next to a beekeeping site.
Avocetta 26:25–31
Moritz RF, Erler S (2016) Lost colonies found in a data mine: global
honey trade but not pests or pesticides as a major cause of regional
honeybee colony declines. Agric Ecosyst Envir 216:44–50
Motten AF (1986) Pollination ecology of the spring wildflower commu-
nity of a temperate deciduous forest. Ecol Monogr 56:21–42
Nyhus P, Fischer H, Madden F, Osofsky S (2003) Taking the bite out of
wildlife damage the challenges of wildlife compensation schemes.
Conserv Pract 4:37–43
Odoux JF, Aupinel P, Gateff S, Requier F, Henry M, Bretagnolle V (2014)
ECOBEE: a tool for long-term honey bee colony monitoring at the
landscape scale in West European intensive agroecosystems. J
Apicultural Res 53:57–66
Paxton RJ (2010) Does infection by Nosema ceranae cause Bcolony
collapse disorder^in honey bees (Apis mellifera)? J Apicultural
Res 49:80–84
Potts SG, Biesmeijer JC, Kremen C, Neumann P, Schweiger O, Kunin
WE (2010) Global pollinator declines: trends, impacts and drivers.
Trends Ecol Evol 25:345–353
Redpath SM, Young J, Evely A, Adams WM, Sutherland WJ,
Whitehouse A, Amar A, Lambert RA, Linnell JDC, Watt A,
Gutiérrez RJ(2013) Understanding and managing conservation con-
flicts. Trends Ecol Evol 28:100–109
Research Group in Conservation Biology-UNEX (2007) La evaluación
del impacto de la ecología del abejaruco (Merops apiaster)sobreel
sector apícola en el ámbito rural transfronterizo de Extremadura.
Informe inédito. Junta de Extemadura-Universidad de Extremadura
Sammataro D, Avitabile A (1998) The beekeeper's handbook. Cornell
University Press, London
Sammataro D, Yoder JA (eds) (2011) Honey bee colony health: chal-
lenges and sustainable solutions. CRC Press, Boca Ratón
Seeley TD (2009) The wisdom of the hive: the social physiology of honey
bee colonies. Harvard University Press, London
Seeley TD (2014) Honeybee ecology: a study of adaptation in social life.
Princeton University Press, Princeton
Seeley T, Heinrich B (1981) Regulation of temperature in the nests of
social insects. In: Heinrich B (ed) Insect thermoregulation. John
Wiley & Sons, New York, pp 159–234
Southwick EE, Heldmaier G (1987) Temperature control in honey bee
colonies. Bioscience 37:395–399
Valera F, Martín-Hernández R, Higes M (2011) Evaluation of large-scale
dissemination of Nosema ceranae spores by European bee-eaters
Merops apiaster. Envir Microbiol Reports 3:47–53
Vanbergen AJ (2013) Threats to an ecosystem service: pressures on pol-
linators. Front Ecol Environm 11:251–259
VanEngelsdorp D, Evans JD, Saegerman C, Mullin C, Haubruge E et al
(2009) Colony collapse disorder: a descriptive study. PLoS One 4:
e6481
Vicens N, Bosch J (2000) Weather-dependent pollinator activity in an
apple orchard, with special reference to Osmia cornuta and Apis
mellifera (Hymenoptera: Megachilidae and Apidae). Environm
Entomol 29:413–420
Woodroffe R, Thirgood S, Rabinowitz A (2005) People and wildlife,
conflict or co-existence? (No. 9). Cambridge University Press,
Cambridge
Yosef R, Markovets M, Mitchell L, Tryjanowski P (2006) Body condi-
tions as a determinant for stopover in bee-eaters (Merops apiaster)
on spring migration in the Arava Valley, southern Israel. J Arid
Environm 64:401–411
Eur J Wildl Res (2018) 64:55 Page 11 of 11 55