Bermuda grass (Cynodon dactylon) as a pollen resource

for honey bee colonies in the Lower Colorado River

agroecosystem1
Eh Erickson, Ah Atmowidjojo

To cite this version:

Eh Erickson, Ah Atmowidjojo. Bermuda grass (Cynodon dactylon) as a pollen resource for
honey bee colonies in the Lower Colorado River agroecosystem1. Apidologie, Springer Verlag,
1997, 28 (2), pp.57-62. <hal-00891402>

HAL Id: hal-00891402

https://hal.archives-ouvertes.fr/hal-00891402
Submitted on 1 Jan 1997

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.

Larchive ouverte pluridisciplinaire HAL, est

destinee au depot et `a la diffusion de documents
scientifiques de niveau recherche, publies ou non,
emanant des etablissements denseignement et de
recherche francais ou etrangers, des laboratoires
publics ou prives.

Original article

as a

Bermuda grass (Cynodon dactylon)

pollen resource for honey bee colonies in the
Lower Colorado River agroecosystem
1
EH Erickson* AH

Carl

Atmowidjojo

US Department of Agriculture, Agricultural Research Service,

Hayden Bee Research Center, 2000 East Allen Road, Tucson, AZ 85705-1596,

USA

(Received 19 April 1996; accepted 14 February 1997)

Summary &mdash; Beekeepers in southwestern Arizona and southeastern California frequently report
the poisoning of numerous honey bee colonies following the application of insecticides to blooming
bermuda grass grown for seed. This study was undertaken to characterize the relative frequency,
intensity, and seasonality of honey bee foraging for bermuda grass pollen. The results show that
bermuda grass pollen was gathered only intermittently at both of two study sites throughout the 54
week study period. Moreover, only a small fraction of all pollen gathered, 1.2% at site A and 1.5%
at site B, was from bermuda grass. Bermuda grass does not appear to be a preferred pollen resource
for honey bees, and it may be foraged only in response to the relative unavailability of other more
acceptable sources. Honey bee losses due to insecticides applied to bermuda grass may be driven by
these foraging dynamics and/or other factors.
Cynodon dactylon / pollen plant / Apis mellifera / pollen foraging / mortality

INTRODUCTION
The southern extreme of the lower Colorado
River agroecosystem includes the YumaWellton area of Yuma County in the southwestern comer of Arizona and parts of Imperial County in the southeastern corner of
1 Use of
a trade
other product.
*

name

does not constitute endorsement

Correspondence and reprints

(1) 520 670 6481; fax: (1) 520 670 6493;

Tel:

California. The region has an abundance of

arable land (> 121 500 ha) and a long growing season. Moderate winters are followed
by hot summers with low annual precipitation (= 5 cm), hence, cropland must be irrigated. The region is intensively cultivated
and agriculturally diverse. Major crops

e-mail:

by

USDA-ARS for its

[email protected]

use over

that of any

include cotton, hay (alfalfa), vegetables, citrus, small grains, and numerous seed crops
(Barmore, 1980). As a result, there is a high
demand for crop pollinators, principally
honey bees, Apis mellifera L in the area.
Bermuda grass, Cynodon dactylon (L)
one of the major seed crops produced in the Yuma-Wellton area with
approximately 5 260 ha grown annually.
Pests of bermuda grass include armyworms,
cutworms, mirids, spider mites, thrips, weevils and whiteflies (Rethwisch et al, 1995).
Pest feeding reduces yields. In response,
insecticides and miticides are often applied
during bloom and seed set. Honey bee foraging for pollen in bermuda grass seed production fields has led to numerous beekeeper
reports of poisoning of honey bee colonies
following the application of insecticides.

Pers, is

There

are

documented instances of

honey bees gathering pollen from bermuda

grass (Oertel, 1980; Schmalzel, 1980;
ONeal and Waller, 1984). Unfortunately,
most reports of bees gathering bermuda
grass pollen are anecdotal and involve
bermuda grass grown as turf or in patches of
weeds. However, Bogdan (1962) reported
that in Kenya, bermuda grass production
fields attracted numerous pollen-collecting
honey bees between 0800 and 1000 h, on
some days, but not others. ONeal and
Waller

(1984) found that bermuda grass

pollen comprised only 0.5% of all pollen

gathered by honey bees in five apiaries at
Tucson, AZ, during the weeks of 3 June and
6 October 1976. They further reported that
0.4% of all pollen gathered by colonies in a
single apiary at Tucson, AZ, over 3 years
(1976, 1978, 1979) was from bermuda grass.
The purpose of this study was to characterize the relative frequency and intensity
of honey bee foraging for bermuda grass
pollen in seed production fields.

MATERIALS AND METHODS

Two sites in the Yuma Valley, separated by about
17.7 km, were selected for study. The entire area
is under cultivation except for wet lands and
banks along the Colorado River. Both sites were
adjacent to and within honey bee flight range of
several fields of bermuda grass grown for seed.
Site A was located on the University of Arizona
Experimental Farm approximately 3.2 km west of
Yuma, site B was approximately 1.6 km east of
Gadsden, AZ. Site A was characterized by mixed
agriculture, in part the result of the diversity of
experimental plots on the farm. In contrast, cropland in the vicinity of site B was devoted principally to citrus and cotton production. They were
also within honey bee flight range of native plants
in uncultivated areas along the Colorado River.

Seven full strength honey bee colonies in two

24.5 cm deep Langstroth hive bodies were placed
at each site in late May 1987. Each colony was
equipped with a combination dead bee/pollen
trap (Pennwalt Corp, Philadelphia, PA). Pollen
samples were removed weekly from each colony
from 6/9/87 to 7/2/88, that is a period spanning
two bermuda grass bloom seasons (54 weeks).
The pollen samples were immediately frozen and
stored for subsequent analyses. Dead bees trapped
from each colony were removed and counted

weekly.
Each of about 700 pollen samples, consisting
of the individual pollen pellets stripped from the
legs of foraging bees returning to the hive, was
individually weighed. The pollen pellets in each
sample were sorted by color and texture. Pollen
pellet types occurring in minor amounts (less
than 2% of the total) were placed in a category
labelled miscellaneous. The data for each type
were averaged over all colonies at each site for
each sampling date and presented as percent of
total sample weight.

Pollen identification

pollen pellet was assumed to be homogewith regard to plant species (ONeal and
Waller, 1984). One or two pellets were crushed
on a microscope slide, stained with basic fuchsin
and mounted in glycerine jelly. The specimens
were covered with a cover slip and placed on a
Each

neous

slide warmer for 24 h. Ten slides were made

from each pollen type. Individual pollen grains on
each slide were identified using an Olympus&reg;

microscope at 600 magnification or1

nification

000 mag-

(oil emersion).

Fresh flowers of bermuda grass, corn, Johnson

grass and sorghum were collected from the study
area for comparison. Reference slides were made
from these flowers using the methods above.
Pollen classes representing more than 2% of each
sample were identified to plant family based on
morphology and size of the individual pollen
grains. Identification was confirmed using the
reference slides and published references (Bassett
et al, 1978; Huang, 1972; Kremp, 1965; Lewis et
al, 1983; Martin and Drew, 1969, 1970; Punt and

Clarke, 1980).

RESULTS AND DISCUSSION

The results of these studies suggest that

honey bees are able to find pollen resources
in the Yuma Valley throughout the year
except for a 4 to 6 week period from late
November through December (fig 1). Data
gaps in February and May 1988, reflect samples damaged or lost due to rain.

The total amount of pollen by weight,

gathered weekly by honey bees at each site
during the study period, was first divided
into three groups: non-Poaceae, Poaceae
other than bermuda grass and bermuda
grass, to analyze bee foraging preferences.
These data, presented in figure I, clearly
show that pollen from grasses other than
bermuda grass predominated over bermuda
grass by a wide margin. As reported by
Bogdan (1962) bermuda grass pollen was
gathered only intermittently at both sites
throughout the study period even though
bermuda grass seed production fields bloom
for 6 to 8 weeks. Converting weights to percentiles, figure 2 again shows that of the
pollen gathered from grasses, only a very
small fraction was from bermuda grass. The
mean incidence of bermuda grass pollen in
all samples for the entire study period was
1.2% at site A and 1.5% at site B. These
data are similar to the findings of ONeal
and Waller (1984). Other Poaceae pollen
comprised means of 13.5% and 17.9%,
respectively, for the two sites, with the bal-

ance (85.3 and 80.1%, respectively) being

non-Poaceae. Bermuda grass pollen
occurred in only 11samples at site A and
12 samples at site B during the study period.
Weeks during which bermuda grass pollen
was gathered coincide with, but do not
encompass entirely, the flowering periods
of the cultivated fields. Bermuda grass
pollen gathered in February and March
likely came from uncultivated plants.

Non-Poaceae pollens comprised the

majority of pollens collected during the
study (fig 2). Thirteen plant families were
represented most frequently in this group.
Not surprisingly, this group was dominated
by the Asteraceae, Fabaceae, Polygonaceae

and Tamaricaceae which, unlike Poaceae

pollen (Stanley and Linskens, 1974), represent some of the most nutritive pollens
for honey bees. Chenopodiaceae and
Rutaceae pollens were also present in sig-

nificant amounts in some samples. The

pollen profiles at both sites were similar.
The number of dead bees trapped per
colony remained low (less than 100 bees
per colony per week) throughout the study
period except for the second week in
September at Gadsden and the third week
in March at Yuma. Although slightly elevated (300-400 dead bees per colony), this
level of mortality was not sufficient to
threaten colony survival.

Honey bees can sometimes be found foraging extensively in blooming bermuda

grass seed production fields in the YumaWellton area (Erickson et al, personal observation). Even so, it is remarkable that so little pollen from bermuda grass, compared to
that from other sources, was gathered by
honey bees in this study. The intermittency
of the intrafield foraging reported here was
also observed by Bogdan (1962). The data

suggest that bermuda grass pollen is not a

preferred resource for honey bees, and may
in fact be gathered only in the absence of
sufficient quantities of other more acceptable pollens. Data from studies on sweet
corn (Erickson et al, unpublished), also a
marginal pollen resource for bees, leads us
to

suspect that, while honey bee losses occur

result of exposure to insecticides applied

bloom, these losses may
be driven by mitigating factors such as the
relative availability of other pollen resources.
as a

to bermuda grass in

ACKNOWLEDGMENTS
The authors wish to thank J Smith for his assistance in study site selection, H Don for colony
management, C Mullis and T Hannen for assistance in collecting pollen samples, and H Don, J
Edwards and K Kehl for their assistance in data

analysis.

Rsum &mdash; Le chiendent (Cynodon daccomme source de pollen pour les

colonies dabeilles dans lagrocosystme
du Bas Colorado. Les apiculteurs du sudouest de lArizona et du sud-est de la Californie font souvent tat dempoisonnement
de nombreuses colonies dabeilles suite aux
traitements insecticides appliqus sur le
chiendent en fleurs, cultiv dans ces rgions
pour la production de semences. Cette tude
vise caractriser la frquence relative,
lintensit et la variation saisonnire de la
rcolte de pollen de chiendent par labeille
domestique. Pour cela sept colonies ont t
places fin mai 1987 dans chacun des deux
sites tudis. Les ruches taient quipes
dune trappe mixte pollen et abeilles
mortes. Les chantillons de pollen ont t
prlevs chaque semaine du 6/9/87 au
7/2/88, priode couvrant deux saisons de
floraison du chiendent, puis congels pour
tre analyss ultrieurement. Les pelotes de
pollen ont t peses et classes selon leur
couleur et leur texture. Pour les types de

tylon)

pelotes prsents

plus

de 2 % dans les

chantillons, les familles botaniques ont t

dtermines daprs la morphologie et la
taille des grains de pollen. Sur chaque lame,
les grains de pollen ont t identifis individuellement au microscope. Le nombre
total dabeilles mortes dans les trappes est
rest faible durant toute la priode dtude.
Toutes les colonies ont pu rcolter du pollen
except durant une courte priode de 4 6
semaines, de fin novembre fin dcembre.
La majorit du pollen ne provenait pas des
Poaceae. Treize familles botaniques ont t
les plus reprsentes, les Asteraceae, Fabaceae, Polygonaceae et Tamaricaceae arrivant en tte. Sur les deux sites dtude, le
pollen de chiendent na t rcolt que sporadiquement au cours des 54 semaines tudies. En outre, seule une trs petite fraction du pollen rcolt au total tait du pollen
de chiendent : 1,2 % sur le site A,1,5 % sur
le site B (fig 1). Il ne reprsente galement
quune faible portion du pollen de gramines rcolt (fig 2). Le pollen de chiendent
ne figure donc pas parmi les sources polliniques prfres des abeilles et semble ntre
rcolt quen labsence de pollens plus intressants en quantits suffisantes. Les pertes
dabeilles dues aux traitements insecticides
sur le chiendent dpendent donc de la dynamique du butinage et/ou dautres facteurs.

Cynodon dactylon / plante pollinifre /

Apis mellifera / butinage / mortalit
Zusammenfassung &mdash; Bermudagras
(Cynodon dactylon) als Pollenquelle fr
Honigbienen im Agro-kosystem des
unteren Coloradoflusses. Die Anwendung
von

Insektiziden whrend der Blte

von zur

Samenerzeugung angebautem Bermudagras

fhrt hufig zur Vergiftung zahlreicher Bienenvlker. Da ber die Beziehung von Bermudagras und Honigbienen sehr wenig
bekannt ist, wurde eine Untersuchung der
relativen Hufigkeit und Intensitt des
Beflugs von Bermudagras durch pollen-

sammelnde Bienen durchgefhrt. Hierzu

wurden zwei Untersuchungsgebiete in einem
landwirtschaftlich divers genutzten Gebiet
nahe Yuma /AZ ausgesucht. Ende Mai1987
wurden an beiden Standorten jeweils 7 Bienenvlker aufgestellt und mit kombinierten
Fallen zur Erfassung toter Bienen sowie
zum Pollensammeln versehen. Die Pollenproben und die toten Bienen wurden vom
6.9.87 bis zum 7.2.88 wchentlich von
jedem Volk entnommen, eingefroren und
spter analysiert. Die Pollenproben wurden
gewogen und nach Farbe und Oberflchenbeschaffenheit sortiert. Jede der Pollensorten wurde mikroskopisch bestimmt. Fr Pollensorten mit einem Anteil von mehr als 2%
in den Proben wurde die Pflanzenfamilie
auf Grund der Morphologie und Gr&szlig;e der
einzelnen Pollenkrner bestimmt. Die
Anzahl von toten Bienen in den Fallen war
whrend des Untersuchungszeitraumes
gering. Alle Bienenvlker konnten whrend
des gesamten Jahres Pollen finden, au&szlig;er
whrend eines kurzen Zeitraums von 4 -6
Wochen vom spten November bis Ende
Dezember. Whrend der gesamten Zeit
stammte der meiste Pollen nicht von Poaceen. Dreizehn Pflanzenfamilien waren
dabei am strksten reprsentiert: Asteraceae, Fabaceae, Polygonaceae und Tamaricaceae. ber den Untersuchungszeitraum
von 54 Wochen wurde an beiden Standorten
nur vereinzelt Pollen von Bermudagras
gesammelt. Darberhinaus entstammte in
beiden Gebieten nur ein sehr geringer Anteil
des gesammelten Pollens dem Bermudagras
(Abb 1; A: 1,2%; B: 1,5%). Andere Graspollen dominierten stark ber den Bermudagraspollen (Abb 2). Dieses stellt daher
keine bevorzugte Pollenquelle dar und wird
mglicherweise nur bei Abwesenheit von
gnstigeren Pollenquellen besammelt. Verluste an Honigbienen auf Grund von Insektizidanwendung auf Bermudagras sind daher
von der Sammeldynamik der Bienen sowie

mglicherweise
abhngig.

auch anderen Faktoren

Cynodon dactylon / Bermudagras /Apis

mellifera / Pollenpflanze/ Pollensammeln
/ Vergiftungen
REFERENCES
Barmore RL (1980) Soil Survey of Yuma-Wellton Area.
USDA-SCS
Bassett IJ, Crompton CW, Parmelee JA (1978) An atlas
of airborne pollen grains and common fungus
spores of Canada. Monograph no 18. Canada
Department of Agriculture

Bogdan AV (1962) Grass pollination by bees in Kenya.

Proc Linn Soc

London, UK,173, 55-60

TC (1972) Pollen flora

Taiwan University. Botany
Taiwan

Huang

of Taiwan. National
Department, Taipei,

Kremp GOW (1965) Morphologic encyclopedia of

palynology. University of Arizona Press, Tucson,
USA

Vinay P, Zenger VE (1983) Airborne and

allergenic pollen of North America. Johns Hopkins University Press, Baltimore & London, USA,

Lewis WH,

UK
Martin PS, Drew CM (1969) Scanning electron photomicrographs of southwestern pollen grains. J Ariz
Acad Sci 5,147-176

(1970) Additional scanning elecphotomicrographs of southwestern pollen

grains. J Ariz Acad Sci 6, 140-161
ONeal RJ, Waller GD (1984) On the pollen harvest by
the honeybee (Apis mellifera L) near Tucson, AriMartin PS, Drew CM
tron

zona

(1976-1981). Desert Plants 6, 81-109

Oertel E (1980) Nectar and pollen plants. In: Beekeeping in the United States. USDA-SEA Agriculture Handbook no 335
Punt W, Clarke GCS (1980) (eds) The northwest European pollen flora II. Elsevier Sci Pub Co, New
York, USA

Rethwisch MD, Natwick ET, Tickes BR, Meadows

M, Wright D (1995) Impact of insect feeding and
economics of elected insecticides on early summer
bermuda grass seed production in the desert southwest. Southwest Entomol vol 20, 187-201

(1980) The diet breadth of Apis

(Hymenoptera: Apidae). MS Thesis, University of

Schmalzel RJ

Arizona, Tucson, USA

Stanley RG, Linskens HF (1974) Pollen biology biochemistry management. Springer-Verlag, New
York, USA