Download - The role of volatile semiochemicals in mediating host location and selection by nuisance and disease-transmitting cattle flies

The role of volatile semiochemicals in mediating hostlocation and selection by nuisance and disease-transmittingcattle flies

M. A. BIRKETT1 , N. AGELOPOULOS1 , K . -M. V . JENSEN2 ,

J . B . JESPERSEN2 , J . A . PICKETT1 , H . J . PRIJS3 , G. THOMAS3 ,

J . J . TRAPMAN3 , L . J . WADHAMS1 and C. M. WOODCOCK1

1Biological Chemistry Division, Rothamsted Research, Harpenden, U.K., 2Danish Institute of Agricultural Sciences, Danish

Pest Infestation Laboratory, Lyngby, Denmark and 3Laboratory for Behavioural Physiology, Institute for Animal Health,

Lelystad, the Netherlands

Abstract. The role of volatile semiochemicals in mediating the location and selection within

herds of Holstein-Friesian heifers by nuisance and disease-transmitting cattle flies was

investigated using coupled gas chromatography–electrophysiology (GC–EAG), coupled gas

chromatography–mass spectrometry (GC–MS), electrophysiology (EAG), laboratory beha-

viour and field studies. Using volatile extracts collected by air entrainment from heifers in the

Netherlands, a number of active peaks were located by coupled GC–EAG for Musca

autumnalis (de Geer) (Diptera: Muscidae) and Haematobia irritans (L.) (Diptera: Muscidae).

Volatile samples were also collected from two heifers in Denmark shown in previous counting

experiments to differ significantly in their fly loads. Coupled GC–EAG using Ha. irritans

antennae revealed differences in the EAG response to the samples, with additional EAG

activity in the sample collected from the heifer with the lower fly load. To identify more EAG

active compounds, volatiles were also collected from 48-h-old urine by air entrainment. In

total, 23 compounds were located and identified by coupled GC–EAG and GC–MS. Further

electrophysiological testing of these compounds with five fly species [M. autumnalis,

Ha. irritans, Hydrotaea irritans (L.) (Diptera: Muscidae), Stomoxys calcitrans (L.) (Diptera:

Musicidae) and Wohlfahrtia magnifica (Schiner) (Diptera: Sarcophagidae)] showed that only

some of the compounds were physiologically active across the range of flies tested. These

included 1-octen-3-ol, 6-methyl-5-hepten-2-one, (Z)-3-hexen-1-ol, naphthalene, and all EAG

active compounds identified from urine. Compounds showing significant EAG activity were

tested for behavioural activity using a wind-tunnel designed for measuring upwind flight

behaviour. At certain concentrations, 1-octen-3-ol, 6-methyl-5-hepten-2-one and 3-octanol

increased upwind flight, whereas naphthalene, propyl butanoate and linalool reduced upwind

flight. In field studies using small herds of heifers ranked according to their fly load, individual

slow-release formulations of 1-octen-3-ol and 6-methyl-5-hepten-2-one, when applied to low

and high fly loading heifers, reduced fly loads on these individuals. This study provides

evidence for the hypothesis that the natural differential attractiveness within herds of Hol-

stein-Freisian heifers, i.e. a single host species, for cattle flies is partly due to differences in

volatile semiochemicals emitted from the host. It is suggested that this phenomenon applies to

other vertebrate host species and their associated insect pests.

Key words. Haematobia irritans, Musca autumnalis, Stomoxys calcitrans,Wohlfahrtia magnifica, air entrainment, behaviour, cattle fly, electrophysiology,6-methyl-5-hepten-2-one, 1-octen-3-ol, semiochemical.

Correspondence: Professor John Pickett, Biological Chemistry Division, Rothamsted Research, Harpenden, Herts, AL5 2JQ, U.K.

Tel.: þ 44 (0)1582 763133; fax: þ 44 (0)1582 762595; e-mail: [email protected]

Medical and Veterinary Entomology (2004) 18, 313–322

# 2004 The Royal Entomological Society 313

Introduction

The behaviour of several species of nuisance, biting and

disease-transmitting cattle flies (Diptera) that settle or

feed on grazing hosts has been shown to lead to

increased disease incidence, reproductive failure and

also reduced meat and milk yields, with significant eco-

nomic losses arising as a consequence (Fraser & Broom,

1990).

Biting and blood-sucking (haematophagous) insects,

including cattle flies, use a wide variety of visual, olfac-

tory, gustatory and physical stimuli in host location and

selection. Of these, volatile semiochemicals play a major

role in mediating such behaviour (Pickett & Woodcock,

1996; Gibson & Torr, 1999). Cattle flies generally show

marked activation and attraction to carbon dioxide

(CO2), universally present in mammalian breath, but

also use other cues emanating from hosts. Numerous

studies have identified semiochemicals involved in host

location and selection for tsetse flies (Diptera: Glossini-

dae) (Hall et al., 1984; Hassanali et al., 1986; Bursell

et al., 1988; Gikonyo et al., 2002). For other cattle flies,

these cues remain largely undefined. Behavioural and elec-

trophysiological responses of Stomoxys spp., e.g. the stable

fly, S. calcitrans (L.), to host odour components have con-

centrated mainly on compounds that have been effective with

tsetse flies, with conflicting reports. In certain cases, traps

baited with 1-octen-3-ol, a component of ruminant breath,

were reported to increase catches of Stomoxys spp. (Hollo-

way & Phelps, 1991; Mihok et al., 1995), whereas in a sepa-

rate study no response was observed (Mullens et al., 1995).

Laboratory behavioural studies with S. calcitrans have

demonstrated anemotactic responses to 1-octen-3-ol and

acetone (Warnes & Finlayson, 1985; Schofield et al., 1997).

Electroantennogram recordings made from S. calcitrans

using host odour components have also been reported.

Responses to 1-octen-3-ol, 3-methylphenol and various alco-

hols, but not acetone, were observed by Schofield et al.

(1995), whereas responses to CO2, acetone and 1-octen-3-ol

were reported by Warnes & Finlayson (1986). For other

major cattle fly pests, including the sheep head fly,Hydrotaea

irritans (L.), the major transmitter for summer mastitis in

cattle, the face fly, Musca autumnalis (de Geer), the horn

fly, Haematobia irritans (L.), and the flesh fly, Wohlfahrtia

magnifica (Schiner), little or nothing is known of the role and

chemical nature of olfactory cues used in host location and

selection.

Cattle fly control is mainly obtained through the

employment of broad-spectrum toxicant insecticides, in

the form of ear-tags, pour-ons, dips, sprays, etc. Due to

concerns over the build-up of insecticide resistance, and

public concern over the possible environmental impact of

such chemicals, alternative control methods are sought. A

more benign approach is to employ insect repellents. Many

natural products derived from essential oils have been

developed, but the active compounds are extremely odifor-

ous, and represent only a weak behavioural drive away

from the potential host. An alternative approach is to

exploit the natural differentiation between members within

the host species. It has been demonstrated for several fly

species that individual animals within herds vary in their

attractiveness (Thomas et al., 1987; Brown et al., 1992;

Jensen et al., 2004). An hypothesis developed to explain

this differentiation, which applies to warm-blooded verte-

brate hosts and their associated insect pests, is that all

members of the host species naturally release volatile

attractants, but those which are less attractive release

additional compounds that interfere with the activity of

the attractants. These compounds, at higher levels, act as

true repellents but at lower, or natural physiological levels,

effectively mask the normal attractiveness of the host spe-

cies (Pickett et al., 1998).

The aim of this study was to test this hypothesis,

leading to the identification of natural repellents, and

attractants, that could form part of novel fly control stra-

tegies, such as the ‘push–pull’ strategy (Miller & Cowles,

1990). The suitability of this control strategy was con-

firmed in a preliminary study preceding this one, which

demonstrated that fly populations could be manipulated

through switching of animals between herds (Jensen et al.,

2004). The adopted approach to semiochemical identifica-

tion was to collect samples of volatile compounds emitted

by cattle that were known to differ in their attractiveness,

and to compare these by coupled gas chromatography–

electrophysiology (GC–EAG), using M. autumnalis and

Ha. irritans as the model insects. Identified EAG-active

compounds were then tested in a laboratory behavioural

bioassay, and in the field, to confirm their biological

activity.

Materials and methods

Volatile collection

Volatiles from Holstein-Friesian cattle, individually

housed in pre-washed stalls in the Netherlands and Denmark,

were collected by air entrainment. Air surrounding

the animals was drawn (1 L/min) through Pyrex glass

tubes (3.9mm o.d., 2.4mm i.d.) containing Porapak Q

(50mg per tube, 50–80 mesh, Supelco, Poole, U.K.) over

a period of several hours. The animals selected in the

Netherlands displayed significantly higher fly loads when

compared to other animals in the same herd. For volatile

collection in Denmark, animals shown in a preceding

study (Jensen et al., 2004) to have very high and low fly

loads (heifers 1257 and 1270, respectively) were selected.

The volatile collections were arranged such that relatively

free movement within the stalls was allowed, thus obviat-

ing any stress in the animals. Samples in the Netherlands

were collected by placing a single Porapak tube at a

fixed position above the animal, halfway along its length

and approximately 15–20 cm above the skin surface

when the animal was standing. In Denmark, tubes were

314 M. A. Birkett et al.

# 2004 The Royal Entomological Society, Medical and Veterinary Entomology, 18, 313–322

suspended above the neck by means of a pulley system

so that they were approximately 10–20 cm above the

animal at all times. In the Netherlands, a total of eight

entrainments, each lasting 24 h, were performed on four

individuals, i.e. two entrainments per individual. In

Denmark, two entrainments, each lasting 9 h, were per-

formed on heifer 1257, and four entrainments, each lasting

12 h, were performed on heifer 1270. Air entrainments of

clean empty stalls were carried out as controls. Porapak

tubes containing trapped volatiles were sent to the U.K.,

then eluted with freshly distilled diethyl ether (500mLper tube). To provide greater amounts of material for

subsequent identifications, samples collected in the Nether-

lands were combined prior to analysis. Samples were

concentrated to 100 mL using a gentle stream of nitrogen

and stored at �20�C in tightly capped microvials until

further use.

Volatiles were also collected from urine from a heifer

in the Netherlands. The sample was collected during

natural urination, and was left at ambient temperature

for 48 h. A portion of the aged urine (700mL) was placed

in a round-bottomed flask (1 L) and volatiles drawn

from the vessel using a purified air flow (1.5 L/min)

through glass tubes containing Porapak Q (50mg per

tube, as described previously). The trapped volatiles were

eluted with freshly distilled diethyl ether (500 mL per tube)

and stored at �20�C in tightly capped microvials until

further use.

Insects

The face fly, M. autumnalis, and the horn fly,

Ha. irritans, were reared at the Danish Pest Infestation

Laboratory (DPIL), Lyngby, Denmark, and sent to the

U.K. and the Netherlands as pupae. The stable fly, S. cal-

citrans, was obtained as pupae from the Novartis Centre

de Recherche Sante Animale, Switzerland. The flesh fly,

W. magnifica, was reared at the University of Granada,

Spain, and sent to the Netherlands as pupae (Soler Cruz

et al., 1996). The sheep head fly, Hy. irritans (L.), was

collected as adults in the field in the Netherlands. For

coupled GC–EAG studies in the U.K., M. autumnalis

and Ha. irritans were kept in controlled environment cabi-

nets at 20�C, LD16 : 8 h until required. The emerged flies

were provided with water prior to use. For EAG and

behavioural studies in the Netherlands, all fly species

were kept in cages at 21�C, 70% humidity, LD 10 : 14 h,

apart from S. calcitrans, which was kept at 24�C. All flieswere fed with sucrose solution (5% v/v) until required, and

were sexed prior to use.

Electrophysiology

For GC–EAG studies in the U.K., recordings from

female flies were made by using Ag/AgCl electrodes

filled with saline solution (composition as in Maddrell,

1969; but without the glucose). The head of the fly was

excised and the indifferent electrode positioned in the

head capsule. The recording electrode was placed over

the cut arista. The signals were passed through a high

impedance amplifier (UN-03b, Syntech, Hilversum, the

Netherlands) and displayed on a chart recorder. For

EAG studies in the Netherlands using authentic com-

pounds, recordings from excised heads of females were

made using glass electrodes filled with Ringer solution

(0.1 M KCl) and 10% PVP (polyvinylpyrrolidone). The

indifferent electrode was placed over the tip of the

antenna and the recording electrode inserted into the

back of the head. The arista of both antennae were

removed to reduce noise during recording from the sti-

mulus puff. The signals generated by the antenna were

passed through an amplifier (AM-02, Syntech) and data

storage and processing were carried out with a custom-

ized software package (Syntech). Stimuli were delivered

via standard disposable Pasteur pipettes containing a

strip of filter paper (40mm� 5mm). Solutions of test

chemicals (20 mL) were applied to filter papers, with the

solvent being allowed to evaporate for 30 s before use,

and delivered through the pipette into a continuous pur-

ified and humidified airstream (1 L/min) using a stimulus

generator (CS-27, Syntech) delivering an air puff (0.8 s,

4 L/min). The interval between stimuli was 1min. Com-

pounds were dissolved in freshly distilled diethyl ether,

and were tested in a random order. As the response of

the preparations declined with time, the mean responses

to compounds were normalized with respect to the mean

response to 1-octen-3-ol, a well-known attractant for

many haematophagous Diptera. To gain an overall pic-

ture of the EAG responses of all identified compounds,

all testing was conducted at one dose (10�2 g). For

comparison of activity, recordings were taken from

known attractants and repellents for other Diptera.

These included a tsetse lure (1-octen-3-ol/m-cresol/

p-cresol in a 1 : 4 : 8 ratio), indole, skatole (3-methylindole),

N,N-diethyltoluamide (DEET), citronellol and propionic

acid. The number of replicates for each fly species varied

according to availability of insects, but was sufficient in

each case for statistical analysis to be performed using

ANOVA.

Gas chromatography (GC)

GC analysis of air entrainment samples was carried out

on two GCs. An AI92 GC equipped with a temperature

programmable on-column injector, flame ionization

detector (FID), fitted with either a 50m� 0.32mm i.d.,

HP-1 non-polar column or a 30m� 0.3mm i.d., BP-20

polar column, and a Hewlett-Packard 5880 GC equipped

with a split-splitless injector, an FID and a 50m� 0.32mm

i.d., HP-1 non-polar column. For the AI92GC using theHP-1

column, the oven temperature was maintained at 40�C for

1min, then programmed at 5�/min to 100�C, then 10�/min

Host location and selection by flies 315


to 250�C. The BP-20 column was maintained at 40�C for

2min, then programmed at 10�/min to 225�C. For the 5880GC, the HP-1 column was maintained at 40�C for 2min,

then 5�/min to 150�C then 10�/min to 250�C. The carrier gasin all cases was hydrogen.

Coupled gas chromatography–electrophysiology

(GC–EAG)

The coupled GC–EAG system, in which the effluent

from the GC capillary column is delivered simultaneously

to the antennal preparation and the GC detector, has been

described elsewhere (Wadhams, 1990). Separation of air

entrainment samples was achieved on an AI93 GC

(AI, Cambridge, U.K.) equipped with a cold on-column

injector and a FID. Two columns were used. The non-

polar column (50m� 0.32mm i.d., HP-1) was maintained

at 40�C for 1min, then 5�/min to 100�C, then 10�/min to

250�C. The polar column (30m� 0.3mm i.d., BP-20) was

maintained at 40�C for 2min, then 10�/min to 225�C. Thecarrier gas was hydrogen. The outputs from the EAG

amplifier and the FID were monitored simultaneously on

a chart recorder.

Coupled gas chromatography–mass spectrometry

(GC–MS)

A Hewlett-Packard 5890 GC was connected to a VG

Autospec mass spectrometer (Fisons, Manchester, U.K.).

Ionization was by electron impact at 70 eV, 250�C, and the

GC, using the HP-1 column, was maintained at 30�C for

5min, then programmed at 5�C/min to 250�C. Compoundstentatively identified by GC–MS were confirmed by

co-injection of authentic samples on GC using non-polar

and polar columns (see above).

Chemicals

All chemicals were bought from the Aldrich Chemical

Company (Gillingham, U.K.). For peak enhancement

and internal reference standards (C9–C17 n-hydrocar-

bons), chemicals were diluted in distilled n-hexane prior

to use. For EAG studies, chemicals were diluted in

freshly distilled diethyl ether. For behavioural studies,

chemicals were dissolved in paraffin oil to provide slow

release formulations. For identified chemicals with

asymmetric carbons, racemic forms were used in

peak enhancement, electrophysiological and behavioural

studies.

Insect behaviour

A glass wind-tunnel (70 cm� 3.5 cm) was used to

monitor upwind flight behaviour of flies in response to

chemical stimuli. Musca autumnalis was selected as the

test insect, as other fly species showed little or no activ-

ity in the tunnel. The tunnel consisted of seven identical

sections of glass tubing, 10 cm long and 3.5 cm diameter.

The middle five sections were designated as the experi-

mental area, with a downwind and an upwind section

attached at either end. For each experiment, batches of

10 flies were introduced into the downwind section 1 h

prior to commencement, to allow acclimatization. The

stimulus, previously dissolved in paraffin oil, was applied

(20 mL) to a filter paper strip (4� 1 cm) and the strip

placed in the upwind section. Purified air was blown

through the tube at a speed of 0.05m/s, and recording

of the fly positions commenced 30 s after stimulus intro-

duction. Each experiment consisted of observations every

10 s for 3min. Chemicals were tested at six doses, using a

decadic series from 10�2 g to 10�7 g. The same batch of

flies was used to test a complete set of doses, each set

starting with the lowest dose. Each chemical was tested

with three batches of flies. To confirm that behaviour

was not affected by exposure to chemicals tested, and

that there was no chemical contamination of the tunnel,

a paraffin control stimulus was tested at the beginning,

middle and end of each series. Each experiment was

separated by an interval of 5min. The tunnel was

cleaned with diethyl ether and dried with purified air

for at least 30min before a new chemical was introduced.

All experiments were carried out between 08.00 hours

and 14.00 hours, as the activity of the flies tended to

drop off after this time. The response of flies to stimuli

was measured by observing the number of flies in the

upwind section of the tube. No significant differences

were observed between the first, middle and last paraffin

controls, and therefore these were pooled and averaged.

The results were analysed by comparing the difference

between the mean number of flies in the upwind section

for the stimuli and the control with Student’s t-test using

matched pairs. To test potential repellence in the tunnel,

the light and position of the tunnel was adjusted such

that flies spent more time upwind. To confirm the valid-

ity of the bioassay technique, the known fly repellent

citronellol (Bartlett, 1985) was also tested across the same

range of doses.

Field studies

Field trials were conducted in Denmark, using small

herds (n¼ 7) of Holstein-Friesian cattle that were selected

from larger herds. A single experiment comprised the use

of one compound per herd. 1-Octen-3-ol and 6-methyl-5-

hepten-2-one were released from slow release formulations

comprising polyethylene bags, protected from potential

damage caused by the animal by being placed in a dispenser

tube made of stainless steel and aluminium (4 cm length,

with 4-mm holes interspersed to allow diffusion). The ends

of the tube were closed with plates, and the tube was



hooked to a belt attached around the chest of the animal,

immediately behind the front legs. The sachets used gave

release rates of 10.5mg/day and 14mg/day for 1-octen-3-ol

and 6-methyl-5-hepten-2-one, respectively. On day 1, the

flies on each animal were identified and the number counted

using a previously established counting method (Jensen

et al., 2004) every 0.5 h, altogether four to five times in

total, and the data were used to rank the animals according

to fly load. Tube dispensers without polyethylene bags were

hung around the necks of the two least and two most fly

attractive individuals, to allow the animals to become accus-

tomed to the devices. On day 2, in the morning, the tube

dispensers on the two least attractive cattle were loaded

with a sachet containing the chemical, with the cattle

being given an hour to settle before counting was initiated.

The flies on each animal in the herd were identified and

counted every 0.5 h, four to five times in total. At the end of

day 2, the dispensers were emptied and cleaned with water

and acetone, and the sachets returned to the freezer. On day

3, the same counting procedure was carried out on the two

most attractive animals, using the same sachets. Statistical

analysis was carried out using a Kruskal–Wallis test fol-

lowed by a multiple-comparison test. Significance indicated

by different letters for each heifer indicates significant dif-

ference between medians of the ranks on a 5% level.

Results

GC analysis of the individual samples collected from

Holstein-Friesian cattle, both in the Netherlands and in

Denmark, showed that the concentration of components

was insufficient for analysis by coupled GC–MS. Samples

were therefore combined and concentrated to 100mL, toprovide separate Dutch and Danish samples, with small

volumes of each individual sample being retained. GC

analysis of the combined samples highlighted the complexity

of the odour profile of cattle (Figs 1a–c), with many compon-

ents present in very low amounts. More volatile material

Fig. 1. Gas chromatograms (GC) of volatiles collected from the headspace and urine of Holstein-Friesian cattle. Peak numbers correlate to

compounds, listed in Table 1, that were located and identified by coupled GC–electroantennography (GC–EAG) and GC–mass spectrometry

(GC–MS) and by peak enhancement with authentic samples: (a) combined and concentrated volatile sample collected from four Dutch cattle;

(b) combined and concentrated volatile sample collected from a Danish heifer (no. 1257) with known high fly load; (c) combined and

concentrated volatile sample collected from a Danish heifer (no. 1270) with known low fly load; (d) volatile sample collected from aged urine

(48 h old) of one Dutch heifer. Traces (a), (b) and (c) were obtained on a non-polar HP-1 column. Trace (d) was obtained from a polar BP-20

column. EAG-active compounds in (a) were located using M. autumnalis and Ha. irritans antennae, and in (b), (c) and (d) using Ha. irritans

antennae. Traces (b) and (c) are provided on the same scale for comparison.



was made available for further experiments by sampling

cattle urine that had been allowed to age (Fig. 1d).

Coupled GC–EAG using recordings from M. autumnalis

located a number of volatiles in the Dutch sample that

elicited an electrophysiological response (Fig. 1a). These

active compounds were identified tentatively by coupled

GC–MS and confirmed by peak enhancement using authen-

tic standards on both non-polar and polar GC columns

(Table 1). These compounds comprised a high proportion

of aromatic compounds, including polar aromatics such as

phenol, all three isomers of cresol (2-, 3- and 4-methylphenol)

and 4-methyl-2-nitrophenol, which were present in signifi-

cantly large amounts, together with non-polar aromatics

such as naphthalene and acenaphthene. Coupled GC–EAG

recordings using Ha. irritans showed a similar number of

active components, but with additional activity from com-

pounds including (Z)-3-hexen-1-ol and 6-methyl-5-hepten-2-

one. Coupled GC–EAG recordings from Ha. irritans,

using volatiles collected from the most and least fly attractive

cattle in ranked herds in Denmark, located a number of

volatiles that elicited an electrophysiological response (Figs

1b and c; Table 1). In the non-attractive sample, more active

peaks were located. Compounds unique to this sample were

identified as the non-polar aromatic compounds propyl-

benzene and styrene, along with camphene, 2-heptanone

and propyl butanoate. EAG active compounds for both the

attractive and non-attractive samples were identified as

phenol and m- and p-cresol. A number of EAG active com-

pounds were located forHa. irritans in the cattle urine sample

(Fig. 1d), including polar aromatics such as m- and p-cresol,

which were located in the whole animal entrainment samples,

along with 2-methoxyphenol. A range of aliphatic alcohols

were also identified, including 3-octanol, 2-nonanol,

2-decanol, 1-nonanol and linalool. Due to the complexity of

the GC traces, the stereochemistry of chiral compounds was

not assigned. Thus, racemic materials were used in further

experiments. Overall, coupled GC–MS showed that a

wide range of compounds were collected from the heifers,

Table 1. Electroantennography responses of cattle flies to volatiles collected from cattle headspace and urine (48 h old).

Haematobia Musca Hydrotaea Stomoxys Wohlfahrtia

Peak no. irritans autumnalis irritans calcitrans magnifica

Type (see Figs 1a–d) Compound (10�2 g) (n¼ 5) (n¼ 10) (n¼ 10) (n¼ 3) (n¼ 3)

Amino acid 3 Propylbenzene NS NS NS NS NS

derivative 5 Phenol NS NS NS NS NS

8 o-Cresol NS NS NS NS –

9 m-Cresol NS NS NS * –

10 p-Cresol NS NS ** ** –

12 Naphthalene *** *** *** *** ***

13 4-Methyl-2-nitrophenol * *** * * NS

14 Acenaphthene NS *** * NS *

16 Styrene NS NS NS NS NS

23 2-Methoxyphenol * *** – ** –

Fatty acid 1 (Z)-3-Hexen-1-ol NS *** ** ** NS

derivative 4 1-Octen-3-ola *** *** *** *** **

7 Decane NS NS NS NS NS

11 Undecane NS *** NS NS NS

15 2-Heptanone * NS *** *** NS

17 Propyl butanoate NS NS NS * NS

19 3-Octanola *** *** – *** ***

21 2-Decanola *** *** – *** –

22 1-Nonanol *** *** – *** –

Isoprenoid 2 a-Pinene NS NS NS NS NS

or derivative 6 6-Methyl-5-hepten-2-one *** *** *** *** **

18 Camphenea NS NS NS NS NS

20 Linaloola *** *** – *** ***

Tsetse lure *** *** – *** –

Indole NS * – NS ***

Skatole NS * – NS ***

N,N-Diethyltoluamide NS NS NS NS –

Citronellol *** *** *** *** *

Propionic acid NS NS * NS –

Compounds listed with no peak numbers are established semiochemicals or repellents for other haematophagous Diptera.aStereochemistry undefined.

*, **, ***: significantly different from control at P¼ 0.05, 0.01 and 0.001, respectively. NS, not significantly different from control at

P¼ 0.05 control¼diethyl ether, 20mL. A single dashed line indicates that the compound was not tested due to unavailability of insects.



including carboxylic acids such as isobutyric acid, but the

majority of these were not proceeded with as they were not

located by coupled GC–EAG.

To confirm the electrophysiological activity of the com-

pounds located by coupled GC–EAG and identified by

coupled GC–MS, compounds were tested at a standard

concentration (10�2 g) using five target fly species

(Ha. irritans, M. autumnalis, Hy. irritans, S. calcitrans and

W. magnifica). No significant differences were observed

between the sexes. Therefore, subsequent testing was carried

out using female flies only (Table 1). Several compounds

gave highly significant EAG responses for all or most of

the flies, including (Z)-3-hexen-1-ol, 1-octen-3-ol, 6-methyl-

5-hepten-2-one, naphthalene and linalool. When presented

individually, a number of the compounds thought to be

EAG active gave little or no response, including phenol

and the cresols (apart from p-cresol). Almost all the

compounds identified from the cattle urine (compounds

no. 19–23) gave significant EAG responses. Of the pre-

viously known semiochemicals tested, only the tsetse lure

and citronellol gave consistently significant responses. In

general, there was a good correlation between the target

species and the responses to the compounds tested.

Compounds that showed EAG activity for all fly species

were tested for behavioural activity against M. autumnalis

using a wind-tunnel bioassay (Fig. 2). Propyl butanoate was

Fig. 2. Upwind flight response of M.autumnalis in a glass wind-tunnel to semiochemicals identified from cattle headspace and urine.

Compounds were tested across a range of doses (n¼ 3). Statistical analysis was carried out with Student’s t-test using matched pairs to

compare differences between test and control data. *P< 0.05. Vertical bars represent standard errors. C¼ control response.



also tested as it was only found in the volatile sample of the

least attractive Danish heifer and was therefore considered

to be a good candidate for repellency. From the urine-

derived compounds, 3-octanol and linalool were selected

on the basis of their structural proximity to 1-octen-3-ol

and established roles as insect semiochemicals. 1-Octen-3-ol,

6-methyl-5-hepten-2-one and 3-octanol showed significant

attraction at certain concentrations, with 1-octen-3-ol

attracting flies at very low levels. Naphthalene and linalool

showed strong repellency at low concentrations, but at

higher concentrations, responses were not significantly dif-

ferent from the control. Propyl butanoate was repellent at

the highest concentration. Surprisingly, the known repellent

citronellol showed no statistically significant activity,

despite displaying high EAG activity.

One small field study was conducted using small, separate

herds of cattle in Denmark, and slow-release formulations

of two selected compounds, 1-octen-3-ol and 6-methyl-5-

hepten-2-one, presented individually (Table 2). On day 1,

for the experimental herd involving 1-octen-3-ol, heifers

1226 and 1170 had the lowest mean fly load, whereas heifers

1180 and 1160 had the highest mean fly load. On day 2,

when 1-octen-3-ol was applied to heifers 1226 and 1170, no

statistical significant change was obtained on either of the

heifers. On day 3, when heifer 1180 and heifer 1160 were

treated, there was a small, but not significant fall in the

percentage of flies on heifer 1180 and heifer 1160 when

compared with day 1.

When 6-methyl-5-hepten-2-one was applied to heifer

1223 and heifer 1178 on day 2, there was no significant

change in fly loads. When applied to heifer 1138 and heifer

1261 on day 3, the fly loads were reduced significantly for

heifer 1261, whereas no effect was observed for heifer 1138,

when compared with day 1.

Discussion

This study reports the use of coupled GC–EAG, GC–MS,

EAG and behavioural studies, in the laboratory and field,

to identify volatile semiochemicals used in host location and

selection by nuisance and disease-transmitting cattle flies.

A number of physiologically active components were

identified, of which 1-octen-3-ol and 6-methyl-5-hepten-2-

one were attractive, and naphthalene, linalool and propyl

butanoate were repellent, in laboratory behavioural assays.

Some of the identified compounds have already been

reported to be present in cattle volatiles and are known as

semiochemicals for dipterous pests. 1-Octen-3-ol, a known

component of ruminant breath, is an attractant for tsetse

flies (Hall et al., 1984). Phenol, m-cresol and p-cresol are

released from cattle urine through microbial degradation,

and also influence tsetse fly behaviour (Hassanali et al.,

1986; Bursell et al., 1988). Mixtures including some of

these chemicals, when released in a specific ratio, have

been used successfully in the field to trap tsetse fly popula-

tions (Vale et al., 1986, 1988; Torr et al., 1995). Surprisingly,

in this study, the phenolic compounds had little or no

impact on electrophysiological or behavioural responses of

cattle flies. 6-Methyl-5-hepten-2-one has been reported

before as a component of waterbuck odour, but was not

electrophysiologically active for tsetse flies (Gikonyo et al.,

2002). 4-Methyl-2-nitrophenol has been identified as a com-

ponent of cattle odour, and is an olfactory stimulant for

cattle tick pests (Stuellet & Guerin, 1994). No reports exist

on the presence of propyl butanoate, (Z)-3-hexen-1-ol and

naphthalene, either as components of cattle odour or as

semiochemicals for haematophagous Diptera in general.

(Z)-3-Hexen-1-ol is associated with induced defence

mechanisms in damaged plants (Agelopoulos et al., 1999),

and naphthalene is a plant-derived volatile semiochemical

mediating maize and sorghum plant interactions with stem-

borer pests (Khan et al., 1997). Aliphatic alcohols have not

been reported previously as part of the volatile profile of

cattle urine. This may be due to differences in the techniques

used in volatile collection in this study and those used in

previously published reports (Hassanali et al., 1986; Bursell

et al., 1988). In this study, the active compounds, i.e.

aliphatic alcohols and cresols, were identified by air entrain-

ment of urine followed by coupled GC–EAG. In previous

studies, urine extracts, prepared by liquid–liquid extraction

using dichloromethane, were fractionated by liquid chro-

matography prior to identification. Fractions with activity

were shown to contain only phenolic compounds, thus

excluding the need for identification of other compounds.

Where the role of volatile semiochemicals in conferring

the differential attractiveness of hosts to cattle flies has

been studied previously, these studies have focused on the

Table 2. The mean percentage of horn flies distributed on herds of

Holstein-Friesian cattle in response to the addition of slow-release

sachet formulations of 1-octen-3-ol (release rate¼ 10.5mg/day)

and 6-methyl-5-hepten-2-one (release rate¼ 14mg/day), used as

individual treatments in separate experiments.

Day of experiment

Treatment Heifer no. Day 1 Day 2 Day 3

1-Octen-3-ol 1226 5� 2ab 0� 0a 6� 2b

1170 4� 1a 2� 2a 4� 3a

1193 11� 3a 12� 8a 22� 7a

1205 24� 4a 22� 4a 18� 7a

1181 9� 7a 10� 2a 14� 4a

1180 20� 2a 12� 5b 14� 3ab

1160 28� 5ab 42� 10a 23� 4b

6-Methyl-5-hepten-2-one 1223 1� 1ab 1� 1a 8� 5b

1178 4� 2a 2� 1a 6� 4a

1225 7� 1a 7� 4a 16� 1b

1168 8� 2a 8� 6a 14� 4a

1196 24� 8a 15� 5a 20� 7a

1138 22� 17ab 44� 5a 23� 9b

1261 35� 12a 23� 14ab 13� 4b

Statistical analysis was carried out using a Kruskal–Wallis test

followed by a multiple-comparison test. Significance indicated by

different letters for each heifer indicates significant difference

between medians of the ranks on a 5% level. Bold type indicates

treated heifers.



differential attractiveness of cattle and hosts belonging to

different taxa to tsetse flies (Vale, 1974a, b; Gikonyo et al.,

2002, 2003). Recent studies have demonstrated that for horn

flies, Ha. irritans, there is natural differential attractiveness

for individuals within herds of Holstein-Friesian heifers, i.e. a

single taxa, and that these differences can be manipulated

following movement of individuals between herds (Jensen

et al., 2004). This study, together with the Jensen et al.

(2004) work, demonstrates, for the first time, the role of

volatile semiochemicals in conferring the differential attrac-

tiveness of these heifers, and the manipulation of fly popula-

tions in the field. The identification of attractants and

repellents in this study supports the hypothesis that attrac-

tiveness of individual hosts is, in part, determined by varia-

tion in the volatile profile of individuals, in particular the level

of volatiles that reduce fly loads. Those semiochemicals that

reduce fly loads may act as outright repellents, or by reducing

the sensitivity of flies to attractants, thus ‘masking’ the pres-

ence of cattle odour. In this study, a tendency towards a

repellent effect was observed when heifers with high fly

loads were treated with two compounds, 1-octen-3-ol and

6-methyl-5-hepten-2-one, which were identified in the labora-

tory as attractants. This activity may have arisen through the

use of slow-release formulations that released inappropriate

levels of the compounds. The average rate of production of

1-octen-3-ol for ox has been determined to be 0.01mg/h (Torr

et al., 1995), which is much lower than the rate used in this

study. Furthermore, higher release rates have been shown to

be repellent for tsetse flies (Vale & Hall, 1985), which may

explain the reduction of fly loads in this study. Thus, further

field-based studies are required to develop the optimum doses

and formulations for these compounds to be attractants, and

to assess the impact of the identified repellent compounds

propyl butanoate, linalool and naphthalene on reducing fly

loads on individual heifers and on whole herds.

In the plant kingdom, it has been established that herbi-

vorous insects avoid energy-wasting visits to feed on or

colonize a host on which they cannot successfully develop,

by responding to volatile components emitted by unsuitable

hosts (Pickett et al., 1998). The identification of attractants

and repellents in this study supports the hypothesis that

similar host/non-host preferences exist for haematophagous

insects in search of an appropriate vertebrate host. Differ-

ences in volatile profiles of cattle are likely to result from

differences in the physiological status of the host, i.e. diet

(Vale, 1981), age, reproductive status and general condition

of the individuals. The genetic background of cattle, i.e. the

breed, is also known to influence host susceptibility. This

has been illustrated by the strong correlation in fly load

between mothers and daughters (Thomas et al., 1987).

Clearly, the identification of naturally occurring semio-

chemicals opens up the way for the development of new

approaches to the monitoring and control of nuisance and

disease-vectoring cattle flies, where attractants and repel-

lents are utilized concurrently as part of a ‘push–

pull’ strategy (Miller & Cowles, 1990). This study also

serves as a model for the identification of new control

agents for other disease-transmitting Diptera affecting

domestic livestock and human beings, e.g. mosquitoes and

biting midges, where the natural differential attractiveness

within a host species can be exploited. The potential for

using volatile, host-derived semiochemicals in reducing

the attractiveness of human hosts to field populations of

malarial mosquitoes, Anopheles gambiae s.l. (Diptera:

Culicidae), has already been demonstrated (Costantini

et al., 2001).

Acknowledgements

The authors wish to thank Alan Todd for statistical advice.

This work was supported by the European Commission

(CEC Contract Mo. AIR3-CT 93–1445). Rothamsted

Research receives grant-aided support from the Biotechnol-

ogy and Biological Sciences Research Council (BBSRC) of

the U.K.

References

Agelopoulos, N.G., Hooper, A.M., Maniar, S.M., Pickett, J.A. &

Wadhams, L.J. (1999) A novel approach for isolation of volatile

chemicals released by individual leaves of a plant in situ. Journal

of Chemical Ecology, 25, 1411–1425.

Bartlett, C. (1985) An olfactometer for measuring the repellent

effect of chemicals on the stable fly, Stomoxys calcitrans (L.).

Pesticide Science, 16, 479–487.

Brown, A.H., Steelman, C.D., Johnson, Z.B., Rosenkrans, C.F. &

Brasuell, T.M. (1992) Estimates of repeatability and heritability

of horn fly resistance in beef cattle. Journal of Animal Science,

70, 1375–1381.

Bursell, E., Gough, A.J.E., Beevor, P.S., Cork, A., Hall, D.R. &

Vale, G.A. (1988) Identification of components of cattle urine

attractive to tsetse flies, Glossina spp. (Diptera: Glossinidae).

Bulletin of Entomological Research, 78, 281–291.

Costantini, C., Birkett, M.A., Gibson, G. et al. (2001) Electro-

antennogram and behavioural responses of the malaria vector

Anopheles gambiae to human-specific sweat components.

Medical and Veterinary Entomology, 15, 259–266.

Fraser, A.F. & Broom, D.M. (1990) Farm Animal Behaviour and

Welfare. Bailliere Tindall, London.

Gibson, G. & Torr, S.J. (1999) Visual and olfactory responses

of haematophagous Diptera to host stimuli. Medical and

Veterinary Entomology, 13, 2–23.

Gikonyo, N.K., Hassanali, A., Njagi, P.G.N., Gitu, P.M. &

Midiwo, J.O. (2002) Odor composition of preferred (Buffalo and

Ox) and nonpreferred (Waterbuck) hosts of some savanna tsetse

flies. Journal of Chemical Ecology, 28, 969–981.

Gikonyo, N.K., Hassanali, A., Njagi, P.G.N. & Saini, R.K. (2003)

Responses of Glossina mortisans mortisans to blends of electro-

physiologically active compounds in the odors of its preferred

(buffalo and ox) and nonpreferred (waterbuck) hosts. Journal of

Chemical Ecology, 29, 2331–2346.

Hall, D.R., Beevor, P.S., Cork, A., Nesbitt, B.F. & Vale, G.A.

(1984) 1-Octen-3-ol: a potent olfactory stimulant and attractant

for tsetse isolated from cattle odours. Insect Science and its

Application, 5, 335–339.

Hassanali, A., McDowell, P.G., Owaga, M.L.A. & Saini, R.A.

(1986) Identification of tsetse attractants from excretory



products of a wild host animal, Synceras caffer. Insect Science

and its Application, 7, 5–9.

Holloway, M.T.P. & Phelps, R.J. (1991) The responses of

Stomoxys spp. (Diptera: Musicidae) to traps and artificial host

odours in the field. Bulletin of Entomological Research, 81,

51–55.

Jensen, K.M.V., Jespersen, J.B., Birkett, M.A., Pickett, J.A.,

Thomas, G. & Wadhams, L. (2004) Variation in the load of the

horn fly, Haematobia irritans (Diptera: Muscidae) in cattle herds

is determined by the presence or absence of individuals. Medical

and Veterinary Entomology, 18, 275–280.

Khan, Z.R., Ampong-Nyarko, K., Chiliswa, P. et al. (1997)

Intercropping increases parasitism of pests. Nature, 388,

631–632.

Maddrell, S.H.P. (1969) Secretion by the Malpighian tubules of

Rhodnius. The movement of ions and water. Journal of

Experimental Biology, 51, 71–97.

Mihok, S., Kang’ethe, E.K. & Githaiga, K.K. (1995) Trials of traps

and attractants for Stomoxys spp. (Diptera: Muscidae). Journal

of Medical Entomology, 32, 283–289.

Miller, J.R. & Cowles, R.S. (1990) Stimulo-deterrent diversion: a

concept and its possible application to onion maggot control.

Journal of Chemical Ecology, 16, 1367–1382.

Mullens, B.A., Peterson, N., Dada, C.E. & Velten, R.K. (1995)

Octenol fails to lure stable fly to insecticide. California

Agriculture, 49, 16–18.

Pickett, J.A. & Woodcock, C.M. (1996) The role of mosquito

olfaction in oviposition site location and in the avoidance of

unsuitable hosts. Olfaction in Mosquito–Host Interactions (ed. by

R. G. Bock and G. Cardew), pp. 109–123. CIBA Foundation

Symposium 200. John Wiley & Sons Ltd, Chichester, UK.

Pickett, J.A., Wadhams, L.J. & Woodcock, C.M. (1998) Insect

supersense: mate and host location by insects as model systems

for exploiting olfactory interactions. The Biochemist, August,

8–13.

Schofield, S., Cork, A. & Brady, J. (1995) Electroantennogram

responses of the stable fly, Stomoxys calcitrans, to components

of host odour. Physiological Entomology, 20, 273–280.

Schofield, S., Witty, C. & Brady, J. (1997) Effects of carbon

dioxide, acetone, and 1-octen-3-ol on the flight response of the

stable-fly, Stomoxy calcitrans, in a wind tunnel. Physiological

Entomology, 22, 380–386.

Soler Cruz, M.D., Vega Robles, M.C. & Thomas, G. (1996) In vivo

rearing and development of Wohlfartia magnifica (Diptera:

Sarcophagidae). Journal of Medical Entomology, 33, 586–591.

Stuellet, P. & Guerin, P.M. (1994) Identification of vertebrate

volatiles stimulating olfactory receptors on tarsus-I of the tick

Amblyomma-variegatum Fabricus (Ixodidae). 2. Receptors out-

side the Hallers organ capsule. Journal of Comparative

Physiology A, Sensory Neural and Behavioural Physiology, 174,

39–47.

Thomas, G., Prijs, H.J. & Trapman, J.J. (1987) Factors contribut-

ing to differential risk between heifers in contracting summer

mastitis. Summer Mastitis (ed. by G. Thomas, H. J. Over,

U. Vecht and P.Nansen), pp. 30–39.MartinusNijhof,Dordrecht.

Torr, S.J., Hall, D.R. & Smith, J.L. (1995) Responses of tsetse flies

(Diptera: Glossinidae) to natural and synthetic ox odours.

Bulletin of Entomological Research, 85, 157–166.

Vale, G.A. (1974a) The responses of tsetse flies (Diptera:

Glossinidae) to mobile and stationary baits. Bulletin of

Entomological Research, 64, 545–588.

Vale, G.A. (1974b) Direct observations on the responses of tsetse

flies (Diptera: Glossinidae) to hosts. Bulletin of Entomological

Research, 64, 589–594.

Vale, G.A. (1981) An effect of host diet on the attraction of tsetse

flies (Diptera: Glossinidae) to host odour. Bulletin of Entomo-

logical Research, 71, 259–265.

Vale, G.A. & Hall, D.R. (1985) The use of 1-octen-3-ol, acetone

and carbon dioxide to improve baits for tsetse flies, Glossina spp.

(Diptera: Glossinidae). Bulletin of Entomological Research, 75,

219–231.

Vale, G.A., Flint, S. & Hall, D.R. (1986) The field responses of

tsetse flies, Glossina spp. (Diptera, Glossinidae) to odours of

host residues. Bulletin of Entomological Research, 76, 685–693.

Vale, G.A., Flint, S. & Hall, D.R. (1988) The olfactory responses of

tsetse flies, Glossina spp. (Diptera, Glossinidae) to phenols and

urine in the field. Bulletin of Entomological Research, 78,

293–300.

Wadhams, L.J. (1990) The use of coupled gas chromatography:

electrophysiological techniques in the identification of insect

pheromones. Chromatography and Isolation of Insect Hormones

and Pheromones (ed. A. R. McCaffery and I. D. Wilson),

pp. 289–298. Plenum Press, New York.

Warnes, M.L. & Finlayson, L.H. (1985) Responses of the stable fly,

Stomoxys calcitrans (L.) (Diptera: Muscidae) to carbon dioxide

and host odours. II. Orientation. Bulletin of Entomological

Research, 75, 717–727.

Warnes, M.L. & Finlayson, L.H. (1986) Electroantennogram

responses of the stable fly, Stomoxys calcitrans, to carbon

dioxide and other odours. Physiological Entomology, 11,

469–473.

Accepted 21 September 2004

