veterinary

parasitology
ELSEVIER Veterinary Parasitology 53 (1994) 109-116

Rotational grazing for control of gastrointestinal

nematodes of goats in a wet tropical environment
I.A. Barger *,a, K. Siale b, D.J.D. Banks c, L.F. Le Jambre a
aCSIRO Division of Animal Health, Private Bag, Arrnidale, N.S. W.. 2350, Australia
bMinistry of Agriculture and Forestry, Nuku'alofa, Tonga
CAustralian Quarantine and Inspection Service, GPO Box 858, Canberra, A. C. T. 2601, Australia
(Accepted 11 August 1993)

Abstract

A preliminary experiment involving contamination of pasture plots with eggs of Hae-

monchus contortus, Trichostrongylus colubriformis and Oesophagostomum columbianum
every month for a year established that in the tropical environment of the Pacific island of
Tongatapu, hatching and development of all species was rapid and continuous, with a short
survival on pasture (3-7 weeks) of the resulting infective larvae. These results indicated
that a rotational grazing system consisting of ten paddocks grazed in sequence for 3.5 days
at a time may permit a reduction in the frequency of anthelmintic treatment of goats. In
comparison with an adjacent set-stocked flock which required treatment on three occa-
sions during the year when mean flock egg counts exceeded 2000 eggs per gram (EPG),
rotationally grazed goats generally maintained mean egg counts of less than 1000 EPG.
Anthelmintic treatment was only given to rotationally grazed goats individually as they
kidded, and there were indications that even this precaution was unnecessary. Because of
the expense of frequent anthelmintic treatment and the resulting selection of strains of
anthelmintic-resistant nematodes, rotational grazing of small ruminants through fencing,
tethering or herding deserves further investigation as a nematode control option in wet
tropical environments.

Key words: Haemonchus contortus; Trichostrongylus colubriforrnis; Oesophagostornum columbianum;

Goat; Grazing management

*Corresponding author.

0304-4017/94/$07.00 © 1994 Elsevier Science B.V. All fights reserved

SSD10304-4017 ( 93 ) 00586-N
110 LA. Barger et al. / Veterinary Parasitology 53 (1994) 109-116

1. Introduction

Several Pacific Island countries have attempted to increase production of small

ruminants in order to reduce their reliance on imported sheep and goat meat.
Under the wet tropical conditions of these islands, nematode parasites, chiefly
Haemonchus contortus and Trichostrongylus colubriformis, cause heavy mortality
and production losses on commercial goat farms (Hussain et al., 1983; Banks et
al., 1990). Control using anthelmintics alone is expensive, requiring treatment
every 3-4 weeks, and has rapidly resulted in widespread resistance to the small
number of broad-spectrum anthelmintics currently available. Methods of con-
trolling gastrointestinal nematode parasites that do not rely on frequent anthel-
mintic treatment are urgently required.
Although climatic conditions permit nematode egg hatching and larval devel-
opment throughout the year in the wet tropics, Banks et al. (1990) found that
survival times of the resulting infective larvae on pasture in Fiji were much shorter
( 5-13 weeks) than the survival times of up to 12 months reported from temper-
ate climates. Similar short survival times of infective larvae on pasture have been
reported during the wet season in Nigeria (Okon and Enyenihi, 1977; Fakae and
Chiejina, 1988), Guadeloupe (Aumont and Gruner, 1989) and north Queens-
land (Fabiyi et al., 1988). Aumont and Gruner (1989) and Banks et al. (1990)
suggested that the short survival times of infective larvae may make control
through rotational grazing practicable in tropical regions. Such methods have not
been successful in temperate climates because the time needed to ensure a signif-
icant reduction in larval numbers through spelling of pastures is agronomically
and economically unacceptable (Donald, 1968; Southcott, 1971; Gibson, 1973 ).
There were two aims of the present study: to see whether the pattern of rapid
egg hatching and larval development, allied with short survival of larvae on pas-
ture, observed by Banks et al. (1990) in Fiji could also be demonstrated in the
cooler climate of Tonga, and if so, whether this could be exploited for parasite
control through rotational grazing.

2. Materials and methods

2.1. Site

The island of Tongatapu (21 ° 10' S, 175 ° 10'W) in Tonga has an mean annual
rainfall of around 1800 mm. Rain can be expected in any month of the year, but
there is a well-defined wet season from October to May. Mean monthly maxi-
m u m air temperatures range from 24 ° C in August to 30 ° C in January, with mean
monthly minima in the same months of 16°C and 20°C, respectively. Experi-
mental plots were situated at the Ministry of Agriculture and Forestry Experi-
mental Farm at Vaini, about 8 km southeast of the capital, Nuku'alofa. Pastures
consisted mainly of Guinea grass (Panicum maximum).
 LA. Barger et al. / Veterinary Parasitology 53 (1994) 109-116 111

2.2. Larval development and survival

Thirteen plots, each 5 m × 2 m were laid out in March 1989 on pasture that had
not been grazed by goats for the previous 6 months. Drainage ditches were dug
between plots to avoid cross-contamination during heavy rain, and the array of
plots was fenced to prevent the entry of animals. From April 1989 to March 1990,
one plot was randomly selected and uniformly contaminated in the first week of
each month with faeces from naturally infected goats. The weight of faeces used
to contaminate each plot was calculated from the current faecal egg count of the
donor goats to contain a total of one million eggs, with known proportions of
Haemonchus, Trichostrongylus and Oesophagostomum as determined on each
occasion by faecal culture. The thirteenth plot was not contaminated, but was
sampled each month to detect any extraneous infection. Plots were sampled at
weekly intervals after contamination by collecting 60 uniformly distributed plucks
of herbage per plot taken at ground level. Infective larvae were recovered from
these samples using the technique of Heath and Major (1968). Sampling was
avoided within 1 day after heavy rain, and was continued until no larvae were
recovered from 2 consecutive weekly samples. Plots were slashed when required
to maintain a similar pasture height to that of an adjacent continuously grazed
goat pasture; clippings remained on the plot.

2.3. Rotational grazing trial

Preliminary experiments conducted in Fiji had shown that infective larvae of

H. contortus and T. colubriformis were first detected in faeces 4 days after egg
deposition, thus defining the maximum period that goats could be allowed to
graze a paddock without risk of immediate auto-infection. The findings of Banks
et al. (1990) in Fiji, together with those of the present larval survival study, sug-
gested that at least 90% of infective larvae had disappeared from pasture by 4-5
weeks after their first appearance following egg deposition. Accordingly, a ten-
paddock rotational grazing system was devised in which each paddock was grazed
for 3.5 days, then spelled for 31.5 days. A grazing period of 3.5 rather than 4 days
was chosen so that stock movements were made at the same times on the same
days of each week.
Does were randomly allocated to two flocks of 20 animals for comparison over
1 year from October 1991 to October 1992. Each flock grazed an area of 3 ha.
The set-stocked flock had access to the whole of their area every day, while the 3
ha available to the rotationally grazed flock was subdivided, using solar-powered
electric fencing, into ten equal paddocks; the flock followed the rotation de-
scribed above. Each flock was housed separately in slatted-floor sheds at night to
prevent theft or dog attack. The areas used for the two flocks were adjacent, sim-
ilar in soil and pasture cover and both had been regularly grazed by goats before
the trial began. All goats in both flocks were treated with ivermectin on the first
day of the trial, and again on Day 15, to avoid heavy contamination of their graz-
ing areas for the first month of the trial and to minimise differences between the
112 LA. Barger et al. / Veterinary Parasitology 53 (1994) 109-116

two areas attributable to pre-experimental contamination. Faecal samples were

obtained from all goats for faecal egg counts and larval identification at 2 week
intervals thereafter.
The anthelmintic treatment programme for both flocks was based on the re-
suits of these 2-week egg counts. If the flock arithmetic mean egg count exceeded
2000 eggs per gram ( E P G ) at any sampling, then each animal in that flock was
treated with ivermectin at a dose rate of 300/tg kg- 1 liveweight. Preliminary ob-
servations had established that this dose reduced egg counts to zero. In addition,
goats in the rotationally grazed flock were individually treated with ivermectin
immediately after kidding. Kidding occurred during December and September
in both flocks, with approximately half of the does kidding at each time. Kids
remained with their dams for the duration of the trial and received any anthel-
mintic treatments that their dams required.
Faecal egg counts were transformed (logX+ 50) for statistical analysis, and
counts from both flocks in Weeks 2, 4, 16, 18, 28, 30 and 38 were excluded from
the analysis. These times followed anthelmintic treatment of one or both flocks,
when most animals in the treated flock had zero counts, leading to abnormally
low or zero within-flock variance. Statistical significance of the difference be-
tween flocks in egg counts was assessed by a multivariate repeated measures anal-
ysis (Wilkinson, 1990).

3. Results

3.1. Larval development and survival

No infective, third-stage larvae (L3) w e r e recovered from the control plot on

any sampling occasion. Third-stage larvae of Haemonchus, Trichostrongylus and
Oesophagostomum were routinely recovered in the first week after plot contami-
nation in any month of the year, with the longest period of recovery being 7 weeks
in May, October and November. Concentrations on pasture of L 3 of the three
genera, corrected for numbers of eggs used to contaminate the plots, are shown
for each month in Fig. 1.
All genera appeared to be remarkably similar in their ability to hatch, develop
to L 3 and survive under Tongan conditions. Concentrations of L3 o n pasture
tended to be higher during the warmer months from November to February, with
survival curtailed to 3 or 4 weeks in December, January and February.

3.2. Rotational grazing trial

Geometric mean faecal egg counts of the rotationally grazed and set-stocked
flocks over the 52 weeks of the trial are shown in Fig. 2. Mean counts over the
full period were higher in the set-stocked flock ( P < 0 . 0 0 1 ) and there was a sig-
nificant time by flock interaction ( P = 0.001 ), indicating that the difference in
egg count between flocks varied with time. The set-stocked flock required three
 LA. Barger et al. / Veterinary Parasitology 53 (1994) 109-116 113

100004 APR 8 9 OCT 89

1000

100

10
X,?'\ \ \
1 m [] [] []

10000 MAY 89 ~ _ ~ NOV 89

1000

100:

~ 10

[] [] [] ± .~. .m

~00oo JUN 89 DEC 89

E
~" 1000 0-
\ ~ ©
100 '\
\ '\ \
~_ lO
\\ ,\ \
>- 1 0 []
~5
0310000 i JUL 89 JAN 90
--Y.
1000:

G I00~ \ 'L
>
\\
\
\
\
\ \
[] [] []

10000 AUG 89 FEB 90

~
-- 1000

tO0

I0
I
I-- [] [] [] [] []

lO000~ SEP 89 . MAR 9 0

lO00J • - - L ' ~ I D ~ ~ 0--0 Haem

~oo i
i ~ Trich
\\ \,.. ",.
Oesoph

i [] [] [] []
\\ [] []
\,
'\ [] [] []
\\ [] .m .m
o 2 4 6 8 2 4 6 8 1'0

Weeks after egg deposifion

Fig. 1. Pasture larval recoveries of infective larvae per kilogram pasture dry matter, corrected for
numbers of eggs deposited, on plots contaminated sequentially from April 1989 to March 1990. Haem,
Haemonchus contortus; Trich, Trichostrongylus colubriformis; Oesoph, Oesophagostomum
columbianum.

additional anthelmintic treatments when mean egg counts exceeded 2000 EPG;
the times of these treatments are shown by the arrows in Fig. 1. Nine of the 20
goats in this flock also required single treatments in Weeks 42-50 for welfare
reasons; these animals were losing condition and scouring. Rotationally grazed
goats required no additional treatments beyond their planned treatment at kid-
ding, and their mean egg count only exceeded 1000 EPG on one occasion (Week
114 I.A. Barger et al. / VeterinaryParasitology 53 (1994) 109-116

3.5 - [ i

3.0
E
o
u 2.5

~ 2.o

E~
~.c 1.0
~v
E 0.5
o

(5 0.0
0 4 8 12 16 20 24 28 32 36 40 44 48 52
Weeks
lira S e t - s t o c k e d ~ Rotation

Fig. 2. Geometricmean faecal egg counts of set-stocked or rotationallygrazed goats from October
1991 to October 1992. Anthelmintictreatmentswere givento all goats in the set-stockedflockat the
times indicatedby the arrows. Rotationallygrazed goats receivedone treatment each followingthe
birth of kids.

28 ). A total of 69 doses of ivermectin was therefore administered to the 20 set-

stocked goats, compared with 20 doses administered to the rotationally grazed
flock.
Faecal cultures of composite samples from each flock every fortnight showed
no consistent differences between flocks in generic composition of nematode egg
output. H a e m o n c h u s and Trichostrongylus comprised 30-60% of larvae re-
covered from cultures, with Oesophagostomum ranging from 5 to 30%. Haernon-
chus tended to predominate from December to March, with Trichostrongylus more
numerous from April to December.
Two deaths occurred during the trial. One set-stocked doe was found dead in
Week 36 at a time when that flock had very high egg counts; the cause of death
was not established. One rotationally grazed doe died in Week 49 of injuries re-
ceived from unknown causes. Autopsy revealed broken ribs and extensive bruising.

4. Discussion

The results of the larval development and survival experiment confirm results
obtained in the hotter climate of Fiji by Banks et al. (1990), and suggest that
similar results would be obtained in other tropical and possibly sub-tropical re-
gions provided that rainfall throughout the year was sufficient to allow develop-
ment to L 3. The survival times observed in the present study, ranging from 3 to 7
weeks, were even shorter than those observed in Fiji, particularly during the cooler
period from April to August. This may be attributable to the more evenly distrib-
uted rainfall in Tongatapu in 199 l-1992 in comparison with that in Koronivia
and Sigatoka in 1986-1987, when the observations of Banks et al. (1990) were
recorded. At both Fijian sites there was a marked dry season at that time, and
there is evidence that dry weather enhances the survival of nematode L 3 in faeces
 LA. Barger et al. / Veterinary Parasitology 53 (1994) 109-116 115

by limiting their opportunities to migrate to pasture, where death rates are higher
(Barger et al., 1984).
The results of the rotational grazing trial, while encouraging, must be regarded
with caution. Because the experiment was not replicated, it was not possible to
reach a firm conclusion about the cause of the lower egg counts in the rotationally
grazed flock. While we believe that they were a result of the rotation of paddocks
at 3.5 day intervals, the lack of replication means that differences between the
grazing areas occupied by the two flocks cannot be excluded as an explanation.
However, both areas carried a similar pasture cover at the beginning of the ex-
periment, were adjacent and had a similar prior grazing history with goats. In
addition, the two ivermectin treatments given to all goats on Days 1 and 15 of
the trial meant that neither area was contaminated with nematode eggs for a pe-
riod of around 5 weeks at the beginning of the trial, by which time residual larval
populations must have declined to very low levels.
Individual anthelmintic treatment of rotationally grazed does following partu-
rition was instituted because of concern about the possible loss of acquired im-
munity of periparturient does to nematode infection. This phenomenon is well-
known in ewes (Connan, 1976 ) but there was no indication in the present study
in either the set-stocked or rotationally grazed does that unusually elevated egg
counts were associated with parturition or lactation. Rahman and Collins ( 1992 )
recently published evidence of a periparturient rise in faecal egg counts in goats;
in contrast to sheep the rise was mild and reached a peak several weeks before
parturition followed by a decline through lactation.
The lower egg counts of the rotationally grazed goats are consistent with results
of the larval survival experiment, and with unpublished observations by D.J.D.
Banks ( 1988 ) in Fiji that nematode parasitism was rarely a problem in tradition-
ally managed village goats, despite the lack of anthelmintic treatment. Under tra-
ditional management, goats were tethered and moved to a new grazing area every
1 or 2 days. Fencing of rotationaUy grazed pastures, as practised in the present
study, is therefore not essential, although subdivisional electric fencing is less ex-
pensive than traditional permanent perimeter fencing. Management systems em-
ploying goatherds to move the flock to new grazing areas every few days should
be equally successful provided that the flock does not return to any area within 4
or 5 weeks of its last grazing. Rotational grazing systems for small ruminants in
wet tropical climates, whether implemented through fencing, tethering or herd-
ing, therefore warrant further investigation as a low-cost means of reducing reli-
ance on anthelmintics.

5. Acknowledgements

We gratefully acknowledge the assistance and co-operation of Aleki Sisifa, Di-

rector of Agriculture, and the Ministry of Agriculture and Forestry in Tonga, who
provided staff and facilities for this study. We also thank the Australian Centre
for International Agricultural Research for generous financial support.
116 1.A. Barger et al. / Veterinary Parasitology 53 (I 994) 109-116

References

Aumont, G. and Gruner, L., 1989. Population evolution of the free-living stage of goat gastrointes-
tinal nematodes on herbage under tropical conditions in Guadeloupe (French West Indies). Int.
J. Parasitol., 19: 539-546.
Banks, D.J.D., Singh, R., Barger, I.A., Pratap, B. and Le Jambre, L.F., 1990. Development and sur-
vival of infective larvae of Haemonchus contortus and Trichostrongylus colubriformis in a tropical
environment. Int. J. Parasitol., 20:155-160.
Barger, I.A., Lewis, R.J. and Brown, G.F., 1984. Survival of infective larvae of nematode parasites of
cattle during drought. Vet. Parasitol., 14:143-152.
Connan, R.M., 1976. Effect of lactation on the immune response to gastrointestinal nematodes. Vet.
Rec., 99: 476-477.
Donald, A.D., 1968. Ecology of the free-living stages of nematode parasites of sheep. Aust. Vet. J., 44:
139-144.
Fabiyi, J.P., Copeman, D.B. and Hutchinson, G.W., 1988. Abundance and survival of infective larvae
of the cattle nematodes Cooperia punctata, Haemonchus placei, and Oesophagostomum radiatum
from faecal pats in a wet tropical climate. Aust. Vet. J., 65:229-231.
Fakae, B.B. and Chiejina, S.N., 1988. Further studies on the development and availability of infective
larvae of bovine gastrointestinal trichostrongylids on pasture in Eastern Nigeria. Vet. Parasitol.,
28: 143-152.
Gibson, T.E., 1973. Recent advances in the epidemiology and control of parasitic gastroenteritis in
sheep. Vet. Rec., 92: 469-473.
Heath, D.D. and Major, G.W., 1968. A technique for the recovery of strongyle larvae from masticated
herbage. J. Helminthol., 42: 299-304.
Hussain, M.Z., Naidu, R., Tuvuki, I. and Singh, R., 1983. Goat production and development in Fiji.
World Anim. Rev., 48: 25-32.
Okon, D.E. and Enyenihi, U.K., 1977. Development and survival ofHaemonchus contortus larvae on
pastures in Ibadan. Trop. Anita. Health Prod., 9: 7-10.
Rahman, W.A. and Collins, G.H., 1992. An association of faecal egg counts and prolactin concentra-
tions in sera of periparturient Angora goats. Vet. Parasitol., 43:85-91.
Southcott, W.H., 1971. Management practices and helminthosis in the lamb. Aust. Vet. J., 47: 170-
174.
Wilkinson, L., 1990. SYSTAT: The System for Statistics. SYSTAT, Evanston, IL, 677 pp.