Above: Plane barrier spraying fipronil near Aukazoabo, SW Madagascar.
Below: Swarms of the Madagascar migratory locust. Photos: C. Tingle
|
Above: Plane barrier spraying fipronil near Aukazoabo, SW Madagascar.
|
|
Madagascar lies in the Indian Ocean
off the east coast of Africa. An isolated and unique tropical ecosystem, the
island supports a rich and diverse flora and fauna, including many endemic
species and families of organisms found nowhere else in the world. The global
importance of biodiversity in Madagascar is internationally acknowledged. This
is a truly special place for natural history and of high conservation value.
The population of nearly 15 million are
mainly farmers, with 78% living in rural areas. Income per head is among the
lowest in the world. The economy is almost entirely dependent on agriculture,
although only about 5% of the total land area of 58.2 million ha. is cultivated.
France accounts for about 30% of the country’s imports and exports and close
links remain with French institutions and companies.
In recent years the island has experienced a
major locust plague. The outbreak of the Madagascar Migratory Locust (Locust
migratoria capito) and the Red Locust (Nomadacris septemfasciata)
began in 1996 and the Malagasy government declared an emergency situation in
1997, with a call for international aid to fund locust control operations.
The operations were oriented entirely around
large-scale application of synthetic chemical insecticides. Given the unique
environment, no chemical spray campaign should be undertaken lightly, and
particular concerns about fipronil, initially the main insecticide applied
aerially on ecologically important non-target species, appear to have been
ignored. Negative impacts of fipronil led the government to withdraw
authorisation for use against locust swarms in February 1999, but even after
this – between March and May 1999 alone – a European Union-supported
operation aerially applied 213,000 litres of the active ingredient (1).
As the Madagascar locust campaign continued,
some donors questioned the strategies employed. Their investigations suggest the
scale of food losses may be exaggerated and that the volume of spray treatment
could be reduced. A UN Food and Agriculture Organisation (FAO) report of June
1999 noted: ‘The impact of the locust infestation on the national production
appears to have been minor largely because the large production areas were not
infested or became only infested after the harvest time.' (2) Prevention
strategies are held to be safer and more appropriate than an annual emergency
response. But in spite of many concerns, funds for an emergency action were
again sought for 1999-2000, and FAO drew up a budget of a further US$17.6
million for locust application and training (3)
Donor response to locusts
Since 1997 pesticides worth more than US$35 million
have been applied from the air and the ground. Emergency funds for spraying
campaigns were provided by the European Commission (EC) and Coopération Française,
with technical assistance from the FAO, the international body charged with
responsibility for coordination of locust control. The World Bank provided
initial funding but withdrew its support in 1998 because of concerns over the
pesticides used (4). Unlike the World Bank, the European Commission has no
policy to govern decisions on pesticides procurement, although guidelines have
been developed recently (5).
Concern over the extent of the spray
campaigns and potential impacts prompted investigations by the US Agency for
International Development (USAID), the German development agency GTZ, and the UK
Department for International Development (DFID). Initially, the only donor to
provide funds for assessment of environmental effects was DFID (and this did not
come as part of the emergency package), despite the fact that all the agencies
involved will report that they assist only development projects that protect the
environment.
In other words, over a two year period about
US$35 million was spent on a massive spraying programme to prevent agricultural
losses which the FAO itself admitted did not materialise because of issues
unrelated to the pesticide spray programme – and without the inclusion by most
donors of an assessment of possible adverse impacts of the pesticides used on
health and the environment. Lack of documentation and poor infrastructure makes
it impossible to confirm the real impacts on farmers, and governments cannot
ignore locust threats. However at the very least emergency control programmes
must go hand in hand with detailed, well-designed monitoring of yield losses,
qualitative and effective spray operations and health and environmental
monitoring.
Living with locusts
Locust attacks have troubled Malagasy farmers for
centuries. While potential crop losses are a serious concern, rural communities
have developed a range of coping strategies. One report points out that: ‘The
Malagasy are practical people who catch and eat the protein and fat-rich
locusts, grind and feed them to their pigs, plant rice at many different times
so that potential losses due to locusts and other constraints are limited.
Malagasy farmers burn fires near their fields to produce noise and smoke to
deter and burn locusts, plant back-up crops, and drive locusts from particularly
susceptible crops.' (6) But once an outbreak reaches plague proportions, none of
these are effective.
Farmers value spraying, but some have
expressed frustration at their exclusion from the recent campaigns. Those
interviewed by a USAID team repeatedly indicated that they would like to see
anti-locust efforts involve their local communities and wanted ‘access to
powder’ for better and more timely control as in the past: some asked ‘Why
pay thousands of dollars for helicopters when our labour is free?’ The
‘powder’ is generally highly toxic (for example propoxur) and is usually
applied without any protective clothing, by pesticide-inexperienced users.
While farmers are prepared to expend energy
on locust control, without regular support they are unlikely to carry out
preventative control measures, which require a dedicated, independent
organisation.
Locusts have an unprecedented place amongst
insects in human perceptions of disaster scenarios. A locust swarm is
awe-inspiring to behold and can be devastating to crops, to the livelihoods of
subsistence farmers and may affect the economy of a country dependent on
agriculture. Locusts occur and cause problems from Brazil to Kazakhstan;
Australia to parts of India: different species in different areas, but locusts
nonetheless.
Until the 1940s and 1950s little could be
done against locust plagues, but with the development of synthetic insecticides,
an effective weapon came to hand. Since then, locust plagues worldwide have been
fought with large (often very large) quantities of insecticides. Locust
emergencies still result in intensive use of pesticides to an extent which would
be generally unacceptable in agriculture in normal circumstances.
However, locusts are merely grasshoppers for
the majority of the time. They only transform into ‘locusts’, given
particular climatic conditions. When this occurs, they change their behaviour
and appearance from solitary grasshoppers to the gregarious locust
‘super-organism’ which can be so damaging. Locust swarms will fluctuate from
year to year and are frequently reduced by poor weather conditions, dry seasons
and lack of moist breeding sites.
The actors in the emergency
Emergency aerial spraying is always costly and never
sustainable. The intention is to reduce locust populations to levels which allow
a sustainable preventative strategy to be re-established. Emergency donations
should not take place without balanced support for long-term monitoring and
prevention capacity. But the recent situation in Madagascar has been
characterised by donor support only for emergency actions.
In the past, locust control in Madagascar
fell under the wing of the Regional Crop Protection Service (DPV) of the
Ministry of Agriculture through its Centre Antiacridien de Betioky-Sud (CAB [the
Anti-Locust Observation Centre]), established in 1932. For over 35 years,
between 1958 and 1994, CAB successfully carried out field surveys to monitor
locust population build up (7). If locusts posed a threat of outbreak, teams of
technicians carried out ground-spray campaigns (or, when necessary, aerial spray
interventions). These regular field surveys are considered the best preventive
approach (8) and during this period the country experienced no major locust
plagues.
In the mid-1990s, a lack of government funds
led to a breakdown in CAB’s operations. According to FAO, which operates an
Emergency Prevention System to provide early warning and coordination on
outbreaks of locusts, good rainfall caused a gradual increase in the population
of the Madagascar migratory locust and the red locust, and an ‘outbreak’ was
declared in 1997. This build up would probably have been prevented if effective
scouting and controls had continued between 1994 and 1997 (9). FAO made
emergency appeals for funding to initiate a pesticide spray programme (10) and
in February 1998 the government established the Comité National de Lutte
Antiacridienne (CNLA) under army control to take over anti-locust operations
from the DPV.
Two helicopters were hired by the European
Union, and two helicopters belonging to the President of Madagascar were brought
into the operations. CNLA and FAO drew up plans for a large-scale aerial spray
campaign of the pesticide fipronil, manufactured by the French company, Rhone
Poulenc (11) (since merged with AgrEvo to form Aventis). For the purposes of the
emergency locust control operations, the island was divided into five
‘Intervention Zones’, covering four-fifths of the landmass.
In November 1998 the Project de Lutte
Anti-acridienne dans l’Aire Grégarigène (PLAAG), was set up by the FAO,
European Union and French cooperation to carry out the spray operations in Zone
5 and to report to CNLA. The PLAAG-controlled operations have been primarily
based on fipronil spraying from helicopters contracted from France. Zone 5 is in
the southwest, where locust outbreaks generally occur before swarming
northwards.
In the CNLA controlled areas, Zones 1-4,
there was a mixture of aerial spraying by a local contractor and a ‘lutte
populaire’, where the army contracts local NGOs (known as Organisations de
Société Civile) to carry out pesticide application. According to one observer,
since late 1998 these local NGOs were asked to work with local farmers, but most
have been established on an entrepreneurial basis only to get the CNLA
contract. As a result few have good contacts with the farmers, and most lack the
technical expertise, outreach networks and administrative experience required to
do the job correctly and effectively.
Kill or overkill?
In the 1998-99 season alone, a total of 780,267 ha
over all zones were treated with total cover sprays, and 2.954 million ha in
Zone 5 were ‘protected’ with insecticides on a ‘barrier’ basis (12),
where aerial spraying is applied in successive strips to kill immature locust
hoppers passing through ground vegetation, whilst leaving large unsprayed areas
between sprayed strips.
A USAID report suggests that the 1998
predictions of locust damage potential were exaggerated and that on a national
scale little crop loss actually occurred. Dry weather reduced locust survival of
the October 1998 breeding season and the plague followed a general pattern of
diminishing naturally after 18 months-2 years (13).
In April/May 1999 more swarms and hopper
(young locusts) bands were reported by FAO to have the characteristics of an
outbreak situation (14). Despite apparently minimal crop losses and fears of
environmental damage caused by pesticides, the spraying continued. While an
emergency response is a poor substitute for the on-going monitoring support in
locust-vulnerable areas, if such a measure becomes necessary, choices can still
be made to ensure that the response is as effective as possible and minimises
adverse side-effects.
 |
 |
 |
Species at risk from fipronil: Large eared tenrec (Geogale aurita), Madagascar bee-eater (Merops superciliosus), Chameleon (Chameleo oustaleti). |
||
Selection of pesticides
Madagascar has registered a number of products for locust
control: fipronil (see p. 20) (restricted in February 1999), and three Insect
Growth Regulators (IGRs) diflubenzuron, hexaflumuron and triflumuron. According
to the FAO report, the pesticides in stock at 20 May 1999 for locust control
included: fipronil; diflubenzuron and triflumuron (Alsystin); the pyrethroids
deltamethrin and phenothrin; chlorpyrifos (an organophosphate); and propoxur (a
carbamate).
Some alternatives to chemical pesticides are
becoming available, with the most promising being Green Muscle based on a fungal
pathogen Metarhizium anisopliae which is specific to locusts and
grasshoppers. A drawback is the slow-acting time for effectiveness, a
characteristic shared, however, by the much-used fipronil.
Although a wide range of insecticides have
been used elsewhere in the world to combat locust emergencies, the early and
mid-phases of the Madagascar campaign placed heavy emphasis on the use of
fipronil, which had never before been used anywhere in a full-scale emergency
campaign. Given the unique environment, no large-scale chemical spray campaign
should be undertaken lightly, but in particular concerns about the impact of
fipronil on ecologically important non-target species appear to have been
ignored.
The World Bank Panel recommended that
deltamethrin replace fipronil because of potential environmental hazards. The
sales of pesticides would clearly be substantial, and, according to FAO, this
change led to ‘heated discussions that are on-going and have overshadowed some
major problems with swarm control that need urgently to be resolved.' (15)
Logistical difficulties – limited
infrastructure and poor roads – led to decisions to spray aerially, but in
recognition of this the World Bank indicated that all aircraft used be equipped
with appropriate agricultural navigation (GPS – Global Positioning System)
systems and detailed monitoring (16) – a recommendation which was also not
followed.
 |
Above: Examining mound of termites (Coarctotermes species). Photo: C. Tingle |
Environmental threats
The widespread use of highly toxic pesticides in this
fragile ecosystem has understandably provoked concern. The World Bank and FAO
sent teams to review environmental concerns and strategy. But efforts to monitor
environmental impacts of pesticides used did not begin until more than a year
into the emergency, with the arrival of specialist ecotoxicologists at the
beginning of 1998.
Funds were provided by DFID and GTZ, and the
UK-based Natural Resources Institute (NRI) carried out the study (17). In
collaboration with DPV, GTZ and in liaison with FAO and the donor community, the
first-ever programme in the world to monitor the environmental impact of an
emergency locust control was established. Funding was extremely limited and
difficult choices had to be made about what could be included. Spray operations
with just two of the many insecticides in use were selected for monitoring and
studies evaluated impacts on non-target terrestrial wildlife including mammals,
birds, lizards and invertebrates. Although the pilots are expected to avoid
spraying water bodies and sensitive habitats, this cannot be guaranteed without
GPS equipment. Water will also be exposed to spray drift and residue runoffs
(18), but funding constraints did not allow this to be investigated.
The NRI study found that fipronil had a
serious impact on termite populations in the sprayed areas. Six months after
spraying, there was evidence of few, if any, healthy colonies in termite mounds
within the barrier spray areas, including those mounds in the inter-barrier
spaces. A follow-up 10 months later in November 1998 found that about 95% of
termite colonies in the fipronil spray barriers were ‘dead’, whilst almost
50% of termite colonies in the sprayed area as a whole were no longer active
(19). The authors point out that: ‘the ecological implications are unknown,
but could be serious given the importance of termites in nutrient cycling, soil
structure and water infiltration, and as a food source for a range of higher
animals.' (20) Destruction of termites on the scale found could be devastating.
In semi-arid areas, termites often fulfil the role taken by earthworms in
temperate and moist areas. They play a key role in the ecological functioning of
the habitat and in the ecosystems food web.
The NRI study also monitored non-target
impacts of barrier spray operations with triflumuron. The findings were that
termites were initially affected to a lesser extent, and within 10 months the
colonies appeared to have recovered. The later USAID field assessment found
scores of dead termite mounds in areas known to have been sprayed with fipronil
in 1998 (21).
The NRI study also found some evidence of
adverse impacts of fipronil on other non-target insects and possibly on some
lizards and birds (22). Furthermore, potential indirect effects on birds and
other wildlife were identified.
In May 1998 the Madagascan Office National
pour l’Environnement (ONE) took over responsibility for the study of the
impact of the locust control operations on non-target species. USAID provided
funding to enable the monitoring programme to continue and to establish further
studies at additional sites, with technical advice, training and quality control
from NRI.
What impact on human health?
Given the extent of spraying in Madagascar and the
potential for pesticides to enter into the food chain, serious questions must be
raised about the health impact on the human population. Yet in spite of this
there has been no systematic medical monitoring of exposed communities. A survey
of 100 people carried out in one area, Mahatsinjo, found that 60% showed
symptoms of pesticide poisoning. Of a further 38 persons tested, 80% showed
reduction in cholinesterase activity due to exposure to organophosphate or
carbamate insecticides (23).
The pigs kept by farmers in some areas roam
freely and are voracious eaters. As pork is widely eaten the exposure risk to
humans from pesticide residues must be considered.
Fipronil has been classified by the US
Environmental Protection Agency (USEPA) as a possible human carcinogen (24) and
the EC has proposed to restrict use on agricultural produce likely to be
consumed by children. Any chronic effects among local people in Madagascar would
not be known for years and even then only a carefully designed and instigated
study would be able to conclude on cause and effect. The health effects of long
term exposure of entire populations from any locust spray campaign is
undocumented (25), and no studies have been conducted of the incidence of cancer
in rural Madagascan populations in order to provide a base line against which to
measure future impacts.
Where farmers are issued with insecticide
powders serious health concerns need also to be addressed. The main insecticides
applied manually by farmers are the organophosphates chlorpyriphos ethyl
(Dursban) and fenitrothion and the carbamate, propoxur (a probable human
carcinogen) (26). One research team observed propoxur dust being applied by
farmers, partly supervised by the Office National pour l’Environnement, with
no or inadequate protective clothing. The dust was supplied in hessian sacks
which were beaten by untrained and largely unprotected operators (wearing only a
dust mask). As a minimum operators should have been provided with a face shield,
gloves, overalls and boots and with a clean supply of soap and water for washing
after dusting or spraying (27).
All the pesticides provided to farmers are
free and may be used for purposes other than locust control, thus opening the
door to a whole range of potential human health problems.
Food shortages questionable
The USAID team spent two weeks in the country in May
1998 and studied two of the locations of high locust concentration, Miandrivazo
and Ihosy, where control helicopters were based and carrying out spray
operations. The team found that the crop losses from locusts in the area were
substantially below anticipated levels. In Miandrivazo the main staple crops
were maize, rice, potatoes, groundnut and sweet potatoes. Individual farmers
interviewed told of losses to some of their rice and maize crop, in some cases
more than 50%. However when asked ‘what will you do to get food and eat’,
82% answered that they had planted cassava as a back up crop, and they would eat
this, or sell it to purchase other foods. Farmers in other rice growing areas
reported it had been a good cropping year. USAID also found that rice and maize
prices remained generally lower than the previous year, indicating adequate
supplies on the market. FAO backed up this observation (28) but nevertheless
sanctioned a continuation of the spraying programme.
Need for education
Farmer knowledge of the health effects of pesticides
and of good practice in pesticide use and handling is very limited. In one area
farmers indicated that they knew they should wash after applying pesticides, and
expressed some concern about the safety of drinking water or of grazing cattle
in sprayed areas. In another survey area farmers were unaware that they should
not graze cattle in areas where aerial spraying had occurred, even though many
of them were herders. The food chain impacts are also unfamiliar to farmers, who
indicated they would not eat contaminated locusts, but would feed these to their
pigs. In a USAID follow up to the May 1998 study, researchers found that in all
surveyed places people were eating locusts. In the south they had not heard
warnings against human consumption of treated locusts (29).
It is clearly unacceptable to provide highly
hazardous pesticides, which will be spread by hand with no or little protective
clothing, exposing farmers to possible acute and chronic toxicity.
Storing up trouble?
Stockpiles of locust pesticides, particularly of
fipronil, are in larger quantities than would be recommended for normal control,
and could lead to future disposal problems. (30) Problems with empty containers
being used to store food and water, which should have been predictable, have
already arisen. Initially there were no mechanisms to retrieve containers, which
posed a health and environmental hazard, but a subsequent operation gathered
over 5,000 empty insecticide drums (31). There is no suitable means of safe
disposal of the drums in the country and the cost of shipping and safely
destroying them is estimated to be around US$132,000. A number of disposal
operations have already taken place in Madagascar, with 43,500 tonnes of
dieldrin being removed in 1993 and a further 100 tonnes in 2000. Contractors are
now requested to include provisions to take back empty pesticide containers.
Conclusion
Decisions to use pesticides on the scale of the recent
locust-control operations in Madagascar must be undertaken with extreme caution.
Donor coordination is essential: with public funds at their disposal, all donors
should reasonably be expected to follow the best environmental and health
practices. The USAID report found ‘no weighing of the pros and cons of
expensive and potentially harmful locust control spray campaigns.' (32)
Most donors express a commitment to environmental protection
and ‘sustainability’ in their development plans which is inconsistent with
the support for widespread spraying of pesticides without the
internationally-acknowledged safeguards. Base line health studies, environmental
monitoring, analysis of the effectiveness of spray campaigns, and independent
crop loss assessments are undoubtedly costly, but they must be seen as an
essential part of large-scale pesticide application. Without them, it is
impossible to demonstrate whether the spraying has been effective and safe, and
the main beneficiaries of large-scale locust spraying programmes, paid for with
public funds, may be the pesticide companies supplying products and not the
‘poor’ farmers who are meant to be protected.
Past experience points to the dangers of
ignoring a precautionary approach. Years of spraying locust outbreaks in the
Sahel with the environmentally-persistent organochlorine dieldrin are now being
paid for in residues throughout the global environment, the food chain and in
the human body, and costly inter-governmental negotiations are seeking a phase
out of production and use of such Persistent Organic Pollutants.
The decision-making processes governing
selection of pesticides for use in emergency locust control need to be
thoroughly reviewed. Why was there such reliance on fipronil, a pesticide with
wide differences in toxicity to different animals in an environment with such
diverse and unique fauna? The pesticide suppliers must bear some responsibility
for the choice. Users are dependent on scientific and technical data from
industry. The product manufacturer, Rhone-Poulenc, was aware that fipronil field
trials had shown adverse impacts on a range of insects, and that it is highly
toxic to certain birds, highly toxic to some fish and most aquatic
invertebrates: does the company value profits over human health, environmental
and social concerns?
Primary responsibility for locust population
monitoring and prevention strategies needs to rest with the government. The CAB
proved able, in the past, to operate an effective preventive control strategy,
and donors need to demonstrate a serious commitment to assisting in rebuilding
and reinforcing local capacity to manage a local migratory pest control network.
There are lessons here to be learnt for all
future locust outbreaks, which are likely to occur until sustainable
preventative strategies are in place in all countries at risk. The
decision-making processes behind locust control operations need to be more open,
transparent and accountable to ensure the strategies adopted are in the best
interests of farmers, tax-payers and the environment. And there must be emphasis
on consistent funding for long-term solutions based on reliable and
participatory strategies for sustainable agriculture.