Insect Pest Agroforestry Systems Strategy

Download as pdf or txt
Download as pdf or txt
You are on page 1of 35

Agroforestry Systems 50: 243–277, 2000.

 2000 Kluwer Academic Publishers. Printed in the Netherlands.

Insect pest problems in tropical agroforestry systems:


Contributory factors and strategies for management

M. R. RAO1, *, M. P. SINGH2 and R. DAY3


1
International Centre for Research in Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya;
2
Central Arid Zone Research Institute (CAZRI), Jodhpur, Rajasthan, India; 3 CAB International,
Africa Regional Centre, P.O. Box 633, Village Market, Nairobi, Kenya (*Author for corre-
spondence, present address: 11, ICRISAT Colony (Phase 1), Akbar road, Secunderabad-
500 009, AP, India; E-mail: [email protected])

Key words: biomass transfer, insect pests, integrated pest management, natural enemies,
sequential systems, simultaneous systems

Abstract. Agroforestry trees are attacked by a wide spectrum of insects at all stages of their
growth just like other annual and perennial crops. Pest management in agroforestry has not
received much attention so far, but recent emphasis on producing high value tree products in
agroforestry and using improved germplasm in traditional systems, and emergence of serious
pest problems in some promising agroforestry systems have increased awareness on risks posed
by pests. Insects may attack one or more species within a system and across systems in the
landscape, so pest management strategies should depend on the nature of the insect and
magnitude of its damage. Although greater plant diversity in agroforestry is expected to increase
beneficial arthropods, diversity by itself may not reduce pests. Introduction of tree germplasm
from a narrow genetic base and intensive use of trees may lead to pest outbreaks. In simulta-
neous agroforestry systems, a number of factors governing tree–crop–environment interactions,
such as diversity of plant species, host range of the pests, microclimate, spatial arrangement
and tree management modify pest infestations by affecting populations of both herbivores and
natural enemies. Trees also affect pest infestations by acting as barriers to movement of insects,
masking the odours emitted by other components of the system and sheltering herbivores and
natural enemies. In sequential agroforestry systems, it is mostly the soil-borne and diapausing
insects that cause and perpetuate damage to the common hosts in tree–crop rotations over seasons
or years. An integrated approach combining host-plant resistance to pests, exploiting alterna-
tive tree species, measures that prevent pest build up but favour natural enemies and biological
control is suggested for managing pests in agroforestry. Species substitution to avoid pests is
feasible only if trees are grown for ecological services such as soil conservation and low value
products such as fuelwood, but not for trees yielding specific and high value products. For
exploiting biological control as a potent, low cost and environmentally safe tool for pest
management in agroforestry, research should focus on understanding the influence of ecolog-
ical and management factors on the dynamics of insect pest–natural enemy populations. Scientists
and policy makers in national and international institutions, and donors are urged to pay more
attention to pest problems in agroforestry to harness the potential benefits of agroforestry.

Introduction

Next to genetic potential of species, soil fertility and climatic conditions, pests
and diseases play a significant role in determining plant productivity. Among
invertebrates, insects form the most dominant group of herbivores that attacks
plants. A long history of co-evolution has led to the development of diverse
244

types of associations between plants and herbivores. Whereas propagation of


some plant species may be at stake in the absence of insects, almost all cul-
tivated species experience insect injury at some stage or the other in their
life cycle. Although agroforestry research has been institutionalised and
considerable research has been done in many countries in the last three decades
to improve the traditional systems and develop new ones, there is little
research on pest problems in agroforestry. This is partly because trees
employed in the traditional agroforestry systems are least improved, they
generally are not as intensively grown as crops, and they produce low value
products. Furthermore, trees exhibit a high degree of resilience to biotic and
environmental stresses and in general experience relatively less economic
damage due to pests compared with crops. There is also limited plant pro-
tection expertise in the national institutions engaged in agroforestry research,
particularly in sub-Saharan Africa and Latin America (Murphy, 1998).
Trees are susceptible to insect damage just like crops, although the extent
of damage depends on the stage at which they are infested and nature of the
insect. They are attacked by leaf eaters in the nursery, termites and root grubs
during the early establishment stage in the field, defoliators, sap-suckers, gall
formers, stem borers, pod or fruit borers, and seed eaters perpetually during
the later stages. Satya Vir and Verma (1996) reported a long list of insect pests
on 17 leguminous tree species widely used in dryland agroforestry in India.
Recent reports on insect pests of some species such as Inga edulis (Ackerman
et al., 1998), Erythrina spp. (Muniappan, 1993), Gliricidia sepium (Opondo-
Mbai, 1995; Singh Rathore, 1995) and Sesbania sesban (Sileshi et al., 2000)
indicate that a large number of herbivores feed on agroforestry trees. Sesbania
sesban, which is widely being tested as short-duration planted fallows for
soil fertility improvement in east and southern Africa, is attacked by 30 species
of insects (Sileshi et al., 2000). A compilation of insects observed feeding on
20 agroforestry trees in sub-Saharan Africa by Singh Rathore (1995) indicate
that some of the species such as eucalypts (Eucalyptus spp.) and pigeonpea
(Cajanus cajan) harbour as many as 160 insect species in different parts of
the tropics. A review of insect pests of fodder tree legumes by Walter and
Parry (1994) emphasised the potential danger from a large number of insects
to both indigenous and exotic tree species. Although many insects that feed
on trees may not attain the pest status, each tree species mentioned in the
above studies has at least three to five major insects with potential to cause
significant economic losses. Although the case for pest management in
agroforestry was recognised in the eighties (Epila, 1986, 1988; Altieri et al.,
1987; Singh and Singh, 1987), few studies had explored the effect of agro-
forestry on pest infestations of trees and/or crops, and strategies for pest
management.
A team of scientists from the International Centre for Research in
Agroforestry (ICRAF) and the Centres for Applied Biosciences-International
(CABI) debated on the potential pest problems in agroforestry and formulated
an action plan in 1988 (Huxley and Greenland, 1989). As a follow up, a
245

visiting fellowship was established at ICRAF in 1991 to collate and


synthesise all the available information on insect pests in agroforestry (Singh
Rathore, 1995). Since then a number of papers on pest problems in agro-
forestry was published (Day et al., 1996; Dix et al., 1995; Dix, 1996; Dix et
al., 1997; Mchowa and Ngugi, 1994; Mukthar et al., 1996; Singh, 1997; Singh
and Parihar, 1997). There has been a growing concern in recent years about
the potential dangers of pests in agroforestry, as some of the promising new
technologies have shown serious pest problems. The promotion of agroforestry
by different stakeholders, including development agencies, has helped to
recognize that the potential benefits of agroforestry cannot be realized unless
the pest problems are tackled.
The extent of insect damage in any system is determined first by the
primary interaction between the plant species and the herbivore and second
by the interactions among components (tree, crop, soil and environment) of
the system. The consequence of these latter interactions may have a positive,
negative or neutral effect on pests of trees and crops depending upon whether
the pests’ activity enhanced, reduced or remains unaffected, respectively.
Many climatic and management variables influence insect activity individu-
ally or collectively and directly or indirectly.
The magnitude of pest problems in agroforestry is determined by whether
the herbivore interacts with (1) individual components of a system, (2) two
or more components of a system and (3) components across systems in the
landscape. The herbivore interactions with multiple species over different
spatial and temporal scales will have implications for pest management
strategies to be followed. The species-specific pests have to be managed
through host-plant resistance, biological control and/or strategic use of
chemicals depending upon the economic importance of the damage. Manage-
ment plays a significant role in overcoming the system-linked pests. At the
scale of landscape, land use pattern, temporal and spatial arrangement of
vegetation, and the presence of alternative hosts influence insect pests and
their natural enemies. Transmigration of pests, predators and parasites to and
from farmland, grassland and woody plants may occur contemporarily or
seasonally. An understanding of the interactive effects of soil–plant–envi-
ronment on pests and natural enemies at different spatial and temporal scales
is essential for developing integrated pest management strategies.
Agroforestry systems can be broadly categorised into four groups depending
on the nature and extent of tree–crop interactions having implications for
pest problems and their management. They are: (1) blocks of sole trees grown
for fodder (fodderbanks), firewood or pulpwood (woodlots), (2) simultaneous
systems in which trees and crops are grown together in the same field, (3)
sequential systems in which trees and crops are rotated over time in the same
field, and (4) biomass transfer systems in which tree foliage produced in one
field is used for green-manuring of crops in another field. Pests that affect
the productivity of fodderbanks and woodlots are essentially those insects that
infest the specific tree species used in the systems. In simultaneous systems,
246

the direct interactions between trees and crops for growth resources exercise
a strong influence on pests of either or both the components of the systems.
In sequential and biomass transfer systems, the indirect interactions that occur
between trees and crops may lead to a different type of pest problems. While
both soil-borne and foliage-pests affect the sequential systems, it is mostly
pests of tree foliage that are transferred to crops in the biomass transfer
systems.
The objective of this review is to systematically analyse the factors that
contribute to insect pest problems in different agroforestry systems and
examine management strategies to cope with them. Pests unique to crops are
not dealt with here, but emphasis is given to the effect of tree component
and its insect pests on the overall performance of the system. For this reason,
a few pests other than insects such as nematodes, vertebrates, or pathogens
are cited to illustrate specific problems arising due to trees in agroforestry.
The review limits itself to tropical agroforestry systems, drawing examples
wherever possible of systems under contemporary research and development.
In the light of the review by Schroth et al. (2000), pests of agroforestry systems
based on tree crops in the humid tropics are not extensively considered here,
except for some examples to illustrate specific cases. The first part of the
paper deals with factors that contribute to the development of pest problems
as a result of herbivore–plant interactions, across all categories of agroforestry
systems at the three levels outlined earlier. The second part of the paper
discusses pest management strategies relevant to all agroforestry systems. In
the concluding section, we indicate the research needs in the pest manage-
ment area for realising the potential benefits of agroforestry.

Factors contributing to insect pest problems in agroforestry

At tree species level

Exotic species
As in the case of crops, introduction of tree species into new areas has been
occurring for many centuries. Movement of tree germplasm across ecolog-
ical zones and continents will continue to be an important element of tree
domestication. Many newly introduced tree species are widely grown in
tropical agroforestry. Some examples are the use of leucaena (Leucaena leu-
cocephala) as barrier hedgerows for soil conservation, fodderbanks and for
improving soil fertility (Asia and Africa), grevillea (Grevillea robusta) as a
shade tree in tea and coffee (east Africa and India), eucalypts (Eucalyptus
spp.) in boundary plantings around croplands (throughout the tropics), and
neem (Azadirachta indica) in windbreaks and boundary plantings (west
Africa). Exotic species may succumb to local pests in their new environment
or introduced pests from their original home, just as exotic pests accidentally
introduced may pose a major threat to both indigenous and exotic trees. There
247

are many examples of these types of insect–host tree interactions that caused
severe economic losses (Table 1).
An exotic pest in a new environment can assume epidemic proportion and
cause severe damage to its host because of the absence of its natural enemies.
The damage could be far greater than in the pest’s native environment until
populations of its natural enemies build up. That is what happened with the
leucaena psyllid (Heteropsylla cubana) when it first spread to Southeast Asia
and South Asia in the eighties and later to east and southern Africa in the
nineties. Although the psyllid was never considered a major pest of leucaena
in its native Central America, it caused severe and extensive damage to
leucaena in Asia and Africa, demoralising both researchers and farmers at a
time when leucaena was widely being promoted in agroforestry. Introduction
and spread of leucaena from a narrow genetic base was the major reason for
such an outbreak of psyllid. A number of pests of eucalypts such as the psyllid
Ctenarytaina eucalypti, the curculionid Gonipterus scutellatus and the cer-
ambycid Phoracantha semipunctata also spread beyond continental barriers
(Cadahia, 1986). However, they did not cause as severe damage to eucalypts
as the leucaena psyllid to leucaena, probably because of the spread of many
eucalyptus species outside the original home Australia.

Intensive use of species


Many agroforestry trees experience fewer pests and diseases compared with
annual crops, probably because of their less susceptibility to pests, limited
domestication, less intensive culture and existence of a balance between insect
herbivores and their natural enemies. The inherent tolerance of a tree species
to pests may be eroded in the course of its genetic improvement for faster
growth, higher yield and/or better quality, as these characters often are neg-
atively correlated with tolerance to pests. Pest problems may also increase if
a species is grown intensively within a system (vertical intensification) and
extensively over large areas (horizontal intensification), as a result of break-
down of the host–herbivore–natural enemy balance. In a number of annual
crops the combination of breeding for higher yields and intensive cultivation
had devastating effects through new and increased pest outbreaks. For
example, in South and Southeast Asia, insects that were considered of no
importance to rice at one time such as rice brown hopper (Nilaparvata lugens)
and rice hispa (Hispa armigera) became the major pests in recent years.
To what extent agroforestry trees would experience increased and/or new
pests as a result of their domestication and intensive cultivation remains a
hypothesis for testing. However, recent events portend that what had happened
in the case of annual crops can also happen with tree species. First, tree
planting on farms is increasing due to promotion of agroforestry by exten-
sion and development organisations, and farmers’ own realisation that they
no longer can depend for tree products on forests (Holmgren et al., 1994).
Second, there are examples of widely grown trees being affected by insect
pests in recent years. The extensively planted grevillea in the Kenyan high-
248
Table 1. Examples of interactions of exotic and indigenous trees and insects.

1. Origin of insect 2. Origin of tree 3. Long-term consequence of interaction 4. Examples of interaction of tree and insect, having
4. species 4. of insect and tree, having origins 4. origins as specified in columns 1 & 2
4. as specified in columns 1 & 2

Indigenous Indigenous A minor pest may assume epidemic Severe defoliation of Sesbania sesban by Mesoplatys
proportion due to the intensive use of ochroptera beetles due to increased use of Sesbania in
the host species in managed ecosystems planted fallows in east and southern Africa (Steinmuller,
1995; Sileshi et al., 2000).

Indigenous Exotic Increased host range and food availability Celosterna scabrator, a minor pest of Acacia spp. attacked
for a local pest. eucalypts over time in India (Sivaramakrishnan, 1986)

Exotic Exotic A new pest gets established, which not The psyllid Heteropsylla cubana caused havoc to Leucaena
only restricts the potential of an introduced leucocephala in Asia, east and southern Africa (Napompeth
exotic species but also may affect other and MacDicken, 1990; Rao, 1995).
local trees over years. The aphid Cinara cupressi devastated Cupressus lusitanica
in east and southern Africa; ‘neem decline’ and oriental
yellow scale Aonidiella orientalis caused severe damage to
Azadirachta indica in west Africa (Murphy, 1998).

Exotic Indigenous Local tree species susceptible to the new The leucaena psyllid adapted to Samanea saman in
pest may severely be damaged, if natural southeast Asia (Braza, 1987);
enemies are not established. Rastrococcus invadens established on fruit trees and other
plants in west Africa (Agounke et al., 1988).
The cypress aphid Cinara cupressi affected Juniperus
procera in eastern and southern Africa (Murphy, 1998).
249

lands, which was introduced from Australia over 100 years ago, has been
found in recent years to be attacked by a stem borer Apate sp. Even the indige-
nous tree Sesbania sesban, which is widely being tested by farmers as 1- to
2-year planted fallows for replenishing soil fertility in east and southern
Africa, has been observed to be severely damaged by defoliating beetles,
Mesoplatys ochroptera and Exosoma sp. (Sileshi et al., 2000). Similarly, in
recent years in western Kenya, Crotalaria agatiflora and C. grahamiana, used
for planted fallows, have been severely defoliated frequently by Amphicalia
pectolina (H. Girma, pers comm.). Although Prosopis cineraria – the wonder
tree of the arid zone of India – was not known to be attacked by many pests
earlier, it is now experiencing severe damage by insects and diseases in parts
of western Rajasthan, India (Singh, 1998).
Efforts are underway for genetic improvement of a number of priority agro-
forstry tree species through germplasm collection, provenance evaluation and
selective propagation (Simons, 1996). At ICRAF, emphasis is given for
improving and promoting indigenous fruit trees and other high value trees to
increase cash earnings for small-scale farmers in Africa (Leakey and Simons,
1998). Furthermore, improved germplasm of some tree species developed for
monoculture plantations is being tested in traditional agroforestry systems
for increasing productivity of the latter (e.g. introduction of high yielding
rubber clones into jungle rubber agroforests in Indonesia and use of improved
bud-grafted Zizyphus mauritiana in India). There is a need to watch whether
these improved materials would encounter any new pests that would restrict
their potential yields in agoforestry.

At system level

Simultaneous systems

In simultaneous systems, the nature and magnitude of tree–crop interactions


and factors that govern them such as component species and their spatial
arrangement, tree density and pruning, and soil and climatic conditions affect
the herbivore–plant interactions. In systems where trees are dispersed in
cropland, the influence of trees on crop pests is probably limited to the crop
under the tree canopy. In shelterbelts and boundary plantings, the interaction
between trees and crops is mostly confined to the tree/crop interfaces, so trees
in these agroforestry systems are unlikely to influence insect fauna several
meters away from them. In contrast, interactions among species in complex
multi-storey systems may influence the growth of each component and fauna
associated with it.
Figure 1 conceptualises the biophysical changes as a consequence of
tree–crop interactions and in turn their effects on insect pests in simultaneous
agroforestry systems. It also lists the general hypotheses that form the basis
for explaining the behaviour of insects in agroforestry leading to increased
250

Figure 1. Effects of simultaneous agroforestry systems on insect pests, and the mechanisms and
hypotheses that explain the observed effects (modified for agroforestry following Hitimana and
McKinlay, 1998).

or decreased pests. The association of trees and crops creates a new envi-
ronment in agroforestry, which may increase the ability of the system to
withstand pests or render it more susceptible to pests (‘associational hypoth-
esis’). Insect pests will be reduced if the new environment increases natural
enemies by providing congenial habitat (‘natural enemies hypothesis’).
Agroforestry increases pests if trees in the system increase and extend food
availability to insects by serving as alternate hosts (‘resource concentration
hypothesis’). Interactions among component species of a system may affect
the physiological status of one or more species, making them either vulner-
able to or robust against insect attack (‘host plant physiology hypothesis’).
Non-host trees in a system may reduce insect populations by diluting the
proportion of the host crop (‘resource concentration hypotheses’), restricting
the movement of insects, interfering with the host-seeking ability of the insect
251

and lowering quality of the host (disruptive crop hypothesis). Other mecha-
nisms that affect pests include synchrony between insect prey and natural
enemies, chemical repellence and some components inhibiting food intake of
insects (Liping, 1991; Altieri, 1994). The different mechanisms are further
discussed in the following sections.

Tree–soil–crop interactions: effects on pests


The tree–crop interactions for growth resources (water, nutrients and light)
in simultaneous systems may have a direct effect on insect activity, particu-
larly if sharing of limited resources by the component species results in
weakening one of them. Both N2-fixing and non-fixing trees compete with
crops for nutrients, and poor quality tree litter in the course of decomposi-
tion immobilises nutrients in the soil (Anthofer et al., 1998). A number of tree
species used in agroforestry (e.g. Acacia polyacantha, Acacia nilotica, Ficus
neriifolia, Juglans regia and Eucalyptus spp.) is known to inhibit crop growth
underneath their canopies due to allelopathic effects of root exudates and/or
litter decomposition products (Anthofer et al., 1998; Bhatt et al., 1997; Suresh
and Rai, 1988). In such agroforestry situations where trees reduce the growth
and vigour of crops, the latter may experience greater pest problems than in
sole systems. Not all the tree–crop interactions lead to increased pests. There
can be fewer pest problems if one species (usually the tree component except
at the seedling stage) shelters the other from unfavourable environmental con-
ditions and improves the growth by efficient nutrient cycling. Schroth et al.
(2000) gave a detailed account of the effects of mineral nutrition on pests
and diseases.
In water-limiting environments, competition for water dominates tree–crop
interactions, which may affect the growth and susceptibility of plants to
insects. Trees and crops experiencing drought are more prone to termite attack.
In semi-arid environments, greater termite attack due to water stress is one
of the major causes for high mortality of tree seedlings planted in croplands.
Similarly, water stress was found to increase the susceptibility of cypress and
palms to bark invasion by Pestalotiopsis funerea (Tuset and Hinarejos, 1985).
Similar infestation levels of the weevil Myllocerus spp. were found to cause
greater injury to Lasiurus sindicus grass in rainfed fields than in irrigated
fields at Jodhpur, India (Singh et al., 1998). Eucalyptus trees affected by
dieback and water stress experienced greater foliar damage by insects than
normal trees (Landsberg and Wylie, 1983). The impact of water stress could
be managed to some extent by thinning the crop stand (Singh and Parihar,
1997) and pruning of trees. Provision of alternative food material to termites
through mulch was suggested as a strategy to divert their attack, but it works
only temporarily and may lead to greater injury after the bait is consumed.

Microclimate
The microclimatic changes caused by trees in mixed systems include shading
of the understorey crop(s), increased humidity, reduced air and soil tempera-
252

ture, and decreased wind speed. These changes may bring about positive or
negative impact on the activity of insects and their natural enemies (para-
sites, predators, and entomopathogenic microorganisms).
The activity of shade loving insects is usually increased under trees. Some
dung beetles were reported to prefer staying under shade (Doube, 1983).
Lower temperatures and higher humidity under trees in warmer regions reduce
desiccation of soft-bodied larvae and sap sucking insects (e.g. aphids) and
increase their activity. In the Sudan zone of Burkina Faso, populations of
Cicadulina sp. – the vector of maize streak virus – were observed to con-
centrate under trees and play a significant role in the initial outbreak of maize
streak (Traoré and Quedraogo, 1997). The coffee berry borer (Hypothenemus
hampei) was found to be favoured by a dense shade (Beer et al., 1998). The
activity of shade-loving insects can be reduced by pruning the branches of
upper-storey trees and by selecting open-canopy species or provenances in
preference to dense-canopy ones. Shade may have a different effect on certain
other insects. For example, density of foliage feeding beetles was lower on
bean intercropped with maize than sole-cropped bean (Risch, 1981). Zhou et
al. (1985) reported greater infestation of Polyura naraea on stands at the edges
of forest canopy facing the sun than on the shaded side. Shade of trees is
known to interfere with host seeking and reproductive behaviour of some
insects (Risch, 1981; Yang et al., 1988).
Many hymenopteran parasites exhibit greater host-searching capacity under
bright light, but moderate shade favours the parasitic wasp (Cephalonomia
stephanoderis) of coffee berry borer (Beer et al., 1998). However, lower
temperatures combined with higher humidity under trees increase parasitism
by the egg parasite Trichogamma sp. (Pu, 1978) and the effectiveness of
entomopathogenic fungi (Jacques, 1983). For an example, the half life and
viability of spores of the fungus Nomuraea rileyi were much longer in the
shade than in the open (Fargues et al., 1988). Similarly, shade increases the
persistence of entomopathogenic fungus Beauveria bassiana on coffee berry
borer (Beer et al., 1998). The length of infective period of Bacillus
thuringiensis was greatly reduced under direct sunlight.

Host range of insects


Plant species belonging to the same or close taxonomic groups in any agro-
forestry system are likely to share common pests because of similar mor-
phological features and/or presence of the same or closely allied biochemicals.
The oligophagous insects feeding on plants of certain biochemical makeup
may adapt easily to closely related species having similar biochemical con-
stitution. For example, the damage caused by the pod borer Etiella zinckelia
in peas was accentuated by the presence of acacia trees in the vicinity (Szeoke
and Takacs, 1984). Acacias also being legumes were suspected to increase
populations of this pest, especially between seasons when the main host pea
was absent in the field. In the genus Heteropsylla, there are psyllids that feed
on a number of leguminous agroforestry trees such as Acacia angustissima,
253

A. villosa, A. farnesiana, A. glomerosa, Albizia adinocephala, Albizia saman,


Calliandra houstoniana, and Prosopis juliflora (Muddiman and Hodkinson,
1992). The same phenomena may also support the populations of natural
enemies of insect pests.
However, many polyphagous insects do feed on taxonomically diverse plant
species. For example, whiteflies, many of which are vectors of diseases, thrive
well on diverse plant types with no apparent taxonomic alliance. In Assam,
India, eucalyptus grown for shade in tea shares the attack of Chrysolampra
flavipes (Gope, 1985). The cassava mite Tetranychus sp. successfully feeds
on banana, and the thrip Retithrips syriacus thrives well on leucaena in
addition to being a pest of cassava (Ghosh et al., 1986). In Hawaii, the thrip
Frankliniella occidentalis, a vector of spotted wilt virus on lettuce, tomato,
and cabbage, was found to have a wide host range, including leucaena (Yudin
et al., 1986). In Brazil, the bagworm Oiketicus kirbyi, a pest of coffee, was
observed to survive on eucalyptus (Arce et al., 1987). In Jaisalmar district
(Rajasthan), India, the polyphagous weevil Myllocerus laetivirens and related
species were found causing severe defoliation on Zizyphus mauritiana by
surviving on nearby forage grasses (MP Singh, pers. obs.).
Insects with a wider host range rapidly multiply and cause greater damage
if many of their hosts are present in a system or nearby in the landscape
(Speight, 1983). A survey of insects on three agroforestry trees (Gliricidia
sepium, L. leucocephala and Senna siamea), crops (bean, maize and
pigeonpea) and wild trees at the ICRAF’s Machakos station in Kenya revealed
that 27 insect species were found specifically on the trees, three on the crops
and 12 on wild trees. Of these, 13 insect species were common to the trees
and the crops, 14 to the trees and wild trees, 19 to the trees and the crops,
and four to wild trees and the crops (Opondo-Mbai, 1995). Mixed systems
with trees and crops having such a high degree of common insects between
them would risk greater insect attack. Some insects thrive on different host
plants in different stages of their life cycle. In arid to semi-arid north-western
India, while the grubs of Scarabaeid beetles feed on seasonal crops, the adults
feed on a number of species including Prosopis cineraria, Azadirachta indica,
and Zizyphus spp. (Singh et al., 1989). These results point out the impor-
tance of knowing the pest spectra of species in designing agroforestry systems.

Masking host plant odours


Many insects are attracted to their hosts through perception of odours emitted
by them. A possible consequence of species interactions in agroforestry is that
odours released by some plants may mask the effect of odours released by
others, making it difficult for insects to locate their host plants. Thus odorif-
erous species in combination with normal hosts of insects can hinder the ability
for host recognition, feeding and reproduction of insects (Schoonhoven, 1968).
The masking effect of odours may also influence the prey-seeking behaviour
of parasitic and predatory insects that depend on the odours of host plants to
identify their prey insects (Monteith, 1960; Shahjahan and Streams, 1973).
254

There could also be cues from companion plants that facilitate the functioning
of natural enemies in one way or the other. Identifying the positive and
negative effects of cues of specific plants on pest–natural enemy combina-
tions will help exploit this phenomenon for pest management, but little is
known at present in agroforestry.

Barrier effect
Hedgerows, boundary tree plantings and windbreaks act as physical barriers
to the dispersal and colonisation of both herbivorous and, predatory and
parasitic arthropods within a field and between fields. The permeability, height
and orientation of tree-rows in relation to wind affect the inward and outward
movement of insects (Bach, 1984; Pasek, 1988). Trees may simply act as
mechanical barriers to the dispersal of insects, and ’nurse trees’ protect other
crops from pest attacks (Kennedy et al., 1959) or have a biological role in
repelling insect pests because of unfavourable morphological features such
as hairy leaves (Levin, 1973). Mahogany (Swietenia macrophylla) planted
between rows of Inga edulis and Schizolobium amazonicum suffered from less
and delayed attacks of lepidopteran shoot borer Hypsipyla grandella
(Ackerman et al., 1998). The dense canopy of Inga shields the mahogany
seedlings, making their location difficult for the adult insects to lay eggs. Trees
being the tall component in agroforestry provide a platform for the weak flying
insects carried by wind to settle for a while before initiating infestation of
the crops around. The infestation of maize stalk borers (Chilo partellus, C.
orichalcociliellus and Sesamia calamiostis) was significantly lower and crop
yields greater in alleycropping between leucaena hedgerows than in sole
cropping. The hedges were suspected to act as mechanical barriers to the
dispersal of the young larvae in search of the host (Ogol et al., 1999).
Tree hedges and boundary plantings serve as reservoirs to both insect pests
and beneficial insects (van Emden, 1965; Solomon, 1981), the most abundant
of the latter category being parasitic-Hymenoptera and Diptera, and predatory
spiders and ladybird beetles. Although tree barriers contribute to maintaining
some populations of insects, most increases in the density of insects are due
to those blown in from elsewhere. The Braconids tend to accumulate on the
leeward side, while the Aphidids, Vespids, large Diptera and Lepidoptera
predominate on the windward side (Pasek, 1988). Systems in which trees block
the entry of injurious insects and hinder the outward movement of natural
enemies of insects will have fewer pest problems.

Spatial arrangement of species


The configuration of a system as determined by species composition, their
canopy colour, structure, height, plant density and spatial arrangement, and
the exposed soil background affect the ability of some insects in recognising
their hosts. Such insects are easily confused by changes in the system con-
figuration. Dempster and Coaker (1974) observed that even small changes in
cropping practices greatly alter the attractiveness of plots to pests and their
255

natural enemies. Increased heterogeneity through spatial arrangement of dif-


ferent species, particularly orientation of trees with respect to wind direction
and movement of sun in a field can restrict the number of invading insects.
Upperstorey trees may camouflage the understorey host crops and prevent
pests from recognising them from a distance (Altieri and Liebman, 1986). The
number of eggs laid by the pyralid moth Cactoblastis cactorum on Opuntia
ficus-indica was affected by plant size, cladode condition, shelter from wind
during oviposition, and height above ground (Robertson, 1987). One of the
reasons for lower herbivore populations in diverse plant communities is
emigration of invading insects in the presence of non-host plants. Studies on
various leaf beetles indicated that they had a shorter residence time in plant
stands with non-hosts (Bach, 1980a, 1980b; Risch, 1980, 1981). Thus the ratio
of host to non-host plants in the system plays an important role in determining
the herbivore abundance.

Sequential systems

Natural bush fallows (five to 20 years), short-duration planted tree fallows


(one to 2 years) or woodlots (three to 10 years) rotated with crops are the most
common sequential systems. The purpose of natural and planted tree fallows
is for replenishing soil fertility, but woodlots are meant for poles, pulpwood,
timber or firewood. One of the reasons for failure of continuous cropping in
humid and subhumid tropics is the build up of pests, diseases and weeds over
seasons. A break in cropping with natural or planted tree fallows using non-
host species adversely affects pests because of their migration or destruction
in the absence of food. The effectiveness of tree fallows in reducing pests
depends on the nature of insects, their feeding habits, biology and ecology,
fallow length and availability of alternative hosts such as grasses and shrubs
in the landscape. To cite an example, intensification of fallow increased
grasshopper populations, while cultivation and afforestation severely reduced
them (Amatobi et al., 1988).
If trees used for planted fallows serve as alternative hosts to insects or
pathogens of crops, then such fallows may increase pests on the subsequent
crops. As Sesbania sesban and Tephrosia vogelii are good hosts of root-knot
nematodes (Meloidogyne spp.), 1- or 2-year fallows of these species increased
the nematode populations and severely reduced yields of the nematode-
susceptible bean in the subsequent cropping season (Desaeger and Rao, 2000;
Desaeger and Rao, unpublished data). The root-knot nematode problem was
particularly severe in light-textured soils. Obviously, Sesbania and Tephrosia
fallows cannot be considered for replenishing soil fertility where any of the
root-not nematode-susceptible crops such as bean, tobacco and cotton is an
integral part of the cropping system. We observed 1.5- to 2-year Sesbania
fallows to increase chaffer grubs, which cut maize seedlings after the fallows
were cleared. However, this problem was not severe and noted only at
Machakos in Kenya. A general principle is to avoid the use of tree species
256

that host polyphagous insects, particularly pests of crops, for planted fallow
technology. Tolerance to pests is not enough in judging tree species for planted
fallows, as even the tolerant trees can perpetuate pests to cause significant
economic losses in subsequent crops.
In the humid tropics, long-term fallows are generally cleared for cropping
by ‘slash and burn’, which destroys most insects and pathogens and permits
growing of crops free from pests for sometime. In the subhumid and semi-
arid tropics where shorter-term planted fallows (one to two years) are gaining
importance, there is no need for burning of fallow biomass because the amount
of biomass harvested is small and wood is a valuable product either as
firewood, poles or stakes for support. Foliage can be managed by either
incorporating it into the soil or spreading on the soil surface (‘slash and
mulch’). The slash and mulch or no-till practices may extend the residual
effects of tree fallows by slowing down carbon mineralisation and preserving
soil structure. Whether this practice will have any negative effects through
pests and diseases needs evaluation before it is recommended for adoption.

Biomass transfer systems

Foliar biomass of trees and shrubs used as green manure or mulch increases
the activity of soil fauna. Apart from suppressing weeds, mulches restrict the
spread of certain soil-borne plant diseases by reducing the splash of soil
inoculum, decreasing soil temperature, increasing soil water, and enhancing
microbial activity. Mulching is known to effectively control the soil-borne
web blight of bean caused by Rhizoctonia solani and its telemorph
Thanetephorus cucumeris (Abawi and Thurston, 1994).
Mulches may have the disadvantage of increasing certain polyphagous
insect pests and soil-borne diseases of crops. The severity of root diseases
caused by Pythium spp., Phytophthora spp. and Sclerotium rolfsii in humid
environments was attributed to specific mulches and management of organic
matter (Abawi and Thurston, 1994). In drier-environments mulching may
increase the activity of litter-feeding insects such as termites and grubs to
become pests on crops. Mulches using foliage of different trees and shrubs
were reported to increase termite activity (Budelman, 1988), although their
effects differed with the quality of the material used. Fresh materials that
decompose rapidly may escape termites and materials that contain anti-feeding
substances such as neem foliage may deter termite activity. In contrast, dried
materials that decompose slowly increase termite activity. Mulches also
provide good niche for rats to multiply. Many farmers in North Lampung,
Indonesia complained about increased rat damage by mulching of field crops,
and they preferred clean cultivation within and outside their upland crops such
as rice and cassava (Gauthier, 1996).
257

At landscape level

In agricultural landscapes, trees and shrubs occur on farms in different spatial


and temporal arrangements with crops and outside farms in communal lands.
The mosaic of systems at any time in the landscape allows frequent transmi-
gration of insect pests and their natural enemies between trees, shrubs, crops
and weeds within and across fields. The vegetation outside the cultivated fields
plays an important role in maintaining populations of herbivores and natural
enemies in agricultural systems. The effects of woody vegetation in the land-
scape on crop pests are least studied. Trees and shrubs being perennials, they
serve as refuge to pests and natural enemies by providing alternative food
sources throughout the year. The herbivore–tree interactions in the landscape
can result in secondary pest infestations, which are little understood.
The crop pests having relatively longer life span and wider host range can
survive on trees in the absence of main food source during the dry periods,
before they re-invade crops in the subsequent rainy season. The larger grain
borer Prostephanus truncatus (Horn) – a stored pest of maize and cassava –
was reported to survive and breed on a number of trees in the natural
vegetation and croplands such as Acacia mellifera, A. polyacantha, Leucaena
diversifolia, L. shanonii, Senna siamea, and Commiphora campestris (Nang’yo
et al., 1993). This implies that planting any of these trees near maize and
cassava storage structures or intercropping maize and cassava between these
trees may aggravate the attack of larger grain borer. The insects that are highly
mobile and complete many generations round the year also move from one
host to another in the landscape. Grasshoppers adapt to cereal crops in the
cropping season and move to grasses after the crops are harvested. Jassids and
leafhoppers also make such cross over movements.
Many trees and shrubs in the landscape can support generalist predators
such as ladybird beetles (coccinellids), syrphid flies (dipteran insects), spiders,
wasps or specific predators and parasites beneficial for biological control of
crop pests. Closer the location of woody vegetation to cropped fields, quicker
would be the migration of these insects to crops. If sufficient populations of
natural enemies are built up on woody vegetation in the landscape before
sowing crops, pest infestations later on crops can be rapidly checked. This is
particularly important in the case of insects that proliferate fast like aphids,
whiteflies, scale insects and mealybugs.
Trees in cropped fields have been thought to increase damage to grain crops
by birds and rats by providing habitats and alternative food sources, leading
to arguments against agroforestry. The damage by these vertebrate pests may
not be confined to the plots containing trees because of their high mobility.
Bird damage is common to isolated and non-seasonal crops that mature early
or late in the season. However, there is no strong evidence to indicate that
trees increase bird damage to crops by harbouring them, especially if crops
in an area mature around the same time. In a study of 50 maize fields 2–7
km from a large blackbird (Agelaius phoenicus) roost, the bird damage was
258

mostly related to the date of silking, greater damage being in earlier maturing
fields. The woods and hedges bordering the cropped fields were weakly cor-
related (Bollinger and Caslick, 1985). On a positive side, birds resting on trees
can predate on insect larvae on crops. Rat damage was observed in experi-
mental plots mostly adjacent to hedgerows immediately after sowing crops
and again close to harvest (Lal, 1989) and at field scale close to tree-rows on
field boundaries (authors’ observations). Periodical cultivation of land close
to hedgerows and tree-rows, and clean weeding within the rows will reduce
rat populations by eliminating weeds, which form the food source for rats
during the dry periods.
Tsetse fly (Glossina spp.), which transmits trypanosomiasis in cattle in sub-
Saharan Africa, is known to be a problem in areas dominated by large-scale
woody vegetation. In the sixties and seveties, trees were deliberately removed
to control the disease in agricultural areas and human settlements, which is
not a desirable practice from both environmental and economic considera-
tions. Recent emphasis on afforestation and agroforestry has raised doubts
about the reappearance of tsetse fly. However, it is unlikely that trees on farm
in continuously cropped areas would increase tsetse fly (Otsyina, 1993). The
sensible approach to control trypanosomiasis is by using recently developed
more effective and cheaper integrated strategies rather than by deforestation
(Jordan, 1986).

Insect pest management in agroforestry

Many trees used by resource-poor farmers in agroforestry are of relatively low


value, so pest management methods must of necessity be low cost to be worth
adopted. However, a number of trees that are grown in such systems (including
eucalyptus, conifers and legumes such as Acacia spp.) are also grown on a
larger scale, sometimes in plantations. In these latter systems more expen-
sive interventions are cost effective, so there are many examples of pest
management methods being used in large-scale plantations of tree species
which are also used in low input agroforestry systems, and we therefore,
include some of those examples here.
Sustainable agroforestry systems should perform with fewer pest problems
and minimum external inputs for protection against pests. Habitat manage-
ment offers a low cost technique for small-scale farmers to minimise by
creating conditions that hinder the pest populations but favour the build up
of natural enemies. Table 2 summarises the broad range of effects of man-
agement factors on insect pests in both simultaneous and sequential systems.
While these general effects of management may hold true in many situations,
exceptions do occur with some systems and insects, which need to be
determined. Removing the source of pest inoculum by producing healthy
planting material and pruning the parts of trees infested by insects such as
scales and mealy bugs, clean weeding at critical stages, and adequate mineral
nutrition are some of the tree-related management practices that help reduce
Table 2. Management variables and their possible range of effects on pest infestations in simultaneous and sequential agroforestry systems.

Management variable Condition(s) of management variable favouring

High pest incidence Low pest incidence

Simultaneous systems
Species diversity Low diversity with closely related species High diversity with widely differing species
Species number Fewer Many
Proportion of host to non-host species High Low
Spatial arrangement of trees and crops Wide with limited interaction among species Close with strong interaction among species
Row orientation Along wind Across wind
Tree density Low density High density
Sequential systems
Nature of tree fallows Monospecies fallows Multispecies fallows
Planting and harvesting of trees and/or
crops in the landscape Asynchronous Synchronous
Residue management Residues used as mulch Residues burnt
Diversity during cropping phase Limited diversity (e.g. monocropping) High diversity (e.g. crop rotation and mixed cropping)
Common to both systems
Field size Large fields Small fields
Distribution of fields practising the
same agroforestry system Aggregated Scattered

259
260

pests. Clean weeding of Sesbania fallows early in the season reduced


Mesoplatys beetle populations, as the larvae succumbed to death in the course
of migrating to weedy vegetation or litter cover for pupation (Sileshi et al.,
2000). Regularly pruned trees can be more vulnerable to insect attack as they
are weakened over time and they also produce young coppice shoots that are
more palatable to insects. Frequent pruning increased the stem borer (Xyleutes
capensis) infestation of Senna siamea (Opondo-Mbai, 1995).
Integrated pest management in agroforestry should take into account some
or all of the following strategies: (1) substituting susceptible tree species with
alternative species that provide similar products and services, (2) strength-
ening species tolerance or resistance to insects, (3) creating conditions that
reduce populations of insect pests but favour natural enemies, (4) strategically
using environmentally safe chemicals and botanical pesticides, and (5)
exploiting biological control through introduction of effective natural enemies.

Substitution of tree species


In some instances, a susceptible tree species can be discarded for a quite
different species that nevertheless performs the same function. A number of
tree species can be considered in any ecoregion for some tree products such
as fodder, poles and firewood, and services such as nutrient cycling, sheltering
and soil conservation. Thus opportunities exist to substitute one unsuitable
tree species because of pests and diseases with another free from maladies.
This strategy probably works only if the alternative tree species gives similar
yields and services without sacrificing quality. In any case, few studies were
conducted in agroforestry to select appropriate tree species taking pest
problems and management opportunities into consideration.
Grout and Stephen (1995) reported that Eucalyptus torelliana was better
suited than others for windbreak of citrus in South Africa, as it was not an
alternative host of citrus thrip (Scirtothrips auranti) and it increased popula-
tions of the predatory mite Euseius tutsi (Amblyseius tutsi). In Papua New
Guinea, leucaena was used as a shade tree for cocoa for a long time. As there
was frequent transfer of leucaena pests to cocoa, it was gradually replaced
with Gliricidia sepium (Laup, 1994). In Costa Rica, Erythrina poeppigiana
was considered inappropriate as a shade tree for coffee as it was found sus-
ceptible to Planococcus citri, a pest shared with coffee (Kettler, 1995). There
are good prospects for replacing the highly psyllid-susceptible L. leucocephala
provenances with psyllid-tolerant or resistant accessions of L. diversifolia and
L. pallida in fodder production systems (Table 3; Austin et al., 1995). In the
central highlands of Kenya, with the advent of the psyllid problem, many
farmers have taken up Calliandra calothyrsus as an alternative fodder to
Leucaena leucocephala (A. Luvanda, Kenya Forestry Research Institute pers.
comm.). In Reunion, the decline of Casuarina equisetifolia, caused in part
by the insect Coelosterna scabrator, is being replaced by various species of
Eucalyptus, Acacia and others (Tassin et al., 1998).
Crotalaria spp. were found to be good alternatives to S. sesban and
261

Table 3. Resistance to psyllid and fodder productivity of some Leucaena species, accessions
and hybrids.

Category 1: Moderate to high resistance and high productivity


L. diversifolia CPI46568, K785, K784
L. pallida K376, K819, K806 x K748, OFI 137/94, OFI 79/92, CQ343
L. leucocephala x L. pallida (KX 2 series) (K636 x K748; K636 x K806; K8 x K376; K584 x
K748 UQ45)
Category 2: Moderate to high resistance but low to moderate productivity
L. lanceolata K952,
L. esculenta K950
L. pallida K953, K748, K806, K156
L. pallida x L. diversifolia (K806 x K156)
L. leucocephala x L. diversifolia (K636 x K156)
L. trichandra OFI 53/88
L. shannonii ssp. Shannonii
L. salvodorensis
L. macrophylla
Category 3: Susceptible but fairly good productivity
L. leucocephala K636, K584, K585
Category 4: Susceptible and low productivity
L. leucocephala K997, K500, Q2521, CPI61227
L. multicapitulata
L. lempirana, L. involucrata, L. pulverulenta
L. trichodes OFI 61/88, OFI 4/91

Source: Glover (1988); Castillo and Shelton (1994) in Wheeler (1988); Rao and Mathuva (1997);
Hughes (1998); Mullen et al. (1998); Wandera and Njarui (1998).

T. vogelii, where the latter cannot be employed as short-duration fallows for


replenishing soil fertility because of their detrimental effect on Meloidogyne-
susceptible crops in the rotation (Desaeger and Rao, 1999, 2000). Planted
fallows of C. agatiflora and C. grahamiana increased the yields of subsequent
maize similar to or better than Sesbania and Tephrosia fallows, and farmers
preferred Crotalaria spp. as they were easy to establish (A. Niang, pers.
comm.). In Indonesia, Wiryadiputra and Priyona (1995) advocated not to grow
Barangan cultivar of banana in coffee for shade in the areas infected by the
nematode Pratylenchus coffeae. In India, Erythrina was discouraged as a
shade tree for cardamom as it increased the root-knot nematode Meloidogyne
incognita (Muniappan, 1993).
The scope for substitution of tree species will be limited where the trees
are considered for specific products such as fruits, nuts, gums, and medicines.
So, alternative strategies have to be employed to combat pests of tree species
that produce such high value products.

Chemical control
Chemical control becomes inevitable and justifiable where all other
approaches fail, if the returns are economical and the chemicals used are
262

environmentally safe and appropriate for the targeted pest. One of the most
frequent uses of chemicals is in the control of termites on eucalyptus. Selander
et al. (1989) reviewed five studies on chemical control of termites on
Eucalyptus in Zambia, and concluded that carbosulfan could be recommended
as an alternative to organophosphates. Chey (1996) found that a single root
drench with chlorpyrifos or chlordane to young Gmelina arborea prevented
all further attack for four years, while aldrin and HCH were almost as
effective. However, Patel and Sahu (1995) found gammaxene ineffective in
protecting E. tereticornis from attack. Chilima (1991) found that controlled
release granular formulations of carbofuran, chlorpyrifos and carbosulfan
significantly reduced termite damage to eucalyptus in Malawi, and recom-
mended them as alternatives to aldrin. In Sri Lanka however, aldrin effec-
tively protected Eucalyptus seedlings from termite attack (Midgley and
Weerawardane, 1986). Mushongahande (1996) recommended fipronil for
termite control in Eucalyptus in Zimbabwe.
In Eucalyptus plantations in Brazil, Thyrinteina (Lepidoptera) is sprayed
with diflubenzuron (Santos et al., 1990). The same chemical is seen as a
substitute for the long used pyrethroids, to control bagworm (Chaliopsis
junodi) in Acacia plantations in South Africa, applied from the air (Atkinson,
1998).
An outbreak of lac insect, Kerria lacca on Santalum album and its host
trees Pongamia pinnata and Casuarina equisetifolia in Karnataka state (India)
was controlled with quinalphos (Remadevi et al., 1997). In Indonesia, melolon-
thids on Paraserianthes falcataria were controlled with carbofuran and
monocrotophos (Intari, 1995).
Chemical control of the leucaena psyllid Heteropsylla cubana typifies the
situation with the use of pesticides in agroforestry. While a number of
chemicals are effective, rarely will they be economically justified or even
affordable by resource poor farmers (Rao, 1995). Another problem with
chemical control is that it also kills natural enemies, disrupting the ecolog-
ical balance that usually prevents insects reaching pest status. Ways round this
problem can sometimes be found; for example Sudarmadji (1988) found that
painting Leucaena trees with systemic insecticide helped control the psyllid
without affecting the introduced coccinellid predator Curinus coeruleus.
Many trees and shrubs are known to contain toxic chemicals of pesticidal
value in their plant parts and a substantial body of literature exists on their
use for pest control (Grainge and Ahmed, 1988). Botanical pesticides are
particularly relevant for poor farmers in the tropics because of their low cost,
and safety to the users, livestock and environment, but their use is often limited
by the unavailability of adequate quantity of plant material. However, trees
such as neem, Melia azedarach, Melia volkensii, Tephrosia vogelii, Euphorbia
tirucalli and Lantana camara can be exploited for their agroforestry and pest
control values.
The trend towards encouraging farmers to plant high value trees that will
provide income, (as well as food in the case of fruit trees), means that chemical
263

control may become more affordable to resource poor farmers in the future.
This could be considered an indication of successful development, but it has
been suggested that the potential pest problems on these trees should be
investigated now, so that farmers have alternatives to chemicals when
problems arise (Day and Murphy, 1998)

Host plant resistance


Host plant resistance (HPR) as a pest management tactic has particular appeal
in low input systems, as once the correct planting material is used, little or
no further intervention is required by the grower. This is also the most
appropriate strategy for agroforestry trees that produce low value products
(e.g. fodder, firewood) and used for service functions (e.g. soil conservation,
improving soil fertility). The first step in developing this tactic for a partic-
ular pest is to determine whether any variability in susceptibility exists in the
host tree. However, while many authors report variability in susceptibility to
pests between provenances of a species, there is much less work reported on
the uptake of these findings. Sasidharan et al. (1997), for example, examined
susceptibility of Albizia lebbek to psyllids (Psylla and Acizzia) and aphids
(Aphis) in a provenance trial, and found none completely resistant, but some
more resistant than others. Similarly, Sharma et al. (1998) screened 30
varieties of Zizyphus mauritiana for resistance to the fruitfly Carpomyia
vesuviana, and found none immune to the pest but a wide range of suscepti-
bility, some lines with < 10% fruits damaged, and others with > 50%. Large
variation was also observed among Zizyphus cultivars for their reaction to
the bark-eating caterpillar Indarbela sp. (Singh, 1984). Cornelius et al. (1996)
reported a Guatemalan provenance of Alnus acuminata to be heavily attacked
by bark beetles (Scolytodes alni), while four Costa Rican provenances were
not attacked. Similarly, Pinus caribaea var. bahamensis is more or less totally
resistant to pine shoot moths (Rhyacionia spp. and Dioryctria spp.) while P.
caribaea var. hondurensis is devastated in Southeast Asia (Baylis and Barnes,
1989).
In some cases different provenances may show tolerance rather than
resistance by antibiosis, allowing good growth despite the presence of pests.
Rorabaugh (1995) reported a superior Prosopis glandulosa accession, which
outperformed other accessions despite attack by psyllids.
For identifying pest resistant material in any tree species, germplasm
screening should be done not only within that species but also across species
in the same genus. This approach was used for H. cubana through range-
wide collection and screening of leucaena germplasm, which lead to identi-
fying many psyllid-resistant or tolerant L. leucocephala provenances and
accessions with good yield potential (Table 3). Many other Leucaena spp.
(L. pallida, L. diversifolia, L. esculenta, L. macrophylla and L. trichandra)
and certain inter-species crosses (e.g. crosses of L. leucocephala K636 x L.
pallida K 748; K 636 x K376) were also found to be resistant to the psyllid.
Some of these materials possessed good agronomic characteristics and can
264

be used directly as alternatives to L. leucocephala in fodder production


systems. Resistance to psyllid is partly associated with the concentration of
condensed tannins in the foliage.
Much work has been undertaken on resistance in eucalyptus to various
pests, including Phoracantha borers (Soria and Borralho, 1997), autumn gum
moth Mnesampela privata, leaf blister sawfly Phylacteophaga froggatti,
psyllids Cardiaspina spp. (Floyd and Goldie, 1997), and termites (e.g.
Meshram et al., 1998; Atkinson et al., 1992).
Again considerable effort has gone into determining the relative resistance
of different species of Eucalyptus to pests, particularly termites. Atkinson et
al. (1992) compared 32 species for resistance to Macrotermes in South Africa
and found E. dunnii the most tolerant, while a number of the most suscep-
tible species was from the subgenus Monocalyptus. Results in South Africa
were very different from those in the literature from Australia. In Zimbabwe,
Mitchell and Boland (1989) compared 41 species of trees from Australia,
Central America and Southern Africa, including a number of eucalypts, for
susceptibility to fungus growing ants Ancistrotermes and Macrotermes. The
standard species E. camaldulensis had 66% mortality due to termites, while
a range of Acacia spp. topped by A. holosericea, and Enterolobium cyclo-
carpum and Senna atomaria had less than 10% deaths caused by termites.
However, in India, E. camaldulensis was the most resistant of seven species
assessed for resistance to Odontotermes (Meshram et al., 1998).
The other genus in which species have been compared for pest resistance
is Sesbania. There seems to be some scope for selecting tolerant material to
Mesoplatys ochroptera beetle among Sesbania spp. and provenances of S.
sesban (Maghembe, J. pers. comm.; Steinmuller, 1995). The tolerance was
related to the deterrence for oviposition and reduced larval growth due to
low digestibility of leaves containing condensed tannins. Some of the
materials suffered less damage because of their ability to quickly recover and
produce compensatory growth after the pest disappeared. In India,
Roychoudhury et al. (1997) assessed the susceptibility of four Sesbania species
to Spodoptera litura, and found S. formosa to be least damaged. Bhalla et al.
(1988) compared six species for resistance to the seed chalcid Bruchophagus
mellipes, and found the species to experience 0% to over 90% infestation.
Screening of a number of Sesbania species and provenances within S. sesban
revealed considerable variability in tolerance to root-knot nematodes, although
none was resistant for use as planted fallows without the risk of affecting
nematode-susceptible crops in rotation (Desaeger and Rao, 1999; J. Bridge,
pers. comm.).
A general conclusion from the literature is that phenotypic assessment of
resistance either within or between species should be repeated spatially and
temporally to give reliable results. Ideally, heritability of the observed
resistance should be determined, though there are relatively few studies of
this nature. Soria and Borralho (1997) for example, found low heritability of
Phoracantha resistance in Eucalyptus (h2 < 0.2), while Kamunya et al. (1997,
265

1999) found higher heritability of Cinara cupressivora resistance in Cupressus


lusitanica (h2 = 0.45–0.76). Breeding for resistant cultivars may be justified
in the case of high value indigenous fruit and timber tree species.

Biological control
The classical biological control, or the introduction of a natural enemy from
the pest’s area of origin, has been implemented for some agroforestry insect
pests, leucaena psyllid probably being the commonest target. Day et al. (1995)
listed 17 countries where the coccinellid Curinus coeruleus had either been
introduced or arrived, seven countries for the coccinellid Olla v-nigrum, and
seven for the encyrtid parasitoid Psyllaephagus yaseeni. In recent years P.
yaseeni has also been introduced to Kenya, Tanzania and Zambia, and the
eulophid Tamarixia leucaenae to Tanzania, Kenya, Zambia and Malawi.
Indications are that the parasitoids are having only a minor effect in Africa
(Day, 1999) and there is no strong evidence of any of the agents having a
major effect in Asia (Day et al., 1995), although qualitative reports claim some
impact (Napompeth, 1994).
More success was achieved in controlling the eucalyptus snout beetle
Gonipterus scutellatus, which was controlled in the 1940s in Africa with the
egg parasite Patasson nitens (Greathead, 1971). Egg parasitoids have also
been released in South Africa and Zambia (O. Shakacite, pers. comm.) for
control of Phoracantha spp., again on eucalyptus.
Psyllaephagus pilosus was introduced to USA for control of the psyllid
C. tenarytaina eucalypti, with a benefit: cost ratio estimated to be in excess
of 9:1 considering only reduced chemical control costs (Dahlsten et al.,
1998), though this is a very different situation from resource poor farmers
in the tropics. Muniappan (1993) reports control of two Erythrina pests
using introduced natural enemies, Icerya purchasi with the parasitoid
Cryptochetum iceryae in Israel, and I. aegyptiaca with the coccinellid Rodolia
pumila in the Pacific, though Erythrina was not the main host plant of the
pests.
A number of parasites (e.g. Perilitus larvicida), including an ento-
mophagous nematode (Hexamermis sp.), and predators (Glypsus conspicuus,
Macrorhaphis acuta and Mecosoma mensor) has been found to attack
Mesoplatys on S. sesban in Zambia and Ethiopia (Sileshi GW, pers. comm.;
Steinmuller, 1995), but their relative importance to check the beetle popula-
tions and ecological conditions favouring their multiplication are not yet
known.
Biological control of the cypress aphid Cinara cupressivora has been
implemented in Malawi, Kenya and Uganda in recent years, using the braconid
parasitoid Pauesia juniperorum. The agent is now established in all the three
countries (C. Z. Chilima, K. E. Mutitu, P. Kiwuso, pers. comm.), but it is not
yet clear what impact it will have.
As with the other examples, biological control of the cypress aphid was
implemented primarily because the pest was causing economic damage to
266

large-scale plantings, rather than because the tree is popular amongst rural
farmers. This pattern is likely to continue until planting of high value agro-
forestry trees on farms becomes more widespread.

Managing vegetation diversity


The diversity inherent in agroforestry systems allows considerable opportu-
nity for manipulations designed to reduce pests by directly restricting build
up of pest populations and indirectly by enhancing the effects of natural
enemies. Whether increasing plant diversity in agroforestry helps reduce pests
is difficult to answer. Diversity in agroforestry varies from a few trees in
parkland systems of the Sahel to well over 100 species in Javanese home
gardens. Furthermore, the simultaneous existence of planned and unplanned
diversity at the farm and the landscape levels makes generalisations difficult
about the effects of plant diversity on pests. As van Emden (1998) pointed
out, diversity by itself will not reduce pests, and some combinations of species
can actually aggravate pest problems. The ecology of tritrophic interactions
in complex cropping systems is not easily understood (Hitimana and
McKinlay, 1998), so although the range of possible factors affecting pest
incidence may be known, predicting what will happen in a particular situa-
tion is difficult (Andow, 1991). Examples of successful manipulations of bio-
diversity within an ecosystem to minimise pest damage have therefore largely
been achieved case by case, rather than starting from a theoretical basis.
However, this is how traditional agroforestry systems have developed, so in
attempting to develop principles on which systems can be designed to suppress
pests, contributions should be sought from indigenous knowledge as well as
theoretical ecology. The challenge is to increase productivity of systems, which
inevitably involves some measure of intensification, without losing function.
There are many examples where this has not been achieved, and Trenbath et
al. (1990) analyzed three, one of which was the increase in scarab and other
beetle attack on eucalyptus in Australia, caused by clearing eucalyptus to
improve pasture.
In Turrialba, Costa Rica, species richness and diversity of hoppers
(Homoptera) were greater in a 2-crop perennial system (coffee-Erythrina
poeppigiana) than pure coffee or a 3-crop system (coffee-erythrina-Cordia
alliodora) (Rojas et al., 1999). In a study by Risch et al. (1983), mixed peren-
nial systems recorded greater populations than their respective monocultures
in 29% of 140 insect species monitored, but mixed annual crop systems
recorded greater populations only in 14%. In the Brazilian Amazone, rice,
bean and maize experienced higher pest infestations when these crops were
intercropped with trees than in their respective pure crops. The increase of
pest infestations depended on the tree species used as follows: Bractris
gasipaes > Euterpe oleracea > Theobroma grandiflorum > Coffee arabica >
Bertholletia excelsa (Fazolin and Estrela, 1999). In India, cassava mite
populations were greater on cassava intercropped with eucalyptus than on
the crop intercropped with banana (Ghosh et al., 1986). In northern Nigeria,
267

grasshopper populations feeding on pearlmillet and sorghum were lower when


these crops were interspersed with neem trees than with Acacia arabica
(Amatobi et al., 1988). Apparently, increased diversity does not automatically
confer protection against insect pests, and much depends on the feeding habit
and mobility of the insects, and species composition of the system.
Mixed cropping of host species with non-host or antagonistic species can
help restrict infestations of some monophagous and sedentary insects. In
contrast to the effect of pure Sesbania sesban fallows, mixed fallows of S.
sesban and Crotalaria grahamiana did not increase the soil populations of
root-knot nematodes, and yields of nematode-susceptible beans following
mixed fallows were not reduced (Desaeger and Rao, unpublished data). Many
Crotalaria species are known to be antagonistic to Meloidogyne (Rodriguez-
Kabana, 1992). However, cropping systems may have little effect on the
infestation and damage potential of species-specific insects that rapidly
multiply, and that are mobile and carried by wind with limited flight control.
For an example, intercropping leucaena with maize, compared to sole
leucaena, did not reduce the leucaena psyllid populations and damage caused
by the insect (Ogol and Spence, 1997). The population build up of such insects
largely depends on weather conditions.
The species composition of a system affects the activity of natural enemies
as much as the prey insects, and consequently the extent of parasitism or
predation of pests. Topper et al. (1998) examined the effect of different
intercrops on pests of cashew in East Africa, particularly Helopeltis. They
found that some species had no effect, but pigeonpea worsened the problem.
Zhou et al. (1994) found planting Ageratum ground cover in citrus orchards
changed the microclimate, which caused some pests to increase in numbers.
But some pests were controlled by an increase in their natural enemies. Tea
grown between Pinus, Acer or Quercus also had fewer pest problems due to
greater natural enemy populations, and Hypsipyla damage to Cedrela odorata
was less when grown in association with sugarcane than with plantain (Patino
Arcos, 1988). However, hedgerow intercropping of maize and leucaena either
unaffected or decreased egg, larval and pupal parasitism of maize stalk borers
in Kenya (Ogol et al., 1998). In Israel, the cotton cushion scale Icerya purchasi
was resistant to predation by Rodolia cardinalis on Spartum junceum and
Erythrina corallodendrum (Mendel et al., 1988). Velayudhan et al. (1988)
reported strong differences in the parasitism of the pentatomid Halys dentata
and the coreid Homoeocerus prominulus by Anastatus ramakrishnai depending
on the tree species on which these pests thrived. While the ranking of trees
for the parasitism of H. dentata was Cassia marginata > Azadirachta indica
> Casuarina equisetifolia, the ranking for the paratism of H. prominulus was
Cassia marginata > Prosopis spicigera > Acacia leucophloea.
Trees and shrubs within and outside the cropped fields attract a variety of
ants, flies, wasps and spiders to their extra floral nectars and offer potential
for biological control of insects detrimental to crops. They provide habitats
to generalist predators (wasps, spiders, birds) and contribute to their survival
268

and/or multiplication by providing alternative preys in the absence of the main


prey host. Hedgerows were found to shelter parasitic- and predatory-wasps
and spiders, particularly during non-cropping periods intervening rainy seasons
in the semi-arid Kenya (Girma et al., 2000). Other attributes being similar,
selection of trees that flower and supply nectar and pollen round the year
will contribute to maintaining the generalist predators. Apart from supporting
natural enemy populations, such trees can be exploited for apiculture to raise
additional revenue. Prosopis species are important sources of pollen during
the dry period in the arid northwestern India. Entomophagous fungi and other
microorganisms are expected to thrive well in agroforestry where trees create
favourable microclimate.
One important element of functional agrobiodiversity is ants. Way and
Khoo (1991) found that coconut planted in cocoa can provide nest sites for
Dolichoderus and Oecophylla, which reduce Helopeltis damage to cocoa.
Unfortunately, Oecophylla also bites people, so this approach has its draw-
backs. Mwaiko (1998) increased Oecophylla populations and improved control
of Pseudotheraptus wayi by using ant bait, as well as intercropping coconut
with citrus, mango or guava. At Machakos, Kenya, we observed that the ant
Pheidole megacephaly attacked the shoot scale Coccus longulus on pigeonpea
and the mealy bug Spilococcus sp. on calliandra. The parasitoids Metaphycus
stanleyi, Eupelmus sp., Cheiloneurus carinatus and Tremblaya minor on
Coccus longulus, and the predators Leucopsis sp. and Nephus sp. and the
parasitoids Anagrys nigrescens, C. carinatus, Psuedectroma sp., Aprostocetus
sp. and Pachyneuron sp. on Spilococcus sp. established only in the absence
of the ants.
Coccinellids are another important group of predators particularly for
homopteran pests. When Xu and Wu (1989) intercropped bamboo with rape,
aphids on the rape attracted coccinellids, which then moved to the bamboo
and reduced scale populations by 97 to 99%.

Conclusion

Information on the biology and ecology of insect herbivores of trees is


essential to determine their pest status. Inadequate attention given to plant
protection in agroforestry is partly because of poor understanding of economic
losses caused by insects to trees. Studying all those insects that attack more
than 2000 tree species documented in agroforestry tree databases is undoubt-
edly a stupendous task. But detailed studies should be targeted only to trees
employed in promising agroforestry systems and major insects that have direct
bearing on productivity. As a first step, research should focus on high
value trees for which germplasm improvement is underway. Regular and pro-
longed monitoring of key insect populations and their natural enemies under
different management situations together with seasonal weather parameters
will help determine the conditions that favour the build up of pests. This
269

information is extremely useful for taking preventive measures against pest


outbreaks.
While introducing new tree species it is essential to collect detailed infor-
mation about the endemic pests of those species and their natural enemies.
Such information will be extremely useful to undertake rapid remedial plans
should a pest from the endemic area reach the new site. Rigorous implemen-
tation of phytosanitary measures can prevent the spread of new pests, but
efforts should simultaneously be made to get new introductions from a wider
genetic base, to exploit the genetic potential of the species and avoid the
devastating effects of pests.
Many herbivores of trees have not been identified, so researchers and
extension staff often fail to correctly identify the pest problems in agroforestry.
Cataloguing all the available information related to insect pests such as host
range, population dynamics, their damage potential, natural enemies and
strategies for management into a database will assist in the choice of appro-
priate tree species for different situations. Incorporating this information into
the existing tree databases further increases their value. Publication of pest
identification manuals will help field staff to quickly recognise pest outbreaks
and implement corrective strategies.
Except for a few reports (Lenné, 1992; Lenné and Boa, 1994; Duponnois
et al., 1999), little is known about diseases and nematodes infesting agro-
forestry trees and their significance to production. The pathogical problems
should not be neglected as some of them can cause serious economic losses.
There is no need for creating a new cadre of agroforestry tree disease or insect
specialists to examine finer details, but even crop entomologists and pathol-
ogists could be relied upon for essential services (Boa, 1998).
Research grants in recent years are mostly directed for short-term impacts,
at the cost of generating basic information of long-term value. This implies
that unless disasters of the leucaena psyllid type occur, funds may not be
available for plant protection research in agroforestry. This is clearly reflected
by limited pest management research in agroforestry in national and interna-
tional institutions. To put into practice the principle that ‘prevention is better
than cure’, there is an urgent need to understand the ecological factors
governing pest populations in agroforestry. This basic knowledge will help
design systems with limited pest problems and develop tools for predicting
pest infestations. If agroforestry is to thrive well as an applied ecological
science offering one of the best strategies for sustainable exploitation of
natural resources, pest management should be based on applying ecological
principles and practical decision tools. Does agroforestry offer any scope for
low-cost pest management, and can potentials be translated into reality? Plant
protection specialists today are confronted with this question for which there
is little hard data. This is similar to the situation agroforesters faced some 20
or 30 years ago about agroforestry potentials for improving soil fertility,
generating higher income and sustainable land management, for which there
is now considerable evidence.
270

References
Abawi GS and Thurston HD (1994) Effect of organic mulches, soil amendments and cover crops
on soil borne plant pathogens and their root diseases: a review. In: Thurston HD, Smith M,
Abawi G and Kearl S (eds) Slash/Mulch: How Farmers Use it and What Researchers Knew
About it, pp 89–100. Cornell University, Ithaca, New York
Ackerman IL, McCallie EL and Fernandes ECM (1998) Inga and insects: the potential for
management in agroforestry. In: Pennington TD and Fernandes ECM (eds) The Genus Inga
Utilisation, pp 117–132. The Royal Botanical Gardens, Kew
Agounke D, Agricola U and Bokonon-Ganta HA (1988) Rastrococcus invadens Williams
(Hemiptera: Pseudococcidae), a serious exotic pest of fruit trees and other plants in West
Africa. Bulletin of Entomological Research 78: 695–702
Altieri MA and Liebman MZ (1986) Insect, weed and plant disease management in multiple
cropping systems. In: Francis CA (ed) Multiple Cropping Systems, pp 183–218. Macmillan,
New York
Altieri MA, Trujillo FJ and Farrell J (1987) Plant–insect interactions and soil fertility relations
in agroforestry systems: implications for the design of sustainable agro-eco-systems. In:
Gholz HL (ed) Agroforestry: Realities, Possibilities and Potentials, pp 89–108. Martinus
Nijhoff Publishers/ICRAF, Dordrecht
Altieri AM (1994) Biodiversity and pest management in agroecosystems. Haworth Press Inc,
Binghamton, New York
Amatobi CI, Apeji SA and Oyidi O (1988) Effect of farming practices on populations of two
grasshopper pests (Kraussaria angulifera Kraus and Oedaleus senegalensis Kraus)
(Orthoptera: Acrididae) in northern Nigeria. Tropical Pest Management 34: 173–179
Andow DA (1991) Vegetational diversity and arthropod population response. Annual Review
of Entomology 36: 561–586
Arce JJC, Peres OP and Berti E (1987) Biology of the bagworm Oiketicus kirbyi (Lands-
Guilding) (Lepidoptera: Psychidae) on Eucalyptus spp. leaves. Anãis da Escola Superior de
Agricultura Luiz de Queiroz 44: 341–358
Atkinson PR, Nixon KM and Shaw MJP (1992) On the susceptibility of Eucalyptus species and
clones to attack by Macrotermes natalensis Haviland (Isoptera: Termitidae). Forest Ecology
and Management 48: 15–30
Atkinson PR (1998) Aerial spraying trials with alternative insecticides and application methods
for the wattle bagworm, Chaliopsis junodi (Heylaerts) (Lepidoptera: Psychidae). ICFR
Bulletin Series No. 6-98. Institute for Commercial Forestry Research, Scottsville, South
Africa
Austin MT, Sorenson CT, Brewbaker JL, Sun W and Shelton HM (1995) Forage dry matter
yields and psyllid resistance of thirty one leucaena selections in Hawaii. Agroforestry Systems
31: 211–222
Bach CE (1980a) Effect of plant density and diversity on the population dynamics of a
specialised herbivore, the striped cucumber beetle, Acalymma vittatta (Fab). Ecology 61:
1515–1530
Bach CE (1980b) Effect of plant diversity and time of colonization on a herbivore–plant
interaction. Oecologia 44: 319–326
Bach CE (1984) Plant spatial pattern and herbivore population dynamics: plant factors affecting
the movement, patterns of a tropical cucurbit specialist (Acalymma innubum). Ecology 65:
175–190
Baylis WBH and Barnes RD (1989) International provenance trials of Pinus caribaea var.
bahamensis. Forest Genetic Resources Information No. 17: 24–25
Beer J, Muschler R, Kass D and Somarriba E (1998) Shade management in coffee and cacao
plantations. Agroforestry Systems 38: 139–164
Bhalla S, Verma BR and Thomas TA (1988) Reaction of germplasm collections of different
Sesbania species to seed chalcid, Bruchophagus mellipes Gahan (Hymenoptera:
Eurytomidae). Journal of Entomological Research 12: 129–131
271

Bhatt BP, Kaletha MS and Todaria NP (1997) Allelopathic exclusion of multistorey crops by
agroforestry trees of Garhwal Himalayas. Allelopathy Journal 4 : 321–328
Boa E (1998) Diseases of agroforestry trees: do they really matter? Agroforestry Forum 9(2):
19–22
Bollinger EK and Caslick JW (1985) Factors influencing blackbird damage to field corn. Journal
of Wildlife Management 49: 1109–1115
Braza RD (1987) Studies on the Leucaena psyllid, Heteropsylla cubana in Surigao del Sur,
Philippines. Leucaena Research Reports 8: 1–6
Budelman A (1988) The decomposition of the leaf mulches of Leucaena leucocephala, Gliricidia
sepium and Flemingia macrophylla under humid tropical conditions. Agroforestry Systems
7: 33–45
Cadahia D (1986) The importance of insect pests of eucalyptus in the Mediterranean region.
Bulletin-OEPP 16(2): 265–283
Castillo AC and Shelton HM (1994) Psyllid resistance, growth, and biomass production of
Leucaena accessions in Queensland, Australia. Nitrogen Fixing Tree Research Reports 12:
8–89
Chey VK (1996) Termiticide trials on young infested Gmelina arborea trees in Segaliud Lokan,
Sabah. Journal of Tropical Forest Science 9: 75–79
Chilima CZ (1991) Termite control in young Eucalyptus plantations in Malawi using controlled
release insecticides. Commonwealth Forestry Review 70: 237–247
Cornelius J, Mesen F, Corea E and Henson M (1996) Variation in growth and form of Alnus
acuminata Kunth. grown in Costa Rica. Silvae Genetica 45: 24–30
Dahlsten DL, Hansen EP, Zuparko RL and Norgaard RB (1998) Biological control of the blue
gum psyllid proves economically beneficial. California Agriculture 52: 35–38
Day RK, Hill MG, Murphy ST and Nyambo B (1995) Candidate agents for biological control
of leucaena psyllid in Africa. In: Ciesla WM and Nshubemuki L (eds) Leucaena Psyllid: A
Threat to Agroforestry in Africa, pp 169–189. FAO, Rome
Day RK, Gichora M, Mititu K and Rao MR (1996) Integrated pest management in agroforestry:
challenges and opportunities. In: Mugah JO (ed) People and Institutional Participation in
Agroforestry for Sustainable Development, pp 208–217. Kenya Forestry Research Institute,
Nairobi, Kenya
Day RK and Murphy ST (1998) Pest management in tropical agroforestry: prevention rather
than cure. Agroforestry Forum 9(2): 11–14
Day RK (1999) Integrated Control of Leucaena Psyllid. Final Technical Report of Project R6524,
Funded by DFID. NR International, Chatham, Kent, UK
Dempster JP and Coaker TH (1974) Diversification of crop ecosystems as a means of control-
ling pests. In: Jones DP and Solomon ME (eds) Biology in Pest and Disease Control, pp
106–114, John Wiley, NewYork
Desaeger J and Rao MR (1999) The root-knot nematode (Meloidogyne spp.) in sesbania fallows
and scope for managing it in western Kenya. Agroforestry System 47: 273–288
Desaeger J and Rao MR (2000) Parasitic nematode populations in natural fallows and improved
cover crops and their effects on subsequent crops in Kenya. Field Crops Research 65: 41–56
Dix ME, Johnson RJ, Harrell MO, Case RM, Wright RJ, Hodges L, Brandle JR, Schoeneberger
MM, Sunderman NJ, Fitzmaurice RL, Young LJ and Hubbard KG (1995) Influences of
trees on abundance of natural enemies of insect pests: a review. Agroforestry Systems 29:
303–311
Dix ME (1996) Pest Management in agroforestry systems: worldwide challenges in the 21st
Century. Journal of Forestry 94: 8–12
Dix ME, Hodges L, Brandle JR, Wright RJ and Harrell MO (1997) Effects of shelterbelts on
the aerial distribution of insect pests in muskmelon. Journal of Sustainable Agriculture 9:
5–24
Doube BM (1983) The habitat preference of some bovine dung beetles (Coleoptera:
Scarabaeidae) in Hluhluwe Game Reserve, South Africa. Bulletin of Entomological Research
73: 357–371
272

Duponnois R, Senghor K, Thioulouse J and Ba AM (1999) Susceptibility of several Sahelian


Acacia to Meloidogyne javanica (Treub) Chitw. Agroforestry Systems 46: 123–130
Epila JSO (1986) The case for insect pest management in agroforestry research. Agricultural
Systems 19: 37–54
Epila JSO (1988) Wind, crop pest and agroforest design. Agricultural Systems 26: 99–110
Fargues J, Rougier M, Goujet R and Itier B (1988) The effect of sunlight on the field persis-
tence of conidiospores of the entomopathogenic hyphomycete, Nomuraea rileyi, on the
surface of vegetables. Entomophaga 33: 357–370
Fazolin M and Estrela JLV (1999) Population and levels of damage of the main pests of bean,
corn and rice in an agroforestry system in the Amazon region. In: Giménez F and Beer J
(comp.) Multistrata Agroforestry Systems with Perennial Crops, pp 107–111. CATIE,
Turrialba, Costa Rica
Floyd R and Goldie J (1997) Breeding resistance in eucalypts to insect attack. Trees and Natural
Resources 39: 16–19
Gauthier R (1996) Wildlife and farming communities in North Lampung: the interactions
between vertebrate pests, farming systems and household livelihood, with implications for
governmental programmes. Final Research Report to Lembaga Ilmu Pengetahuan, Bogor,
Indonesia
Ghosh SP, Pillai KS and Thankappan M (1986) Cassava based multiple cropping systems: 2.
Incidence of pests and diseases. Journal of Root Crops 12: 83–89
Girma H, Rao MR and Sithanatham S (2000) Insect pests and beneficial arthropod populations
under different hedgerow intercropping systems in semiarid Kenya. Agroforestry Systems
50: 279–292
Glover N (1988) Evaluation of Leucaena species for psyllid resistance. Leucaena Research
Reports 9: 15–18
Gope B (1985) A fore warning: new host record of Chrysolampra flavipes Jacoby. Two and a
Bud 32: 40
Grainge M and Ahmed S (1988) Handbook of plants with pest-control properties. Resource
Systems Institute, East-West Centre, Honolulu, Hawaii, John Wiley & Sons, New York
Greathead DJ (1971) A review of biological control in the Ethiopian region. Technical
Communication No. 5. Commonwealth Institute of Biological Control, UK
Grout TG and Stephen PR (1995) New windbreak tree contributes towards integrated pest
management of citrus. Citrus Journal 5: 26–27
Hitimana N and McKinlay RG (1998) The effects of intercropping on phytophagous pests: a
review. Agroforestry Forum 9(2): 9–11
Holmgren P, Masakha EJ and Sjöholm H (1994) Not all African land is being degraded: a recent
survey of trees on farms in Kenya reveals rapidly increasing forestry resources. Ambio 23:
390–395
Hughes CE (1998) Leucaena: A genetic resource handbook. Technical Forestry Paper 37, Oxford
Forestry Research Institute, Oxford University, UK
Huxley PA and Greenland DJ (1989) Pest management in agroforestry systems: A record of
discussions held at CAB International, Wallingford, UK 28–29 July 1988. Agroforestry
Abstracts 2(2): 37–46
Intari SE (1995) An experiment to control beetle pests attacking Paraserianthes falcataria in
Citeureup District, Bogor (in Bahasa). Duta Rimba 20: 30–35
Jaques RP (1983) The potential of pathogens for pest control. Agriculture, Ecosystems &
Environment 10: 101–126
Jordan AM (1986) Trypanosomiasis Control, Longman
Kamunya SM, Day RK, Chagala E and Olng’otie PS (1997) Variation and inheritance of
resistance to cypress aphid Cinara cupressi (Buckton) in Cupressus lusitanica (Miller).
Annals of Applied Biology 130: 27–36
Kamunya SM, Olng’otie PS, Chagala E, Day RK and Kipkore WK (1999) Genetic variation
and heritability of resistance to Cinara cupressi (Buckton) in Cupressus lusitanica (Miller).
Journal of Tropical Forest Science 11: 587–598
273

Kennedy JS, Booth CO and Kershaw WJS (1959) Host finding by aphids in the field: 1.
Gynoparae of Myzus persicae (Sulzer). Annals of Applied Biology 47: 410–423
Kettler JS (1995) A new insect pest of Erythrina poeppigiana in Costa Rica. Nitrogen Fixing
Tree Research Reports 13: 51–53
Lal R (1989) Agroforestry systems and soil surface management on a tropical alfisol. 1. Soil
moisture and crop yields. Agroforestry Systems 8: 7–29
Landsberg J and Wylie FR (1983) Water stress, leaf nutrients and defoliation: a model of dieback
of rural eucalypts. Australian Journal of Ecology 8: 27–41
Laup S (1994) Pests and diseases of shade trees and their relation to cocoa in Papua New Guinea.
Papua New Guinea Journal of Agriculture, Forestry and Fisheries. 37(2): 76–85
Leakey RRB and Simons AJ (1998) The domestication and commercialization of indigenous
trees in agroforestry for the alleviation of poverty. Agroforestry Systems 38: 165–176
Lenné JM (1992) Diseases of multipurpose woody legumes in the tropics: a review. Nitrogen
Fixing Tree Research Reports 10: 13–32
Lenné JM and Boa ER (1994) Diseases of tree legumes. In: Gutteridge RC and Shelton HM
(eds) Forage Tree Legumes, pp 292–307. CAB International, Wallingford, UK
Levin DA (1973) The role of trichomes in plant defense. Quarterly Review of Biology 48: 3–21
Liping Z (1991) Biological control of insect pests and diseases in agroforestry systems. In: Avery
ME, Cannell MGR and Ong CK (eds) Biophysical Research for Asian Agroforestry, pp
73–90. Winrock International, USA
Mchowa JW and Ngugi DN (1994) Pest complex in agroforestry systems: the Malawi experi-
ence. Forest Ecology and Management 64: 277–284
Mendel Z, Blumberg D and Assael F (1988) Biological control of the cottony cushion scale,
Icerya purchasi, on ornamental plants in Israel. Hassadeh 69: 350–351
Meshram PB, Farooqui UM and Bharmal CD (1998) Preliminary screening of Eucalyptus species
and provenances against termites Odontotermes spp. Indian Forester 124: 573–577
Midgley SJ and Weerawardane NDR (1986) Termites and their control in Eucalyptus planta-
tions of the Community Forestry Project, Badulla. Sri Lanka Forester 17(3–4): 111–115
Mitchell MR and Boland DJ (1989) Susceptibility to termite attack of various tree species planted
in Zimbabwe. In: Trees for the Tropics: Growing Australian Multipurpose Trees and Shrubs
in Developing Countries, pp. 215–227. ACIAR Monograph No. 10, Canberra, Australia
Monteith LG (1960) Influence of plants other than the food plants of their host on host finding
by tachinid parasites. Canadian Entomology 92: 641–645
Muddiman SB and Hodkinson ID (1992) Legume-feeding psyllids of the genus Heteropsylla
(Homoptera: Psylloidea). Bulletin of Entomological Research 82: 73–117
Mukthar A, Bhandari RS and Lalji Prasad (1996) Insect problem in agroforestry tree species
and its management. Van Vignan 34(3): 84–99
Mullen BF, Gabunada F, Shelton HM, Stür WW and Napompeth B (1998) Psyllid resistance in
Leucaena. In: Shelton HM, Gutteridge RC, Mullen BF and Bray RA (eds) Leucaena–
Adaptation, Quality and Farming Systems, pp 51–60. ACIAR Proceedings No. 86, Canberra,
Australia
Muniappan R (1993) Pests and diseases of Erythrina: a review. Journal of Coffee Research 23:
1–13
Murphy ST (1998) Protecting Africa’s trees. Unasylva 192 49: 57–61
Mushongahande M (1996) Regent (fipronil): a new insecticide for termite control in Eucalyptus.
ForMat 8: 3–4
Mwaiko W (1998) Role of Oecophylla longinoda (Formicidae) and Amdro (hydramethlynon)
ant bait in the integrated pest management of Pseudotheraptus wayi (Coreidae) on coconuts
in Tanzania. In: Topper CP, Caligari PDS, Kullaya AK, Shomari SH, Kasuga LJ, Masawe
PAL and Mpunami AA (eds) Trees for Life – The Key to Development, pp 452–455.
BioHybrids International Ltd, Reading, UK
Nanga’yo FLO, Hill MG, Chandi EA, Chiro CT, Nzeve DN and Obiero J (1993) The natural
environment as a reservoir for the larger grain borer Prostephanus truncatus (Horn)
(Coleoptera: Bostricidae) in Kenya. African Crop Science Journal 1: 39–47
274

Napompeth B and MacDicken KG (eds) (1990) Leucaena psyllid: problems and management.
Proceedings of an international workshop held in Bogor, Indonesia, 16–21 January 1989.
Bangkok: F/Fred Project Winrock International, 214 pp
Napompeth B (1994) Leucaena psyllid in the Asia-Pacific Region: Implications for its man-
agement in Africa. Regional Office for Asia and the Pacific (RAPA) Publication 1994/13,
Bangkok, Thailand
Ogol CKPO and Spence JR (1997) Abundance, population dynamics and impact of the Leucaena
psyllid Heterosphylla cubana Crowford in a maize–Leucaena agroforestry system in Kenya.
Insect Science Application 17: 183–192
Ogol CKPO, Spence JR and Keddie A (1998) Natural enemy abundance and activity in a
maize–Leucaena agroforestry system in Kenya. Environmental Entomology 27: 1444–1451
Ogol CKPO, Spence JR and Keddie A (1999) Maize stem borer colonization, establishment
and crop damage levels in a maize-Leucaena agroforestry system in Kenya. Agriculture,
Ecosystem and Environment 76: 1–15
Opondo-Mbai ML (1995) Investigations of arthropods associated with agroforestry in Machakos.
PhD thesis, University of Oxford, UK
Otsyina RM (1993) Would agroforestry and afforestation risk tsetse invasion? Agroforestry
Today 5(1): 6–8
Pasek JE (1988) Influence of wind and windbreaks on local dispersal of insects. Agriculture,
Ecosystems and Environment 22/23: 539–554
Patel M and Sahu AK (1995) Control of termite attack on some high pulp yielding trees. Indian
Forester 121: 295–299
Patino-Arcos H (1988) Evaluacion de associaciones agroforestales para la zona aluvial del rio
Calima, Buenaventura Valle, Columbia. Serie Tecnica Cooperacion Nacional de Investgacion
y Fomento Forestal No. 28, Bogota, Columbia
Pu Z (1978) Principles and Methods for Biological Control of Insect Pests. Beijing Academic
Press
Rao MR (1995) Leucaena psyllid in Kenya and experience with chemical control. In: Ciesla
WM and Nshubemuki L (eds) Leucaena psyllid: A Threat to Agroforestry in Africa, pp
136–142. FAO, Rome
Rao MR and Mathuva MN (1997) Resistance to psyllid and productivity of some Leucaena
species and their hybrids in Kenya. In: Fungoh PO and Mbadi GCO (eds) Focus on
Agricultural Research for Sustainable Development in a Changing Economic Environment,
pp 611–620. Kenya Agricultural Research Institute, Nairobi, Kenya
Remadevi OK, Raja Muthukrishnan, Rao AR, Sivaramakrishnan VR, Santhakumaran LN and
Muthukrishnan R (1997) Epidemic outbreak of lac insect, Kerria lacca (Kerr.), on Santalum
album (sandal) and its control. Indian Forester 123: 143–147
Risch SJ (1980) The population dynamics of several herbivorous beetles in a tropical agro-
ecosystem: the effect of intercropping corn, beans and squash in Costa Rica. Journal of
Applied Ecology 17: 593–612
Risch SJ (1981) Insect herbivore abundance in tropical monocultures and polycultures: an
experimental test of two hypotheses. Ecology 62: 1325–1340
Risch SJ, Andow D and Altieri MA (1983) Agroecosystem diversity and Pest Control: data,
tentative conclusions and new research directions. Environmental Entomology 12: 625–629
Robertson HG (1987) Oviposition site selection in Cactoblastis cactorum (Lepidoptera):
constraints and compromises. Oecologia 73: 601–608
Rodriguez-Kabana R (1992) Cropping systems for the management of phytonematodes. In:
Grommers FJ and Maas PWT (eds) Nematology from Molecules to Ecosystem. Proceedings
of the Second International Nematology Congress, 11–17 August 1990, Veldhofen, pp
219–233. European Society of Nematologists, Dundee, UK
Rojas L, Godoy C, Hanson P, Kleinn C and Hilje L (1999) Diversity of hoppers (Homoptera:
Auchenorhyncha) in coffee plantations with different types of shade, in Turrialba, Costa
Roca. In: Jiménez F and Beer J (com.) Multistrata Agroforestry Systems with Perennial
Crops, pp 216–219. CATIE, Turrialba, Costa Rica
275

Root RB (1973) Organisation of plant–arthropod association in simple and diverse habitats: the
fauna of collards (Brassicae olevaceae). Ecological Monograph 43: 95–124
Rorabaugh JC (1995) A superior accession of western honey mesquite (Prosopis glandulosa var.
torreyana) for riparian restoration projects. Desert Plants 11: 32–40
Roychoudhury N, Kalia S and Joshi KC (1997) Spodoptera litura (Fabricius) Boursin
(Lepidoptera: Noctuidae), a major pest of Sesbania spp. and its species preferences. Indian
Journal of Forestry 20: 67–69
Santos GP, Anjos N dos, Zanuncio JC, Zanuncio TV and Dos Anjos N (1990) Eficiencia de
diflubenzuron a lagarta-parda do eucalipto, Thyrinteina arnobia Stoll, 1782 (Lepidoptera:
Geometridae), em condicões de laboratorio e campo. Anais da Sociedade Entomologica do
Brasil. 19: 345–354
Sasidharan KR, Rishi RR and Deeparaj B (1997) Response of Albizia lebbeck provenances to
the infestation of psyllids and aphids. Uttar Pradesh Journal of Zoology 17: 177–182
Satya Vir and Verma SK (1996) Insect pests of agroforestry leguminous trees in India. Annals
of Arid Zone 35: 349–359
Schoonhoven LM (1968) Chemosensory bases of host plant selection. Annual Review of
Entomology 13: 115–136
Schroth G, Krauss U, Gasparotto L, Duarte Aguilar JA and Vohland K (2000) Pests and diseases
in tropical agroforestry systems – a review. Agroforestry Systems 50: 199–241
Selander J, Loyttyniemi K and Chendauka BS (1989) Control of termites in eucalypt
plantations: testing protection methods and evaluating carbosulfan as an alternative to
organochlorines. Research Note No. 46, Division of Forest Research Kitwe, Forest
Department, Zambia
Shahjahan M and Streams FA (1973) Plant effects on host finding by Leiophron pseudopal-
lipes (Hymenoptera: Braconidae), a parasitoid of the tarnished plant bug. Environmental
Entomology 2: 921–925
Sharma VP, Lal OP, Rohidas SB and Pramanick PK (1998) Varietal resistance in ber (Zizyphus
mauritiana Lamk.) against the fruitfly, Carpomyia vesuviana Costa (Diptera: Tephritidae)
under field conditions. Journal of Entomological Research 22: 61–67
Sileshi G, Maghembe JA, Rao, MR, Ogol CKPO and Sithanantham S (2000) Insects feeding on
Sesbania species in natural stands and agroforestry systems in southern Malawi. Agroforestry
Systems 49: 41–52
Simons AJ (1996) Delivery of improvement for agoforestry trees. In: Dieters MJ, Matheson AC,
Nickles DG, Harood CE and Walker SM (eds) Tree Improvement for Sustainable Tropical
Forestry, Vol 2, pp 391–400. Queensland Forestry Research Institute, Gympie, Australia
Singh MP (1984) Field screening of some jujube cultivars against the attack of bark eating
caterpillar, Indarbela quadrinotata (Walker). The Madras Agricultural Journal 71: 416–417
Singh Rathore, MP (1995) Insect Pests in Agroforestry. Working Paper No. 70. ICRAF, Nairobi,
Kenya
Singh MP (1997) Insect pest management in agroforestry. In: Gupta JP and Sharma BM (eds)
Agroforestry for Sustained Productivity in Arid Regions, pp 12–132. Scientific Publishers,
Jodhpur, India
Singh MP and Parihar DR (1997) Insect pest management in silvipastoral systems. In: Yadav
MS, Manjity Singh, Sharma, SK Tiwari, JC and Burman U (eds) Silvipastoral Systems in
Arid and Semi-arid Ecosystems, pp 371–385. UNESCO/CAZRI, Jodhpur, India
Singh MP (1998) Injurious Hexapoda associated with Prosopis. In: Tewari JC, Pasiecznik NM,
Harsh LN and Harris PJC (eds) Prosopis Species in the Arid and Semi-arid Zones of India,
pp 105–108. Prosopis Society of India/The Henry Doubleday Research Association, UK
Singh MP, Bhansali RR, Kaul RK and Tripathi RS (1998) Integrated pest management in peren-
nial plants. In: Faroda AS and Manjit Singh (eds) Fifty Years of Arid Zone Research in India,
pp 375–392. Central Arid Zone Research Institute, Jodhpur, India
Singh MP, Satyavir and Parihar DR (1989) A note on the Coleopterous pests of forest plants
in the Indian Desert. Indian Journal of Forestry 12: 330–331
Singh P and Singh S (1987) Pest and pathogen management in agroforestry systems. In: Khosla
276

PK and Khurana DK (eds) Agroforestry for Rural Needs, pp 153–177. Indian Society of Tree
Scientists, Solan, India
Sivaramakrishanan VR (1986) Note on recent outbreak of Celosterna scabrator Fabricius
(Lamiidae: Coleoptera) on Eucalyptus in Karnataka. Myforest 22: 103–105
Solomon MG (1981) Windbreaks as a source of orchard pests and predators. In: Thresh JM
(ed) Pests, Pathogens and Vegetation: The Role of Weeds and Wild Plants in the Ecology
of Crop Pests and Diseases, pp 273–283, Pitman Books, London
Soria F and Borralho NMG (1997) The genetics of resistance to Phoracantha semipunctata attack
in Eucalyptus globulus in Spain. Silvae Genetica 46: 365–369
Speight HR (1983) The potential of ecosystem management for pest control. Agriculture,
Ecosystem and Environment 10: 183–199
Steinmuller N (1995) Agronomy of the N2-fixing fodder trees Sesbania sesban (L.) Merr. and
Sesbania goetzii Hamms in the Ethiopian highlands. Verlag Ulrich E. Grauer, Stuttgart,
Germany
Sudarmadji D (1988) The effect of contact and systemic insecticides on a predator of Leucaena
psyllid, Curinus coeruleus Mulsant (in Bahasa). Menara Perkebunan 56: 73–75
Suresh KK and Rai RSV (1988) Allelopathic exclusion of understorey by a few multipurpose
trees. International Tree Crops Journal 5: 143–151
Szeoke K and Takacs L (1984) Damage caused by lima bean borer (Etiella zinckella Tr) in peas.
1. Study of habits. Novenyvedelem 20: 433–438
Tahvanianen JO and Root RB (1972) The influence of vegetational diversity on the population
ecology of specialised herbivore, Phyllotreta crucifera (Coleoptera: Chrysomelidae.
Oecologia (Berl.) 10: 321–346
Tassin J, Michon R and Orazio C (1998) Substitution of tree-species in declining Casuarina
equisetifolia stands with a protection role in Reunion (in French). Bulletin-Technique Office
National des Forets 36: 91–96
Topper CP, Pearce M, Grunshaw J and Anthony J (1998) Preliminary observations on the role
of common intercrops in the population dynamics of Helopeltis and damage to cashew. In:
Topper CP, Caligari PDS, Kullaya AK, Shomari SH, Kasuga LJ, Masawe PAL and Mpunami
AA (eds) Trees for Life – the Key to Development, pp 182–184. BioHybrids International
Ltd, Reading, UK
Traoré S and Ouedraogo I (1997) Influence of the microclimate on the dispersion of MSV vector
leafhoppers (Cicadulina spp.) in Burkina Faso. In: Proceedings International Conference on
Pests in Agriculture. La Corum, Montpellier, France. Vol 3: 927–934, Associacion Nacionale
pour la Protection des Plantes, France
Trenbath BR, Conway GR, Craig IA and Gliessman SR (1990) Threats to sustainability in
intensified agricultural systems: analysis and implications for management. In: Gliessman
SR (ed) Agroecology: Researching the Ecological Basis for Sustainable Agriculture, pp
337–365. Ecological Studies vol 78, Springer Verlag, New York, USA
Tuset JJ and Hinarejos C (1985) Influence of tissue moisture content on susceptibility of cypress
and palms to attack by two Pestalotiopsis species. Anales del Instituto Nacional de
Investigaciones Agrarias Agricola 28 (supp): 211–234
Vandermeer J (1989) The ecology of intercropping. Cambridge University Press, Cambridge
van Emden HF (1965) The role of uncultivated land in the biology of crop pests and beneficial
insects. Scientific Horticulture 17: 121–136
van Emden H (1998) Effects of complex mixed cropping systems on insect pests. Agroforestry
Forum 9(2): 6–9
Velayudhan R, Senrayan R and Rajadurai S (1988) Parasitoid – host interactions with
reference to Anastatus ramakrishnae (Mani) (Hymenoptera: Eupelmidae) in relation to
pentatomid and coreid hosts. Proceedings of the Indian National Science Academy: B.
Biological Sciences 54: 145–153
Walter GH and Parry WH (1994) Insect pests of forage tree legumes. In: Gutteridge RC and
Shelton HM (eds) Forage Tree Legumes in Agriculture, pp 309–321, CAB International,
Wallingford, UK
277

Wandera FP and Njarui DMG (1998) Psyllid tolerance and productivity of Leucaena species
and hybrids in semi-arid Kenya. In: Shelton HM, Gutteridge RC, Mullen BF and Bray RA
(eds) Leucaena – Adaptation, Quality and Farming Systems, pp 136–140. ACIAR
Proceedings No. 86, Canberra, Australia
Way MJ and Khoo KC (1991) Colony dispersion and nesting habits of the ants, Dolichoderus
thoracicus and Oecophylla smaragdina (Hymenoptera: Formicidae), in relation to their
success as biological control agents on cocoa. Bulletin of Entomological Research 81:
341–350
Wheeler RA (1988) Lucaena psyllid trial at Waimanalo, Hawaii. Leucaena Research Reports
9:25–29
Wiryadiputra S and Priyono D (1995) Studies on the use of banana trees (Musa spp.) for coffee
and cocoa shading V. Development of Pratylenchus coffeae on some banana cultivars derived
from tissue culture. Pelita Perkebunan 11: 132–139
Xu FY and Wu DX (1989) Control of bamboo scale insects by intercropping rape in the bamboo
forest to attract coccinellid beetles. Chinese Journal of Biological Control 5: 117–119
Yang J, Jin L, Qu S, Luo G and Ceng W (1988) Effects of mixed forest of Chinese fir and
Homana on population of insect pests. (Chinese) Journal of Ecology 7(1): 45–47
Yudin LS, Cho JJ and Mitchell WC (1986) Host range of western flower thrips, Frankliniella
occidentalis (Thysanoptera: Thripidae), with special reference to Leucaena glauca.
Environmental Entomology 15: 1292–1295
Zhou TY, Xu WJ and Zhong GQ (1985) A preliminary investigation on Polyura narcaea
Hewitson. Insect Knowledge Kunchong-Zhishi 23(1): 24–25
Zhou CA, Zou JJ and Huang SD (1994) The effects of Ageratum interplanting in hilly citrus
orchards in Hunan Province on citrus mite and insect populations. Acta Phytophylacica Sinica
21: 57–61

You might also like