Cookies on Plantwise Knowledge Bank

Like most websites we use cookies. This is to ensure that we give you the best experience possible.

Continuing to use means you agree to our use of cookies. If you would like to, you can learn more about the cookies we use.

Plantwise Knowledge Bank
  • Knowledge Bank home
  • Change location
Plantwise Technical Factsheet

coffee berry borer (Hypothenemus hampei)

Host plants / species affected
Bertholletia excelsa (Brazil nut)
Coffea (coffee)
Coffea arabica (arabica coffee)
Coffea canephora (robusta coffee)
List of symptoms/signs
Fruit  -  internal feeding
Fruit  -  lesions: black or brown
Fruit  -  premature drop
Seeds  -  internal feeding
Seeds  -  rot

Attack by H. hampei begins at the apex of the coffee berry from about eight weeks after flowering. A small perforation about 1 mm diameter is often clearly visible though this may become partly obscured by subsequent growth of the berry or by fungi that attack the borer. During active boring by the adult female, she pushes out the debris, which forms a deposit over the hole. This deposit may be brown, grey or green in colour.

Infestation is confirmed by cutting open the berry. If the endosperm is still watery, the female will be found in the mesoderm between the two seeds, waiting for the internal tissues to become more solid. If the endosperm is more developed, the borer will normally be found there amongst the excavations and irregular galleries that it has made. The borer sometimes causes the unripe endosperm to rot, most commonly by species Erwinia, causing it to turn black (Sponagel, 1994) and the borer to abandon the berry.

Prevention and control


Phytosanitary Measures

Transportation of seeds containing the H. hampei has been the main cause of its spread worldwide. A few coffee-producing countries or zones within countries are still free from this insect and in these cases stringent quarantine precautions are strongly recommended. Hollingsworth et al. (2013) found that treatment of infested coffee berries at a temperature of approximately -15°C for 48 h provided 100% control of all life stages.


Cultural Control

Harvesting coffee berries is itself an important control measure. Rigorous collection of remnant berries after harvest, both from tree and ground, can substantially reduce infestations as it breaks the cycle and leaves little substrate for immigrating H. hampei. Collected berries should be boiled or buried if infestation levels are high. If processed, they should be placed in a drier, or if sun-dried, placed under netting smeared with grease or oil to capture escaping borers. These methods are most successful when done carefully by resource-poor farmers (Le Pelley, 1968). However, such manual collection methods are laborious, especially the collection of fallen berries or those on the lower branches. Studies in Colombia have shown that farmers tend to leave many berries after harvest, especially low down on the trees and that the older the tree, the harder the farmers find it to remove the berries (Baker, 1997). Many experiments have been carried out in Colombia to accelerate decomposition of the fallen berries and on the feasibility of collecting them by manual or machine methods. So far no practical progress has been achieved (Baker, 1999).

There are some suggestions that populations of H. hampei tend to be lower under shade trees rather than in full sun. Armbrecht and Gallego (2007) recorded significantly more predation under shade than full sun coffee. Others however (e.g. Bosselmann et al., 2009) have found the reverse. It is likely therefore that the effect of shade is highly dependent on a number of local factors, for example, Jonsson et al. (2015) found higher levels of H. hampei under unshaded than shaded coffee in Uganda, whereas the reverse was true for the white stem borer (Monochamus leuconotus).

Biological Control

The two bethylid parasitoids, Cephalonomia stephanoderis and Prorops nasuta have been introduced from Africa to India and many Latin-American countries in the 1980s and 1990s. The few studies undertaken on their effectiveness suggest that in general they have only a moderate controlling effect and that it is rare to find more than 5% of perforated berries parasitized one or more years after releases were made (Barrera, 1994). However a follow-up study seven years after a campaign to rear and release large numbers of C. stephanoderis in different coffee growing areas of Pulney Hills, Tamil Nadu, India, recorded 16-45% parasitism from five different areas (Roobakkumar et al., 2014). Generally low parasitism may be because the berries are harvested before the wasps have a chance to emerge, though more studies are needed to explain their scarcity in the field. Both species parasitize only one berry: the female enters and stays with her brood, rather similar to the borer's maternal behaviour. From the point of view of biocontrol this is unfortunate as a parasitoid that lays eggs in many berries might be more effective. Mass release studies of C. stephanoderis in Colombia and other countries have been carried out but the costs of mass production are uneconomical and likely to remain so because of the high cost of the diet (coffee beans) for the borer host.

Phymastichus coffea was seen as a promising biocontrol agent because it attacks adults and thus might help to prevent establishment of the borer in the endosperm, where economic damage is caused. It can also parasitise borers from more than one berry and the few studies on this in the field have suggested that it may be more effective control agent than the bethylids (Baker, 1999). However, to date there are no follow up field studies that suggest it is having any suppressive effect on the borer in the field.

The fungus Beauveria bassiana is found naturally wherever H. hampei is present. In humid climates infection may reach more than 50% and is probably the most significant natural control agent of H. hampei. Pascalet (1939) found it prevalent in the forest zone of Cameroon and concluded that conditions favourable to an outbreak were a dense borer population, 20-30°C temperature, sufficient rain to produce the humidity necessary for vigorous sporulation, followed by one or two sunny days to induce an even distribution of spores, followed by light rains to favour development of spores on the bodies of the borers. Intensive efforts in Colombia, Nicaragua, Mexico and Ecuador have been made to develop an effective mycopesticide based on B. bassiana. Results have been very variable with sprays (with varying concentrations of fungal spores/tree) causing anything from 10-86% mortality (Lacayo, 1993; Sponagel, 1994; Bravo, 1995; Bustillo and Posada, 1996; Baker, 1999). High field mortality of H. hampei in the entry canal of the berry (80%+) have been achieved but only at uneconomically high doses. At lower doses the mortality is usually between 20-50% of adult females entering the berry. Further problems relate to the viability and virulence of commercially prepared formulations of the fungus and the product requires careful quality control and monitoring to ensure acceptable standards. Currently in Colombia, despite a concerted research and extension effort over many years, few farmers still apply the fungus. Benavides et al. (2012) suggest that applying a mix of B. bassiana strains may improve virulence. Another approach has been to inject B. bassiana into coffee in the hope that it might establish inside the plant and act as an endophyte to attack the borer when it drills into the berry (Vega et al., 2005).

More recently efforts to increase the virulence of Metarhizium anisopliae (a fungus which occasionally attacks H. hampei), by inserting a scorpion toxin gene through genetic engineering (Pava-Ripoll et al., 2008).

Vega et al. (2002a) have also studied the presence of Wollbachia in H. hampei, a bacterial infection that may be the cause of its skewed sex ratio. However to date there seems to be no practical way to use this knowledge to devise a novel control method.

In general nematodes would be difficult to apply to coffee trees, but might be easier to apply to the ground under the trees where the microclimate might be very suitable for them. The fallen berries under the tree are known to be a very important reservoir of re-infestation and yet difficult to control either by chemicals, fungi or manual collection and experimental releases of parasitoids suggest that few of them attack fallen berries. Hence what is needed is something that could actively search for an infested berry and tunnel its way into the berry to attack the coffee berry borer inside. Lopez-Nuñez of Cenicafé, Colombia, working with Steinernema carpocapsae (All strain), S. glaseri and Heterorhabditis bacteriophora has achieved infection and mortality of H. hampei in laboratory and small-scale field trials (Baker, 1999). Efforts continue to evaluate its performance in larger field trials.

In recent years there have been a number of studies to evaluate the effect of bird predation (e.g. Johnson et al., 2010; Karp et al., 2013 ) which through exclusion cage experiments show significant control effects in heavily infested field conditions. The presence of H. hampei in the diet of some birds has been confirmed through DNA analysis of faecal samples (Karp et al., 2014), however less than 10% of birds tested positive for H. hampei. Exclusion studies have also been carried out with ants (e.g. Solenopsis geminata;Trible and Carroll, 2014) which show a significant predation effect. To date though, no long term field experiments have been performed which demonstrate reliable and significant predation from a range of naturally occurring predators. The main difficulty is that generalist predators tend to search for high density prey and may switch away from H. hampei at levels above an acceptable economic threshold.

Thus despite intensive efforts over the last 25 + years, the impact of biocontrol on H. hampei continues to be disappointingly low.

Chemical Control

Wherever possible, chemical control should be done in foci (hot-spots) of infestation, before they grow and coalesce. This however requires regular visual monitoring of the coffee fields.

Insecticides can be effective if they are applied when the female is in the entry tunnel before she penetrates the endosperm. They are not effective at controlling mature infestations, especially on fallen berries. As coffee trees are frequently densely planted and taller than the persons spraying them, serious contamination is likely; pesticide poisonings and deaths are reported from Colombia (P.S. Baker, CABI Bioscience, personal communication, 1996). The worldwide ban on endosulfan means that most farmers now depend on less toxic alternative pesticides such as fenitrothion, fenthion and pirimiphos methyl (Villalba et al., 1995) though they are often less effective. This has led to renewed interest in ways to reduce pesticide use (Watts and Williamson, 2015). However, full-scale independent field trials followed through to harvest damage assessment for currently available chemicals and biological alternatives are still lacking.

A newer combination insecticide registered for use in Colombia and recommended for use by the Colombian Coffee Federation is a mixture of chlorantraniliprole and thiamethoxam (Arcila et al., 2013). Currently however, the costs of these chemicals mean that they are too expensive for many farmers. A study of the effects of neem, a botanical insecticide and repellent, concluded that efficacy of various commercial products in field against H. hampei is very low due to rapid degradation (Vijayalakshmi et al., 2014).

Integrated Pest Management

A crude version of IPM is employed by many farmers, involving some cultural control and insecticidal spraying. Different schemes, based on sampling and economic thresholds have been developed (Decazy and Castro, 1990), but it is difficult to establish simple thresholds on a perennial crop with a prolonged flowering period and a long berry development period. Further, if a chemical control option is selected, it needs to be carried out many weeks (16 or more) before harvest when the borers are in their most susceptible stage (Decazy et al., 1989; Barrera, 1994). Establishment of an economic threshold is equally difficult when the coffee farmer is unsure of the impact of the post-harvest borer population on the next harvest many months hence. Extensive studies of Colombian farmers attest to the difficulty of adoption of complex IPM regimes (Duque and Chaves, 2000). In many cases a value of 5% berries damaged is used as a ‘rule-of-thumb’ action threshold.

The main issue is that there is no simple and cheap method to control this insect. This has led to the promotion of a very wide range of combinations of control elements which has sometimes resulted in quite complex IPM schedules that farmers, especially smallholders, find difficult to adopt. It is frequently not clear that each added element exerts a significant or cost-effective increment to control. To an extent this is due to the complex nature of the pest, which is cryptic and may have multiple overlapping populations growing on several populations of berries resulting from different flowerings. This situation demands extensive and multi-year research studies which are frequently beyond the budgets of small research facilities of most coffee countries. The prospects for IPM of H. hampei are dealt with in detail in Baker (1999).

Host-Plant Resistance

Chevalier (cited in Le Pelley, 1968) found Coffea liberica almost immune to H. hampei followed by C. excelsa, C. dewerei, C. canephora and C. arabica in increasing order of attractiveness to the borer. Villagran (1991) found that. H. hampei had difficulty in penetrating the hard exterior of C. liberica berries. However, Roepke (in Le Pelley, 1968) states that C. liberica is preferentially attacked. Extensive studies by Kock (1973) reported C. canephora variety Kouilou (or Quoillou) is attacked less than the Robusta variety.

Villagran (1991) found C. kapakata supporting very significantly fewer immature stages of the borer than other varieties and some tendency for C. arabica variety Mundo Novo also to support fewer progeny. Olfactometry tests by Duarte (1992) showed C. kapakata to be significantly less attractive. C. kapakata appears to be one of the most resistant coffee species currently known but this is not a commercial variety and neither the berries nor the plant resemble a coffee plant to the casual observer.

Romero and Cortina-Guererro (2004) in laboratory studies in Colombia found no difference in levels of antixenosis (deterrence to attack coffee in field tests) of various coffee varieties (including C. arabica Caturra, various Ethiopian accessions as well as C. liberica). However Romero and Cortina-Guererro (2007) did find differences in antibiosis (expressed as fecundity) with Ethiopian accession CC532 and C. liberica both yielding significantly fewer borer progeny.

Gongora et al. (2012) confirmed the inhibitory effects of C. liberica through a functional genomics study using ESTs libraries, cDNA microarrays and an oligoarray containing 43,800 coffee sequences. The results allowed for a comparison of C. liberica vs. C. arabica berry responses to H. hampei infestation after 48h. Out of a set of 2,500 plant sequences that exhibited differential expression under H. hampei attack, twice the number were induced in C. liberica, than in C. arabica. One of the identified biochemical pathways was the one that leads to the production of isoprene. The authors studied the effect of isoprene on H. hampei by monitoring the development of the insect from egg to adult, using coffee-artificial diets amended with increasing concentrations of isoprene. Concentrations of isoprene above 25 ppm caused mortality and developmental delay in all insect stages from larva to adult, as well as the inhibition of larvae moulting.

Hence it seems certain that varying amounts of resistance or antibiosis to the borer exists within species of Coffea. Such resistance to attack or even moderate antibiosis is worthy of further study because an increase in development time and/or decrease in fecundity could have a pronounced effect on infestation levels. Conventional breeding to introduce such inhibition from outside the Arabica genome might be difficult however, hence genetic engineering may be increasingly considered in the future.

A team of CIRAD scientists were the first to succeed in producing a transgenic coffee plant with Bt resistance to leafminers but there is no information about its effect on H. hampei (Leroy et al., 2000). Scientists from Brazil and Colombia (Barbosa et al., 2010) transformed C. arabica by introducing an enzyme inhibitor from the common bean (Phaseolus vulgaris). Beans have evolved an amylase enzyme inhibitor (or ‘starch blocker’) to make them less palatable to attacking insects. They demonstrated that crude seed extracts from genetically transformed C. arabica plants expressing the α-amylase inhibitor-1 gene (α-AI1) under the control of the common bean P. vulgaris seed-specific promoter PHA-L, inhibited 88 % of H. hampei α-amylases during in vitro assays. Since then, offspring from these GM coffee plants have been cultivated under greenhouse conditions to study the heredity, stability and expression of the α-AI1 gene. Subsequently Albuquerque et al. (2015) carried out in vivo assays of H. hampei development in berries of the transformed plants. A 26-day assay showed that the lifecycle of H. hampei was still completed, though significantly fewer offspring developed than on non-transformed control beans. Other tests showed that gene expression occurred only in the endosperm tissue. Commercial interest in developing transgenic coffee resistant to pests and diseases is still low however and might meet considerable consumer resistance.


Theoretically it would be possible to develop a forecasting model to predict upsurges of H. hampei, because under some conditions, especially after a long dry spell with high temperatures, large populations develop on fallen berries which then emerge after early rains. This however would require regular field monitoring and dissections of sampled berries and the costs of mounting such an exercise are probably too high. However, even occasional and non-intensive monitoring of borer during the post-harvest dry season, could give field technicians a feel for the build-up of populations that could be translated into warnings to farmers in exceptional circumstances. Recent El Niño events which cause prolonged hot and dry conditions, almost invariably give rise to an upsurge in infestations.

Traps with a 1:1 ethanol + methanol lure can be used to trap flying borer. Numbers caught relate quite closely to nearby total populations (Mathieu et al., 1999) so could be used to monitor borer populations. However, the traps placed outside an infested plot catch very few insects, so the power of the trap is low. This means that its use to detect borer flying into a quarantined zone is questionable. For that purpose simply checking coffee trees for infestations is probably quicker, more sensitive and cheaper. This is probably also true for routine monitoring of borer populations. Traps are now used sometimes as part of an IPM control strategy, i.e. for control rather than monitoring (Dufour and Frérot, 2008). Spectacular catches have been achieved in El Salvador (Dufour et al., 2004) and were related to measured declines in infestation. However results can be very variable. Fernandes et al (2014) deployed 900 traps in four coffee farms and achieved a 57% reduction infestation, but levels were still above an economic loss threshold. It seems likely that traps can be effective in specific conditions, when placed after early rains when borers are emerging and when there are few berries to compete for the traps’ attractiveness. However the proportion of borers trapped to total infestation levels is always low (<5%) so it is questionable whether traps are cost effective, especially since they need regular servicing to replenish the lure, clear debris etc., something that most farmers are not good at. Hence the traps need to be evaluated for specific coffee-growing conditions and results weighed against costs of the traps, their regular servicing and farmers’ willingness to service them regularly.

Related treatment support
Plantwise Factsheets for Farmers
Tangara, J. E.; CABI, 2006, Spanish language
Alanoca Murga, E.; CABI, 2006, Spanish language
Pérez, A.; Danielsen, S.; CABI, 2007, Spanish language
Pest Management Decision Guides
Magina, F.; Mbuba, A.; Mboya, A.; CABI, 2013, English language
UK, CABI; CABI, 2015, Spanish language
UK, CABI; CABI, 2016, English language
Silva, R.; Sanabria, R.; Silva, A.; Bellorin, D.; CABI, 2013, Spanish language
Rutikanga, A.; CABI, 2015, English language
External factsheets
TNAU Agritech Portal Crop Protection Factsheets, Tamil Nadu Agricultural University, English language
TNAU Agritech Portal Crop Protection Factsheets, Tamil Nadu Agricultural University, Tamil language
Pestnet Factsheets, Pestnet, English language
PlantVillage disease guide, PlantVillage, Spanish language
PlantVillage disease guide, PlantVillage, English language
Zoomed image