Tanzania

Discussion Group

Ayuda de traducción - Aide de traduction - Assistenza di traduzione - Übersetzungsunterstützung

Last modified: 3 December 2005

Contacts or sources of information - Agricultural Research Institute, Mikocheni, Dar es Salaam

Occurrence - The disease seems to have occurred in Tanzania many years ago, even before the 19th century. The first reliable reference to LD in the country was by Stein in 1905 during German colonial rule when the disease was called "Herzfäule" (heartrot).

Spread - Most serious LD outbreaks occurred in Kilwa, Rufiji and Kisarawe districts in the 1940's, but did not cause significant losses before 1960's. (Schuiling et al.,1992a).  It occurs in virtually the whole of coastal belt of mainland Tanzania, and it was first noted on the island of Mafia some 10 years ago, but the island of Zanzibar is still free from infection (other Indian ocean coconut producing islands like the Seychelles, the Comoros, Mauritius, Madagascar &Reunion are also free of the disease).

Currently active areas - Lethal disease of coconut is still prevailing in the coconut growing districts along the coastal belt of Tanzania where it is estimated to have killed 38-40% of the coconut palms since 1960's.  Striking differences in disease incidence exist among the affected areas with 56% of the palms killed in the southern regions, compared to only 8.5% in the north (Schuiling et al.. (1992a). 

Suspected new outbreaks - Possible causes for these differences are thought to be genetic variability within the palm populations, different insect vectors, or different strains of the pathogen.

Other palm/plant hosts

New hosts, new vectors, new strains or suspected loss of resistance - confirmed or unconfirmed

Research projects - High priority has been put on intensifying research to identify and control LY-type diseases of coconut with emphasis on:

Since 1981 disease resistance trials have been setup at 6 sites where disease has been rampant using eleven dwarf, fifteen tall, thirty sub-populations of the East African Tall (EAT) and twenty two F1 crosses (Schuiling et al., 1992a). The original selection of the introduced coconut varieties and hybrids planted from 1981 to 1983 was based on their production records, on results from resistance trials in Jamaica, and on the composition of the agronomy trials planned for Tanzania.

Palms that survived the disease after more than 10 years of field exposure were used in a crossing programme to produce breeders' test materials (BTM) for further testing. These were planted from 1994 to 1996 at a spacing of 10 x 14 m at Kifumangao (high disease incidence), Chambezi (moderate disease pressure), Mkuranga (disease free) and in a number of on-farm trials.

Of the imported dwarfs planted under high disease pressure, Malayan Red Dwarf (MRD) and Sri Lankan Green Dwarf (SGD) showed the best survivor rate (15%), followed by MYD with 5%. The two locally collected dwarfs, Pemba Red Dwarf (PRD) and Malayan Green Dwarf (MGD), and the semi-tall King Coconut that were planted between 1985 and 1988) appear to be promising with 36%, 63% and 30%, respectively still surviving at Kifumangao.  The Cameroon Red Dwarf (CRD) and Equatorial Guinea Dwarf (EGD) appear to be the most susceptible dwarf varieties because all test palms have been killed by LD under high disease pressure.

The introduced tall coconut populations and hybrids succumbed to LD earlier than the dwarf varieties. All (100%) of West African Tall (WAT), Polynesia Tall (PYT) and, Malayan Tall (MLT) and more than 98% of Rennell Tall (RLT) and Cambodia Tall (CBT) palms have been killed by LD at Kifumangao.  The disease incidence in the Vanuatu Tall (VTT) and Thailand Tall (THT) populations which, four years ago, had been thought to be promising (Kullaya et al., 1995), has now increased, and only 11 and 8.5% of the palms are still alive.  On the other hand, LD losses in the tall populations (Kar Kar Tall, KKT; Panama Tall, PNT; Tagnanan Tall, TGT and Rotuma Tall, ROT) planted in 1989 at Kifumangao range between 55.5% and 88.0%.

Schuiling et al. (1992b) reported that resistance in the Vanuatu Tall crosses was marginally better than in all other crosses.  However, now, seven years later, all the EGD x VTT and MYD x VTT palms planted at Kifumangao have succumbed to LD.  There are only two MRD crosses that still appear promising at Kifumangao.  These are MRD x EAT with about 26%% and MRD x MLT with 33% of the palms planted in 1985 still remaining. However, these levels of resistance are still insufficient for commercial exploitation, and it is difficult to predict what the situation will be like after another few years.

The majority of different sub-populations of the local EAT appear to be very susceptible to LD under high disease pressure.  The EAT sub-population from the Livestock Breeding Station (LBS), which has so far shown only 16.3% LD incidence after nineteen years under low disease incidence at Pongwe, did not perform so well at Kifumangao where 70% of the palms planted 10 years ago have been killed by the disease.  However, the collections from Mwambani, Boza, Kimanga and Vuo in Tanga, as well as those from Ng'apa, Mtoni and Songomnara in Lindi region suffered lower losses of between 36% and 56% at Kifumangao, and the disease progressed at a much lower rate compared to the other EAT sub-populations.  Three sub-populations from Kenya and one from Kilwa, which were planted between 1990 and 1992 at Kifumangao, also seem promising, but it is not certain whether all these will maintain high levels of resistance over time.

Under low to moderate disease pressure at Pongwe and Chambezi , the performance of most EAT sub-populations was similar to the trend observed at Kifumangao, although losses to disease were much lower at these sites.  Sub-populations which performed well at Kifumangao, did equally well at Chambezi and Pongwe.  The EAT sub-population from Chambezi also seems resistant at these latter sites.  Losses in this collection over a period of eighteen years is about 38.3%, an increase of only 1% during the past four years.

Results on the performance of the breeders' test materials (BTM), which were produced by crossing individual palms of different varieties that had survived the disease after more than ten years, show that overall, the RLT and VTT crosses appear to die much faster than the EAT crosses, but there are no indications to suggest that any of the BTM would appear to be highly resistant under high disease pressure. 

These results show that varieties resistant to Lethal Yellowing (LY) in Jamaica, such as MYD or those with intermediate resistance like the Panama Tall and Cambodia Tall (Been, 1981), are highly susceptible to high LD pressure in Tanzania.  Similarly, these results are in contrast to results reported by Sangare et al. (1992) and Batugal P.A. and V. Ramanatha Rao (1998) which indicated that the SGD, VTT and MYDxVTT were showing some degree of tolerance to the Cape St. Paul Wilt Disease in Ghana.  We think this could partly be due to the fact that the phytoplasmas associated with LY-type diseases in the Caribbean and West Africa are similar, but not identical to LD-phytoplasma in Tanzania (Harrison et al. (1994).

The pathogen - Phytoplasma has been confirmed to be the causal pathogen for LD in all the three East African countries of Kenya, Tanzania and Mozambique.  However, the striking differences in disease incidence between different affected districts within a country and between countries have led to the conclusion that either genetic variability within palm populations, different vectors, or different strains of the pathogen are responsible.  An interaction of either two or more of these factors with the environment could also be a possible cause.  Our main concern is to determine the cause of these outstanding differences in disease incidence, because this could enable us to better understand the mechanism of spread and devise suitable control strategies.

By 1992, lack of a quick, specific, and sensitive method for detection of the LD pathogen in the field was identified as the main factor limiting understanding of the etiology and epidemiology of LD in Tanzania.  Prior to this, routine identification was based on staining of razor-cut sections of meristematic tissue from LD-infected palms with a fluorochrome, 4'-6-diamidino 2-phenylindoe 2HCl (DAPI) and linking the fluorescence in tissues to the presence of phytoplasmas in phloem shown by electron microscopy (Schuiling and Mpunami, 1990).  Though sensitive, these techniques had the disadvantage in that DAPI staining is not specific for phytoplasmas alone, and the electron microscope requires a lot of time to prepare samples and is very expensive to maintain.  Consequently, the molecular approach to pathogen detection was adopted. 

This was implemented as part of the EU-funded collaborative project entitled "Investigations on the etiology and control of Lethal yellowing-type diseases in Africa" that was coordinated by NRI (1992 to 1997).

Sensitive disease diagnostic techniques including DNA probes and PCR primers were developed and evaluated in advanced laboratories of collaborating partners (Eden-Green et al., 1994-97) and introduced to MARI for disease detection and other applied studies.  By use of these techniques, infection could be readily detected in tissues of LD affected palms, without having to destroy the whole palm.  Information about the distribution of the pathogen in affected palms was also obtained, and incubating infections could be detected up to two months prior to the onset of disease (Mpunami, 1997).  This information is useful for understanding disease epidemiology.

By use of DNA probes, PRC, RFLP and rDNA sequencing, studies on genetic relationships between phytoplasmas associated with the LY-type diseases in Africa and elsewhere revealed that the LY-type diseases in the Caribbean and Florida, West Africa and East Africa are similar but not identical (Harrison et al., 1994).  Similarly, the use of these techniques confirmed the causal pathogens for the LYD in Tanzania, Kenya and Mozambique to be phytoplasma.  However, the phytoplasmas in Kenya and Tanzania were indistinguishable although both were distinct from the isolate from Mozambique, which was shown to be more similar to the West African strain (Mpunami et al., 1999).   Furthermore, genetic comparisons were made between several Tanzanian phytoplasma isolates from regions of low, medium and high disease incidence, but no strain difference was detected.

Failure to detect strain differences between the phytoplasmas from Tanzania is not totally conclusive since comparison was based on a 500 base pair product of the phytoplasma ribosomal DNA amplified from LD infected specimens.  This product was based on the sequences from the conserved region of the 16S rRNA gene, hence the probability of detecting differences is low.  Additional tests using primers whose sequences are based on unique regions of the phytoplasma, like the 16S-23S spacer region on ribosomal RNA, or LD phytoplasma genomic DNA are required to confirm the previous findings. This is one area that still requires attention, and is considered a priority research topic for the near future.

The possibility of phytoplasmas other than the LYD phytoplasma being present in the coconut population as recently reported by Dr. Harrison (the Mexican case), cannot be overruled for East Africa.  In Tanzania for instance, we have been observing occasional palms dying from atypical LD symptoms, that we usually ascribe to either nutritional disorders or drought effects.  In some years (1996 and 1998) a substantial number of palms was lost to the disorder in specific locations, but the malady eventually died out.  In the light of such intriguing research findings, we will have second thoughts over such atypical LD symptoms and investigate them more thoroughly in future.

Vectors studies - The search for vectors of LD has been another big challenge.  Earlier insect studies identified more then 30 different plant hopper and leafhopper species on coconut palms, but the LY vector, Myndus crudus was not found in Tanzania.  Several of the predominant insect species were evaluated in transmission trials, but none could transmit the disease during a four year trial period (Anonymous, 1984-1989).  Transmission tests were therefore abandoned, until after development of molecular techniques.  These techniques have been utilized in the search of insect vectors for LD. More than 5000 homopterous insects were screened by polymerase chain reaction (PCR) and phytoplasma detected in a few individuals.  This made it possible to narrow down the list of potential vectors to only two species.

Additional tests have confirmed that these two species can ingest the phytoplasma from infected palms. Future transmission tests are necessary to confirm whether these insect species are LD vectors.

Rehabilitation/replanting programmes - The big differences in performance of the different EAT sub-populations at different locations suggest that recommendations for planting resistant varieties should be location specific.  Varieties which may be very susceptible to LD under conditions of high incidence but that tolerate infection where disease incidence is low, should not be discarded as susceptible, but can be planted in the low incidence areas if more resistant material is not available.

The differences observed among the East African Tall sub-populations with respect to LD resistance can partly be due to differences in the genetic make up of these sub-populations.  However, this may not be the only reason because some populations, which resist infection at one locality, succumb completely at another. Other factors like rainfall, ground water, vegetation, insect vector and races of the pathogen may also contribute significantly to the differences.  Plans are underway to study and confirm if there are more than one LD phytoplasma strains in Tanzania. 

The relatively high LD incidence in the BTM suggests that some of the palms that were used to produce these crosses after having survived the disease for more than ten years were not resistant, but had merely escaped infection.  And this renders the conventional strategy of breeding for LD resistance by screening different coconut populations in the field and then using the survivors in a cross breeding programme not always effective.

There are a number of factors that render breeding for resistance/tolerance to LD less effective. These include:

Economic importance/threat - In Tanzania, 22 million coconut trees (Cocos nucifera) are planted along the mainland coast and in the islands of Zanzibar, Pemba and Mafia. 


This page is under continuous review. If you have an idea or an opinion to improve the contents of the page or the site, tell the editor. If you disagree with anything, say so. If you don't see or get a satisfactory response in a reasonable time contact other participants. At all time keep in contact with other individuals; this site is not a substitute for person to person contact.


Go back one page

Go to CICLY main page

Return to Top

Join the CICLY discussion group

Powered by groups.yahoo.com