RESULTS
 In the first chemical trial (PO 86) it is shown that among the
chemical treatments, application of Endrin significantly prevents
the spread of hartrot disease (Table 1).

TABLE 1. Incidence of hairot in coconut palms in trial one (PO 86) in different
chemical treatments.


 Treatments lumber treated % dead

 Bndrtn (Insecticide) 96 12
 Furadan (lematictde) 96 41
 Chestnut compound (Fungictde) 92 35
 Control 88 44


 ++ Significant at 1% level L.S.D. = 21.5 for treatments


 In trial 2 (VP 548), the Endrin treatment again proves to be
highly effective in controlling hartrot disease. In both the above
trials, Furadan, Chestnut compound and Sevin are not effective
and are similar to the controls (Table 2).

TABLE 2. Incidence of hartrot in coconut palms in trial two (VP 548) in Endrin
and Sevin treated plots.



 Treatments lumber treated % dead

 Endrin 43 7"'
 Sevin 44 64
 Control 40 70


 ++ Significant at 0.1% level L.S.D. 50 for treatments


 The results of the first weed control trial (PO 86) indicate that
 complete weed control in coconut plots is also very effective in
 checking the disease (Table 3).


TABLE 3. Incidence of hartrot in coconut palms in weed control trial one (PO 86).



 Treatments lumber treated % dead

 Complete weed control 64 7"
 1o weed control 60 62


 + Significant at 1% level L.S.D. = 2.5 for treatments



VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


 In the second weed control trial (VP 599), with similar
treatments as above, the results indicate that complete weed con-
trol in coconut gardens significantly decreases the incidence of
hartrot (Table 4).

TABLE 4. Incidence of hartrot in coconut palms in weed control trial two (VP 599).



 Treatments lumber treated % dead

 Complete weed control 52 36.5+
 No weed control 54 51.4


 + Significant at 5% level L.S.D. = 11.1 for treatments


DISCUSSION
 The results of several trials to control hartrot or fatal wilt of
coconut indicate the possibility of effective control of this malady
by applying certain chemical and cultural practices.
 Among the chemicals, the insecticide Endrin significantly
reduced the incidence of the disease. This may be due to the type
of chemical, for Endrin is a chlorinated hydrocarbon, while Sevin
is carbamate insecticide. Furadan is also a carbamate compound
(Thomson, 1977) but used as a nematicide, and Chestnut com-
pound is fungicide. Endrin is more persistent in its effects than
Sevin and Furadan and may, therefore, better eliminate the vec-
tor or vectors involved in transmission of the disease. Further, En-
drin has been applied at more frequent intervals, i.e., once in
two months, Sevin once in three months and Furadan once in
four months. The ineffectiveness of Furadan and Chestnut com-
pound may be due to non-involvement of nematodes or fungi in
the cause of this disease. The vector, or vectors, and the infective
stage-like nymph may be confined to subterranean feeding on
the roots of grasses and coconut (Eden-Green, 1978), and thus
the frequent soil application of Endrin may have been more ef-
fective than Sevin applied on the crown.
 This finding is quite in agreement with that of Lopez et al.
(1975) for controlling 'Marchitez sorpresiva,' a wilt disease of oil
palm in Colombia associated with Phytomonas sp., a sieve
restricted protozoan flagellate, by Endrin application. Further,
Desmier de Chenon (1984), who suspects Lincus StMl. as the
possible vector, mentioned that this species does not appear
capable of flying and is mainly terricolous. Thus, when both
nymphae and adults move down the stem after the appearance of
the first disease symptom on the palm, they pass through the cir-
cle treated with Endrin and the insecticide becomes effective
against them.
 Similarly, weed control also is effective in controlling the
disease. The role of weeds in the etiology of the disease is not
clear yet. They may serve as host plants for the pathogen, pro-
bably Phytomonas staheli (Kastelein et al., 1984). Weeds may
provide the natural habitat for the vector or vectors (Griffith,
1980). The presence of weeds may have some beneficial
physiological effects on the flagellates in the palms, especially in
the roots. Thus, by removing these, better control of the disease
is obtained.
 In practical application of these results, it has been found that
a combination of Endrin application and weed control more or
less permanently stops this disease in areas where it was endemic.
In this case, Endrin is applied only to surrounding trees of an in-


43