Trinidad. Generally, the groves on the leeward size of both
islands from north to south are affected before the mite migrates
across the island to the windward side. In other words, the mite
moves slowly against the wind. In fact, it moves locally with the
moving air mass from north to south in the northeast trades. In
St. Vincent, this was observed during a two-year period. A similar
pattern has been reported between 1982-1984 in St. Lucia. The
movement outward appears to be from mite population pressure
within the grove. A parallel situation occurred over an eight-year
period in Trinidad, which has a surface area of 1,978 sq. mi. com-
pared with 150 sq. mi. for St. Vincent and 233 sq. mi. for St.
Lucia.

Movement oftthe Coconut Mite within Groves
 The first site of invasion in the coconut grove is the female
flower on the inflorescence. From this position, the mites proceed
to multiply and inhabit the calyx of the developing fruit. Within
the year, 2-3 other inflorescences may show attack, and these
developing fruit show scarification or deformity from the secon-
dary spread. Characteristically, the following features have been
observed on newly infested trees with only one or two bunches af-
fected:
 1. When younger fruits are infested, older fruits nearby may
 be free from attack (85% frequency).
 2. When older fruits are found infested, most younger fruits
 nearby are attacked (54% frequency).
This is taken to imply that although infestation begins normally
from the young female flower, yet there cannot be an immediate
succession of infestation to developing fruits on nearby in-
florescences. The most likely methods of dispersal of the mite are
wind, insect and water.

Wind Dispersal
 A minimum period for infestation to occur from infected trees
 is not yet known. Such depends on the density of the mite
 population on the infested tree and the availability of mites for
 wind dispersal. Mites available for transmission by wind become
 available from persisting infested calyces from dropped nuts and
 mites migrating from under the calyx which remain exposed to
 the surface of the fruit.
 Persisting calyces on infested inflorescences may contain
 thousands of living female mites and eggs. These may be readily
 blown off by the wind to be deposited on neighboring flowers.
 Experimental demonstrations of wind transmission have deter-
 mined this method of dispersal.
 Migrating mites often leave the ecological niche of the calyx
 and move out, generally during darkness, onto the exposed sur-
 face of the fruit. Preliminary observations suggest overcrowding
 as a major factor under the calyx. However, this migratory
 phenomenon in the mite could be genetically predisposed and
 normal. Experimentally, such exposed mites are easily washed off
 by moving water.

 Insect Dispersal
 The tall variety of palms found in most estates in Trinidad are
 cross-pollinated, wind being an essential agent. However, the
 presence of nectaries attract several species of insects to the
 flower. Adult mites have been found on several bees which visit
 the open flower. The source of such mites would have to be an in-
 fested flower. As such, it would appear that their significance in
 economic dispersal of the mite is small.

 Water Dispersal
 The fact that moving water can wash mites off from the expos-
 ed surface of the fruit allows for some degree of water dispersal to
 nearby opened flowers during gentle rains. On the other.hand,
 rain, more than likely, would wash these exposed mites down to


the soil.

VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


MATERIALS AND METHODS
Laboratory Demonstration of Wind Dispersal
of E. guerreronis
 The experiment was carried out in a walk-in growth chamber
held at 28C and 80% 90% RH. Over the 24 h period of the ex-
periment, the chamber was in darkness. Mite infested coconut
bunches of different age groups, for separate experiments, were
hung vertically from horizontal piping in such a way that they
were in the direct path of wind created by electric fans with
variable speeds. An anemometer placed between the fan and the
bunches gave an accurate record of the simulated wind speed
reaching the bunches. Wind speed varied between 5 and 6 mph
during the experimental periods. The wind transported mites
were collected from black polythene sheets previously smeared
with vaseline. Such sheets were tacked onto the shelf and wall
space directly below and behind the bunches at a distance of 1 m.
The mites were examined and counted after every 24 hours of ex-
perimentation. The bunches were divided into two groups, those
less than 7 months and those older.
Wind Dispersal Results: The results demonstrated the
feasibility of wind transmission. Mites were lifted from all in-
fested bunches, carried by the air current and deposited on the
polythene sheets within 24 h of experimentation. Bunches of
nuts less than 7 months appeared to give a smaller number of
mites attached to the polythene sheets. No eggs were found on
the sheets from either source.

Laboratory Demonstration of Water Dispersal
of E. guerreronis
 Infested coconut bunches of varying age groups were immersed
 completely for one measured minute in a container with 50 L of
 water at 25*C. The bunches were allowed to drain completely in
 the remaining water. Twenty such infested bunches were treated
 over a period of 4 h and the water filtered through Whatman No.
 1 filter paper, Whatman No. 1, one litre at a time. The successive
 filter papers were soaked in 1 1 of water and the liquid refiltered
 through lens paper for examination.
 Results of Water Dispersal: The lens paper contained several
 hundred mites which had been washed off by the water. No eggs
 were found. The experiment demonstrated the ease with which
 mites present on the surface of the coconut fruit could be washed
 off. Such an experiment simulated the effects of heavy rainfall on
 the coconut palm.

 PRELIMINARY CONCLUSIONS
 For the most part, the coconut mite is dispersed by wind, both
 within the coconut grove, from infested fruit to female flower,
 and from grove to grove, and country to country. More than like-
 ly, heavily infested groves would contribute high densities of
 mites for wind dispersal during the dry season. During the wet
 season, mite population and dispersion would decrease due to
 heavy rains. Periods of prolonged drought or succession of short
 rainy seasons would tend to assist in mite population growth and
 wind dispersal.
 From a farmer's point of view, during long periods of heavy
 rainfall the mite would have appeared to vanish from the estate.
 What really happens, however, is that existing infested nuts
 maintain only their survival population of mites under the calyx
 until the unfavorable weather has passed. The author has en-
 countered an identical situation in Brazil where the mite popula-
 tion which increased towards the end of the dry season, seemingly
 vanished during the period of rains.

 An Approach to the Control of the Coconut Mite
 The natural habitat of E. guerreronis on coconut palms is the


protected area between the undersurface of the calyx and the ex-
ocarp of the fruit. The mite population develops for more than 12


129