stance, the exposure of an applicator or mixer-loader on the backs
of the arms and legs is not known. Whether the lower arms
receive more exposure than the upper arms is seldom monitored.
In many of our experiments, the lower arms received significantly
more exposure than the upper arms, but the generality of this
result is unconfirmed. If these data were available, comfortable
protective suits utilizing relatively open mesh areas might be cer-
tifiably protective at this time. And, actual exposure estimates
might, in fact, be reduced through their use. Certainly, taking
these additional data adds extra work and expense to an exposure
study, but the long-term benefits might be substantial.

Mixer-loader and Applicators vs. Harvesters
 Applicators and mixer-loaders certainly receive different levels
 and types of exposure than do harvesters. The mixer-loaders are
 exposed to concentrate a well as drift; applicators are primarily
 exposed to drift and the tank mixed material. The reentering
 harvester is exposed to a presumably homogeneous application of
 pesticide on fruit, leaf, and soil surfaces. Both groups may also be
 exposed by working on or around contaminated machinery and
 in or around contaminated loading areas.
 For harvesters, different sources of variation exist, but these
 may not be extreme. Theoretically, harvesters are exposed only to
 the residues remaining in the field, and most heavily when work-
 ing in that field. The experiment appears simple. Pads are placed
 on the body of the harvester at various locations, the residues on
 leaves, fruit, and soil are measured, and the appropriate correla-
 tions are made.
 For the experimenter, however, there are all sorts of possible
 constructions. Where should the pads be located? Should they be
 placed inside or outside the clothing? Will clothing chosen by the
 worker suffice or should standard clothing be issued? Was the
 field sprayed during 1 h, 1 day, or 3 days? If the spraying took
 longer than one day, where should the workers start working?
 Will they overlap sprayed sections as the work progresses? How
 many daily residue samples should be taken as a consequence?
 Should pads with a surgical gauze front be used or would.
 polyurethane foam be satisfactory? How should these pads be
 assessed for residue loss? How long should the worker wear the
 pads? Is the pesticide converted in the field into a toxicologically
 important metabolite? Can it be extracted and analyzed? How
 should the urine be collected: 24-h urines or a timed grab sam-
 ple? And finally, how many sampling periods (days) should the
 experiment entail in order to make the results statistically useful?
 We offer these suggestions: for the initial experiment, the pads
 should be placed inside the clothing for lower and upper arm,
 chest, back, shoulders and shin exposure. For the upper body,
 the pads can be conveniently pinned inside an issued shirt. They
 can also be pinned inside the pants, but it should be noted
 whether the worker wears the same pants each day. For later ex-
 periments, a reduction in the number of pads may be possible. It
 is a mistake, however, to simply observe a harvest operation and
 decide a prior that only leg patches are necessary.
 The time the pesticide application was made is important for
 several reasons. If the purpose of the experiment is to correlate
 field residues with worker exposure, then knowing the pesticide
 used and its application date can be crucial. An experiment of
 this type should begin at the legal reentry time and extend
 through at least two pesticide "half-lives." This insures the validi-
 ty of the correlation of residues with exposure because a broad
 range of both has been utilized. This sampling time may last one
 week or longer. The area to be sprayed may be large. We have,
 for instance, used three spray machines simultaneously in order
 to assure a 1-day application. All harvesters are then exposed to


the same daily residue over the sampling period.
 When a "blind" harvester experiment is conducted and the ap-
 plication was made over a few days, the number of each type of

VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


residue sample should be doubled and taken from where the
harvesters are working that particular day. This will help with the
overlap problem. Even if the experiment is only a 1- or 2-day ex-
periment, reentry should commence as soon as possible after ap-
plication. This assures some results at least, from an analytical
standpoint, that may fit an existing model. If the workers reenter
a field after ten days and the analytical chemist detects no resi-
dues because of low levels, little has been accomplished except
the expense of time and money.
 The most commonly used exposure pad for monitoring
harvester exposures is faced with surgical gauze, backed with
a-cellulose and glassine weighing paper. This pad has proven un-
comfortable for the worker, difficult to attach, and takes time to
prepare. We know of one instance where polyurethane foam pads
were used (Brady, E., personal communication). They were con-
venient and may be efficient. However, there is no good method
for assessing the residue collecting efficiency of these devices for a
harvester exposure experiment. In spite of years of research in this
area, the transfer process of field surface residues to the body of
the harvester is not known with certainty. Probably foliar and
field dust are primarily involved. How, then, is the efficiency of a
collection device for a harvesting operation measured? The resear-
cher is presently confined to the application of pesticide-laden
dust or a pesticide solution to the exposure pad, followed by a
disappearance study. Although the disappearance study may in-
dicate a 50% loss from a pad in, say, 2 h, the pads may have to be
worn longer. The reason is because exposure for a harvester is
generally low and enough residue must be collected for analyses.
We attach the pads just before workers enter the field in the mor-
ning and remove them 4-5 h later at the noon break. The amount
detected on the pads can be corrected according to the disappear-
ance experiment, but this correction is not entirely reliable since
the pesticide may disappear at a different rate when attached to
dust, as may have been the case in the field.
 The presence of a toxic metabolite on foliage or in soil and the
 possible consequence to harvesters have been reviewed (Gunther
 et al., 1977; Nigg and Stamper, 1982). We mention this con-
 sideration because of its importance to harvesters and because the
 urine analyses may have to account for the excretion products of
 these metabolites. Urine collection from harvesters is not dif-
 ficult. A timed grab sample from the start of work until the noon
 break has provided excellent correlations between residue levels
 on foliage and urinary metabolites in harvester (Nigg et al.,
 1984). We attribute this to the greater likelihood of homogene-
 ous exposure to a harvester than to an applicator or mixer-loader.
 Worker Methods and Work Rates
 For the applicator, mixer-loader group the type of equipment
 used, the number of tanks applied per unit time, the concentra-
 tion of the tank mix, and the loading method all affect the ex-
 posure process. This has been known for years and is described in
 many published reports (Davis, 1980; Nigg et al., 1984).
 For harvesters, there are only a few field experiments described
 in the literature. The crop harvesting method has been studied and
 some reports exist which can be compared; What seems apparent
 from these reports is that the exposure process is similar for the
 harvesting of such tree fruits as citrus and apples. At least, the pro-
 portion of harvester exposure to pesticide on the leaf surface is the
 same. For other types of crops this proportion may be different.
 Regardless of crop type, the work rate appears to be related to
 exposure. This means that the number of boxes picked, crates
 loaded, tassels removed, etc., is confounded with residue levels in
 affecting exposure. The individual worker's production delimits
 the contact with the plant, a subject which has been studied us-
 ing movies and time analysis (Wicker and Guthrie, 1980), and
 estimated with surveys (Wicker et al., 1980). Therefore, work rate


data should be gathered for each subject; it may explain variation
in urinary or dermal exposure unaccounted for by field residues.


229