and daily urinary arsenic concentration correlated at the 99%
confidence level (Wojeck et al., 1982). Franklin et al. (1981)
found a correlation between 48 h excretion of azinphosmethyl
metabolites and the amount of active ingredient sprayed. A
significant correlation could not be made, however, between 48 h
excretion and an exposure estimate. In the Franklin et al. (1981)
experiment, a fluorescent tracer had been added to the spray mix-
ture. Qualitatively, unpatched areas (face, hands, neck) also
received significant exposure, perhaps leading to a weak correla-
tion between the patch estimate and urinary metabolites.
Winterlin et al. (1984) monitored the dermal exposure of ap-
plicators, mixer-loaders, and strawberry harvesters to captain us-
ing exposure pads. Although the applicator, mixer-loader group
showed higher dermal exposure, no metabolite was detected in
their urine while harvester urine had detectable levels.
 The complexity of the urinary excretion kinetics of pesticides
may render useless any search for a simple linear correlation be-
tween estimated dermal dose and urinary metabolites. Some ex-
perimenters have investigated this area. Drevenkar et al (1979)
studied the excretion of phosalone metabolites in one volunteer.
Excretion reached a maximum in 4-5 h, but was not complete in
24 h. Funckes et al. (1963) exposed the hand and forearm of
human volunteers to 2% parathion dust. During the exposure,
the volunteers breathed pure air and placed their forearm and
hand into a plastic bag which contained the parathion. The ex-
posure took place for 2 h at various temperatures. There was an
increased excretion of paranitrophenol in urine with increasing
exposure temperature. More importantly, paranitrophenol could
still be detected in the urine 40 h later. In another human experi-
ment, Kolmodin-Hedman et al. (1983b) applied methylchloro-
phenoxy acetic acid (MCPA) to the thigh. Plasma MCPA reached
a maximum in 12 h and MCPA appeared in the urine for five
days with a maximum at about 48 h. Given orally, urinary MCPA
peaked in 1 h with about 40% of the dose excreted with in 24 h.
In a rat experiment, seven different organophosphates at two
doses were fed to two rats per compound (Bradway et al, 1977).
The rats were removed from exposure after the third day and
blood and urine collected for the next ten days. The percent of
total dose excreted in urine over ten days averaged (high and low
doses): dimethoate, 12%; dichlovors, 10%; ronnel, 11%;
dichlofenthion, 57%; carbophenothion, 66%; parathion, 40%;
and leptophos, 50%. Very little of this excretion occurred beyond
the third day after exposure. Parent compounds of ronnel,
dichlofenthion, carbophenothion, and leptophos were found in
fat on day 3 and day 8 after exposure. In another rat experiment,
animals were given dermal and intramuscular doses of azin-
phosmethyl (Franklin et al., 1983). About 78% of the dermal
dose had been excreted in urine in 24 h. Its rate of excretion
reached maximum within 8-16 h, continued at about the same
rate for another 16 h, and declined to a steady level 16 h
thereafter. There was a linear relationship between dermal dose
and urinary excretion. The intramuscular dose was excreted much
more rapidly than the dermal dose. No apparent relationship ex-
isted between the intramuscular dose and urinary excretion.
 Because these experiments illustrate the excretion differences
 between dermal, intramuscular, and oral dosing, the differences
 between compounds, and also raise questions about which urinary
 metabolite to monitor (Franklin et al., 1983), a very comprehensive
 experimental design would be necessary to correlate dermal ex-
 posure, absorption, and urinary metabolite levels. Statistical con-
 siderations centering around the large variation encountered
 among replicates probably make the economics of such an experi-
 ment in small animals, and certainly humans, prohibitive.
 Several approaches to monitoring respiratory exposure are
 available. One common method was developed by Durham and
 Wolfe (1962) and employs a respirator with the collection pads
 protected by cones from direct spray. Another method uses the


personal air sampler with a pump carried by the worker and a col-
lection device in the general breathing zone. In his review, Davis
(1980) argued in several ways that the two methods suffer from
the lack of efficacy data. That is, the trapping efficiency of a
respirator or personal air sampler is seldom checked for the collec-
tion of both aerosols and vaporized materials. We agree with this
assessment. Even when the collection efficacy is checked, the
pesticide may have been applied in a solvent, and air subsequent-
ly drawn through the device. This is imprecise at best. Collection
techniques which include both aerosols and vapor are more ac-
curate and would better reflect actual field situations. The
respirator with collection pads is the simpler method. It requires
no adjustment for breathing rate, but does require a tightly fit-
ting respirator.

Dermal Exposure Pads
 Dermal exposure pads have been constructed of a-cellulose,
cloth, polyurethane foam, and combinations of these materials.
They are designed to collect spray materials and have been used
for pesticides formulated as emulsifiable concentrates and wet-
table powders. These pads appear to work well. However, these
collection devices are almost never assessed for pesticide loss. The
question is, if the pad is left on the worker for 6 h, how much
material evaporated or degraded in those 6 h? There would appear
to be two ways to answer this question. Davis (1980) reviewed the
practice of fortifying pads with the same pesticide-water mixture as
in the field experiment. Alternatively, unfortified pads could be
applied to the worker, with some left for 1 h and then removed,
some left for 2 h and removed, and so forth. This latter test
assumes that the exposure to the pads is equal for all exposure
periods and ignores temperature effects. Fortified pads placed in an
out-of-doors holding device appears to be the better approach. By
removing and analyzing these pads at intervals, the approximate
length of time a pad should be worn by a worker in the field can be
determined. This is an important criterion toward "unques-
tionable" worker data, i.e., the ability to account for the behavior
of the pesticide on the pad during the experiment.
 Once the time of exposure and laboratory recovery studies
arecompleted, storage stability should be determined. The sim-
ple expedient of storing one or two fortified pads with each
worker's pad set will determine storage stability as extraction and
analyses proceed. Resulting recoveries also serve as a check on the
accuracy of laboratory extractions. The required number of these
fortified pads depends on the size of the experiment. The
criterion is to allow for enough measurements to statistically
validate the quality of both storage and extraction. We use a
minimum of three fortified pads per exposure day.
 Placement of the pads on the body of the subject has many
ramifications. If a total body exposure estimate is to be made, the
calculation method should be considered. Davis (1980) and
Popendorf and Leffingwell (1982) are good sources for these
methods. Pads will be placed so as to optimize the total body
estimate. Obviously, if the calculation requires a leg value, ac-
curacy would dictate placing a pad on the leg. For areas where
pads are inconveniently worn, such as the face, combinations of
shoulder and upper body pad residues may be used as approx-
imations. However, there is at least one published study in which
face exposure was estimated (Franklin et al., 1981). Development
of methods for measuring face exposure and for measuring ex-
posure to other body areas where pads are not normally placed
would be helpful to this research area.
 The location of pads on the subjects is an important considera-
 tion. For those geographical areas where temperatures can be very
 warm during a spray season, data on where the greatest exposure
 occurs to the worker's body would be very helpful. These data
 would allow development of protective suits which do not at-
 tempt total coverage and might prove more comfortable. For in-


PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY-VOL. XX


228