Impacts of carbofuran on birds in Canada and the United State - Carbofuran and Wildlife Poisoning: Global Perspectives and Forensic Approaches - page 212

Such actions are atypical for usual planting situations where farmers only have a limited time to
plant their fi elds while soil and weather conditions are amenable.
Despite these efforts, there was still substantial bird mortality. As in the Utah study, there
were no signifi cant differences between the kill rates in the 10G and 15G plots. Carcass counts were
103 individuals of 17 species on the 69 hectares in Illinois and 29 individuals of 11 species on
the 124.5 hectares in Iowa.
There are various reasons why these fi gures are minimum estimates of mortality. First, granule
incorporation was uncharacteristically high due to the special efforts taken by FMC personnel to
incorporate the granules. Second, FMC did not select sites with high bird populations. FMC chose the
Iowa sites as being typical of the intensively cropped United States Corn Belt, and the nearest wood-
lot was more than 1.6 kilometres from any of the plots. Nesting habitat was therefore very limited
for any tree-nesting species; these would therefore not frequent these specifi c fi elds. Furthermore,
carcass searches were only conducted every three days. In cornfi eld habitats in Maryland, Balcomb
(1986) found a very high disappearance rate of planted carcasses, ranging from 62 to 92% within
the fi rst 24 hours. Also, investigators on the Iowa plots confi ned their search to the fi elds themselves
and not adjacent areas, whereas observations indicate that poisoned birds will reach shelter if they
can (Mineau and Collins 1988). There was evidence in this study and the Utah study that some poi-
soned birds left the fi eld and died elsewhere. Taken together, these factors suggest that the calculated
mortality rate based on fi nding carcasses alone is a gross underestimate of the actual kill rate.
Mineau (2005) attempted a semi-quantitative analysis of the Iowa, Illinois and Utah studies
described above. Dividing the carcass fi nding results between fi eld edge and fi eld interior and cor-
recting the results for carcass fi nding rates and scavenging, he calculated rates of primary poison-
ing to be 3.05 birds per hectare from the Iowa study and 15.9 birds per hectare for the Illinois
study. Based on the popularity of the product in the late 1970s to the mid 1980s, this represented a
mortality of 17 to 91 million songbirds in US maize alone. The same calculation was not possible
with the Utah study, but there was an estimated 52 dead horned larks per hectare. This estimate far
exceeds normal densities for this species, indicating that the birds were drawn into the fi elds from
surrounding habitat to feed on the granules.
8.2.1.4 FMC 1986b: corn, Florida and Texas, in-furrow application
This study (and the next one discussed, by Balcomb and colleagues) is interesting because granules
were applied in-furrow (at 110 g ai/100 m of furrow) rather than banded (FMC 1986b). This appli-
cation method leaves the fewest granules on the surface. However, one typically still fi nds many
granules in turn rows and elsewhere in the fi eld, particularly where there are irregularities in the soil
such as rocks or residuals from the previous crop.
It appears that the plots chosen by FMC for this study did not attract the usual guilds of fi eld edge
birds. The Texas plots were rotated from cotton, one of the most insecticide-intensive crops, and the
availability of insects for insectivorous bird species was likely very low. The Florida plots provided
very poor bird habitat and, more importantly, received 30 applications of insecticide (permethrin or
methomyl) during the two months after planting and carbofuran application. Despite a reportedly
high scavenging rate, the Texas site had an uncorrected kill rate of 0.74 birds per hectare of planted
fi eld. An 'uncorrected' kill rate is based on the raw number of carcasses found by investigators.
A 'corrected' kill rate means that the kill rate was adjusted upwards to refl ect the fi nding and scav-
enging rates. This kill rate for the Texas site is within the range established from the Iowa and Illinois
studies. Field personnel found 148 dead fi sh crows ( Corvus ossifragus ) on a Florida plot. Since
the crows' gastrointestinal tracts contained no intact carbofuran granules, FMC hypothesised that the
birds had been poisoned by other pesticides. EPA reviewers concluded that the lack of carcasses in
the Florida plots was due to very low use of the fi elds by birds (US EPA 1989).