Aerial insects were sampled with a 30 cm diameter sweep net
[23] in three positions along the central strip of each plot. Samplings
were carried out between 10 A.M. and 3 P.M. under weather
conditions allowing moderate to high insect activity (i.e., sunny
days, temperature above 15 ◦C, and null to moderate wind speed).
Four net sweepings were made at each sampling point. Captured
insects were killed in situ and pooled into a single sample per
plot. Insects were taxonomically determined at order, family, and
morpho-species levels in all cases, considering that differences
between number of morpho-species and taxonomic species are
in many cases very small [24]. Abundance of each insect morphospecies
was obtained by counting the total number of individuals
captured per plot. Each insect morpho-species was classified into
herbivores and non-herbivores according to its habits and food
preferences during the crop cycle, based on anatomical characteristics
and bibliography [25]. All these insect specimens are
preserved in the authors’ insect collection in the School of Agriculture,
University of Buenos Aires. Frequency was calculated as
the percentage of plots of a particular treatment containing a given
species. Species richness was estimated as the total number of
species per plot.