Page 18 - Volume 2,Issue 2, July 2009

Albano

et al.

62

tiges of the biofungicide presence (Table 2).

Spearman correlation analyses per-

formed on data obtained from the bumble

bee assay showed a positive correlation be-

tween the number of CFU per bumble bee

forager and the number of CFU per corre-

spondent visited flower (r

s

= 0.45; n= 28; P=

0.01). A positive correlation was also found

between the total length of visits and the

number of CFU per flower (r

s

= 0.33; n= 42;

P= 0.03).

Discussion and conclusions

Inoculum picked up by pollinators exit-

ing the hive-mounted dispenser

The average load of inoculum carried by

each honey bee exiting the hives-mount-

ed dispensers was lower than that men-

tioned in other works for

T. harzianum

(17,

19, 27). Although the sample size was not

very large, the amount of inoculum on bum-

ble bees exiting the hive was comparable

to that obtained in other work with

Bom-

bus terrestris

using another dispenser mod-

el [OP-dispenser- 4.3 x 10

4

CFU per bumble

bee (21)].

However, the results of the studies men-

tioned above cannot be directly compared

to ours, since (i) most of the studies used dis-

penser models different from ours (ii) some

of them do not mention the quantity of in-

oculum that is placed in the dispenser, and

(iii) some do not indicate the length of time

between the filling of the dispenser with

biofungicide and insects capture. Moreover,

in other studies (17) small nuclear honey bee

hives were used while in the present study

we used standard hives. As mentioned by

Bilu

et al.

(5), some of these factors may in-

fluence the dispenser performance, and,

therefore, in case these factors are not con-

sidered and measured, correct comparison

of the effectiveness between distinct dis-

penser types is not possible.

Bee vectoring dynamics in open field

and greenhouse conditions

According to the results, the load of in-

oculum on the bees while exiting from hives

dispensers is apparently more rapidly lost to

the environment when the vector is a hon-

ey bee than when it is a bumble bee. In the

honey bee assay the inoculum could have

been lost in several ways: (i) losses during

flight activity and grooming behaviour, (ii)

transfer of inoculum to the flowers of the tar-

get crop, or (iii) transfer of inoculum to com-

petitor flowers. Since the honey bee assay

was carried out in open field conditions and

within a more extensive intervention area,

honey bees probably lost the inoculum de-

posited on their bodies more rapidly as they

were successively transferring it to the flow-

ers throughout their foraging trip. As a re-

sult, only 40% of the honey bees captured

during their foraging activity had vestiges

of the inoculum on their body, and quanti-

ties were much smaller than those found on

the bumble bee foragers. Probably for the

same reason, only 33.33% of flowers freely

exposed to honey bee visits were shown to

carry the inoculum. The high percentage of

non-colonised flowers in the honey bee as-

say could be due to the lack of honey bee

visits or because visits were depleted of in-

oculum. Yu and Sutton (31) point out the

same explanation for the absence of prop-

agules of

Gliocladium roseum

that was re-

corded in 5-20% of the flowers exposed to

pollinator visits.

In both assays the inoculum density that

was deposited on flowers exposed to polli-

nators was very variable. The large variabili-

ty in the number of deposited CFU per flow-

er, reflected in high standard deviations,

imposes some caution in this data interpre-

tation. Other studies also detected this vari-

ability and several factors were pointed out

for explaining that: heterogeneity of the in-

oculum loads in bees, bee vectoring dynam-

ics, orientation of flowers, redistribution of

inocula by the bees, other organisms, air

currents, rain and other factors (2, 31) and, fi-

nally, microclimatic conditions that could in-

fluence the attractiveness of flowers (27).

The mean density of inoculum detected

on freely exposed flowers, in the assay with

honey bees, was considerably lower than