II.3. BRASSICA CROPS (BRASSICA SPP.) – 201
reducing the chances of cross fertilisation between plants of the new field and fields
previously visited (Cresswell, 1994). Honey bees are also more efficient pollinators than
wind-borne pollen over longer distances. This is to be expected since to effect
fertilisation, wind-borne pollen must fall from the sky and land on an unfertilised stigma.
Using published measurements of pollen dispersal, Hayter and Cresswell (2006)
estimated that when bees are scarce, wind can contribute to pollination of fields 1 km
distant at a level of up to 0.3%, but only up to 0.007% when bees are abundant. However,
with a non-GM pollen source 500 m from a beehive and a GM field 800 m from the same
hive, Ramsay et al. (1999) detected some pollen grains from the GM field in largely
non-GM pollen loads. They concluded that there was either switching between fields or a
long persistence of pollen grains on the bees, or there was pollen mixing within the hive.
Ramsay et al. (1999) also found that honey bee colonies can forage up to 2 km from their
hive, indicating a potential for pollen transfer around the hive covering an area 4 km in
diameter. The maximum 4 km distance for pollen dispersal by bees corresponds closely
with the 4 km maximum for the wind-borne pollen model reported by Timmons et al.
(1996).
A number of models have been developed to predict the level of gene flow that might
be expected among B. napus fields and feral populations as well as interspecific crosses
with B. rapa (among others, Bateman, 1947a, 1947b; Lavigne et al., 1998; Colbach et al.,
2005; Klein et al., 2006; Devaux et al., 2007; Graziano Ceddia, Bartlett and Perrings,
2007). However, as many biotic and abiotic factors affect gene flow, the models currently
only provide an approximation. Further, the models have tended to focus on pollen
dispersal and its arrival on the stigma, and have paid little attention to hybridisation and
introgression.
Outcrossing in the field
Although B. napus is self-compatible (autogamous), pollen from neighbouring and
distant B. napus plants compete with the plant’s own pollen to effect fertilisation.
There are no genetic or morphological barriers to cross-pollination among B. napus
plants, so crossing between fields does occur (Becker, Damgaard and Karlsson, 1992;
Becker et al., 1991; Rakow and Woods, 1987). The outcrossing rate within fields varies
considerably, averaging between 20% and 40%, mainly depending on the environmental
conditions during flowering (see Becker, Damgaard and Karlsson, 1992 and references
therein). It is estimated that one hectare of spring oilseed rape produces 9.3 ± 0.5 kg of
pollen each 24 hours during a 17-day flowering period with B. rapa fields producing
20.2 kg/ha/day, more than twice that of B. napus (Szabo, 1985). Most of the crossing
occurs between neighbouring plants (Rakow and Woods, 1987), but long-distance pollen
transfer can occur by both wind and insects (primarily bees). The measurement of pollen
flow via wind or insects, or estimating the amount of outcrossing using male sterile or
emasculated bait plants, provides information on the potential for outcrossing; however, it
is not an accurate indicator of the actual outcrossing level that can occur between fully
fertile oilseed rape crops. In reality, male sterile plants would normally be growing in
association with fully fertile plants, so data from male sterile bait plants significantly
overestimate the level of outcrossing that would normally be expected. Ramsay,
Thompson and Squire (2003) concluded that bait plants over-estimate the outcrossing
level by at least one order of magnitude.
Numerous experiments have been undertaken in recent years to determine the
frequency of outcrossing that occurs between two populations of B. napus, with
increasing distance between the pollen donor and recipient populations. The availability