pdf, 57.71Mb - Entomological Society of Canada

The number of seeds
per head vs. head
size and the number
of U. sty/ala
larvae present (a) Methods
Cirsium vulgare (Savi) Ten.. 151
recorded for each plant, as well as the number of heads attacked, and the number of U.
stylolo larvae per gall. Attack by U. stylala enlarges the receptacle but does not affect
the base where it is attached to the involucre, so its diameter was used to compare the
size of the productive heads on the attacked and un attacked plants.
(b) Results
An average of 88% of the productive heads was galled on the attacked plants. This
percentage is higher than that obtained from the transect samples (Table 35), but as a
range of thistle sizes was selected from near the center of the colony, it was not a random
sample. The 564 galled heads contained an average of 10.3 ± 0.3 SEM larvae. This is also
slightly higher than the mean in Table 35.
Plant weight accounted for most of the variation in the number of seed heads on both
the attacked and unattacked plants (Equations 1 and 2. Table 36). The level of attack on
individual plants ranged from an average of 1. 7 larvae/head to 18.1 larvae/head, but the
attack level was not affected by plant size (Equation 9, Table 36). The slopes of the
regression lines (Equations 1 and 2) suggest that attack reduced the number of heads:
plants of 100 g had 9.5% fewer and those of 200 g had 14.6% fewer than the unattacked
plants. However, the slopes were not significantly different and the differences in the
averages were partly balanced by a slightly larger head size on the attacked plants
(average diameter 13.7 ± 0.07 mm compared with 12.8 ± 0.06 for the unattacked plants).
It is concluded that the receptacle area in the two groups of plants was practically identical.
This was in contrast to the results of the seed-head Urophora spp. on diffuse knapweed
(Harris 1980) where the number of heads and plant size were reduced by attack.
Heads from both attacked and unattacked plants were bagged shortly before the pappus
was ready to fly. Seed head size was measured as the radius of the receptacle attachment
to the involucre as previously described, and the plump seed. separated with a blower,
was counted for each head. Plump seed rather than germinated seed was used for
regression analysis: both produced similar trends but the correlations were higher with
the former. Part of the problem was a high variability of germination between heads.
Anderson (1968) reported 40% germination at 20°C although this was increased to 80%
with temperatures alternating between 20-30°C. We obtained 66.2 ± 2.0% germination
of plump seed from unattacked plants at room temperature and 52.4 ± 4.8% from
attacked plants.
(b) Results
On unattacked plants, large heads produced more plump seed than did small ones
(Equation 3, Table 36) as is normal in composites. In contrast on the attacked plants,
seed production decreased with head size (Equation 4, Table 36) although the slope and
the correlation coefficient were so low that the production of 62 seeds for an average
head of 13.7 mm diameter can be used as a constant for all heads. This means that for the
plant it is now more productive to develop several small heads rather than a few large
ones.