The root hemiparasite M. arvense was grown at different densities with
out a host and with Hordeum vulgare as host. Increased density had no
effect on parasites without a host but significantly reduced the growt
h of parasites grown with a host. All M, arvense seedlings grown witho
ut a host died after a few weeks and attained a biomass only 60% highe
r than mean seed mass. In a second experiment M. arvense was grown wit
h Medicago sativa. The shoots of parasites and hosts were separated in
parts of the pots by an aluminium foil to exclude above-ground compet
ition. In addition, the host plants were killed by cutting off their s
hoots after 4, 7, 10 and 16 weeks of growth. The continued availabilit
y of M. sativa as a host increased the biomass of M. arvense by more t
han two orders of magnitude. Across all treatments there was a strong
allometric relationship between root mass and shoot mass of M. arvense
, but root mass was much less stimulated by the host than shoot mass.
The parasite, however, survived the killing of its host plants at all
stages of development. Parasites, whose host had been killed, continue
d to grow for several weeks but their biomass was reduced in compariso
n to parasites grown with a living host. There was no significant effe
ct of above-ground interactions on parasite biomass, but parasites gro
wn unseparated from host shoots were taller. The negative effect of th
e hemiparasite M, arvense on its host M. sativa was much stronger than
that of another host individual, and disproportionate to its size: 1
mg of Melampyrum biomass caused a reduction in host biomass of 2.4 mg.
M. arvense also reduced the total productivity per pot, indicating a
lower efficiency in resource utilization by the parasites. Root growth
of parasitized M. sativa was more strongly reduced than shoot growth,
resulting in a lower biomass allocation to roots.