Photosynthetic gas exchange has been surveyed in 22 of the 23 species
currently placed in the genus Oryza and constituting the wild relative
s of cultivated rice. Unimproved, wild germplasm of a number of specie
s showed light-saturated assimilation rates in atmospheric air at leas
t as great as cultivars and elite breeding material of the cultigen O.
sativa. One of these species was O. australiensis, different accessio
ns of which were significantly (P = 0.001) superior in assimilation ra
te to the O. sativa genotypes tested, including representatives of the
Indica, Japonica and Javanica subgroups. Amongst species, assimilatio
n rate was correlated positively with light saturation and with carbox
ylation efficiency. The wild species fell into two distinct groupings
according to whether they originated from sun or shade habitats, with
the higher assimilation rates being associated with the sun species. A
ssimilation rates were also higher in diploids than in tetraploids and
this was associated with the fact that all sun species are diploids a
nd all tetraploids are shade species. The carbon dioxide compensation
concentrations ranged from 28 to 43 mu mol mol(-1) with the two lowest
values (28 and 32 mu mol mol(-1)) coming from accessions of O. rufipo
gon. The mean value for the absolute quantum yield of photosynthesis m
easured on attached leaves was 0.060. There was a large range in the a
ctivity of phosphoenolpyruvate (PEP) carboxylase with a number of spec
ies having rates several times those of C-3 species. Some species with
the highest assimilation rates were assessed for photorespiratory los
ses and these were generally around 30% and similar to O. sativa culti
vars. However, a range of O. rufipogon accessions had photorespiration
rates significantly (P = 0.01) lower than the O. sativa genotypes tes
ted. No species in the genus possessed C-4 photosynthetic metabolism t
hough some did overlap with compensation concentrations and phosphoeno
lpyruvate carboxylase activities reported for C-3-C-4 intermediate spe
cies. The potential value of wild relatives to the improvement of cult
ivated rice is discussed.