Original Paper Abstract: A growth analysis was conducted with 24 central Eu- ropean grass species in full daylight to test whether traits un- derlying interspecific variation in relative growth rate (RGR) are the same in full daylight as they are at lower light, and whether this depends on the ecological characteristics of the studied species, i.e., their requirements with respect to nutrient and light availability. In contrast to studies with herbaceous species at lower light, net assimilation rate (NAR) contributed more than leaf area ratio (LAR) or specific leaf area (SLA) to interspe- cific variation in RGR. This was associated with a larger inter- specific variation in NAR than found in experiments with lower light. Without the two most shade-tolerant species, however, the contribution of LAR and its components to interspecific var- iation in RGR was similar or even higher than that of NAR. Leaf dry matter content correlated negatively with RGR and was the only component of LAR contributing in a similar manner to var- iation in LAR and RGR. There was a positive correlation between NAR and biomass allocation to roots, which may be a result of nutrient-limited growth. RGR correlated negatively with bio- mass allocation to leaves. Leaf thickness did not correlate with RGR, as the positive effect of thin leaves was counterbalanced by their lower NAR. Low inherent RGR was associated with spe- cies from nutrient-poor or shady habitats. Different compo- nents constrained growth for these two groups of species, those from nutrient-poor habitats having high leaf dry matter content, while those from shady habitats had thin leaves with low NAR. Key words: Leaf area ratio, net assimilation rate, nutrient re- quirements, Poaceae, relative growth rate, shade tolerance, tis- sue mass density. Introduction The inherent relative growth rate (RGR) of a plant species is closely associated with its ecological behaviour. Fast-growing species prefer fertile and disturbed sites, whereas slow-grow- ing species are restricted to infertile or shady habitats (Grime and Hunt, 1975 [15] ; Poorter and Remkes, 1990 [27] ; Reich et al., 1998 [32] ). The slow growth of species from unproductive habi- tats is probably not a selective advantage in itself, but rather an unavoidable consequence of trade-offs necessary to increase fitness in such environments (Lambers and Poorter, 1992 [17] ). Thus, knowledge of plant characteristics associated with the variation in RGR in different environments helps us to under- stand the importance of different traits for the adaptation of species or genotypes to specific environmental conditions. There is strong evidence that, among herbaceous plants, leaf area ratio (LAR, the area of CO 2 assimilating surface per plant dry mass) is more important than net assimilation rate (NAR, the rate of total dry mass increase per leaf area and time) in de- termining interspecific variation in RGR (Lambers and Poorter, 1992 [17] ; Poorter and Van der Werf, 1998 [29] ). Of the compo- nents of LAR, specific leaf area (SLA, leaf area per unit leaf dry mass) is more closely associated with the variation in RGR than is leaf mass ratio (LMR, the proportional biomass allocation to leaf blades; Poorter and van der Werf, 1998 [29] ). Of the compo- nents of SLA, the amount of dry mass needed per unit tissue volume or fresh mass (leaf tissue mass density or leaf dry mat- ter content, respectively), at least among grasses, correlates better with RGR than does leaf thickness (Ryser and Lambers, 1995 [36] ; Ryser and Aeschlimann, 1999 [34] ). This indicates that species with a high NAR, a high biomass al- location to leaves or thin leaves may also occur under condi- tions which strongly limit growth, whereas species with low tissue densities are more strictly limited to productive condi- tions. A conclusion from this would be that environmental fac- tors associated with a low RGR constrain tissue mass density to a larger extent than they constrain NAR, LMR or leaf thick- ness. However, the interrelationships between plant traits de- pend on growth conditions. For example, variation in SLA among herbaceous seedlings at low growth irradiance (200 mmol quanta m ±2 s ±1 ) is determined mainly by leaf tissue density, but at high growth irradiance (1100 mmol quanta m ±2 s ±1 ) mainly by leaf thickness (Meziane and Shipley, 1999b [23] ). The generally observed minor contribution of NAR, LMR and leaf thickness to interspecific variation in RGR may thus be just a result of the low growth irradiance in most of the growth analyses, as, for example, the potential for a high photosyn- thetic rate cannot be fully expressed. This hypothesis was in- vestigated by Poorter and Van der Werf (1998 [29] ) in a compila- tion of published literature on herbaceous species. The authors could not find confirmation of an increased importance of NAR Interspecific Variation in RGR and the Underlying Traits among 24 Grass Species Grown in Full Daylight P. Ryser and S. Wahl Geobotanisches Institut ETH Zürich, Zürich, Switzerland Received: February 11, 2000; Accepted: May 5, 2001 Plant biol. 3 (2001) 426 ± 436  Georg Thieme Verlag Stuttgart ´ New York ISSN 1435-8603 426