Tree seedling responses to in situ CO2-enrichment differ among species anddepend on understorey light availability

Seedlings of six major European temperate forest tree species (Fagus sylvat ica, Acer pseudoplatanus, Quercus robur, Taxus baccata, Abies alba, Pinus s ylvestris) were exposed to 360, 500, and 660 mu L CO2 L-1 in the understore y of a 120-y-old forest over two growing seasons. Seedlings rooted in the n atural forest soil within 36 open-top chambers (12 OTCs per CO2 treatment), each with a different known quantum flux density (QFD) ranging from 0.36 t o 2.16 mol m(-2) d(-1) (= 0.8% to 4.8% of full sun). In contrast to a frequ ent assumption the natural CO2 concentration in the understorey is close to the ambient concentration in the free atmosphere during daytime. The CO2-e ffect on seedling growth differed greatly among species and was strongly co determined by microsite-specific QFD. Biomass production in the deep-shade tolerant species Fagus and Taxus increased by 73% and 37% under elevated CO 2 in low QFD microsites but was not significantly different among CO2-treat ments in high QFD microsites. The less shade-tolerant species Acer, Quercus , and Abies showed no significant response to elevated CO2 in low QFD micro sites, but increased their biomass by 39%, 25%, and 55% in high QFD microsi tes. In the shade-intolerant Pinus, seedling survival was too low for a saf e conclusion. Our data showed that the largest relative responses to increa sing CO2 occurred at a comparatively small increase from 360 to 500 mu L L- 1 with only small and non-significant changes with a further increase to 66 0 mu L L-1. Subtle shifts in the availability of light can totally reverse interspecific differences in the CO2 response. Given these different respon ses, we conclude that increasing atmospheric CO2 is likely to induce change s in species composition of temperate forests due to altered chances of rec ruitment. However, these shifts will depend on light patterns in the unders torey, and thus on canopy structure, disturbance patterns and forest manage ment.