Effect of grasses on native tree seedling establishment along a water stress gradient: results of forest and greenhouse experiments

Abstract

Because of anthropogenic uses, such as cattle raising or forest clearings, some native forests show 17 an early successional grass stratum that can reduce or nullify native tree regeneration. We 18 hypothesize that increasing grass coverage lowers or nullifies native seedling establishment by 19 reducing or preventing water intake by the seed or seedling. The aim of this work is to assess the 20 effect of grasses over tree seedlings under varying soil water stress levels. We performed a field 21 experiment consisted of 36 experimental exclosures in two northwest Patagonian valleys, covering 22the regional gradient of altitude, rainfall, and tree canopy openness. We also conducted a 23 greenhouse experiment, consisted of 36 experimental pots with four levels of grass coverages, 24 under three levels of water stress. Correspondence has been observed between both experiments. 25 In the field experiment we found that the probability of finding an alive tree seedling is about 2.78 26 times higher in ungrassed sites compared to grassed ones. In the greenhouse experiment we found 27 that the presence of grasses reduced on average 43% the final number of established seedlings 28 over all irrigation levels, suggesting that a water competition process is occurring. Management 29 practices that induce the invasion of grass layer would negatively impact on tree regeneration in 30 forests that have not co-evolved with large herbivores. This situation can be exacerbated in regions 31 that show a water limitation in the growing season or where climate change can intensify water 32 stress conditions.