Response to below-ground interaction only
In our study, at the first stage, plants were not large and branchy enough to interfere each other above ground, thus they should have mainly experienced below-ground interactions, with negligible above-ground interactions. During this period, we found no response in total mass and two kinds of responses in root: shoot ratio under two soil conditions respectively (Fig. 1 and Table 3a): (1) increased root: shoot ratio with higher densities in fertile soil; (2) canalized root: shoot ratio in infertile soil. When total mass did not respond to interaction, we consider it experienced low level of interaction. In this scenario, response of root allocation to below-ground interaction varied with soil conditions.
Plant-plant interaction can produce either competitive or facilitative results (Callaway 2007, Callaway et al. 1997), depending on plant growth stages or stress levels (Callaway 1995, Callaway et al.2000, Callaway and Walker 1997). Soil conditions may affect root responses via modifying the intensity of below-ground interaction. Low to moderate level of interaction is more likely to produce facilitative effects (Casper and Jackson 1997), as in studies of Gersani et al. and O’Brien et al. (Gersani et al. 2001, O’Brien et al. 2005). It also implied facilitative effects of below-ground interaction can only be detected when above-ground interaction is negligible, which usually occurs at early stage of plant growth and when resources are abundant. Increased root allocation in response to density in fertile soil at 30 days of plant growth in this study suggested a facilitative effect of below-ground interaction. By contrast, no response in root allocation in infertile soil suggested below-ground interactions were not strong enough to affect root allocation, or otherwise facilitative effect was counteracted by competition, which was aggravated due to resource deficiency. As resource availability decreased, intensity of below-ground competition generally increases (Cahill 1999, Casper and Jackson 1997, Schenk 2006). We also found a decrease in root:shoot ratio by intraspecific interaction in dry soil, though response of total mass indicated facilitation. It suggested water deficiency may intensify below-ground competition, leading to decreased root allocation, even when the overall result of interaction is facilitative.
No significant facilitative effects on total mass may be explained by a shift of plant strategy in dealing with environmental stress. For example, under plant-plant interaction, plants may prefer to alter biomass allocation at first, then alter total mass as interaction intensity increases. As interaction intensity continued to increase, total mass can decreased at high level of interaction, when root:shoot ratio responded to interaction (Li et al. 2016) or kept stable (Cahill 2003). As plants grew larger, the strength of competition first increased then decreased (Hutchings and Budd 1981). Plants in fertile soil first altered biomass allocation then kept it stable, due to the attenuation of interaction intensity (Wang et al. 2017). However, we do not know why root allocation increased with interaction in this situation (Li et al. 2016). Of course, the intensity of below-ground interaction can also be affected by a variety of other factors, such as pot space, number of neighbors (or growing density), sizes of species (due to age or genetic nature), thus we also referred to information on these factors.