Our study demonstrated the complex interactions among N deposition, precipitation amount and precipitation frequency in affecting the growth of R. pseudoacacia. Growth of R. pseudoacacia was observed to be promoted by increased precipitation amount, increased N deposition, as well as high precipitation frequency under medium or high precipitation level. Water stress made R. pseudoacacia more sensitive to be promoted by N deposition.
Increased precipitation amount promoted growth of R. pseudoacacia, so does high precipitation frequency except for under low precipitation amount
Higher precipitation amounts significantly promoted the growth of R. pseudoacacia under both watering frequencies (Fig. 3). This was due to the water stress, first, inhibited leaf expansion (Hsiao 1973), and then photosynthesis (Tezara et al. 1999). With a higher precipitation amount, plants had a significantly lower root to shoot ratio under both F1 and F2, as have demonstrated in optimal partitioning theory, that plants allocate more resource to organ that acquires the most limiting resource (Ledo et al. 2018).
High precipitation frequency also significantly promoted plant total biomass and other plant growth traits with a relatively high precipitation amount (W2 and W3 in Fig. 3). So, as plants tend to increase their net photosynthesis rate after each precipitation (Yang et al. 2014), we suspect that R. pseudoacacia may have accumulated more biomass when plants received more times of precipitation. The above may also explain why plants had more growth under high precipitation frequency compared with low precipitation frequency with the same precipitation amount. Thus for R. pseudoacacia, high precipitation frequency promoted plant growth more effectively than a large precipitation event followed by a few dry days, with the same or even lower precipitation amount. With the same precipitation amount, plants received higher precipitation frequency would have better growth (Coe and Sparks 2014). On the other hand, plants responded similarly under both precipitation frequencies with low precipitation amount, as plants may keep high water-use efficiency even after precipitation due to the low precipitation level. So, we suggest that global climate change, especially extremely low precipitation frequency, regardless of the amount, would lead to a lower productivity of R. pseudoacacia forests, and how the extreme precipitation would affect the invasion of R. pseudoacacia still need further study. Meanwhile, precipitation frequency change didn’t affect plant growth under drought conditions thus won’t affect the invasion of R. pseudoacacia especially during the seedling stage.
High N deposition promoted the growth of R. pseudoacacia, and water stressed plants were more sensitive to N deposition
Robinia psueduacacia had significantly higher total biomass under N4 (33% and 45% higher respectively, Fig. 1B) than under N1 and N2 after combining different precipitation regimes. This may be different from previous studies on the effect of N fertiliser on R. psuedoacacia, in which biomass of R. pseudoacacia was not significantly affected by N fertiliser (Wang et al. 2021a; Wang et al. 2018b). However, unlike previous studies that are watered in accordance with plants requirement, our plants were supplied with various precipitation regimes, and high precipitation may have magnified the effect of N to some extent. In the present experiment, like many non-fixing species such as Calluna vulgaris and Senna surattensis (Carroll et al. 1999; Wang et al. 2018b), the growth of R. pseduoacacia was significantly promoted by N deposition.
Compared with the soil before starting experiment (87.7 mg kg−1), soil after plant growth still contained about 60 - 70 mg kg−1 plant-available N on average (Fig. 2D). Like invasive species Bromus tectorum, R. pseduoacacia also lowered the soil N during growth (Evans et al. 2001). Though R. pseduoacacia is an N2-fixing species, its N2-fixation still cannot compensate the N loss in soil and absorbing N from soil is still needed, which may in turn lower the risk of other invasions due to a lower soil N (Perry et al. 2010).
Under low precipitation amount, plants received higher N deposition had significantly higher chlorophyll concentration and higher leaf N:P ratio than received lower N deposition, and a balanced N:P ratio around 15 (Koerselman and Meuleman 1996) under high N deposition. This stronger response of leaf traits under low precipitation amount may be explained by the fact that water availability affect ion mobility in soil, thus plants growing in dry places are more likely to be nutrient limited than in wet places, and nutrient supply under dry area are more effectively promoting plant growth (Chapin 1991; Kreuzwieser and Gessler 2010). So, plants under low precipitation amount were more sensitive to N deposition.
Nodule mass fraction of R. pseudoacacia significantly decreased with more N deposition under all precipitation amounts. Nitrogen fixation, which costs more energy compared with absorbing N from soil (Lambers et al. 2002), is only favourable when absorbing N from soil is becoming more expensive for satisfying the plant N needs (Rastetter et al. 2001); nitrate in soil can also inhibit nodulation and N2 fixation (Cowie et al. 1990; Hansen et al. 1992). In condition of facultative N2 fixation, plant allocation to bacterial symbionts is controlled by the cost of fixation compared with directly absorbing N with root (Menge et al. 2009). So, when more N was added in our experiment, plants allocated less biomass to nodule thus decreased N2-fixation rate. As to the effects of precipitation amount on N2 fixation, nodule mass fraction of R. pseudoacacia was significantly higher at a high precipitation amount (Fig. 3F). This might be because increased precipitation amount increased plant relative growth rate (Hsiao 1973; Tezara et al. 1999) thus increased the demand of N and the investment of nodule biomass. Plants under low precipitation amount had more dramatic decrease in nodule ratio as N deposition increased compared with under medium and high precipitation amount, so high N deposition may cause a stronger inhibition of nodule growth and N2 fixation under dry condition than plants under well-watered condition.
Phosphorus-utilisation efficiency was affected by both N deposition and precipitation frequency
Leaf P concentration in our experiment significantly increased when plants received higher precipitation frequency (Fig. 4C). Leaf P concentration equals to the reciprocal of P-utilisation efficiency, so in our experiment (Wang et al. 2021b), plant P-utilisation efficiency increased when precipitation frequency decreased. Interestingly, precipitation amount did not significantly affect P-utilisation efficiency of R. pseudoacacia. It is also worth noting that leaf P concentration in our experiment significantly decreased when more N was added to plants (Fig. 4B). So, in our experiment, plant P-utilisation efficiency increased when N deposition increased. With higher precipitation, P would be a limitation for R. psuedoacacia, rather than water and N. Moreover, many previous studies have shown that P is very important in N2 fixation because nodules are P sinks (Schulze et al. 2006; Wang et al. 2018a) and N2-fixing processes are P consuming (Raven 2013). A lack of P in legumes would lead to cease of N2 fixation (Sa and Israel 1991; Schulze et al. 2006). Only N limitation in soil would promote N-fixation thus won’t lead to N limitation in legumes, but both N and P limitation in soil would lead to N limitation. So, further studies should still be conducted in N2-fixing species on the relationship between the P-utilisation efficiency and water-utilisation efficiency.
Altered N absorbing mechanism of R. pseudoacacia under various N and water status
Leaf N concentration of R. pseudoacacia under N1W2F1, N1W3F1 and N1W2F2 was significantly higher than that under N3W2F1, N2W3F1 and N2W2F2 (Fig. 4A). The total biomass of plants under both N1 and N2 group was very similar (Fig. 1B), so R. pseudoacacia with no N deposition (N1) absorbed more N than that with low N deposition (N2). Leaf N:P ratio of R. pseudoacacia under N2W2 and N2W3 was also slightly lower than that in other groups (Fig. 2B). However, as plant-available soil N was very similar under N1 and N2 with medium and high precipitation amount, the more N that R. pseudoacacia absorbed under N1 was mainly from N2 fixation. This meant that extra N added to plants in N2 not only had no significant promotion effect on its growth, but also decreased its N2-fixation rate and N concentration in leaves. However, as more N added in group N3 and N4, growth of R. pseudoacacia significantly increased and leaf N concentration was similar to or higher than that in N2 especially under medium and high precipitation amount. This was due to the fact that N2-fixation is more energy costing, and is less likely to be undertaken when more N is in soil (Lambers et al. 2002). This result gave us a new insight of the effects of low N deposition in environment, which may not only decrease plant N2-fixation rate, but also did not provide enough N for plants. Slightly N deposition might inhibit N2 fixation rate, but haven’t reached the threshold of promoting plant growth under medium or high precipitation amount in our experiment. Plants under this condition might have lower leaf N concentration and leaf N:P ratio, compared with plants received no or high N supply, thus may led to a limited growth due to N limitation at least in the early growing stage. Due to this strategy, under the level of low N deposition (N2), which is similar to the N deposition level in China right now, R. pseudoacacia as an alien species receiving medium or high precipitation amount might not be facilitated, and its invasion won’t be promoted until N deposition reached certain level; while very high N deposition would not only increase soil N pool but also increase the risk of R. pseudoacacia invasion.
However, plants under low precipitation levels had another story. Nodule biomass was more significantly decreased when receiving more N deposition, and leaf N concentration and leaf N:P ratio was not significantly affected by N deposition. No significant low leaf N were observed under slightly N supply, which might indicate a different N utilisation strategy compared with plants under medium and high precipitation level.