Precipitation is the main factor affecting the variation of foliar nitrogen isotope composition in two leguminous shrub species of northwestern China

Abstract

An increase in foliar nitrogen isotope composition (δ¹⁵N) with decreasing precipitation has been shown to occur widely in non-N₂-fixing plant species. However, similar patterns have not been identified in N₂-fixing species. Here, we tested the relationships of foliar δ¹⁵N with local environmental factors and leaf properties in two leguminous shrub species (Caragana korshinskii and Caragana liouana) sampled from 30 populations. Results indicated that the mean annual precipitation (MAP) primarily accounted for the variation of foliar δ¹⁵N in the two species. Further analysis revealed strong negative correlations between foliar δ¹⁵N and MAP within and across species. These results suggest that the foliar δ¹⁵N of leguminous shrub species also shift along precipitation gradients, which augments our understanding of the relationships between foliar δ¹⁵N and climatic factors.

1. Introduction

The foliar nitrogen (N) stable isotope composition (δ¹⁵N) integrates several fundamental biogeochemical processes and can be an indicator of N availability [1]. Assessing the variability of foliar δ¹⁵N of plants across climatic gradients would thus enhance our understanding of the spatial and temporal patterns of N cycling and how disturbances alter the N cycle [2–4]. Research on this topic has increased over the past decade [1–4]. Both local and global studies have shown that foliar δ¹⁵N increases with decreasing mean annual precipitation (MAP) [4–6]. However, this general trend has only been considered in non-N₂-fixing species [4]. To our knowledge, no such information is available on how foliar δ¹⁵N varies with precipitation within and across N₂-fixing plant species, although many previous studies conducted at large spatial scales have considered legumes at the functional group level without considering interspecific differences [4–6].

The genus Caragana belongs to the family Leguminosae and is well known for its role in controlling soil erosion and land desertification, and for its economic values as a honey source, fuel and fodder in China [7]. Because of the great environmental and economic benefits, this genus has attracted increasing attention in ecological research [7–9]. In this study, we selected two of the most common Caragana species, Caragana korshinskii and Caragana liouana, which are two dominant shrub species in the desert regions of northern China. We asked whether foliar δ¹⁵N of these two Caragana species showed negative relationships with precipitation gradients, as reported for non-N₂-fixing species.

2. Material and methods

We studied two Caragana species (C. korshinskii and C. liouana) sampled from a total of 30 wild populations (15 for C. korshinskii and 15 for C. liouana) across 28 sites (electronic supplementary material, table S1 and figure S1). Caragana korshinskii and C. liouana are dominant shrub species in the desert steppe regions of northwestern China. At each sampling site, fully expanded sun-exposed leaves were collected from four different individual plants 4 m apart from each other and pooled into one sample. A minimum of three replicates were collected from each population and all samples were collected from robust mature plants that grew in unshaded habitats. Plant samples were placed in paper envelopes and dried at 65°C to a constant mass upon returning to the laboratory. δ¹⁵N values were determined from 5 to 6 mg samples of homogeneously ground material of each replicate using an isotope ratio mass spectrometer (Finnigan MAT, Bremen, Germany). Stable N isotope ratios (δ¹⁵N) are reported per mil (‰) relative to atmospheric N₂: δ¹⁵N = (R_sample/R_atmos−1) × 1000 (‰), where R_sample is the isotope ratio (¹⁵N/¹⁴N) in plant samples, and R_atmos the isotope ratio (¹⁵N/¹⁴N) for atmospheric N₂. Analytical precision, based on repeated measurements of a laboratory plant standard and two protein standards, was within 0.3‰ for δ¹⁵N. These analyses also yielded N and carbon (C) concentrations on a mass basis.

Geographical information and altitude were recorded using a global positioning system. Meteorological data of each sampling site were provided by the IWMI (International Water Management Institute) online climate summary service portal (http://wcatlas.iwmi.org/Default.asp), including MAP, mean annual temperature (MAT), daily temperature range (DTR), relative humidity (RH), sunshine hours (SH), wind run (WR), moisture availability index (MAI), Penman–Monteith (PM), mean temperature of warmest month (MTWM) and mean temperature of coldest month (MTCM).

A one-way ANOVA was used to explore the difference in leaf δ¹⁵N between populations and between species. A principal component analysis (PCA) was used to comprehensively define the relationships between foliar δ¹⁵N, the interrelated changes in geographical and climatic factors and leaf properties. Finally, we performed stepwise multiple regressions to test for the relative importance of these variables on foliar δ¹⁵N.

3. Results

Foliar δ¹⁵N showed substantial variability among populations for each of the two species (electronic supplementary material, figure S2), but the differences between species were not significant (electronic supplementary material, figure S3). Foliar δ¹⁵N of C. korshinskii ranged from 0.29 to 2.78‰ with a mean of 1.35‰ and foliar δ¹⁵N of C. liouana ranged from −0.37 to 2.86‰ with a mean of 0.91‰ (electronic supplementary material, figure S2).

The 17 variables were examined using PCA. The first PCA axis accounted for 39.48% of the total variation, while the second axis accounted for 33.16% (figure 1). MAI, MAP, DTR, PM, MTWM, MAT and C : N influenced PC1, while PC2 was influenced primarily by latitude, altitude, MTCM, WR, SH and foliar N and C concentrations (figure 1). Further analysis performed by multiple stepwise regressions revealed that, after accounting for species, variation in foliar δ¹⁵N was solely controlled by MAP (y = 2.91–0.01MAP, R² = 0.50, p < 0.001).

Figure 1.

PCA based on environmental factors and leaf properties of two Caragana species from 30 populations. Geographical factors: environmental factors: MAP, mean annual precipitation; MAT, mean annual temperature; DTR, daily temperature range; RH, relative humidity; SH, sunshine hours; WR, wind run; MAI, moisture availability index; PM, Penman–Monteith; MTWM, mean temperature of warmest month; MTCM, mean temperature of coldest month. Leaf properties: δ¹⁵N, nitrogen isotope composition; N, foliar nitrogen concentration; C, foliar carbon concentration; C : N, the ratio of foliar carbon to nitrogen.

For both species, foliar δ¹⁵N was significantly and negatively correlated with MAP (figure 2a). Within each species, the foliar δ¹⁵N of both C. korshinskii and C. liouana were significantly negatively correlated with MAP as well (figure 2b,c).

Figure 2.

Correlation between foliar δ¹⁵N and MAP for both species (a), C. korshinskii (b) and C. liouana (c). Shown are linear fits, correlation coefficients and p-values.

4. Discussion

We studied two Caragana species along environmental gradients in the arid and semi-arid regions of northwestern China and found that the mean foliar δ¹⁵N values of the two species all fell within the range of δ¹⁵N commonly reported for leguminous species (0 ± 2‰) [10]. These relatively low values of leguminous species have been attributed to both the capacity of legumes to obtain N from atmospheric N₂, which generally has a δ¹⁵N value near 0% [11] and the lack of isotopic discrimination during biological fixation of atmospheric N₂ [10,12]. Consistent with previous studies [13–15], we also found a significant difference in foliar δ¹⁵N among populations within each species (electronic supplementary material, figure S2). This could be due to the large variation in soil δ¹⁵N from the east to west regions of northern China [5] and the tight correlation between foliar and soil δ¹⁵N [10,16].

As previously observed for non N₂-fixing species [4,6], we found that foliar δ¹⁵N declined with increasing MAP for all populations (figure 2a). Similarly, Wang et al. [5] also found that foliar δ¹⁵N of the genus Caragana significantly decreases with increasing aridity. Interestingly though, we also found that foliar δ¹⁵N was strongly negatively correlated with MAP for each of the two Caragana species (figure 2b,c). For non-N₂-fixing plant species, the increase in foliar δ¹⁵N with decreasing MAP has been explained as a response to the dominant gaseous nitrogen loss pathways in low-MAP environments, which leaves the remaining soil pool enriched in ¹⁵N [4,6]. By contrast, N₂-fixing plant species can obtain N from both soil N and atmospheric N via symbiotic relationships [17]. However, in arid and semi-arid regions where water and N availability are limited, N₂-fixing plants rely heavily on atmospheric N [17–19]. Still, symbiotic nitrogen fixation is highly sensitive to changes in water availability [20]. As water availability decreases, the formation of new nodules is suppressed, the size of the nodules decreases and the specific activity of the nodules is reduced [19–21]. This likely leads to these species using their developed root systems, taking up water from the deep soil where δ¹⁵N values are generally higher in deeper portions of the profile [22]. This could explain the negative relationship between foliar δ¹⁵N and precipitation and also reflect a strategy where Caragana species balance the allocation of photosynthates between energy-intensive nodule biosynthesis and structural maintenance within plant organs, such as roots, leaves, stems and seeds [23–24], so as to optimize available N.

Although our study highlights the generality of foliar δ¹⁵N patterns across climatic gradients, given that the foliar δ¹⁵N of leguminous and non-leguminous species vary with precipitation, further studies are also needed to reveal the mechanisms underlying patterns of foliar δ¹⁵N, and explain such findings as the differences in gaseous N losses, N₂-fixing bacteria and mycorrhizal associations and plant N metabolic processes across climate gradients.

Data accessibility

The data are provided in the electronic supplementary material.

Authors' contributions

F.M. and T.-T.X. conceived and designed the research; F.M., M.L. and Z.-J.S. collated the data; F.M. and J.-L.L. carried out the analyses; F.M., T.-T.X. and Z.-J.S. wrote the paper. All authors edited and revised the manuscript and agreed to be held accountable for the content therein.

Competing interests

We declare we have no competing interests.

Funding

This study was supported by the National Natural Science Foundation of China (grant nos. 31660188, 31760056, 31260166).