Studies of NH
 4
 +
 and NO
 3
 −
 uptake ability of subalpine plants and resource‐use strategy identified by their functional traits

The leaf economics spectrum (LES) is based on a suite of leaf traits related to plant functioning and ranges from resource-conservative to resource-acquisitive strategies. However, the relationships with root traits, and the associated belowground plant functioning such as N uptake, including nitrate (NO 3 −) and ammonium (NH 4 +), is still poorly known. Additionally, environmental variations occurring both in time and in space could uncouple LES from root traits. We explored, in subalpine grasslands, the relationships between leaf and root morphological traits for three dominant perennial grass species, and to what extent they contribute to the whole-plant economics spectrum. We also investigated the link between this spectrum and NO 3 − and NH 4 + uptake rates, as well as the variations of uptake across four grasslands differing by the land-use history at peak biomass and in autumn. Although poorly correlated with leaf traits, root traits contributed to an economic spectrum at the whole plant level. Higher NH 4 + and NO 3 − uptake abilities were associated with the resource-acquisitive strategy. Nonetheless, NH 4 + and NO 3 −-uptake within species varied between land-uses and with sampling time, suggesting that LES and plant traits are good, but still incomplete, descriptors of plant functioning. Although the NH 4 + :NO 3 − uptake ratio was different between plant species in our study, they all showed a preference for NH 4 + , and particularly the most conservative species. Soil environmental variations between grasslands and sampling times may also drive to some extent the NH 4 + and NO 3 − uptake ability of species. Our results support the current efforts to build a more general framework including above-and below-ground processes when studying plant community functioning.


Abstract:
The leaf economics spectrum (LES) is based on a suite of leaf traits related to plant functioning and ranges from resource-conservative to resource-acquisitive strategies.
However, the relationships with root traits, and the associated belowground plant functioning such as N uptake, including nitrate (NO 3 -) and ammonium (NH 4 + ), is still poorly known. Additionally, environmental variations occurring both in time and in space could uncouple LES from root traits. We explored, in subalpine grasslands, the relationships between leaf and root morphological traits for 3 dominant perennial grass species, and to what extent they contribute to the whole-plant economics spectrum. We also investigated the link between this spectrum and NO 3 and NH 4 + uptake rates, as well as the variations of uptake across four grasslands differing by the land-use history at peak biomass and in autumn. Although poorly correlated with leaf traits, root traits contributed to an economic spectrum at the whole plant level. Higher NH 4 + and NO 3 uptake abilities were associated with the resource-acquisitive strategy. Nonetheless, NH 4 + and NO 3 uptake within species varied between land-uses and with sampling time, suggesting that LES and plant traits are good, but still incomplete, descriptors of plant functioning. Although the NH 4 Introduction Functional traits have been widely used to describe different plant strategies. One major axis of specialisation involves a trade-off between conservation of resources in well protected and long lived tissues, and acquisition of resources by tissue with high use-efficiency and turnover, and commonly referred as the leaf economic spectrum (LES, Wright et al. 2004). More specifically, species with an exploitative strategy share similar leaf attributes such as high specific leaf area (SLA) and nitrogen concentrations (LNC) that have been associated with short leaf life-span, high photosynthetic capacity as well as high decomposability (Reich 2014, Cornwell et al. 2008, and dominate in nutrient rich environments, while slow-growing conservative species carry opposite trait values and are more common in poor or harsh conditions (Chapin 1980, Ordonez et al. 2009). Despite some evidences of a similar contribution of root traits to the plant strategy (Roumet et al. 2006, Freschet et al. 2010, Fort et al. 2013, the significance of root traits is less understood than the one for leaf traits, mainly because weak correlations between analogous leaf and root traits have been reported (Craine et al 2005, Tjoelker et al. 2005, Freschet et al. 2010, and also because root functioning is often overlooked compared to leaves in field conditions. Nutrient uptake ability, one of the main functions provided by roots (Hodge 2004, James et al. 2009), is both influenced by anatomical and physiological adjustments such as specific root length or maximal uptake rate (Vmax, but see Bassirirad 2000). Among nutrients, nitrogen is one of the best studied mineral nutrients and its uptake by plants under both the ammonium (NH 4 + ) and nitrate (NO 3 -) forms is influential for plant and ecosystem functioning. However, rarely have morphological and physiological properties of root been assessed simultaneously in field conditions, whereas some information come from species grown in standardized conditions (Maire et al. 2009, Grassein et al. 2015 To assess NH 4 + and NO 3 uptake patterns over the growing season, the same sampling design was repeated twice during 2010. At each date for each species and grassland, we sampled the roots and soil (approximately: 25x25x25 cm) of five individuals (genetically distinct individuals at least 2m apart). The first sampling corresponded to the peak biomass and targeted flowering onset (just before anthesis), and the second sampling corresponded to autumn after last management activities occurred. For D. glomerata in TMF and B. erectus in TU, the two sampling dates were mid-June and mid-September. For F. paniculata in UM and the three species in UU, the sampling dates were: early July and early September. These two dates are called "Summer" and "Autumn" hereafter. As much as possible, species were sampled at the same time during the day to avoid any diurnal variation in N uptake (Gessler et al. 1998). In total, we have sampled 12 points (3 species*2 seasons*2 habitats per species).

Soil nitrogen pools
At each date and for each grassland, soil nitrogen concentrations were measured from six soil cores (dimensions 4.5 cm Ø, 10 cm deep) kept on ice in the field and maintained at 4°C upon return to the laboratory (within 2h). Soils were sieved through a 5.6 mm mesh to remove roots and stones. A subsample of 10g fresh sieved soil was prepared for extraction of inorganic N in 0.5M K 2 SO 4 , and analysed using a colorimetric analyser (FS-IV autoanalyser (OI-Analytical, College Station, TX, USA) (following Bowman et al. 2003) to measure soil concentrations of ammonium (NH 4 + ), nitrate (NO 3 -) and Total Dissolved Nitrogen (TDN). Soil aliquots were used to determine soil water (7 days at 70°C) and soil organic matter contents (550°C during 4 hours). Finally, soil subsamples were air-dried to measure soil pH, or ground to a fine powder for measurements of total carbon (C) and N contents using an elemental analyser (FlashEA 1112, Thermo Fisher Scientific Inc., Waltham, MA, USA).
At each date, five individuals of each species, with roots and soil, were excavated from each field, transferred within half an hour to the laboratory located at the Lautaret Pass (Station Alpine Joseph Fourier) and kept at 4°C until the NO 3 and NH 4 + uptake rate measurements to maintain the functional integrity of the roots. Living young fine roots were washed with deionised water, cut to 2-cm length and then, rinsed in 1mM CaSO 4 at 4°C for 3 min. The NO 3 and NH 4 + uptake rates were measured during the first hour following plant harvest as described by Louahlia et al. (2000). The optimal conditions for uptake measurements by excised root determined by Lainé et al. (1993) were used in the present study.  Finally, the NH 4 + :NO 3 uptake ratio was calculated as the ratio between NH 4 + Vmax and NO 3 -Vmax.
A principal component analysis (PCA) was performed using all plant functional traits at the individual level to describe their functional strategy based on leaf and root traits. To investigate the relationships between functional traits of leaves and roots, and NH 4 + and NO 3 uptake ability (hypothesis 1), we used Pearson correlation coefficients. Relationships between the functional strategy and uptake of NH 4 + and NO 3 at the root level were tested using regression analyses between the N uptake rates (Vmax) and the first PCA.
Comparisons of NH 4 + and NO 3 uptake rates for species (hypothesis 2), fields and date (hypothesis 3) were conducted with ANOVA followed by Tukey tests to compare species and grasslands. In details, the effects of sampling time and fields on plant traits within each species were tested using two-ways ANOVA. Similarly, the effects of sampling time and fields on maximal NH 4 + and NO 3 uptake rates within each species were tested using twoways ANOVA. The effects of fields and sampling time on NH 4 + :NO 3 ratio within each species were tested using two-ways ANOVA. Then, we tested only in UU grasslands, the species effect using one-way ANOVA. Finally, we used a two-ways ANOVA and Tukey post hoc test to test soil parameters differences between fields and dates. Data were logtransformed when necessary to achieve normality and heteroscedasticity. All statistical analyses were performed using the software R 3.4.4, with multivariate analyses (PCA) being performed using the package Ade4 (Dray & Dufour 2007).

Results
We observed large variations for leaf and root functional traits in spite of a restricted number of species in our study ( Table 2). The range of variation was similar to, and sometimes even This functional axis was positively correlated to NH 4 + and NO 3 -Vmax (Fig 2a and 2b) and negatively to NH 4 + :NO 3 uptake ratio (Fig 2c) indicating a more pronounced preference for NH 4 + at lower values of axis 1. Except RDMC, all traits taken separately were poorer predictors of the NO 3 and NH 4 + maximum uptake rates than this functional axis, although the first PCA axis was significantly correlated with all functional traits (Table 3).
In UU grassland, NH 4 + Vmax in summer was similar for the three species (Fig. 3a)  compared to the other grasslands, and overall greater values for F. paniculata (Fig. 4).
Within species, a limited number of traits were significantly different between grasslands ( Since all species occurred in the UU grasslands, we choose to focus on soil parameters from these grasslands. UU and UM only differ for SWC in autumn (Table 4), all other soil variables were similar between these two grasslands, which had similar past land-use history (Table1). UU had consistently higher SWC and SOM, and lower soil pH and CN ratio than TMF and TU. All grasslands had similar soil NH 4 + concentrations. During the summer, we observed higher TDN and NH 4 + :NO 3 soil ratio, and lower soil NO 3 concentration in UU compared to TMF and TU, but we did not find these differences in autumn. Discussion:

Relationships between leaf and root traits
In the aim to find parallels between above and below-ground organs ( 1973, Kronzucker et al. 1999, Maire et al. 2009). The preferential uptake for an inorganic N form could also be influenced by environmental and physiological factors (Britto & Kronzucker 2013). Our results did not support any trade-off in the intrinsic ability of plant species to take up both N forms, even after removing possible environmental conditions or interactions between inorganic N forms. Although we could not directly test for the relationship between soil parameters and plant N uptake rates, differences between grasslands in the N uptake within species highlight that management practices may have important effects on plant N uptake, likely through N cycling changes and the quality of the N pool available as already pointed out by previous studies (Zeller et al. 2000, Robson et al. 2007).
Other studies have suggested that N preference could be dependent on the soil availability of the different N forms (Näsholm et al. 2009, Stoelken et al. 2010). While our results partially supported this hypothesis, with variation within species between different grassland, the different species sampled in the same grassland showed differences in their NH 4 + :NO 3 uptake ratio, supporting the hypothesis that this "preference" is partially related to the strategy of species, or at least to species identity. But overall, more exploitative species with higher maximum uptake rates for one inorganic N form are also likely to have high uptake rates for other N forms as previously found (Kastovska & Santruckova 2011). beyond the possibility in our study. Nonetheless, the variability we observed in the ratio of uptake between the inorganic N forms suggested that, to some extent, plant physiology was adjusted to match the soil conditions where species occurred. Yet, differences between species with different strategies remain, with higher uptake rate for both N forms associated with a more exploitative strategy, and we hypothesised that this should be also the case for organic N sources (Kastovska and Santruckova  Variations of N uptake ability during the growing season Plant N uptake ability also varies during the growing season, with N uptake increasing (Stahl et al. 2011) or decreasing (Jaeger et al. 1999) depending on the ecosystems investigated. In the UU grassland, NH 4 + uptake was higher for all species during the autumn than during the summer, and the same was found for NO 3 uptake by B. erectus. Plant activity is usually considered to slow down during the autumn compared to the peak biomass in summer, an assumption supported by higher LDMC and lower SLA for all species related to the senescence of leaves. However, we did not observe any changes for root traits, suggesting that roots could remain active during this time of the growing season, especially in the process of resource storage, an important feature for subalpine/alpine plants (Kleijn et al. 2005).
Additionally, studies have reported an increase of grassland N cycling rate in the autumn that could be explained by more favourable soil conditions (first rains and mild temperature), and associated with still active N uptake by plants as observed in our study (Miller et al. 2009, Larsen et al. 2012). This could also be related to the better retention of NH 4 + vs.NO 3 in wet soils during autumn, making NH 4 + more available for plant uptake (Brady and Weil 2001).
Despite the fact that only few soil parameters differed between the two investigated seasons in the UU grassland, the N uptake increase in autumn was more likely a site-dependent effect related to soil conditions (Miller et al. 2009, Stahl et al. 2011, Legay et al. 2013), rather than a species response since all species did show the same pattern in the other grasslands. Yet, a multiple-year study remains necessary to better conclude on these seasonal patterns.    1) and Vmax for NH 4 + (a), NO 3 -(b) and NH 4 + : NO 3 uptake ratio (c). The three 192 relationships were significant (p-values<0.05) assuming a polynomial relationship of order=2, and the resulting R² are indicated on each graph.