Relationship between photosynthetic-water and nitrogen use efficiencies in young Pinus taeda L. trees at two contrasting sites

Background: Nitrogen and water are important limiting factors to forest productivity. At the plant level, there is contrasting empirical evidence about the trade-off between water use efficiency (WUE) and nitrogen use efficiency (NUE). Methods: A study was conducted on 3-year-old loblolly pine ( Pinus taeda L.) trees sampled at two contrasting sites (Virginia Piedmont (VA) and North Carolina Coastal Plain (NC)) in the southeastern United States. We investigated the leaf-level relationship between water and nitrogen use efficiency (i.e., WUE ins versus PNUE). Both parameters were measured in the fall, three years after outplanting. Results: WUE ins and PNUE were higher at VA than NC. At both sites, WUE ins increased from August to November, while PNUE showed a consistent decline for the same period only at NC, with no clear pattern observed at VA. The Pearson’s coefficient of correlation (r) between WUE ins and PNUE was


Introduction
Photosynthetic water (WUE) and nitrogen use efficiency (PNUE) are important variables typically assessed in tree ecophysiological studies because water and nitrogen are the most limiting resources in many ecosystems (Reich et al. 1989;Galloway et al. 2003).These variables may be useful for the selection of desirable traits associated with N acquisition and the selection of genotypes most suitable for individual physiological components (Li et al. 1991).At the leaf level, these parameters can be derived from instantaneous measurements of leaf gas exchange and nitrogen Keywords: Instantaneous water use efficiency, photosynthetic nitrogen use efficiency, loblolly pine variations of PNUE and how this trait is related to WUE (Warren & Adams 2006).Renninger et al. (2015) mentioned that a trade-off between WUE and PNUE may occur when there is spatial segregation of water and nitrogen (e.g., water available at higher depths than nitrogen), which may alter the uptake of one resource at the expense of another.This trade-off has been observed in evergreen species (Field et al. 1983), deciduous species (Broeckx et al. 2014;Reich et al. 1989), annual crops (Cabrera-Bosques et al. 2007), and grasses (Gong et al. 2010).It has also been observed in deep-rooted species, such as pines, which may exhibit higher WUE and lower NUE compared to shallow-rooted species (Renninger et al. 2015).Arguably, these variables may not be related under optimal resource availability (Broeckx et al. 2014;Palmroth et al. 2013).This suggests that resourceuse efficiency depends on the most limiting resource (Reich et al. 1989).
Nitrogen is one of the main factors limiting loblolly pine productivity in the southern United States (Fox et al. 2007).However, reductions in the precipitation are also projected in the future in the western edge of the species range, which might compromise further gains in productivity (Maggard et al. 2016).One of the approaches to increase productivity under this type of scenario is properly matching specific genotypes (i.e., varieties) with management prescriptions, but the growth responses will depend on soil resources and sites characteristics (Yáñez et al. 2015).In this study, we use two 3-year-old loblolly pine stands located at two sites of contrasting plant productivity in the southeastern United States to investigate the relationship between instantaneous water use efficiency (WUE ins ) and photosynthetic nitrogen use efficiency (PNUE).Both sites exhibit high differences in tree growth at young ages, which has been reported to be more related to limitations of soil moisture than nitrogen content (Yáñez et al. 2015;Yáñez et al. 2017).Thus, we hypothesised that a potential tradeoff of WUE and PNUE for the species in these sites is mostly driven by the soil moisture than by the nitrogen availability.Understanding the trade-off between proxies of NUE and WUE may guide the design of silvicultural systems that could optimise both water and nitrogen use in sites with different productivity and potentially threatened by increased periods of drought.

Study sites and tree selection
This study was conducted in two field trials established in 2009 in the southern U.S. Yáñez et al. (2015) described details of the study sites and trials.Briefly, one site was located in the North Carolina Coastal Plain (NC), at the Bladen Lakes State Forest (34° 49' 49.63''N, 78° 35' 18.52''W) in the native range of loblolly pine, while the other site was located at the Virginia Piedmont (VA) at the Reynolds Homestead Forest Resources Research Center (36° 38' 35.32''N, 80° 09' 18.84''W) in an area where loblolly pine grows successfully although outside the native range of the species.The planting material and study design were the same at the two sites and corresponded to a factorial combination of six genetic entries, two silvicultural intensities and four clonal varieties, and one open pollinated (OP) and one control mass pollinated (CMP) family (Yáñez et al. 2015).This study was originally designed to assess the potential productivity of loblolly pine genotypes under different silvicultural intensities and planting densities.The sites represent two contrasting edaphoclimatic areas where loblolly pine is planted.The soil at the VA site was a welldrained Fairview Series (sandy clay loam, kaolinitic, mesic Typic Kanhapludults), while the soil at the NC site was a poorly drained Rains series (fine sandy loam, siliceous, semiactive, thermic Typic Paleaquults).Table 1 shows the soil characteristics at the study sites (slope and physical properties obtained from websoilsurvey.se.egov.usda.gov).The annual average temperature and precipitation at VA are 13 °C and 1,159 mm, respectively, while at NC, those values are 16.9 °C and 1,171 mm, respectively (from NOAA online weather data http:// sercc.com/nowdata.html).

Photosynthetic water and nitrogen use efficiency
Three years after planting (the year 2012), we selected a subsample of 36 healthy trees at each site, representing the variation in tree size at both sites.Selected trees were established at a planting density of 1,235 trees ha -1 (3.66 between rows and 2.21 m within rows).Leaflevel physiology and nitrogen content were measured in the months of August, October, and November.At each measurement date, three fully-sun exposed fascicles were randomly selected in the upper third of the crown.Fascicles selected were fully expanded and taken from the last current year foliage.Light-saturated photosynthetic rate (A sat , at 1,600 µmol m -2 s -1 ) and transpiration (E, mmol m -2 s -1 ) were measured between 9.00 and 13.00 hrs.(local time) using a portable photosynthesis system (LI-6400, LiCor Inc., Lincoln, NE, USA).Gas exchange measurements were carried out on detached fascicles avoiding overlapping.The instantaneous water use efficiency (WUE ins ) was determined as the A sat to E ratio (Medrano et al. 2015).Block temperature and relative humidity in the chamber were set at ambient conditions,  while the carbon dioxide concentration was set to 385 µmol mol -1 .Afterward, in the laboratory, the sampled fascicles were oven dried, weighed, and ball-milled to a fine powder.Specific Leaf Area (SLA) was determined as the leaf area to mass ratio.Nitrogen concentration (N) was measured by dry combustion with a VARIO MAX CN Analyzer (Elementar, Hanau, Germany).Then, the photosynthetic nitrogen use efficiency (PNUE, µmol CO 2 mol -1 N s -1 ) was determined as the A sat to N ratio (Guo et al. 2016) and adjusted by SLA.

Statistical analysis
To assess the variation of each variable over time at each specific site, we run a one-way repeated measures analysis with date as the main factor.Temporal correlation among residuals were accounted for using the unstructured function of PROC MIXED of SAS version 9.4 (SAS Institute, Cary, NC, USA).Post-hoc mean comparisons were based in the Tukey's mean comparison method.Additionally, we calculated the mean and coefficient of variations (CV) for WUE ins and PNUE on gas exchange traits.The relationship between WUE ins and PNUE was explored by site and measurement date using the Pearson's coefficient of correlation.Significant differences were considered at an alpha level of 0.05.

Results and Discussion
Water and nitrogen are the main factors limiting the productivity of many ecosystems.Thus, improving the use efficiency of those scarce resources is essential to improve the productivity of commercial species such as loblolly pine.In this study, WUE ins increased from August to November (from 2.3 to 6.8 µmol mmol -1 at NC and from 3.5 to 7.6 µmol mmol -1 at VA), which is due to a higher decline rate in E relative to A sat through the autumn months (Table 2).During the studied period, the mean temperature at the sites decreased from 26 to 9 °C at NC and from 23 to 7 °C at VA, which likely had a greater effect on the transpiration than photosynthetic rates.Similarly, Tang et al. (2003) showed decreasing air temperature, light intensity, and leaf transpiration from the summer to winter, being that gradient less pronounced for photosynthetic rates for loblolly pine.In contrast, PNUE decreased at NC during the same study period, but a higher decline (38%) was observed from October to November (Table 2).PNUE decreased by 47% from October to November at VA, while the PNUE value was intermediate between these two extremes during August (Table 2).A drastic decrease in PNUE from October to November for both sites might be explained by a higher decline in the photosynthetic rate during this period (due to low light intensity and temperature) relative to the small decline of foliar N associated with the starting of nutrient remobilisation (Yáñez et al. 2017).
Moreover, for a particular site and measurement date, we observed a moderate variation among trees in both WUE ins and PNUE (Table 2).Coefficients of variation for WUE ins varied from 12.7% to 25.3% at NC, and from 12.5% to 39.4% at VA, whereas for PNUE ranged from 15.2% to 39.9% at NC, and from 18.3% to 28.5% at VA.The within-site variability of these parameters could be associated with the spatial variability of water and nitrogen at the study sites, which may also vary with time.The topography is irregular at the VA site (slopes of 2 to 15%, Table 1), with different grades of erosion typical of Piedmont sites, likely creating gradients in the soil water and nutrient retention.In contrast, the soil at the NC site is poorly drained (slope of 0 to 2%) with variations in the microtopography, thus with variations in the water table (Yáñez et al. 2015).
Both sites were similar in N content and available water capacity (Table 1), and according to Yáñez et al. (2017) the trees at both sites were not N deficient (foliar nitrogen > 1.1 %), thus the lower soil moisture at VA during summer and the excess of water at NC likely limited growth at both sites.The results of this study showed that the relationship between PNUE and WUE ins varied between sites and measurement dates (Figure 1).The Pearson's coefficient of correlation (r) between WUE ins and PNUE was negative (r=0.50) at VA in August, positive (r=0.77) at NC in October, and non-significant for the other measurement dates.At the VA site, the negative relationship between PNUE and WUE ins found in August might be explained by the spatial segregation of resources mentioned by Renninger et al. (2015).This month is the end of summer; thus, less water is available in the soil to meet the plant requirements.Moreover, Yáñez et al.

Site
(2015) reported a higher presence of weeds at the VA than NC (particularly Rubus spp.).This likely increased the water stress (i.e., stomatal closure) of loblolly pine trees at VA, and triggered the trade-off between WUE ins and PNUE.Stomatal closure due to water stress has a faster-decreasing effect on stomatal conductance and transpiration than photosynthesis, with no impact in foliar nitrogen in the short-term, which explains the trade-off between WUE ins and PNUE (Warren & Adams 2006).The lack of significant relationships between WUE ins and PNUE in August and November at NC and October and November at VA suggests that neither water nor nitrogen were simultaneously limiting at those dates (Palmroth et al. 2013;Soolanayakanahally et al. 2009).While a trade-off between WUE and NUE has been found in loblolly pine (Palmroth et al. 2013) and other species (Renninger et al. 2015;Dijkstra et al. 2016), the coefficient of correlations between those parameters are moderate to low.In our study, foliage acclimated quickly within the chamber (less than 3 min), and stomatal conductance did not depress within this measurement time.However, we recommend verifying this acclimation period in future studies since foliage water status seems to have a higher effect on WUE versus PNUE trade-off than nitrogen.

FIGURE 1 :
FIGURE 1: Relationship between instantaneous water use efficiency (WUE ins ) and photosynthetic nitrogen use efficiency (PNUE) for each site (NC = North Carolina Coastal Plain, VA = Virginia Piedmont) and date.* and ns denotes significant and non significant correlations at an alpha level of 0.05, respectively.

TABLE 2 :
P-values for the analysis of variance per site (NC = North Carolina Coastal Plain, VA = Virginia Piedmont) on light-saturated photosynthetic rate (A sat ), transpiration (E), nitrogen concentration (N), instantaneous water use efficiency (WUE ins ), and photosynthetic nitrogen use efficiency (PNUE).The coefficient of variation (CV) is presented for WUE ins and PNUE.At each site, different letters within a column indicate significant differences according to the Tukey's mean comparison test.