Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
Knowledge of determining factors for nitrogen uptake preferences and how they are modified in changing environments are critical to understand ecosystem nitrogen cycling and to predict plant responses to future environmental changes. Two 15N tracer experiments utilizing a unique differential labelled nitrogen source were employed in both African savannas and greenhouse settings. The results demonstrated that nitrogen uptake preferences were constrained by the climatic conditions. As mainly indicated by root δ15N signatures at 1:1 ammonium/nitrate ratio, in the drier environments, plants preferred nitrate and in the wetter environments they preferred ammonium. Nitrogen uptake preferences were different across different ecosystems (e.g. from drier to wetter environments) even for the same species. More significantly, our experiments showed that the plant progeny continued to exhibit the same nitrogen preference as the parent plants in the field, even when removed from their native environment and the nitrogen source was changed dramatically. The climatic constraint of nitrogen uptake preference is likely influenced by ammonium/nitrate ratios in the native habitats of the plants. The constancy in nitrogen preference has important implications in predicting the success of plant communities in their response to climate change, to seed bank use and to reforestation efforts.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The assimilation of NH4+ is more energetically efficient when compared with NO3-, because NH4+ can be directly incorporated into glutamate via an NH4+ assimilation pathway. Nitrate, on the other hand, must first be modified via a reduction pathway before assimilation (Engels & Marschner 1995). However, NO3- is usually more available for uptake in many ecosystems, owing to its higher mobility (Brady & Weil 1999). Nitrate can be incorporated into organic compounds in both root and leaf tissues whereas NH4+ is only synthesized into amino acids in the root tissues near the site of uptake to avoid toxic accumulation (Engels & Marschner 1995). Either NH4+ or NO3- can dominate the inorganic N pool of an ecosystem. For example, in most mature undisturbed forests, the soil inorganic N pools are dominated by NH4+ (Kronzucker et al. 1997). In well-aerated agricultural soils or other frequently disturbed sites, NO3- is the principal inorganic N source (Brady & Weil 1999). In arid and semiarid ecosystems such as African savannas, nutrient availability varies spatially and temporally, and nutrients are considered to be a major limiting factor for plant growth when water limitation alleviates (Scholes et al. 2002; Aranibar et al. 2004). The Kalahari Transect (KT) in southern Africa traverses a dramatic aridity gradient through Zambia, Botswana, Namibia, and the Republic of South Africa and is essentially composed of homogeneous soils – the deep Kalahari sands (Fig. 1) (Shugart et al. 2004; Wang et al. 2007). The rainfall variability along the KT ranges from less than 200 mm mean annual precipitation (MAP) in southwest Botswana to over 1000 mm MAP in the north (i.e. western Zambia) (Shugart et al. 2004). Therefore, the KT provides ideal conditions (homogenous soils and gradients in rainfall) for studying, at subcontinent scales, the association between N uptake preferences and aridity, without confounding soil effects. This uniquely designed N uptake study was accomplished by using 15N as a differential labelled source of N, in both greenhouse and field-based setting, to test two hypotheses. The first is that N preferences are constrained by climatic conditions of the native habitat of the plants. Following the observations of Houlton et al. (2007) in Hawaiian rainforests, we hypothesized that plants prefer NO3- in the drier environments, and plants prefer NH4+ in the wetter environments. We further hypothesized that the climatic constraints on N uptake preference is influenced by the NH4+/NO3- ratios in the habitats; and that – the second hypothesis – plants may inscribe the adaptation of the N uptake preference in their progeny. All greenhouse and field plants were fertilized with NH4NO3. There were two treatments, one with 14NH415NO3 and the other with 15NH414NO3, allowing for direct monitoring of N uptake preferences. Whenever possible, a control (without fertilizer addition) was also utilized.
MATERIALS AND METHODS
For the greenhouse experiment, a total of six different grass species from four sites of their native habitat along the KT precipitation gradient (Fig. 1) were used in this study (Table 1). Seeds of four grass species were collected in August 2004, and seeds of two other grass species (Schmidtia pappophoroides Steud. ex J.A. Schmidt from Pandamatenga and Eragrostis spp. from Mongu) were collected in March 2005. In Mongu (897 mm MAP), the wettest area in this mega transect, seeds of Eragrostis spp. was collected. In Pandamatenga (698 mm MAP), the second wettest area in the transect, seeds of S. pappophoroides, Pogonarihria squarrosa (Licht, ex Roem. & Schult.) Pilg. and Leptocarydion vulpiastrum Stapf were collected. In Ghanzi (424 mm MAP), Botswana, a relatively dry site with intermediate rainfall conditions, seeds of S. pappophoroides and Eragrostis lehmanniana Nees were collected. In Tshane (365 mm MAP) of southern Botswana, the driest site of this transect (Table 1), Enneapogon cenchroides (Licht.) Roem. & Schult. ex C.E. Hubbard and E. lehmanniana were collected. Based on our own field observations, the number of grass species increased from the very dry area to wetter area and decreased from wetter area to the wettest area (Table 2, Wang et al. 2010). The number of species used in the greenhouse and field experiments generally followed the pattern of species distribution in the field. The greenhouse experiment was conducted at the University of Virginia from February to October 2005. The two additional grass species (S. pappophoroides from Pandamatenga and Eragrostis spp. from Mongu) were grown in the greenhouse from February to June 2006.
Table 1. Field characteristics of five study sites along the Kalahari Transect and species list for the field and greenhouse experiments
Table 2. The grass richness data of the four sites along the Kalahari Transect
The data was modified from a field survey conducted in the wet season 2005. The species richness is based on the number of species from 20 1 × 1 m2 plots from each site and the details of field sampling can be found in Wang et al. (2010).
The seeds were initially germinated in pans. The seedlings were then transferred into plastic pots (one individual per pot) containing commercial sand. Plants were equally well-watered during their germination and growth (volumetric soil water content was between 20 and 40%). During an initial two-week period of adjustment after transferring from germination pans to individual pots, a low concentration (300 ppm) commercial soluble fertilizer [Peters N (2.1% NO3-, 17.9 urea)-P (phosphate)-K (potash): 20–20-20] was applied to facilitate survival. A couple of weeks after adjustment from transplanting, a 15N-labelled fertilizer (NH4NO3, 50‰ in 15N) was applied to each individual plant to establish an N concentration of 15 µg N/g dry soil. The total amount of N applied as fertilizer was comparable with the abundance of mineral N at the African field locations (Feral et al. 2003; Wang et al. 2007). The 15N-labelled fertilizer solution was evenly sprayed around the base of each plant (2 cm radius) at a very slow rate to prevent diffusion of the fertilizer exceeding the 2 cm radius, and the solution infiltrated about 20 cm of the potting sand (the roots were about 20–30 cm deep and the root distributions were similar for all the species). The 15N-labelled fertilizer solution was applied twice in 48 h. The two treatments (NH415NO3 and 15NH4NO3) received the same amount of total N fertilizer. In order to distinguish the uptake preferences, the 15N was labelled at different locations in the N compounds (15NH4+ versus 15NO3-). NH4+ and NO3- have the same molar concentrations because of the chemical composition of NH4NO3. If plants have any preference for either NH4+ or NO3-, we could measure it through the differences in δ15N signatures between NH4+ labelled treatment and NO3- labelled treatment. The comparisons were always between roots or between leaves (e.g. no cross-comparisons of roots and leaves). If the δ15N is higher in the NO3- labelled treatment than the NH4+ labelled treatment, the plants prefer NO3-; if the δ15N is higher in the NH4+ labelled treatment, the plants prefer NH4+.
Small isotopic fractionation of NH4+ and NO3- uptake with high external N concentrations has been observed in plants grown in nutrient solutions (Högberg et al. 1999; Yoneyama et al. 2001). Because equal molar NH4+ and NO3- was applied to all the species and to all the sites, the fractionation was assumed to have minimal effects on the observed δ15N signature patterns. Many studies (e.g. Drew & Saker 1978; Jackson, Manwaring & Caldwell 1990; Wang, Mou & Jones 2006) have shown that plants can respond to nutrient addition within hours of applications. Therefore short application duration (one injection per 24 h interval and 48 h in total) was employed in this study, which can minimize biases caused by the potential N transformation processes in the potting sand. Four individuals were used for each species × treatment combination (three individuals of E. lehmanniana from Tshane were used for 15NH4NO3 treatment, owing to lower seedling availability). Three individual plants of each species were used as controls, and were grown without fertilizer (except the low initial soluble fertilizer to facilitate survival) or labelled 15N additions. To evaluate the effect of higher soil N content on N preference, double amount of fertilizer (but the same 15N signature) was applied to P. squarrosa and L. vulpiastrum in addition to the field-level N treatment mentioned above. All grasses were harvested 24 h after the second fertilizer application. The roots were carefully rinsed with tap water to remove excess N fertilizer that was not taken up. Leaves and roots from each individual were separated and stored in paper bags.
In January and February 2006, field 15N labelling and uptake experiment was carried out in four locations along the KT where the grass seeds were obtained. The main objective of the field experiment was to test the plant N uptake preference in situ and to explore the mechanisms of N uptake preference by comparing field (parents) and greenhouse (F1 generation) results. The species selected were essentially the same as those used in the greenhouse experiment (Table 1) and the locations were similar though not exactly the same because of field logistics. The non-exact match of field and greenhouse sites may not be ideal, however, three sites overlapped between the field and greenhouse experiments and these overlaps included both dry and wet locations – the major contrast in this study. The field N uptake experiment was conducted in exactly the same way as the greenhouse experiment except a higher 15N signature for both NH4+ and NO3- (100‰15NH4NO3 and NH415NO3) was used. The 15N-labelled fertilizer solution was evenly sprayed around the base of each plant (2 cm radius) at a very slow pace and the water infiltrated about 10–20 cm of the soil. The root distributions were similar for all the species but the root sizes varied depending on species. We chose individuals of similar sizes for 15NH4 and 15NO3 treatments to avoid complications because of root size differences. The 15N-labelled fertilizer solution was applied twice in 48 h before harvesting. At Pandamatenga, control plants (without nutrient or water addition) of S. pappophoroides, Panicum maximum and L. vulpiastrum were collected and analysed for comparison.
All samples were then oven-dried at 60 °C for 72 h. After drying, foliar and root samples were ground to powder. Nitrogen isotope analysis was performed using a GV Micromass Optima isotope ratio mass spectrometer (IRMS) coupled to a Carlo Erba elemental analyser (EA). The 15N content of the plants was reported in the conventional notation:
where (15N/14N)sample is the N isotope ratio of samples, and (15N/14N)standard is the N isotope ratio of the standard material. The standard for N stable isotopes is atmospheric molecular N. Reproducibility of the measurements was approximately 0.2‰.
The existence of N uptake preference was determined by the δ15N difference between 15NH4+ and 15NO3- treatment in either the roots or the leaves. Considering the different N assimilation locations for NO3- and NH4+ (both roots and leaves for NO3- versus roots only for NH4+) and the short duration of the experiments, only root δ15N differences between 15NO3- and 15NH4+ treatment were used to indicate NH4+ preference whereas both the foliar and root δ15N differences were used to indicate NO3- preference. A one-way ANOVA and Tukey post hoc test (at α = 0.05 significance level) was used to evaluate the significance of the differences detected in the roots or leaves for all species from each site. To better visualize the results and compare parent versus progeny N uptake preference, the degrees and forms of N uptake preference at each site were summarized against MAP of each site for both field and greenhouse settings (e.g. Figs 2e and 4e). The degrees and forms of N uptake preference were indicated by 15N signature differences between 15NH4+ and 15NO3- treatment either in root (for NH4+ and NO3- preference) or in leaves (for NO3- preference). For NO3- preference, if the 15N signatures in both the root and foliar samples were significantly different between 15NH4+ and 15NO3- treatment, the larger difference value was used to indicate the maximum NO3- preference. In the summary figures (e.g. Figs 2e and 4e), if the difference (15NO3- − 15NH4+) was a positive value, the plant preferred NO3-; if negative, the plant preferred NH4+. A value of 0 indicated that no significant preference was detected.
For the greenhouse experiment, the individuals in relatively dry areas (E. cenchroides and E. lehmanniana from Tshane, and S. pappophoroides and E. lehmanniana from Ghanzi) exhibited a preference for NO3- (Fig. 2a,b,e), indicated by higher foliar δ15N signatures in 15NO3- treatment than in 15NH4+ treatment. Whereas others in the wetter habitats (S. pappophoroides, P. squarrosa and L. vulpiastrum from Pandamatenga and Eragrostis spp. from Mongu) had a preference for NH4+, indicated by higher root δ15N signatures in 15NH4+ treatment than in 15NO3- treatment (Fig. 2c,d,e). For the two species (P. squarrosa and L. vulpiastrum) with additional treatment of higher concentrations of NH4NO3 (but the same 15N signature), their uptake preferences did not change (Fig. 3).
The field experiment generally showed the same trends with one exception. The individuals in relatively dry areas (S. pappophoroides and E. lehmanniana from Tshane, and S. suniplumis from Ghanzi) exhibited a preference for NO3- (Fig. 4a,b,e), indicated by higher δ15N signatures in 15NO3- treatment than in 15NH4+ treatment either in roots (S. pappophoroides in Tshane and S. suniplumis in Ghanzi) or in leaves (E. lehmanniana in Tshane). The exception was E. lehmanniana from Ghanzi, which did not show any preference (Fig. 4b). This was different from what was observed in the greenhouse experiment for this species of this site (Figs 2b and 4b). The individuals in the wetter areas (S. pappophoroides, P. maximum, L. vulpiastrum and Digitaria spp. from Pandamatenga) had a preference for NH4+, indicated by higher root δ15N signatures in 15NH4+ treatment than in 15NO3- treatment (Fig. 4d,e). The individuals (S. pappophoroides) in Maun (with an intermediate rainfall level) did not show preference for either NH4+ or NO3- (Fig. 4c,e).
The hypothesis that N preferences are influenced by climatic conditions of the native habitat of the plants is supported by the data from both the greenhouse and field N uptake experiments. In both settings, plants showed preference for NO3- in the drier environments and showed preferences for NH4+ in the wetter environments. More convincingly, the species (S. pappophoroides) that appeared in different environments (e.g. from drier to wetter environments) showed different N uptake preferences ranging from NO3- in drier environments to NH4+ in wetter environments in the greenhouse condition (Fig. 2e); and ranging from NO3- in the driest site to no preference in the intermediate site and then to NH4+ in the wettest site in the field condition (Fig. 4e). The variation in N preferences clearly tracked the environmental condition under which they originally grew.
The absorbed N will eventually end up in the leaf/shoot issue and the foliar 15N signature is a good indicator for N uptake preference. However, our method of using both leaf and root tissue 15N signatures to indicate N uptake preference is valid for this particular experimental design based on the following reasons. NH4+ is only synthesized into amino acids in the root tissues (Engels & Marschner 1995). It is also known that up to 95% (with a mean of 58%) of the labelled 15N remained in root systems of various species after 48 h of 15N labelling in an earlier 15N labelling experiment with similar experimental setup (Wang, Mou & Jones 2006). In this study, we limited the N labelling duration to 48 h to minimize the effect of N mineralization and nitrification on 15N signatures. Therefore the relatively short labelling duration may not be long enough to allow assimilated N to transport from root tissues to shoot tissues and it is not surprising to see that the NH4+ preference and some of the NO3- preference (S. pappophoroides from Tshane, and S. suniplumis from Ghanzi) were shown only in the root 15N signatures.
The climatic switch of plant N uptake preference is likely determined by the NH4+/NO3- ratios of the habitats. Several lines of evidences support this argument. Firstly, previous field observations along the KT showed that NH4+/NO3- ratios are higher at the wettest end (Supporting Information Fig. S1) (Aranibar 2003; Feral et al. 2003). This NH4+/NO3- ratio gradient is likely controlled by both biological and hydrological factors. In the wetter end, there are higher decomposition rates (more NH4+ availability) and more nitrate leaching (Scanlon & Albertson 2003), leading to a higher NH4+/NO3- ratio. Our observed N uptake preference switches between the dry and wet habitats match the NH4+/NO3- ratio gradient of the KT. Secondly, there are several laboratory and field observations from distinct geographic regions showing NH4+/NO3- ratios of the sites affect plant N uptake preferences. For example, by using kinetic and compartment-analysis techniques with the radiotracer 13N to compare the efficiency of N acquisition from NH4+ and NO3- sources in the seedlings of Picea glauca (white spruce) of western Canada, it was found that the uptake of NH4+ is 20 times higher than that of NO3- from an equimolar solution; cytoplasmic concentrations of NH4+ are 10 times higher than that of NO3- (Kronzucker et al. 1997). Similarly, it was found that the N uptake by Pinus sylvestris (Scotch pine) in Scandinavia from a 15N-labelled solution of NH4+ is much higher (about 10-fold) than from a 15N-labelled solution of NO3- (Wallander et al. 1997). The chief form of inorganic N available for P. glauca and P. sylvestris, in their native habitats, is NH4+. On the other hand, Populus tremuloides (trembling aspen) and Pinus contorta (lodgepole pine), two early successional species, were found to show a very high NO3- utilization rates at high external NO3- concentrations, and their common habitats are disturbed sites where available N is predominant by NO3- (Min et al. 1998). In addition, Houlton et al. (2007) found that from drier sites to wetter sites the plants switch N source from NO3- to NH4+ and such switch is in accordance with the NH4+/NO3- ratio changes in Hawaii tropical forests. Based on our direct monitoring of N uptake preference and evidences from previous work, we think it is probably a ubiquitous phenomenon that soil NH4+/NO3- ratio in plants' native habitat determines plant N uptake preference across different ecosystems.
More significantly, through the comparison of the summary of field results (Fig. 2e) and of the greenhouse F1 generation results (Fig. 4e), our observations suggest that the plants may maintain the adaptation in N preference, at least for the first generation. The field experiments demonstrated that plants along the KT showed N uptake preferences that tracked the environmental conditions. When seeds from each location (with different environmental conditions) were collected, germinated and grown in the greenhouse with equal amounts of water and N being applied (which were different from their field conditions), the plants showed the same N preferences as the parent plants did in the field (Figs 2e and 4e), indicating that plants appear to maintain the parental adaptation (at least for the F1 generation) with respect to N uptake preference. This constancy in N preference is further supported by the additional greenhouse observations that when higher concentrations of NH4NO3 were applied to P. squarrosa and L. vulpiastrum, the uptake preferences for the form of N did not change (Fig. 3). Although the current findings do not distinguish genetic effects from maternal effects for the ‘imprint’ of N uptake preference, such consistent results imply that, for example, during climate change scenarios, plant communities (at least for herbaceous plants) may keep their original physiological traits and pass them on to future generations. Longer-term studies of plant communities are clearly necessary to infer the physiological responses to climate change. The similar responses in the greenhouse and the field settings also indicate that the observed patterns are not caused by bonding differences of soil particles and NH4+ among sites as the soil substrate was the same in the greenhouse experiment.
The ammonium nitrate fertilizer with differential labelling treatment provides a unique and direct way to test N uptake preference as shown in this study, though the preference observations were only tested for a 1:1 ratio. Previous studies showed that NO3- uptake can be depressed by the supply of NH4+ (e.g. Lee & Drew 1989), and the negative effect of NH4+ supply on NO3- uptake depends on the utilization of energy during NH4+ assimilation or the inhibitory effect of assimilatory products rather than on NH4+ per se (Engels & Marschner 1995). However, the NH4+ suppression effect on NO3- uptake was reported to be minimum when NH4+ concentration is under 1 mM (Breteler & Siegerist 1984; Engels & Marschner 1995). In the current experiment, the labelling solution was only around 0.02 mM, therefore the NH4+ suppression effect on NO3- uptake should be small, if any.
Amino acids and dissolved organic nitrogen (DON) are other potential nitrogen sources for plant uptake and plant uptake of amino acids have been found in various boreal ecosystems (e.g. Schimel & Chapin 1996; Persson & Nasholm 2001). In arid and semi-arid environments, however, there are very limited data on plant DON and amino acid uptake or even on DON and amino acid distributions. The water diffusivity in dry soils can range between 10−6 and 10−7 m2/year (for values of soil moisture typical of the root zone), such low diffusivity will probably limit the movement of DON and amino acid and therefore the plant uptake of DON and amino acid in arid and semi-arid environments. Although this study did not assess the role that either DON or amino acid plays in these ecosystems, it is worthwhile to explore their potential roles in future research.
In summary, this study has illustrated a climatic constraint on the preferential uptake of the different forms of inorganic N by plants and indicated that the N uptake preference is likely influenced by the NH4+/NO3- ratios in the native habitats of the plants. From an evolutionary point of view, it supports the local adaptation concept (Endler 1986; Chapin 1988). It is also along the line that the most abundant species rely on the most abundant N forms as shown in the study of McKane, Grigal & Russelle (1990). In addition, the results suggest that ‘imprinting’ of such preference exists in seeds. It showed that plant seeds retain the adaptation towards the N uptake preference of their parents, even when the abundances of NO3- and NH4+ changed. Practically, the maintenance of plant N uptake preference across generations provides an underlying mechanism explaining why, for example, a majority of the replanted conifer species in western Canada failed to survive as observed in a study of Kronzucker et al. (1997), that is, the newly replanted trees have the potential genetic imprint of NH4+ uptake preference whereas the soil in the new environment after clear cutting is dominated by NO3-. The maintenance of plant N uptake preference also has implications to seed bank use.
The project was funded by NASA-IDS2 (NNG-04-GM71G). We greatly appreciate the assistance in seedlings identification from Chris Feral and Kebonyethata Dintwe. Ms. Wendy Crannage provided tremendous help in the greenhouse maintenance. We also appreciate the teamwork and field assistance from Lydia Ries, Thoralf Meyer and Paolo D'Odorico at University of Virginia, Billy Mogojw at Harry Oppenheimer Okavango Research Center, University of Botswana. We thank Dr Howard Epstein for the thoughtful discussion at the beginning of the experimental design. The clarity of this manuscript is significantly improved by comments from Dr Manuel Lerdau, Dr Jin Wang, the ecology and isotope geochemistry discussion groups in Department of Environmental Sciences of the University of Virginia and Dr Lars Hedin from Princeton University. We are grateful for the comments from two anonymous reviewers and editor Dr Werner Kaiser.