Nutrient availability and nutrient use efficiency in plants growing in the transition zone between land and water

Introduction

The transition zone between terrestrial and freshwater habitats is highly dynamic, with large spatial and temporal fluctuations in environmental characteristics (Naiman & Decamps 1997; Ward et al. 2002; Hoffmann et al. 2006). These characteristics are closely coupled to the dynamics of the flow regime that mediates lateral instability in the plan form of the stream channel and to periodic floods with sediment deposition within the riparian areas (Kronvang et al. 2009). Over time, these processes have created a mosaic of habitats from lotic in-stream habitats, over a variety of water bodies or backwater habitats that are almost lentic in character, to a wide range of terrestrial habitats. The environmental characteristics of the various habitats in the transition zone differ, as do the constraints that shape the plant communities. Among these constraints is mechanical stress from wind or water movement, light and nutrient availability, and substrate characteristics, which are all considered very important for species composition across the land–water ecotone (Bornette & Puijalon 2011).

Different life forms are associated with habitats in the transition zone, including submerged species living permanently under water, amphibious species able to live both submerged and emerged, and terrestrial plants living solely emerged (Sand-Jensen & Frost-Christensen 1998). The availability of nutrients for these different life forms is likely to vary across and within the transition zone, reflecting processes taking place in the catchment and in the soil matrix that moderate the availability of nutrients (Cey et al. 1999; Smith et al. 2008). Nevertheless, we expect that the availability of nitrogen (N) and phosphorus (P) is generally higher for species living submerged within the stream channel than for species living emerged in the land–water ecotone, both because concentrations are expected to be higher due to continued supply with the flowing water and because nutrients can be taken up through the entire surface area of the plants (Madsen et al. 1998).

The photosynthetic nutrient use efficiency (PNUE), defined as the ratio of CO₂ assimilation rate to leaf organic N content (Poorter & Evans 1998), reflects the ability of the plants to utilise nutrients in the photosynthetic process, and can therefore be used to obtain insight into resource allocation in plants. The PNUE is independent of morphological plasticity and directly related to resource availability (Field & Mooney 1986). Low nutrient use efficiency is found in species that are stress-tolerant (Chazdon & Field 1987; Reich et al. 1994), whereas high nutrient use efficiency is found in competitive species that are normally fast growing and thus able to allocate high amounts of nutrients into the photosynthetic apparatus. We use the PNUE and the photosynthetic phosphorus use efficiency (PPUE) as measures of the nutrient use efficiency of the plants.

Besides being influenced by nutrient (i.e. N and P) availability, we believe that differences in the availability of inorganic carbon (C) across the land–water ecotone may affect PNUE and PPUE. Generally, the availability of inorganic C is substantially lower in the aquatic environment because of highly reduced diffusion rates in water and thick boundary layers (Raven 1984; Sand-Jensen & Frost-Christensen 1998). Therefore, the C fixation rate per unit nutrient may be low, consequently lowering the PNUE and PPUE of species living permanently submerged compared to species living partly submerged or species living emergent.

In the present study, we investigate the nutritional status of different plant types growing in the water ecotone in lowland stream areas in Denmark, and how the photosynthetic performance of these plant types is related to their nutritional status. Specifically, we hypothesised that: (i) tissue nutrient content differs among submerged, amphibious and terrestrial species, with higher content in submerged species; and (ii) the PNUE gradually increases from submerged species over amphibious to terrestrial species, reflecting differences in the availability of N and P relative to inorganic C across the land–water ecotone.

Material and Methods

Data analyses

The photosynthesis rates for aquatic species were calculated on a shoot dry weight (DW) basis and then converted to an area-based rate to be comparable with photosynthesis rates measured on terrestrial and amphibious species. To convert photosynthesis rates from DW to area, we used existing data on specific leaf area of the different species (A. Battrup-Pedersen, unpublished data). Photosynthesis rates for aquatic species were measured as production of O₂, and, again to be comparable with the rates measured on terrestrial and amphibious species, these were converted into uptake of CO₂ using the following expression:

The PNUE (μmol·CO₂ mol⁻²·N s⁻¹) was calculated from photosynthesis rates and tissue N content using the following expression:

Similarly, the PPUE (μmol·CO₂ mol⁻²·P s⁻¹) was calculated from photosynthesis rates and tissue P content:

Differences among life forms in C, N, P, photosynthesis rates, PNUE and PPUE were compared using anova analysis, and to test homogeneity we used Levene′s test. All statistical analyses were performed in R (R Foundation for Statistical Computing, Vienna, Austria).

Results

Carbon, N and P tissue content of plants of the submerged, amphibious and terrestrial groups differed significantly (Fig. 1). The C content of plants increased from submerged over amphibious to the terrestrial species group, both expressed on a shoot/leaf dry weight basis (anova P < 0.05) and on a shoot/leaf area basis (anova, P < 0.001; Fig. 1, Table 1). In contrast to the C content, N and P tissue contents were higher in submerged species compared to amphibious and terrestrial species (anova P < 0.001; Fig. 1, Table 1). The highest N and P tissue contents were found in Ranunculus peltatus and Potamogeton crispus, respectively, whereas the lowest N and P contents were in Glyceria maxima, an amphibious species (Table 1).

Nutrient limitation was assessed by analysing C:N as a function of percentage tissue N of the DW, and C:P ratios as a function of percentage tissue P of the DW (Duarte 1990). The C:N ratio declined markedly with increasing N tissue content, from a minimum of 0.78% DW to about 2% DW tissue N (x-axis), followed by a more gradual decline in the C:N ratio in the range of 2% DW to about 4% DW tissue N (Fig. 2), indicating that individuals with tissue N < 2% were N-limited, whereas individuals with tissue N from 2 to 4% had intermediate N availability. The C:P ratio also declined rapidly as P tissue content increased from a minimum of 0·018% DW to about 0.2% DW, followed by a more gradual decline to about 0.6% DW (Fig. 2), indicating that individuals with tissue P < 0.2% were P-limited, whereas individuals with tissue P content between 0.2 and 0.6% had intermediate P availability. On an individual basis, we found that 4% of the individuals had N content <2%, indicating N limitation, 8% had P content <0.2%, indicating P limitation, and 7% had lower content of both N (<2%) and P (<0.2%), suggesting co-limitation. There were no systematic differences in C:N and C:P ratios among the three plant groups as a whole. However, only a few aquatic specimens had tissue N and tissue P contents that could indicate limitation (<2% and <0.2%, respectively; data not shown), whereas about one-third of the terrestrial and amphibious specimens had tissue N and tissue P contents that could indicate limitation. Therefore, nutrient availability seems to be saturating for growth of most aquatic plant individuals but to a lesser extent for amphibious and terrestrial plant individuals.

Photosynthesis rates were lower in submerged species compared to amphibious and terrestrial species (anova, P < 0.001; Fig. 3, Table 1). Callitriche sp. had the lowest photosynthesis rate of all submerged species (average 0.29 μmol·CO₂ m⁻²·s⁻¹) and R. peltatus had the highest rate (average 3.12 μmol·CO₂ m⁻²·s⁻¹), although still lower than the rates found in terrestrial and amphibious species. The photosynthesis rate of amphibious species ranged from 9.08 μmol·CO₂ m⁻²·s⁻¹ (average for G. maxima) to 23.46 μmol·CO₂ m⁻²·s⁻¹ (average for Equisetum fluviatile), while the rate of terrestrial species ranged from 8.20 μmol·CO₂ m⁻²·s⁻¹ (average for Carex nigra) to 13.9 μmol·CO₂ m⁻²·s⁻¹ (average for Epilobium angustifolium; Table 1). PNUE and PPUE were also lower in submerged species than in amphibious and terrestrial species (anova P < 0.001; Fig. 3). Callitriche sp. had the lowest PNUE among all species (8.87 μmol·CO₂ mol·N⁻¹ s⁻¹) and P. lucens the lowest PPUE (183 μmol·CO₂ mol·P⁻¹ s⁻¹), while E. angustifolium, a terrestrial species, had both the highest PNUE (166.97 μmol·CO₂ mol·N⁻¹ s⁻¹) and the highest PPUE (2596 μmol·CO₂ mol·P⁻¹ s⁻¹; Table 1) measured.

Discussion

As predicted, tissue nutrient content was generally higher in submerged species than in amphibious and terrestrial species. The tissue content measured was within ranges previously reported for submerged species (1.2–6.3 N w/w; Baattrup-Pedersen et al. 2013a; 0.05–0.6% P; Demars & Edwards 2008). The higher content in submerged species likely reflects that these species can take up nutrients from both the continually renewed interstitial water in the sediment and from open water (Cedergreen & Madsen 2002), and therefore have access to higher amounts of nutrients than amphibious and terrestrial species, despite the fact that these species groups live side-by-side in the land–water ecotone.

To investigate potential nutrient limitation in the three species groups, we assessed threshold values derived from C:N to tissue N relationships and C:P to tissue P relationships, as suggested by Duarte (1990). We found that the three species groups overlapped in C:N to tissue N relationships and in C:P to tissue P relationships and, additionally, that threshold values indicating limitation was <2% for tissue N and <0.2% for tissue P. These threshold values were very close to previously reported thresholds for nutrient limitation of 1.8% for N and 0.2% for P in seagrasses using the same approach (Duarte 1990). They were also very close to values reported by Gerloff & Krumbholz (1966) for angiosperm aquatic plants: 1.3% for tissue N and 0.13% for tissue P. Using the derived threshold values, the number of individuals observed that could possibly suffer from nutrient limitation was comparable among the three plant groups. Taken together, we only found indications of potential N limitation in 4% of individuals, potential P limitation in 8% of individuals, and potential co-limitation by N and P in 7% of individuals. Submerged species had the lowest number of individuals experiencing potential nutrient limitation, as expected from their high nutrient tissue content compared to the contents measured in amphibious and terrestrial species.

We also considered the possibility of assessing nutrient limitation by applying tissue N:P ratios, as suggested by Koerselman & Meuleman (1996). The N:P ratios of leaves can, however, be higher compared to that of whole plant tissues (T. Riis & A. Baattrup-Pedersen, unpublished data), reflecting the particular high content in photosynthetically most active tissues (Feller et al. 2008). This is primarily due to Rubisco, which is the predominant protein in leaves of C₃ plants and may contribute up to 50% of the soluble leaf protein (Spreitzer & Salvucci 2002). Consequently, ratios obtained here based on leaf N and P are not directly comparable with the suggested ratios of Koerselman & Meuleman (1996) that are based on whole plant content. Furthermore, it has also been observed that optimal N:P ratios for plant growth may change over time, depending on plant growth rate and age of the tissues, and also may vary over seasons and with the availability of light (Gusewell & Koerselman 2002), which all add some uncertainty to the use of N:P ratios in assessing nutrient limitation. Finally, C:N and C:P ratios are likely to better reflect the relative resource availability than N:P ratios in general, since C:N and C:P threshold values take into account the amount of C in the plant tissues, which is normally less variable than the N and P content.

As predicted, we found that the PNUE declined from terrestrial over amphibious to submerged species. Accordingly, the higher amount of nutrients likely allocated to supporting structural tissues in the terrestrial and amphibious species groups than in the aquatic species group (Poorter & Evans 1998; Onoda et al. 2004) seems to be more than offset by more optimal allocation of N to and between photosynthetic components in these species compared to submerged species. We infer that this difference in PNUE among the groups can be mediated by contrasting CO₂ availability in the terrestrial and aquatic environments. Although the concentration of inorganic C in water can be high compared to that in air, the availability of inorganic C for photosynthesis will be low because of 10⁴ times slower diffusion of gases in water and a long diffusion path through the boundary layer (100–500 μm; Madsen & Sand-Jensen 1991). Therefore, to obtain the same flux of CO₂ across the boundary layer in air and water and, hence, the same rate of photosynthesis, the CO₂ concentration should be more than 10⁴ higher in water. However, the concentration of CO₂ seldom exceeds 25 times equilibrium concentrations (Rebsdorf et al. 1991; Sand-Jensen & Frost-Christensen 1998). The lower PNUE of submerged species may thus reflect that these species are better acclimated to lower availability of inorganic C than terrestrial species, yielding differences in nutrient allocation patterns to and between the various photosynthetic components among the species groups. It was previously shown that high availability of CO₂ can lower the amount of N allocated into Rubisco without negatively affecting carboxylation capacity of the plants, probably due to down-regulation of the photosynthetic apparatus (Ge et al. 2012). Consequently, the higher PNUE of terrestrial and amphibious species than of submerged species may reflect the fact that these species allocate proportionally more N into processes of importance at high CO₂ availability, i.e. enzymes involved in substrate regeneration. Furthermore, many submerged species possess C concentrating mechanisms that serve to improve the availability of inorganic C in the aquatic environment (Raven et al. 2011), which may also lower the PNUE of species in the submerged species group. Thus, the capacity and efficiency of bicarbonate (HCO₃⁻) uptake, which is by far the most widespread C concentrating mechanism in aquatic plants (Sand-Jensen 1983), are largely dependent on the amount of tissue N in the plants, indicating that the N cost associated with HCO₃⁻uptake can be substantial (Madsen & Baattrup-Pedersen 1995; Baattrup-Pedersen et al. 2013b).

In addition to differences in nutrient allocation to and among photosynthetic components, it is also possible that the lower PNUE of submerged species reflects that higher amounts of nutrients are stored in this species group compared to the amphibious and terrestrial species groups because of the relative availability of N and P to inorganic C is higher in the aquatic environment. Most perennial plants can store N when uptake from the external environment exceeds immediate growth requirements. This has been interpreted as an adaptation whereby growth is better matched to N availability (Chapin 1980; Proe & Millard 1994). Nitrogen may be stored in a variety of forms in plants, but in foliage it is most often stored as inorganic N (nitrate or ammonium), free amino acids or proteins (Millard 1988), which may all lower the PNUE. Similarly, excess P can be stored in vacuoles (Bieleski 1973), which may lower the PPUE since it is not used in compounds affecting the photosynthetic performance of the plants, e.g. nucleic acids, sugar phosphates, ATP and phospholipids, which all have important roles in the photosynthesis process (Hidaka & Kitayama 2009).

Finally, lower PNUE in submerged species could also reflect that species within this group are adapted to lower light availability and therefore invest proportionally more N into light capture than terrestrial and amphibious species, translating into higher light use efficiency. The restricted availability of light in the aquatic environment compared to the terrestrial environment also implies that photosynthesis saturates at a lower irradiance in submerged species (ca. 250 μmol·m⁻²·s⁻¹) compared to amphibious and terrestrial species (ca. 1800 μmol·m⁻²·s⁻¹). Consequently, at saturating N the PNUE of amphibious and terrestrial species will be higher.

In conclusion, we found that ca. 20% of plant individuals growing in the transition zone between land and water were nutrient-limited and that these were primarily amphibious and terrestrial species. Additionally, we found that the nutrient use efficiency increased from submerged over amphibious to terrestrial species. This pattern likely reflects differences in nutrient allocation to and within the photosynthetic apparatus among the species groups, mediated by large differences in the relative availability of N and P to inorganic C in the aquatic and terrestrial environments. Taken together, our results indicate that enhanced nutrient loading may affect the relative abundance of the three species groups in the land–water ecotone of stream ecosystems. Thus, species belonging to the amphibious and terrestrial groups are likely to benefit more from enhanced nutrient availability in terms of faster growth, compared to aquatic species, and also that this can be detrimental to aquatic species growing in the land–water ecotone, e.g. Ranunculus and Callitriche. Enhanced growth of e.g. G. maxima may intensify competition for light in the transition zone for aquatic species and lead to a decline in their relative abundance, in particular in streams subjected to low summer flows and with minimal physical stress – a situation that is likely to be intensified in the coming years as lower summer precipitation associated with climate change becomes more common.