- Top of page
- Materials and methods
Dramatic, but predictable, changes in light availability occur beneath the canopy of temperate deciduous forests through the seasons (Hicks & Chabot 1985). Herbaceous plants exhibit two distinct strategies for coping with the deep shade cast by the overstorey canopy in summer. Spring ephemeral herbs avoid shade stress by exploiting the narrow window of full sun and adequate temperatures prior to canopy closure whereas summergreen and evergreen species have adapted to tolerate the long summer shade period (Hicks & Chabot 1985). Spring ephemeral herb abundance tends to be associated with high soil fertility (Al Mufti et al. 1977; Rogers 1982; Host & Pregitzer 1991), possibly because the rapid photosynthetic rates characteristic of this life history require an abundance of water and mineral nutrients (Grime 1979, 1994). In contrast, deciduous forests on infertile sites are expected to be dominated by summergreen and evergreen herbs (Grime 1994).
Spring ephemerals might therefore be expected to possess a suite of characteristics associated with plants from high-resource habitats (i.e. rapid growth rate, high physiological capacity for resource capture and low nutrient-use efficiency) whereas shade-tolerant herbs would show the opposite traits, typically associated with low-resource habitats (Grime 1979, 1994; Vitousek 1982; Schulze & Chapin 1987; Chapin 1991). Above-ground resource acquisition, reflected by photosynthetic capacity, is indeed higher in spring ephemerals (Sparling 1967; Taylor & Pearcy 1976; Rothstein 1999). Patterns of mineral nutrient capture should parallel those of photosynthetic capacity as these two processes are believed to be interdependent (Bloom et al. 1985; Campbell et al. 1991; Chapin 1991; Grime 1994). However, to the best of our knowledge, relationships between life history and the physiology of mineral nutrient capture by deciduous forest herbs have never been investigated.
The sugar maple-basswood–Osmorhiza ecosystem type of north-western lower Michigan, USA, is characterized by both the highest rates of N mineralization and nitrification in the region (Zak & Pregitzer 1990), and an abundant and diverse herbaceous plant community (Host & Pregitzer 1991). We investigated patterns of N use in three members of this community: Alliumtricoccum Ait., a spring ephemeral; Violapubescens Ait., a summergreen species; and Tiarellacordifolia L., a semievergreen (sensuMahall & Bormann 1978). We therefore compared N-use among plant species adapted to temporal niches of variable light availability, while holding soil fertility constant, rather than among plant species adapted to habitats differing in soil fertility (as in Ingestad 1976; Aerts & Caluwe 1994; Vázquez de Aldana & Berendse 1997). We hypothesized that N uptake capacity would parallel differences in photosynthetic capacity among species (Tiarella < Viola < Allium), as well as the seasonal changes in photosynthetic capacity observed in Viola and Tiarella (Rothstein 1999). We also hypothesized that nitrogen-use efficiency (NUE; or the efficiency of biomass production with respect to nitrogen) would be inversely related to leaf N concentration and positively related to leaf longevity, and species would thus be ranked in the reverse order to N uptake. In particular, we expected that the components of NUE (sensuBerendse & Aerts 1987) would vary between species, with nitrogen productivity decreasing, and mean residence time increasing, from Allium to Viola to Tiarella.
- Top of page
- Materials and methods
There were significant differences among species and seasons in both total leaf N and leaf NO3− concentration (Fig. 2). Spring-time leaf N concentrations in Allium and Viola were approximately 70% greater than those of Tiarella (Fig. 2a), but while leaf N dropped dramatically in Viola from spring to summer, there were no seasonal changes in Tiarella. A substantial portion of the N in leaves of Viola and Tiarella was in the form of NO3− (1–16%), but NO3− was barely detectable in the leaves of Allium (Fig. 2b). Differences in leaf N concentration paralleled differences in photosynthetic rates (Rothstein 1999), such that there were no differences among species in springtime PNUE (Table 1; P = 0.653).
Figure 2. Seasonal and interspecific patterns of (a) leaf N and (b) leaf NO3− concentration in Allium tricoccum (●), Viola pubescens (▪) and Tiarella cordifolia (▴). Allium data are confined to the spring because of its limited leaf display. Values represent means (±1 SE) of five ramets of each species. Means within a panel with the same letter are not significantly different according to Tukey’s HSD test (P > 0.05).
Download figure to PowerPoint
Table 1. Components of nitrogen-use efficiency in Allium, Viola and Tiarella: springtime photosynthetic nitrogen-use efficiency (PNUESPR), springtime nitrogen productivity (ASPR), annual nitrogen productivity (A), mean residence time of nitrogen (MRT), and nitrogen-use efficiency (NUE). Values are means ± 1SE. Means in the same column followed by the same letter are not significantly different according to Bonferroni’s multiple comparison test
|Species||PNUESPR (µmol CO2 g N−1 s−1)||ASPR (g biomass g−1 N day−1)||A (g biomass g−1 N day−1)||MRT (years)||NUE (g biomass g N−1)|
|Allium||8.6 ± 0.2a||0.98 ± 0.16a||12.4 ± 2.1a||2.84 ± 0.36a||37.6 ± 8.8a|
|Viola||7.8 ± 0.5a||0.42 = 0.05b||19.4 ± 2.5a||2.05 ± 0.21a,b||42.3 ± 8.3a|
|Tiarella||8.8 ± 1.2a||1.09 = 0.11a||31.2 ± 3.7b||1.69 ± 0.21b||55.9 ± 9.2a|
The roots of all three species took up less 15NO3− (Fig. 3a) than 15NH4+ (Fig. 3b), although their relative preferences for NH4+ (i.e. the difference between rates of NH4+ and NO3− uptake) diminished from Allium to Viola to Tiarella. Nitrate and NH4+ uptake both followed the same pattern among species, decreasing from Tiarella to Viola to Allium. Specific uptake capacity of both ions tended to decrease over time for all species, although these changes were, for the most part, not statistically significant.
Figure 3. Seasonal and interspecific patterns of (a) 15NO3− and (b) 15NH4+ uptake. Values represent means (±1 SE) of five ramets of each species. Means within a panel with the same letter are not significantly different according to Tukey’s HSD test (P > 0.05). Symbols are as follows: Alliumtricoccum (●), Viola pubescens (▪) and Tiarellacordifolia (▴).
Download figure to PowerPoint
In situ N uptake by transplants of Viola and Tiarella (Fig. 4a) closely mirrored biomass accumulation (Fig. 4b). Allium, however, accumulated little N in the spring despite a very rapid relative growth rate, and even maintained net N uptake in the summer, despite a decline in biomass during this period. The negative relative N uptake in June was probably the result of N loss during leaf senescence.
Figure 4. (a) Relative biomass growth and (b) relative N uptake of Allium, Viola and Tiarella through 1 full year. Values represent means (±1 SE) of 10 plants, except for the final three samples of T.cordifolia, which were of five plants each. Final growth and N uptake values with the same letter are not significantly different according to Bonferroni pairwise comparisons. Symbols are as follows: Alliumtricoccum (●), Viola pubescens (▪) and Tiarellacordifolia (▴).
Download figure to PowerPoint
Despite the fact that Tiarella roots had the greatest specific uptake capacity for NH4+ and NO3− (Fig. 3), Viola exhibited the most rapid rate of relative N uptake in situ (Fig. 4). This is probably due to the higher proportion of Viola’s total biomass in roots (averaged across all seasons, root mass ratio was significantly greater in Viola[0.46 ± 0.04] than in Allium[0.24 ± 0.02] or Tiarella[0.21 ± 0.04]). After 1 full year of growth, net relative biomass gain was significantly greater in Tiarella compared with Allium (P = 0.032), while Viola was intermediate, and not significantly different from either. Relative N uptake by Allium was significantly lower than that of Viola (P = 0.002), and marginally lower than that of Tiarella (P = 0.072). Values for the latter two species were equivalent.
These seasonal patterns of growth and N uptake gave rise to differences among the three species in the components of NUE (Table 1). Viola had a significantly lower springtime N productivity than the other two species (P = 0.001), although PNUE did not vary at this time. Over an entire year of growth, however, N productivity increased significantly from Allium to Viola to Tiarella (P = 0.001). Mean residence time of N followed the opposite pattern decreasing from Allium to Viola to Tiarella (P = 0.023). Therefore total NUE, the product of A and MRT, did not differ between species (P = 0.388).
Both Allium and Viola showed seasonal shifts in N allocation associated with leaf mortality, with N moving from shoot to storage pools (Fig. 5). In contrast, shoots remain the dominant pool of N in Tiarella throughout the year. In the spring, when all three species had green leaves and there was direct light reaching the forest floor, Allium (0.69 ± 0.01) and Tiarella (0.62 ± 0.02) allocated proportionately more N to their shoots than did Viola (0.50 ± 0.03; P < 0.001).
Figure 5. Seasonal patterns of N allocation to shoots (leaves, stems and flowers), roots and storage (bulbs or rhizomes) in (a) Alliumtricoccum, (b) Violapubescens, and (c) Tiarellacordifolia. Values represent means (+1 SE) of 10 ramets of each species, except for Tiarella in the autumn, in which means are of five ramets. All N values are expressed relative to the initial biomass of each plant.
Download figure to PowerPoint
Patterns of 15N partitioning in the same plants (Table 2) are consistent with patterns of total N allocation. In all three species, 15N that was in roots immediately following labelling had moved to leaves by the second harvest. After this, 15N in Allium remained as a discrete pool (significantly the largest; Table 2) whereas the isotope dispersed rapidly throughout the entire plant in the other species so that no tissue pool could be distinguished statistically from any of the others after September 1997. After leaf senescence in Allium, the bulk of the label was translocated back to the bulb, which is consistent with the apparent resorption of 70% of green-leaf N in free-growing Allium plants (Table 3), and the increase in total N in the storage pool over this same period (Fig. 5). The relatively low resorption observed in Tiarella (Table 3) is consistent with leaves remaining the largest pool of isotope, even if not significantly so (Table 2). Viola roots always contained a relatively large 15N pool, although not always the largest and not significantly so after the initial harvest.
Table 2. Changes in the distribution of 15N label in Allium, Viola and Tiarella through 1 year of growth. Values are means (±1 SE of the 15N label in each tissue, adjusted for initial plant size (µg 15N g−1 initial dry weight). Shoot includes leaves, stems and reproductive tissues. Storage includes bulbs and rhizomes. Values in bold type are significantly greater than all non-bold values within that sample date, using Tukey’s HSD multiple comparison procedure (P ≤ 0.05)
| ||Distribution of 15N label (µg 15N g−1 initial dry weight)|
| ||April 1997||May 1997||June 1997||September 1997||October 1997||November 1997||April 1998|
|Allium|| || || || || || || |
|Shoot|| 0.5 ± 0.11|| 83.9 ± 12.59|| 6.38 ± 1.14*||–||–||–|| 8.5 ± 1.18|
|Storage|| 7.1 ± 0.77|| 13.8 ± 1.63|| 68.5 ± 7.21||–||71.0 ± 19.55||–||41.3 ± 3.26|
|Roots|| 57.2 ± 8.05|| 13.9 ± 1.63|| 12.7 ± 1.47||–||19.1 ± 6.90||–||12.2 ± 1.42|
|Viola|| || || || || || || |
|Shoot||–||–||108.7 ± 8.65||67.4 ± 3.92||–||–||10.0 ± 1.14|
|Storage|| 32.5 ± 2.84||–|| 11.0 ± 1.21||14.0 ± 1.45||31.0 ± 4.10||–||19.8 ± 1.76|
|Roots||109.0 ± 7.00||–|| 22.5 ± 2.03||22.7 ± 2.00||35.4 ± 4.6||–||32.3 ± 2.81|
|Tiarella|| || || || || || || |
|1997 shoot||–|| 32.4 ± 7.71||131.1 ± 15.60||–||92.8 ± 13.50||64.4 ± 11.25||41.8 ± 15.44|
|1998 shoot||–||–||–||–||–||–||14.0 ± 1.96|
|Rhizome||–|| 47.6 ± 6.16|| 16.7 ± 2.12||–||32.8 ± 13.56||12.5 ± 2.84||16.0 ± 2.05|
|Roots||–||121.1 ± 12.43|| 35.3 ± 5.63||–||25.6 ± 4.91||10.1 ± 1.90||14.0 ± 2.86|
Table 3. Apparent N resorption from leaves of Allium and Tiarella. The differences in N content between green and shed leaves were significant for both species using a Student’s t-test (P < 0.001; n = 5). Values for leaf N are means ±1 SE
| ||Green leaf N (mg g−1)||Shed leaf N (mg g−1)||Resorption efficiency (%)|
|Allium||49.8 ± 2.24||14.8 ± 1.34||69.8|
|Tiarella||29.1 ± 0.57||23.1 ± 0.84||20.6|
- Top of page
- Materials and methods
We initially proposed that interdependence of above- and below-ground resource capture would lead to N uptake capacity and photosynthetic activity varying in parallel, both within and among species. However, we found that the species showed opposite rankings for rates of N uptake and photosynthesis, and patterns within species were inconsistent. We also proposed that NUE should increase with decreasing leaf N concentrations and increasing leaf longevity, but overall NUE did not differ among species. Moreover, its components (A and MRT) differed in the opposite direction to our predictions. Above- and below-ground physiology can therefore be uncoupled, and a variety of combinations of assimilation rates, allocation patterns and tissue phenologies used to balance above- and below-ground resource capture. Patterns of N uptake and NUE also varied widely in response to ecological and life-history factors unrelated to N availability.
Species rank for photosynthetic capacity (Tiarella < Viola < Allium;Rothstein 1999) was the reverse of their rank for specific N uptake capacity (Fig. 3). Within species, Viola alone exhibited the predicted parallel response, with both photosynthetic capacity (Rothstein 1999) and N uptake capacity (Fig. 3) declining by a factor of two from spring to summer. Specific N uptake in Tiarella did not change from one season to another whereas its photosynthetic capacity decreased from spring to summer and then increased in autumn (Rothstein 1999). Perhaps the most unusual relationship between N uptake and photosynthesis was in Allium, which, despite the fact that it was leafless, sustained its NH4+ uptake capacity throughout the summer. Above- and below-ground resource capture, in terms of instantaneous physiological activity, are clearly not interdependent in Allium and Tiarella.
These disjunctions disappear for Tiarella when N uptake is measured in situ, with N uptake closely tracking biomass growth throughout the year (Fig. 4). The disjunction was, however, reinforced by in situ measurements on Allium, where net uptake of N continued over the summer months despite a loss of biomass. Clearly, the rapid photosynthetic rates necessary to sustain the spring ephemeral life history (Sparling 1967; Taylor & Pearcy 1976; Rothstein 1999) require a large investment of N. However, Allium meets this demand from rapid remobilization of N from its bulb, accumulated as a result of uptake throughout the year, rather than by rapid springtime uptake. Indeed, even though Allium shows efficient resorption of N, spring uptake of N is outweighed by losses of N due to leaf senescence. However, in other spring ephemeral species, root senescence coincides with leaf senescence (e.g. Erythroniumamericanum; D. Rothstein personal observation), and sufficient N uptake must therefore occur in the spring.
Predictions of interdependence are based on the premise that, to optimize growth, plants must maintain a balance between acquisition of above- and below-ground resources (Bloom et al. 1985). Although in Tiarella and Allium physiological processes measured over short time scales appear to violate this assumption, the whole plant over its entire annual growth cycle does not. The very low physiological activity of Tiarella leaves relative to its roots is compensated for by a greater allocation to leaves relative to roots. In contrast, Allium maintains its balance via differences in above- and below-ground phenology, and an ability to store and remobilize N (Tables 2 & 3).
While both high resorption efficiency and extended leaf lifespan are viewed as mechanisms of increasing MRT (Eckstein et al. 1999), differences here were determined primarily by differences in resorption efficiency, and MRT was actually inversely related to leaf lifespan. Despite having the shortest leaf lifespan, Allium’s efficient resorption and storage of leaf N resulted in very low annual N losses. In contrast, Tiarella’s poor resorption efficiency resulted in large losses of leaf N, even though it holds its leaves for the entire growing season. Although we are limited by a lack of data regarding leaf shedding in Viola, it appears from the changes in 15N pools (Table 2), and the autumnal increase in storage pools of N (Fig. 5), that Viola falls between Allium and Tiarella in its efficiency of leaf N resorption, contributing to its intermediate MRT. The unexpected inverse relationship between leaf lifespan and MRT may be explained by the fact none of these species has leaves that persist through more than one growing season, thus constraining any positive effects of leaf longevity on MRT. Furthermore, longer leaf display may actually increase the opportunity for N losses from the plant via leaching, particularly given the high leaf concentrations of soluble NO3− in Viola and Tiarella.
Patterns of tissue phenology and N allocation clearly exert more influence over A than does leaf N concentration. High springtime A in Allium and Tiarella is associated with a higher proportional allocation of N to shoots, relative to Viola (Fig. 5), rather than higher leaf N concentration (similar in Viola and Allium, lower in Tiarella). Differences in springtime A should approximate differences in maximal, instantaneous A, given that this period has light, temperature and moisture conditions that are the most conducive to growth, but in this rapidly changing environment, annual and springtime N productivity followed very different patterns among species. Not surprisingly, extended, slow C assimilation (Tiarella) resulted in greater annual N productivity than rapid C assimilation confined to a brief window of time (Allium).
Berendse & Aerts (1987) postulated that there were evolutionary trade-offs between traits leading to high A and those leading to high MRT, such that low-nutrient habitats should be dominated by plants with high MRT and low A whereas high-nutrient habitats should be dominated by plants with high A and low MRT. Even though our three species co-occur in a single high-nutrient habitat, their differences in life history still produced the expected negative relationship between A and MRT (Table 1). In fact, we found nearly as much variation in A and MRT among co-occurring species of differing phonologies as other researchers have found among species associated with habitats of widely varying N availability. Compare the 2.5-fold difference in A, and 1.7-fold difference in MRT, between Tiarella and Allium to the 2.3-fold difference in A, and a 1.9-fold difference in MRT, between a perennial grass species associated with acidic, nutrient-poor meadows and one associated with heavily fertilized pastures (Vázquez de Aldana & Berendse 1997).
Overall, the below-ground resource acquisition by deciduous-forest herbs does not relate to above-ground resource acquisition in nearly as clear-cut a manner as predicted, i.e. physiological rates of N uptake do not simply mirror rates of photosynthesis. However, patterns of allocation to, and phenology of, above- vs. below-ground structures appeared to compensate for disjunctions in physiology, resulting in relatively balanced C and N acquisition in all three species. Finally, life-history adaptations that appear to be related primarily to light availability also have important consequences for plant nitrogen economy.