Impact of Nitrogen Deposition on the Species Richness of Grasslands

See allHide authors and affiliations

Science  19 Mar 2004:
Vol. 303, Issue 5665, pp. 1876-1879
DOI: 10.1126/science.1094678


A transect of 68 acid grasslands across Great Britain, covering the lower range of ambient annual nitrogen deposition in the industrialized world (5 to 35 kg Nha–1 year–1), indicates that long-term, chronic nitrogen deposition has significantly reduced plant species richness. Species richness declines as a linear function of the rate of inorganic nitrogen deposition, with a reduction of one species per 4-m2 quadrat for every 2.5 kg Nha–1 year–1 of chronic nitrogen deposition. Species adapted to infertile conditions are systematically reduced at high nitrogen deposition. At the mean chronic nitrogen deposition rate of central Europe (17 kg Nha–1 year–1), there is a 23% species reduction compared with grasslands receiving the lowest levels of nitrogen deposition.

Conservation of biodiversity underpins some of the largest and most ambitious environmental legislation in the world (1). Most of the focus of this legislation is on mitigating damage to ecological communities from direct environmental degradation such as land clearing. Although chronic “low-level” stresses such as air pollution are also considered to damage biodiversity, their actual importance is not well understood.

Because nitrogen (N) is the limiting nutrient for plant growth in many terrestrial ecosystems (2), atmospheric deposition of reactive N has the potential to reduce plant species richness (the number of species in a given area—an important component of biodiversity) through favoring species better adapted to high nutrient levels (3). Through intensive agriculture and fossil fuel combustion, humans have greatly enhanced the global emission and deposition of fixed N over the past 50 years (4). Nitrophilous species have increased, and N-sensitive vegetation has declined in European peatlands, heathlands, grasslands, and forests since the mid-20th century (3, 5, 6). Similar changes can be induced by experimental addition of high levels of N (>25 kg N ha–1 year–1) (710). Despite these findings, there has as yet been no clear evidence that enhanced deposition of a nonpoint pollutant over a large region has had any impact on terrestrial biodiversity.

We surveyed a grassland community in Great Britain to determine whether any significant variability in plant species richness could be detected and, if so, whether it was related to regional-scale levels of atmospheric pollution (inorganic N, inorganic S) deposition. The location allowed us to consider as large an area as possible while remaining within one climatic region with well-described vegetation communities. We chose an Agrostis-Festuca grassland (11), because this vegetation type is common throughout Europe, Australia, and North America and makes up economically valuable pastureland. Sixty-eight sites were sampled in 2 m by 2 m quadrats (5 replicate quadrats per site) along the gradient of atmospheric N deposition (5 to 35 kg N ha–1 year–1 wetfall and dryfall) during the summers of 2002 and 2003 (Fig. 1A) (12, 13). Most (a mean of 70%) of the inorganic N deposited is as reduced N (NH3, NH4+) (Fig. 1B). Plant species presence and abundance were measured in each quadrat, and from these data we calculated site species richness. We also considered the Shannon diversity index (H), which takes into account both species number and relative abundance (13, 14).

Fig. 1.

(A) Total inorganic N deposition (kg N ha–1 year–1) to the United Kingdom (12), with field site locations shown. (B) The deposition of reduced (NH3, NH4+) and oxidized (NO, NO2, NO3) N across the sites.

For each site, we compiled a data set of the potential drivers on plant species richness, including all of those described as globally important (15): nine chemical environmental factors [deposition of reduced inorganic N (NH3, NH4+), oxidized inorganic N (NO, NO2, NO3), and total inorganic N; deposition of SO42–; acid deposition (total inorganic N + SO42–); topsoil (A or O horizon) pH and subsoil (30 to 40 cm) pH; topsoil percentage of N; and topsoil C:N ratio]; nine physical environmental factors (mean annual temperature and precipitation, actual and potential evapotranspiration, soil moisture deficit, litter cover, altitude, slope, and aspect); and two human modifications (grazing intensity and enclosures) (table S1). These variables were entered into a stepwise multiple regression with site species richness as the dependent variable.

Species richness showed large variability, ranging from a mean of 7.2 to 27.6 species per quadrat. Species richness was highest in four sites in western Scotland, then showed high variability from west to east along an overall negative trend (fig. S1A). There was a clearer latitudinal gradient, with some sites of high species richness in the extreme south, and a gradual increase in species richness from south to north (fig. S1B). N deposition showed the opposite trends (fig. S1, C and D).

Of 20 variables measured to account for the variability in species richness, total deposition of inorganic N (Ndep, kg N ha–1 y–1) was the most important predictor, explaining more than half of the variation in the number of species per quadrat (Fig. 2A and Eq. 1). The trend was linear and negative, indicating that for every 2.5 kg ha–1 year–1 of inorganic N deposited on an acid grassland, a mean of one additional species is excluded from a randomly placed 4-m2 quadrat: Embedded Image Embedded Image Embedded Image(1) After accounting for N deposition, mean annual precipitation (MAP, mm) explained an additional 8% of variability in species richness. A further 5% was explained by the A horizon soil pH (Top pH, Fig. 2B) and 3% by altitude (Alt, m). In total, 70% of the variability in species richness could be explained by these four variables: Embedded Image Embedded Image Embedded Image Embedded Image(2) The secondary variables—precipitation (+), pH (+), and altitude (–)—have all been identified in previous studies as potential influences on vegetation biodiversity (15, 16). A closer examination of the effect of these variables revealed that the addition of mean annual precipitation and altitude is due to a group of seven grasslands in western and central Scotland, with a mean annual precipitation exceeding 2700 mm. Removing these sites from the database and re-running the stepwise regression with the 20 predictor variables eliminated both MAP and altitude as significant, leaving N deposition (r2 = 0.44, P < 0.0001) together with pH as the most significant model (r2 = 0.50, N = 61, P < 0.02). No other variable was significant in this regression.

Fig. 2.

(A) Acid grassland species richness plotted against N deposition for 68 field sites visited in the summers of 2002 and 2003. The regression line shown is Eq. 1. (B) Plant species richness versus N deposition and topsoil pH. The regression equation shown is: Plant Species Richness = 6.63 + 3.40(Top pH) – 0.316(Nin); r 2 = 0.61, N = 68, P < 0.004.

Topsoil pH, which ranges from 3.7 to 5.5, shows a fairly uniform distribution across both longitude and latitude and across the gradient of N deposition (Fig. 2B). Soil pH can be considered to modify the effect of N deposition: At any level of N deposition, sites with a higher soil pH are predicted (based on this analysis) to show higher species richness. Soil pH is in part due to the local site characteristics (parent material, organic matter content, management history) and in part to the long-term rate of acid deposition (16). If soil pH is “forced” as the first variable in the regression, N deposition is still highly significant (P < 0.007).

All of the other independent variables in our study were either not significantly correlated to species richness or were not significant after accounting for the effect of N deposition (table S1). Temperature and evapotranspiration have been shown in other studies to positively correlate with global or regional species richness (15). Our study found the opposite effect, with mid- and lower-latitude sites generally showing lower species richness than higher latitude sites (fig. S1B). This indicates that temperature is not responsible for the observed variability in species richness.

Somewhat surprisingly, there was no relationship between species richness and topsoil percentage of N or C:N, and there was no relationship between these variables and the rate of N deposition. This may be due to the wide variability in organic matter content of these soils, which ranged from organic-poor mineral soils to peat (%C = 2.3 to 40%). There was no relationship between species richness and percentage of litter cover, aspect, slope, presence/absence of enclosures, or level of grazing. SO42– deposition was weakly correlated to species richness, but again was not significant if N deposition was accounted for. Following this, acid deposition (SO42– + N) was a poorer predictor of species richness than N deposition alone.

The statistical analyses together with examination of individual trends show that, of the variables considered, the observed variability in species richness is best explained by N deposition. We next determined whether certain species are systematically missing from the low species-richness plots, and if so, if these species share any common characteristics. We determined, using canonical correspondence analysis (13), and confirmed by reviewing the relevant data, that there is in fact a consistent loss of certain species: the forbs Plantago lanceolata (ribwort plantain), Campanula rotundifolia (harebell), and Euphrasia officinalis (eye-bright); the grass Molinia caerulea (purple moor grass); the shrub Calluna vulgaris (heather); and the moss Hylocomium splendens (mountain fern moss). These six species constitute 21% of the maximum species number we observed. The only plant found to substantially increase in occurrence and cover at higher levels of N is Hypnum cupressiforme, a moss that can survive relatively high levels of air pollution (17).

Campanula, Euphrasia, and Calluna are indicators of infertile sites (18), and Campanula is recognized as intolerant of competition with vigorous grasses (19). Plantago (7), Calluna (20), Molinia (20, 21) and Hylocomium (22) have all been shown to decline under experimental N applications. The identity of these species suggests that the decline in species richness is probably due to N-demanding species growing at the expense of less competitive species (23).

Equation 1 gives the mean species richness of an acid grassland affected by the lowest levels of N we measured (5 kg N ha–1 year–1) to be 21.3 species per quadrat. Using this as a conservative estimate of the species richness of a “pristine” grassland, and substituting into Eq. 1 the approximate mean N deposition in the eastern United States [7.7 kg N ha–1 year–1 (24)] and the United Kingdom or central Europe [17 kg N ha–1 year–1 (25)], gives a current average reduction of 5.2% and 23% of acid grassland species richness, respectively. The currently accepted critical load for N deposition to grasslands in Europe [25 kg N ha–1 year–1 (26)] gives a reduction of 38.4%.

Experimental addition of nitrogen to grasslands in Great Britain (7) and North America (10) has resulted in a loss of 3 to 4% of species richness for each 100 kg N ha–1 added over the course of the experiments (27). If we make the conservative assumption that N deposition has a fully cumulative effect on vegetation [i.e., no N is lost because of seepage, denitrification, etc. (28)], and taking a value of 3.5% species richness loss for each cumulative 100 kg N ha–1 deposited, the total amount of N required to decrease species richness by 25% is 714 kg N ha–1. At the mean nitrogen deposition of central Europe, the time to reach 714 kg N ha–1 is 42 years at a constant level of deposition, and longer if N deposition has increased from lower levels in the past or if some N is not taken up by vegetation. This time frame is consistent with the enhanced emission of reactive N that began around the start of the 20th century and accelerated in its latter half (29). Although N deposition is beginning to decline in many areas of Europe (29, 30), the potential importance of cumulative N deposition and the alteration of soil pH (which may take many decades to recover) may mean that a return to an “unpolluted” condition of high species richness may not be achieved in the foreseeable future.

Supporting Online Material

Materials and Methods

Fig. S1

Table S1


References and Notes

View Abstract

Navigate This Article