Variability of [[Delta].sup.15]N in soil and plants at a New Zealand hill country site: correlations with soil chemistry and nutrient inputs.
Abstract: This study investigated [sup.15]N enrichment and nutrient cycling in hill country used for semi-extensive pastoral agriculture, at a site where pre-European seabird breeding occurred. Soil (0-15 cm) and plant samples were taken from 18 ridgeline and sideslope transects. Three stock camps (locations which grazing animals frequent) were identified within the study area, two on the ridgeline and one on the sideslope. Soil [sup.15]N enrichment was greatest at stock camps, and lowest where stock input was minimal. Soil natural abundance [sup.15]N ([[Delta].sup.15]N) was therefore an index of stock nutrient inputs. Soil [[Delta].sup.15]N increased with decreasing C:N ratio, consistent with N loss through volatilisation and/or nitrate leaching from net mineralisation. Plant [[Delta].sup.15]N from stock camps was lower than its associated soil, implying that [sup.15]N enrichment of plant-available N was lower than that of total soil N. However, the correlation between plant [[Delta].sup.15]N and soil [[Delta].sup.15]N varied between stock camps, indicating differences in N cycling. Olsen P was higher at stock camps, although again differences were found between stock camps. Total P and N were correlated neither with stock camps nor topography, but were higher than expected from parent material concentrations and literature results, respectively. It is postulated that significant contributions of both elements from former seabird breeding remain in the soil.

Additional keywords: C:N ratio, nitrogen, nitrogen-15, natural abundance, phosphorus, stock camping.
Subject: Soil chemistry (Research)
Soils (Nitrogen content)
Livestock (Environmental aspects)
Sea birds (Environmental aspects)
Author: Hawke, D. J.
Pub Date: 03/01/2001
Publication: Name: Australian Journal of Soil Research Publisher: CSIRO Publishing Audience: Academic Format: Magazine/Journal Subject: Agricultural industry; Earth sciences Copyright: COPYRIGHT 2001 CSIRO Publishing ISSN: 0004-9573
Issue: Date: March, 2001 Source Volume: 39 Source Issue: 2
Geographic: Geographic Scope: New Zealand Geographic Name: New Zealand Geographic Code: 8NEWZ New Zealand
Accession Number: 73023637
Full Text: Introduction

Since human contact around 2000 years ago (Holdaway 1996), a series of environmental disturbances derived from human activity and natural events have occurred in New Zealand. Natural events include a devastating volcanic eruption (Taupo in ~200 AD), and major earthquakes at intervals of a few hundred years (Wells et al. 1998). Human induced changes include large-scale deforestation, extensive faunal extinction (Holdaway 1989), and conversion of much of the landscape to pastoral agriculture or horticulture.

Soil development in New Zealand has generally been stable since the end of the Otiran glaciation (Molloy 1993). Present day soils are therefore old enough to have witnessed these disturbances. Thus, we see a tephra layer from the Taupo eruption in many North Island soils, charcoal fragments from deforestation, and modifications resulting from the use of soil for food production (plate 15.2: Molloy 1993; McIntosh et al. 1999).

Among the disturbances just listed was faunal extinction. Palaeontological evidence shows that extensive seabird breeding colonies existed on mainland New Zealand (Holdaway 1989) until ~300-700 years BP (Hawke et al. 1999). Given the large nutrient inputs typical of seabird breeding colonies, it follows that soil-based evidence of seabird breeding might also remain. Studies supporting this conjecture have used stable carbon (C) and nitrogen (N) isotopes as well as conventional soil chemistry analyses (Moors et al. 1988; Mizutani et al. 1991; Hawke et al. 1999).

The effects of contempory agricultural activities are also likely to be superimposed on soil parameters. Present-day stock camps (the phenomenon where grazing animals (stock) favour particular locations in the landscape) commonly occur along ridgelines, the same features probably occupied by former seabird breeding colonies. Stock camping in hilly country enriches the soil in N, phosphorus (P), and potassium (K) at flat sections of ridgelines, impoverishing the remainder of the landscape (Saunders and Auld 1969; Gillingham and During 1973). Effects of stock camping on [sup.15]N enrichment in hill country grazing systems are unknown, since landscape-scale research involving natural [sup.15]N enrichment has mostly involved relatively flat land. However, variability of [sup.15]N enrichment in plants has been attributed to patchy inputs of dung (Kerley and Jarvis 1996). Other potentially relevant correlations involving [sup.15]N enrichment of hill country soil include slope position (Koba et al. 1998), and redox potential differences originating from drainage effects (Sutherland et al. 1993; Clay et al. 1997).

The present study investigated a hill country landscape with a humid climate and acid soils where pre-European seabird breeding is known to have occurred. The aims of the study were to examine the impact of stock on soil [sup.15]N natural abundance ([[Delta].sup.15]N) within the context of former seabird breeding, and to relate the [sup.15]N enrichment of plant material to both stock activity and soil [sup.15]N enrichment.

Materials and methods

Study area

The study area overlooks Akaroa Harbour on the south-eastern corner of Banks Peninsula at 43 [degrees] 49'S, 173 [degrees] 0.5'E on the east coast of the South Island of New Zealand. A map showing the location of the study area is given in Powell and Hawke (1995). The ridgeline of which the study area is a part is derived from the crater rim of a Tertiary age volcano. The ridge leading to Purple Peak (Otepatatu; 646 m) lies at the northern end of the study area, and Stony Bay Peak (Taraterehu; 690 m) is at the southern end. The slopes on both sides are steep, with intermittent bluffs particularly on the Akaroa Harbour (western) side. Soils in the study area are Orthic Recent Soils (New Zealand Soil Classification; Hewitt 1992) derived from greywacke loess and basalt. Soils are moderately acidic, with pH decreasing with altitude from 5.4 to 4.8 (D. J. Hawke unpublished data).

The study area has a temperate humid climate with an altitude-dependent annual rainfall of ~1800 mm. The sampling period coincided with a fairly severe drought that started with the 1997-98 El Nino event, with 1998 rainfall reduced to 1353 mm (H. Wilson, pers. comm.) and an increased frequency of hot northwest winds. The study area is either semi-extensive pastoral farmland or former farmland destocked over the past decade since inclusion in Hinewai Reserve. Present-day vegetation is mostly exotic pasture, with scattered gorse (Ulex europaeus), broom (Cytisus scoparius), and native shrub species. At higher altitude, native tussock grasses (Chionochloa spp.) become more important. Fertiliser input has been minimal or absent for at least 2 decades. Prior to European land-clearance during the 19th century, Banks Peninsula was mostly covered in podocarp forest. However, it is likely that the ridges were burnt by the Maori prior to European colonisation (Wilson 1992).

Cattle and sheep are the main grazing animals. However, feral goats were abundant on the slopes of Stony Bay Peak until the present decade. At low altitude, domestic geese contribute to stock camp inputs. Nutrient inputs from pre-European seabird breeding are indicated by Maori tradition (see Andersen 1927, p. 145).

Soil sampling along transects

Between April and July 1998 a total of 18 transects were set up from the summit plateau of Stony Bay Peak down the ridge to Purple Peak Saddle, then down the fall-line of the slope between 2 streams. The 8 transects from Stony Bay Peak to Purple Peak Saddle were 100 m apart at right angles to the ridgeline. The 3 transects from Purple Peak Saddle toward Purple Peak were 50 m apart, again at right angles to the ridgeline. The remaining 7 transects were set up from the last transect on the ridgeline, 50 m apart, down the fall-line of the slope between 2 streams. The estimated elevation of the lowest transect was 420 m, giving an altitude interval of 270 m to the top of Stony Bay Peak (690 m). On each transect 3-10 soil cores of 15 cm depth and 5 cm diameter were taken 6 m apart and pooled for subsequent analysis. Cores were not taken if soil depth was [is less than] 15 cm. Soil from each core was sieved (3.5 mm) and air-dried at 40-50 [degrees] C.

Dung deposits, indicating well-used stock camps, were found at 3 contrasting locations: within Purple Peak Saddle, at the northern end of the ridgeline, and at the toe of the sideslope. Seabird breeding is likely to have occurred at the ridgeline stock camps, but is unlikely at the sideslope stock camp due to lack of takeoff sites. This provided the opportunity to assess the contribution of former seabird breeding to the soil chemistry.

Vegetation and associated soil sampling

Vegetation and associated soil samples were collected from the 3 stock camping sites in January and March 1999 (the austral summer). The sites were Purple Peak Saddle, the northern end of the ridgeline, and the toe of the sideslope. At each site, 5-11 cores of 5 cm diameter and 15 cm depth were taken randomly with the restriction that each core was [is greater than] 15 cm away from any legume plants. Plant material (roots and shoots) was thoroughly washed in water to remove soil and decaying organic matter before being dried at 60 [degrees] C. The soil from each core was processed according to the method described earlier.

Chemical analysis

Soil samples were analysed for total P, Olsen P, total C, total N, major oxides, and natural abundance [sup.15]N. Total P and major oxides were analysed using wavelength-dispersive X-ray fluorescence (XRF) by Spectrachem Analytical Ltd (Wellington, New Zealand), after ignition at 1000 [degrees] C to determine organic matter content. Olsen P was determined according to Blakemore et al. (1987).

Total N, [sup.15]N, and total C were determined by the Institute of Geological and Nuclear Sciences Ltd (Wellington, New Zealand). Analyses were carried out simultaneously on finely ground ([is less than] 250 [micro]m) soil or plant matter. The procedure involved converting sample C and N to [CO.sub.2] and [N.sub.2] using an elemental analyser interfaced via GC to a mass spectrometer (Europa Geo 20/20). The nitrogen isotope ratio was calculated as a per mil (%) deviation from atmospheric N:

[[Delta].sup.15]N (%) = 1000 x [[([sup.15]N/[sup.14]N).sub.sample] - [([sup.15]N/[sup.14]N).sub.air]/[([sup.15]N/[sup.14]).sub.air]

Results were validated during the analytical run by replicated comparisons with flour (Europa Scientific; [[Delta].sup.15]N = 3.01%) and N- 1 ([[Delta].sup.15]N = 0.4%) standards. The standard error of repeated analysis of standards for [[Delta].sup.15]N was 0.3%; agreement between duplicate sample analyses was [+ or -] 0.3% or closer. This level of precision is comparable to that reported elsewhere (e.g. Kerley and Jarvis 1996). The total C and total N results were used to calculate C: N ratio.

Statistical analyses

Effects of stock camping, stock camp position, and transect location on soil [[Delta].sup.15]N, total N, total P, Olsen P, and C:N ratio were examined using 2-tailed t tests. Linear correlations between soil [[Delta].sup.15]N and other soil analytical quantities were also determined as part of elucidating the processes controlling soil [[Delta].sup.15]N. The plant and associated soil [[Delta].sup.15]N results were compared using a paired t-test. The effect of different stock camps on the mean [[Delta].sup.15]N values of plants and their associated soil from different stock camps were compared using an F-test. The relationship between soil [[Delta].sup.15]N and plant [[Delta].sup.15]N within each stock camp was determined by non-parametric (Spearman's Rank) correlation.

Valid t-tests require normal distributions. Normality was tested using the Kolmogorov-Smirnov Test, results being consistent with normality for all variables (P [is greater than] 0.10). The comparisons involving stock camp position had small sample sizes ([is less than] 5), for which the Kolmogorov-Smirnov Test is invalid. Normality was therefore assumed for these comparisons.

Results

Soil N and P

Mean total P and total N showed no significant differences (P [is greater than] 0.05; Table 1) when comparing ridgeline v. sideslope, stock camping v. no stock camping, and ridgeline v. sideslope stock camps. Mean Olsen P was significantly higher at stock camping sites than non-stock camping sites (P [is less than] 0.05; Table 1), but differences were not significant (P [is greater than] 0.05; Table 1) when comparing ridgeline v. sideslope and ridgeline v. the sideslope stock camps. One notable feature of the stock camp data was low values of Olsen P (9-12 mg/kg) at the stock camp at the northern end of the ridgeline. Here, the soils are relatively thin and, so, dry out easily; Rowarth et al. (1985) showed that low moisture levels inhibit dung remineralisation.

[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]

The relative contribution of weathering and external inputs to total P concentrations was assessed by constructing a mixing line between the loess and basalt parent materials based on Si[O.sub.2] content. The Si[O.sub.2] content of 36 basaltic rock fragments (mass range 0.24-10.10 g; mean 1.62 g) collected from soil cores in the process of sieving was 51.9%; total P was 2043 mg/kg. Loess obtained from a road cutting contained 71.5% Si[O.sub.2] and 603 mg/kg total P. The Si[O.sub.2] content of soil minerals was 47.8-67.0% Si[O.sub.2], while total P was 1750-4560 mg/ kg. The Si[O.sub.2] and total P concentrations from each soil sample (corrected for organic content) were plotted (Fig. 1). All data lay well above the mixing line between the loess and basalt end members, indicating additional source(s) of P to the system. Organic matter content was significantly correlated with total P (P [is less than] 0.01; r = 0.63) and total N (P [is less than] 0.001; r = 0.91) concentrations. These correlations imply that organic matter cycling contributes to both total P and total N. The potential contribution of organic cycling to total P was assessed by collecting samples from a transect across intensively farmed pasture developed on loess on the nearby Canterbury Plains. The pooled sample contained 1200 mg/kg P and 73.1% Si[O.sub.2] (D. J. Hawke unpublished data), well below the soil data in

[Figure 1 ILLUSTRATION OMITTED]

Soil [[Delta].sup.15]N

Values of [[Delta].sup.15]N (Fig. 2) ranged from 3.5-7.0% (mean 5.3%; standard deviation 1.0%) were at the top end of the range of -1.1 to 6.8% (mean 3.2%) reported from 61 sites on `improved grasslands' from throughout New Zealand by Steele et al. (1981). Ridgeline and sideslope means were not significantly different (P [is greater than] 0.05; Table 1), implying similar controls on N-cycling. Mean stock camp [[Delta].sup.15]N was significantly greater than [[Delta].sup.15]N from non-stock camping areas (P [is less than] 0.001; Table 1). Consistent with this, the 5 highest values (all [is greater than] 5.8%) occurred in the 3 stock camps. The 2 lowest values (3.5%o and 4.1%) were found along the ridgeline on a steep rocky slope and a cold, shady location among gorse, respectively, where neither former seabird breeding nor stock camping are likely to have occurred. The mean [[Delta].sup.15]N from the stock camp at toe of the sideslope (where former seabird breeding is unlikely) were not significantly different (P [is greater than] 0.05; Table 1) from the mean of the ridgeline stock camp data (where former seabird breeding is likely). Therefore, former seabird breeding does not contribute significantly to present-day soil [[Delta].sup.15]N. Instead, the soil [[Delta].sup.15]N distribution indicates the intensity of stock input, with greatest enrichment where stock inputs are greatest.

[Figure 2 ILLUSTRATION OMITTED]

Mean C:N ratios (Table 1) were significantly lower at stock camping sites than non-stock camping sites (P [is less than] 0.01), and higher on the ridgeline v. sideslope (P [is less than] 0.05). Differences were not significant when comparing ridgeline v. sideslope stock camp data. Soil [[Delta].sup.15]N increased significantly (P [is less than] 0.001) with decreasing C :N ratio (r = -0.74; Fig. 3). As noted above, C:N was lowest at the stock camps, where substantial ammonia volatilisation is likely.

[Figure 3 ILLUSTRATION OMITTED]

Plant and soil [[Delta].sup.15]N at stock camping sites

Plant [[Delta].sup.15]N data from the 3 stock camping sites (Table 2) ranged from 0.8% to 5.8%o. The null hypothesis, that sample means from the 3 sites were indistinguishable, was rejected (F = 3.996; P [is less than] 0.05).

Table 2. Correlation between plant and soil [[Delta].sup.15]N from stock camping sites

(A) Spearman's Rank correlation coefficient. (*) P < 0.05.

The [[Delta].sup.15]N data for soil associated with the plant material (Table 2) ranged from 5.2% to 8.7% and so were consistent with high levels of stock input. As was the case for the plant data, the null hypothesis that the sample means from the 3 sites were indistinguishable was rejected (F = 3.867; P [is less than] 0.05).

The correlation between soil and plant [[Delta].sup.15]N (Fig. 4) was only significant at P = 0.05 (Spearman's Rank correlation coefficient = 0.50). The correlation was then tested for each stock camp (Table 2). One result, showing a negative correlation, was not statistically significant. A second stock camp data set had a relatively low correlation coefficient. Although the third data set had a high correlation coefficient, it had a small sample size. Thus, the results suggest that there are different processes linking soil to plant [[Delta].sup.15]N between stock camps.

[Figure 4 ILLUSTRATION OMITTED]

Notwithstanding the inconsistent correlations between soil and plant [[Delta].sup.15]N, plant [[Delta].sup.15]N was always less than [[Delta].sup.15]N of the associated soil. The null hypothesis, that the mean of the difference between plant and soil [[Delta].sup.15]N values was zero, was rejected (paired t test: t -14.144, 25 degrees of freedom, P [is less than] 0.001). The 95% confidence interval of (soil [[Delta].sup.15]N - plant [[Delta].sup.15]N) was 3.2 [+ or -] 0.5%.

Discussion

The 2 main potential contributors to soil [sup.15]N enrichment at the study site are present-day stock input, and former seabird breeding. Results showed that [[Delta].sup.15]N distribution reflected the intensity of stock input, with greatest enrichment where stock inputs are greatest. Former seabird breeding was not a significant contributor. The results therefore justify a reexamination of the conclusions of Hawke et al. (1999) regarding the use of [Delta][sup.15]N as a tracer of former seabird breeding colonies. At this point, it is worth noting that Maori tradition (on which the presumption of former seabird breeding at the study area was based) is generally regarded as being concordant with archaeology (e.g. Finney 1994).

An understanding of the mechanism of stock-derived soil [sup.15]N enrichment comes from isotopic fractionation patterns in the N-cycle. The [Delta][sup.15]N of urine is low (e.g.-1.0%o; Erikson and Hogh-Jensen 1998) due to the preferential excretion of [sup.15]N-depleted N. However, nitrate leaching, volatilisation of ammonia, and gaseous losses of [N.sub.2], NO, or [N.sub.2]O all enrich the remaining soil N in [sup.15]N (Delwiche and Steyn 1970; Yoneyama 1996); these processes are well known in pastoral farming (Sherlock and Goh 1984; Luo et al. 1999; Silva et al. 1999). High values for soil [Delta][sup.15]N at stock camps compared with samples outside the stock camps provide independent support for the idea that N loss by volatilisation and leaching is particularly significant at such places. Volatilisation of ammonia is especially important at urine patches (Haynes and Williams 1993) and hence probably at stock camps. Faeces are enriched in [Delta][sup.15]N relative to the foodstuff by ~3%o (Steele and Daniel 1978; Kerley and Jarvis 1996); the [Delta][sup.15]N of ryegrass shoot material from the nearby Canterbury Pains is ~2%o (D. Hawke and K. Bodger, unpublished data).

Soil [Delta][sup.15]N increased significantly with decreasing C:N ratio. In experiments simulating a cropping system, Turner et al. (1983) also found that [Delta][sup.15]N increased as plant material decomposed (implying mineralisation). In contrast, a positive correlation between [Delta][sup.15]N and C :N was found by Selles et al. (1984, 1986) on cultivated land, at C :N ratios similar to that of my study. Isotopic fractionation during net N immobilisation appears to be affected by the source of N available to microorganisms, with enrichment of organic N most likely to occur under low ammonia concentrations (Yoneyama 1996). Although inorganic N speciation was not measured in any of these studies, stock inputs at my study site may have caused the soil ammonia concentration to be much higher than at the cropping sites studied by Selles et al. (1984, 1986).

Results from sampling plants and associated soil showed that [sup.15]N enrichments at the 3 stock camps were not identical. Recalling that the 3 stock camps each occupied different features in the landscape (a saddle, a dry ridgeline, and the toe of the sideslope), changes in soil redox potential arising from differing drainage are a potential cause (Sutherland et al. 1993; Clay et al. 1997). Consistent with this explanation, urine application (important at stock camps) leads to emission of NO and [N.sub.2]O (Williams et al. 1998). The significance of [N.sub.2]O emission from soil increases with water saturation (Wolf and Russow 2000), so that well-drained stock camps are likely to emit less of this gas. Such considerations may be important in attempting to model [N.sub.2]O (an important greenhouse gas) emissions due to pastoral agriculture.

Within stock camps, plant [Delta][sup.15]N was always less than that of its associated soil. This reflects a low [Delta][sup.15]N of the plants' N source (Sutherland et al. 1993), and/or isotopic fractionation during uptake of ammonia by roots (Yoneyama 1996). As noted earlier, sampling was carried out during the second summer of a drought and avoided legumes. Given that low pH, soil moisture, and temperature inhibit nitrification (Haynes 1986a), and that high ammonia levels (as expected in a stock camp) inhibit nitrate uptake (Haynes 1986b), it seems reasonable to conclude that nitrate was not the predominant N source for the plants sampled. Support for this conclusion was a study by Thomas et al. (1988) in acidic Scottish upland soils showing high nitrate concentrations only in summer. In addition, recent results (Koba et al. 1998) for the [Delta][sup.15]N of nitrate in forest soils were much lower than the plant [Delta][sup.15]N from my study, whereas the ammonia [Delta][sup.15]N was much closer at ~+5%. Haynes and Williams (1993) proposed that inorganic N may cycle through labile organic pools before being taken up by plants. However, the correlation between soil [Delta][sup.15]N and plant [Delta][sup.15]N varied between stock camps. This implies a location-dependent balance between remineralised soil organic matter and ammonia from urine and faeces as the main N sources.

Total P and Olsen P results showed an interesting contrast. The very low ratio of Olsen P to total P ([is less than] 1%) implies that P cycling involved only a small proportion of total P with any conversion to plant-available P being very slow. Olsen P results were significantly higher at stock camps, implying stock-mediated transfer to these locations. This phenomenon has been recognised for many years (Saunders and Auld 1969; Gillingham and During 1973), and is minimised in more intensive pastoral systems by grazing management practices. Similar to the results for soil and plant [Delta][sup.15]N, differences were found between stock camps. While the Purple Peak saddle data stood out for [Delta][sup.15]N, it was the stock camp at the northern end of the ridgeline that was differentiated for Olsen P. Soil moisture was the likely cause for differences in Olsen P, whereas soil drainage was the likely cause for [Delta][sup.15]N. These results reinforce the conclusion that nutrient cycling varies between stock camps over relatively short spatial scales.

Total P concentrations showed no correlation between presence and absence of stock camps, ridgeline v. sideslope, or ridgeline v. the sideslope stock camps. Calculations using a simple mixing model showed that total P concentrations were too high to be accounted for by weathering of soil parent materials. There is no known history of fertiliser application to the study area. Atmospheric P fluxes are negligible (Graham and Duce 1979) and the soil parent materials are relatively well constrained. The most likely explanation is input from former seabird breeding. The major limitation of this approach is that only the top 15 cm of soil was sampled, so that some of the excess P could be accounted for by recycling of organic material. While an inventory of total P for the study area would resolve the uncertainty, the site has a rather complex (and often inaccessible) topography. However, comparison with total P from an intensively farmed site supported the conclusion that total P concentrations were anomalously high at the study site.

The range of total N was high on the rating scale established for New Zealand agricultural soil by Blakemore et al. (1987). Contributions from soil parent material were negligible. The results were comparable with those found for both present-day and former seabird breeding areas (Hawke et al. 1999 and references therein). The results were also much higher than found by Hawke et al. (1999) at a hill country site with no former seabird breeding, but similar rainfall and elevation. Although it is not easy to separate contributions from organic matter cycling and [N.sub.2] fixation, the lowest total N was at a shady ridgeline site where there was active legume (gorse) growth. Therefore, as for total P, the results are consistent with an external source of N. Former seabird breeding is the most likely explanation.

Given that the comparison between ridgeline and sideslope stock camps inferred no seabird input for either total P or total N, it appears that much of the former seabird input has been redistributed. The most likely agents are leaching and erosion due to the humid climate and steep topography, and stock movement. It is not known when extensive seabird breeding ceased in the study area. The tradition documented by Anderson (1927) was recorded in 1917 with the informant (H. Tikao) being quite old. Thus, seabird breeding almost certainly ceased before the end of the 18th century.

Conclusions

Results from the present study show that soil [Delta][sup.15]N is an index of stock input, with greatest enrichment where stock inputs are greatest. While use of surface soil [Delta][sup.15]N as an indicator of former seabird breeding is not necessarily invalid, stock input must be considered as a potential confounding factor (and vice versa). High soil [Delta][sup.15]N at stock camps provides independent support for the idea that N loss by volatilisation and leaching is particularly significant at such places.

While plant [sup.15]N enrichment was always less than that for associated soil at stock camps, significant differences occurred between stock camps for plant [Delta][sup.15]N and soil [Delta][sup.15]N. Together with rather variable correlations between plant [Delta][sup.15]N and soil [Delta][sup.15]N, it follows that N cycling differs between stock camps. In particular, emission of [N.sub.2]O (a greenhouse gas emitted in substantial quantities from urine patches) will probably vary depending on drainage effects.

Total P was higher than expected from mixing between loess and basalt parent materials, consistent with input from former seabird breeding subsequently redistributed by stock and/or erosion. Total N concentrations were also high compared to expected values for other agricultural land in New Zealand, and to literature results for a non-seabird breeding area. Effects of former seabird breeding at the study area therefore remain as elevated levels of N and P.

Acknowledgments

Thanks to G. Curry and the Maurice White Native Forest Trust for access to their properties. I am grateful for sampling assistance by J. Causer, J. & M. Hawke, R. Holdaway, and Q. Ma. Laboratory assistance was by H. Barlow and N. McKeown. The study benefited from discussions with J. Causer, K. Dodds, R. Holdaway, G. Otway, G. Reveley, A. Robinson, and H. Wilson, while referee comments substantially improved the paper. Literature relevant to the study was kindly provided by G. Lambert and P. Rolston.

References

Andersen JJ (1927) `Place names of Banks Peninsula.' (Government Printer: Wellington, NZ; 1976 reprint by Capper Press: Christchurch)

Blakemore LC, Searle PL, Daly BK (1987) Methods for chemical analysis of soils. New Zealand Soil Bureau Scientific Report No. 80, Government Printer: Wellington, NZ.

Clay DE, Chang J, Clay SA, Ellsbury M, Carlson CG, Malo DD, Woodson D, DeSutter T (1997) Field-scale variability of nitrogen and d[sup.15]N in soil and plants. Communications in Soil Science and Plant Analysis 28, 1513-1527.

Delwiche CC, Steyn PL (1970) Nitrogen isotope fractionation in soils and microbial reactions. Environmental Science and Technology 4, 929-935.

Erikson J, Hogh-Jensen H (1998) Variations in the natural abundance of [sup.15]N in ryegrass/white clover shoot material as influenced by cattle grazing. Plant and Soil 205, 67-76.

Finney B (1994) Experimental voyaging and Maori settlement. In `The origins of the first New Zealanders'. (Ed. DG Sutton) pp. 52-76. (University of Auckland Press: Auckland)

Gillingham AG, During C (1973) Pasture production and transfer of fertility within a long-established hill pasture. New Zealand Journal of Experimental Agriculture 1, 227-232.

Graham WF, Duce RA (1979) Atmospheric pathways of the phosphorus cycle. Geochimica et Cosmochimica Acta 43, 1195-1208.

Hawke DJ, Holdaway RN, Causer JE, Ogden S (1999) Soil indicators of pre-European seabird breeding in New Zealand at sites identified by predator deposits. Australian Journal of Soil Research 37, 103-113.

Haynes RJ (Ed.) (1986a) Nitrification. In `Mineral nitrogen in the plant-soil system', pp. 127-165. (Academic Press: Orlando, FL)

Haynes RJ (Ed.) (1986b) Uptake and assimilation of mineral nitrogen by plants. In `Mineral nitrogen in the plant-soil system', pp. 303-378. (Academic Press: Orlando, FL)

Haynes RJ, Williams PH (1993) Nutrient cycling and soil fertility in the grazed pasture ecosystem. Advances in Agronomy 49, 119-199.

Hewitt AE (1992) `New Zealand soil classification.' DSIR Land Resources Scientific Report No. 19. (Government Printer: Wellington)

Holdaway RN (1989) New Zealand's pre-human avifauna and its vulnerability. New Zealand Journal of Ecology (Supplement) 12, 11-25.

Holdaway RN (1996) Arrival of rats in New Zealand. Nature 384, 225-226.

Kerley SJ, Jarvis SC (1996) Preliminary studies of the impact of excreted N on cycling and uptake of N in pasture systems using natural abundance stable isotopic discrimination. Plant and Soil 178, 287-294.

Koba K, Tokuchi N, Yoshioka T, Hobbie EA, Iwatsubo G (1998) Natural abundance of nitrogen-15 in a forest soil. Soil Science Society of America Journal 62, 778-781.

Luo J, Tillman RW, Ball PR (1999) Grazing effects on denitrification in a soil under pasture during two contrasting seasons. Soil Biology and Biochemistry 31, 903-912.

McIntosh PD, Gibson RS, Saggar S, Yeates GW, McGimpsey P (1999) Effect of contrasting farm management on vegetation and biochemical, chemical, and biological condition of moist steepland soils of the South Island high country, New Zealand. Australian Journal of Soil Research 37, 847-865.

Mizutani H, Kabaya Y, Moors PJ, Speir TW, Lyon GL (1991) Nitrogen isotope ratios identify deserted seabird colonies. Auk 108, 960-964.

Molloy L (1993) `Soils in the New Zealand landscape, the living mantle.' (New Zealand Society of Soil Science: Lincoln, New Zealand)

Moors PJ, Speir TW, Lyon GL (1988) Soil analyses and [sup.13]C/[sup.12]C ratios identify sites of deserted Rockhopper penguin colonies. Auk 105, 796-799.

Powell HKJ, Hawke DJ (1995) Free aluminium and aluminium complexation capacity of natural organic matter in acidic forest soil solutions from Canterbury, New Zealand. Australian Journal of Soil Research 33, 611-620.

Rowarth JS, Gillingham AG, Tillman RW, Syers JK (1985) Release of phosphorus from sheep faeces on grazed, hill country pastures. New Zealand Journal of Agricultural Research 28, 497-504.

Saunders WMH, Auld MW (1969) Topdressing within a hill country farm. Ruakura Farmers' Conference Proceedings, 16-25.

Selles F, Karamonos RE, Bowren KE (1984) Changes in natural [sup.15]N abundance of soils associated with tillage practices. Canadian Journal of Soil Science 64, 345-354.

Selles F, Karamonos RE, Kachanoski RG (1986) The spatial variability of nitrogen-15 and its relation to the variability of other soil properties. Soil Science Society of America Journal 50, 105-110.

Sherlock RR, Goh KM (1984) Dynamics of ammonia volatilization from simulated urine patches and aqueous urea applied to pasture I. Field experiments. Fertiliser Research 5, 181-197.

Silva RG, Cameron KC, Di HJ, Hendry T (1999) A lysimeter study of the impact of cow urine, dairy shed effluent, and nitrogen fertiliser on nitrate leaching. Australian Journal of Soil Research 37, 357-369.

Steele KW, Wilson AT, Saunders WMH (1981) Nitrogen isotope ratios in surface and sub-surface horizons of New Zealand improved grassland soils. New Zealand Journal of Agricultural Research 24, 167-170.

Steele KW, Daniel RM (1978) Fractionation of nitrogen isotopes by animals: A further complication to the use of variations in the natural abundance of [sup.15]N for tracer studies. Journal of Agricultural Science 90, 7-9.

Sutherland RA, van Kessel C, Farrell DE, Pennick DJ (1993) Landscape-scale variation in plant and soil nitrogen-15 natural abundance. Soil Science Society of America Journal 57, 169-178.

Thomas RJ, Logan KAB, Ironside AD, Bolton GR (1988) Transformations and fate of sheep urine-N applied to an upland U.K. pasture at different times during the growing season. Plant and Soil 107, 173-181.

Turner GL, Bergersen FJ, Tantala H (1983) Natural enrichment of [sup.15]N during decomposition of plant material and soil. Soil Biology and Biochemistry 15, 495-497.

Wells A, Stewart GH, Duncan RP (1998) Evidence of widespread, synchronous, disturbance-initiated forest establishment in Westland, New Zealand. Journal of the Royal Society of New Zealand 28, 333-345.

Wilson HD (1992) Banks ecological region. Protected Natural Areas Programme Survey Report No. 21, Department of Conservation, Christchurch.

Williams PH, Jarvis SC, Dixon E (1998) Emission of nitric oxide and nitrous oxide from soil under field and laboratory conditions. Soil Biology and Biochemistry 30, 1885-1893.

Wolf I, Russow R (2000) Different pathways of formation of [N.sub.2]O, [N.sub.2] and NO in black earth soil. Soil Biology and Biochemistry 32, 229-239.

Yoneyama T (1996) Characterization of natural [sup.15]N abundance of soils. In `Mass spectrometry of soils'. (Eds TW Boutton, S Yamasaki) pp. 205-223. (Marcel Dekker: New York)

Manuscript received 2 August 1999, accepted 9 August 2000

D. J. Hawke

School of Science, Christchurch Polytechnic Institute of Technology, PO Box 22-095, Christchurch 8032, New Zealand.
Site           Plant [[Delta]   Soil [[Delta]
                .sup.15]N(%)    .sup.15]N(%)    Correlation(A)

                 Mean   s.d.     Mean   s.d.

Purple Peak      2.5    1.1      6.1    0.6        -0.32
Saddle
  (n = 10)
North end of     4.0    1.9      7.1    0.9         1.00(*)
ridgecrest
  (n = 5)
Bottom of        3.9    1.0      6.8    0.8         0.60(*)
sideslope
  (n = 11)
Gale Copyright: Copyright 2001 Gale, Cengage Learning. All rights reserved.