Amador, José A.; Glucksman, Andrew M.; Lyons, Jodi B.; Görres, Josef H.
Riparian forest zones are used for mitigation of agricultural and urban nonpoint source pollution. Phosphatase activity is ubiquitous in soil and sensitive to environmental perturbations, and hence, it may serve as an indicator of soil quality in riparian areas. We examined the relationship between phosphatase activity and selected soil physical and chemical properties and the spatial distribution patterns of phosphatase, organic matter, and moisture in a soil drainage catena within a riparian forest zone. Mean (%CV) phosphatase activity for the whole catena was 1.220 (40.6) µmol p-nitrophenol released/g/h (p-NP/g/h), with activity ranging from 0.152 to 2.544 µmol p-NP/g/h (n = 271). Poorly (PD), somewhat poorly (SPD), and moderately well drained (MWD) soil had significantly different (P < 0.01) mean (%CV) phosphatase activity of 1.516 (25.1), 1.327 (34.9), and 0.800 (39.9) µmol p-NP/g/h, respectively (n = 87-94). Phosphatase activity in each drainage class showed a significant, positive, and unique linear relationship with soil organic matter and moisture, and activity in MWD soil was also correlated positively with pH and water-extractable inorganic P levels. When data for the whole catena were considered, phosphatase activity was found to increase disproportionally with soil organic matter and moisture. Analysis of root mass distribution in a limited number of soil samples indicated that in SPD and PD soils roots constituted a significantly higher proportion of soil mass than in MWD soil. Furthermore, when mean phosphatase activity for each drainage class was considered, enzyme activity appeared to increase proportionally with the fraction of soil mass constituted by roots.
Spatial structure of phosphatase activity was evaluated using spatial semivariance models. The fraction of total variance in phosphatase activity explained by the semivariance model, a measure of level of spatial structure exhibited by this variable, decreased in the order SPD>MWD>PD. The range value, representing the distance over which phosphatase was autocorrelated, differed markedly among drainage classes, with values of 12.2, 2.7, and 0.8 m for MWD, SPD, and PD soil, respectively. The range for organic matter and moisture was similar to that observed for phosphatase activity in SPD and PD soil; organic matter had a range similar to that of the enzyme activity only for MWD soil. Kriged maps of phosphatase distribution showed that spatial co-occurrence of phosphatase activity, soil organic matter, and moisture levels was most apparent in MWD soil, less so in SPD soil, and no spatial co-occurrence of these properties was apparent in PD soil. The spatial distribution of phosphatase activity in riparian forest soils appears to be partly controlled by position in the landscape, with organic matter and moisture as important factors in determining the distribution of phosphatase in moderately well and somewhat poorly drained areas, and less so in poorly drained areas. The distribution of roots may also be an important determinant of the spatial variability of phosphatase activity in riparian forest soils.
A good ecological indicator must meet a number of criteria, including that it show a prompt and accurate response to perturbation, reflect some aspect of the functioning of the ecosystem, be readily and economically accessible, and be universal in distribution yet show individual specificity to temporal or spatial patterns in the environment (Holloway and Stork 1991). Soil phosphatases (phosphomonoesterases) (EC 3.1.3) catalyze the hydrolysis of organic phosphate esters to ortho-phosphate and thus constitute an important link between biologically unavailable and mineral P pools in the soil (Speir and Ross 1976). Phosphatases are produced by microorganisms, plant roots, and earthworms (Ramirez-Martinez 1968; Skujins 1976; Neal 1973) and thus are ubiquitous in soil. Soil phosphatase activity is sensitive to a number of environmental perturbations (Gianfreda and Bollag 1996), including organic amendments (Bonmati et al. 1985; Spiers and McGill 1979), waterlogging (Pulford and Tabatabai 1988), compaction (Dick et al. 1988), fertilizer additions (Pang and Kolenko 1986; Spiers and McGill 1979), tillage (Doran 1980), heavy metal inputs (Juma and Tabatabai 1977; Mathur and Rayment 1977; Mathur and Sanderson 1978), and pesticides (Schaffer 1993). The sensitivity of phosphatase activity to temporal changes in soil caused by environmental and management factors makes it a potential indicator of changes in soil quality (Dick 1994).
For soil phosphatase activity to be used as an effective indicator of soil quality, the temporal and spatial variability of this activity-and the factors that control this variability-must be ascertained. Harrison (1979) and Harrison and Pearce (1979) have shown that acid phosphatase activity varies seasonally in forest soils, with up to 19% of the total variation in phosphatase activity over the course of a year attributable to seasonal effects (Harrison and Pearce 1979). Coefficients of variation for phosphatase activity in agricultural ecosystems range from 13% in a grazed pasture (Speir et al. 1984) to 35.6% in an ungrazed grass-legume system (Bonmati et al. 1991) and can be as high as 101% in forest soils (Harrison 1979). This variability is likely driven-at least in part-by the spatial distribution of relatively static soil properties known to influence soil phosphatase activity, such as soil organic matter, available P content, pH, and clay content, and that of microorganisms, plant roots, and soil fauna, which produce these extracellular enzymes. The biotic and abiotic components of the ecosystem interact to produce the phosphatase activity observed at a particular point in space. An understanding of the influence of spatially variable soil properties on phosphatase activities may help elucidate the factors driving the spatial distribution of phosphatase activity, allowing for better design of sampling schemes for its assessment.
Riparian areas are the interfaces between terrestrial and aquatic ecosystems (Gregory et al. 1991) and include both upland and wetland areas. These areas are used widely to remove nonpoint source pollution from agricultural and urban runoff and thus serve to improve surface water quality (e.g., National Research Council 1993; Jacobs and Gilliam 1985; Peterjohn and Correll 1984; Lowrance et al. 1984). Sediment trapping is an important mechanism of nutrient removal in riparian areas (Lowrance et al. 1984; Cooper et al. 1987). Riparian wetland areas, because of their flatter slopes and high surface roughness, tend to accumulate sediment-bound P that originates from upland areas (Lowrance et al. 1984; Peterjohn and Correll 1984; Vought et al. 1994). For this reason, riparian wetland areas are believed to be important in mitigation of phosphorus present in runoff (Lowrance et al. 1986). In addition, sediments transported in agricultural or urban runoff may have high levels of pesticides, heavy metals, and other toxins (Lowrance et al. 1985). Riparian forest areas adjacent to agricultural and urban land uses are thus likely to receive elevated contaminant inputs that may interfere with the biological processes responsible for the ability of riparian forest soils to assimilate pollutants, affecting the water quality functions of these landscape features adversely. Development of an indicator of soil quality for riparian areas may allow for better management of these landscape features with respect to mitigation of nonpoint source pollution.
The question of variability of soil properties in forest soils and its implications for sampling has long been recognized, particularly in forest plantations (e.g., Mader 1963; Riha et al. 1986). Riparian forests are likely to present a particularly difficult challenge in terms of spatial heterogeneity relative to agricultural or forested upland ecosystems. Differences in slope and drainage in these transition areas occur at scales of meters to tens of meters. This physical heterogeneity results in small-scale variability in plant communities (Allen et al. 1989; Ehrenfeld 1987), horizontal and vertical distribution of root biomass (Golet et al. 1993; Ehrenfeld 1987), and soil organic matter and inorganic P levels (Allen et al. 1989; Mitsch and Gosselink 1993; Wright and Sautter 1988) both among and within drainage classes. These differences are likely to result in a high degree of variability in the spatial distribution of phosphatase activity in riparian forest soils.
Although the problem of spatial variation of soil phosphatase activity has been recognized by others (e.g., Harrison and Pearce 1979) and attempts at quantifying this variability have been made in the past (e.g., Bonmati et al. 1991; Speir et al. 1984), the problem has not been addressed using geostatistical techniques. Unlike first-order, univariate statistics-which assume variables are randomly distributed in space-geostatistical analysis allows for quantitation of the spatial dependence, i.e., autocorrelation, of a variable and incorporates this information into estimates of the spatial distribution of variables, reducing the uncertainty in spatial interpolation (Burgess and Webster 1980).
The present study constitutes the first landscape-scale analysis of phosphatase activity in a deciduous riparian forest in the Northeastern U.S. Using an exhaustive data set, we evaluated the relationship between phosphatase activity and static soil properties thought to be important in determining the distribution of this activity, such as organic matter, moisture, pH and inorganic P levels in soil from different drainage classes within a catena. Using geostatistical analysis, we assessed the spatial dependence and the spatial distribution of phosphatase and of soil properties that were found to be correlated with this activity. To our knowledge this is the first study addressing soil phosphatase activity in a riparian forest and constitutes the first evaluation of spatial heterogeneity of soil phosphatase activity using geostatistical analysis in any ecosystem.
Soil samples were obtained from a riparian forest within the Peckham Farm research area of the University of Rhode Island in Kingston, RI (41°30' N, 71°45' W). The canopy is predominantly red maple (Acer rubrum). The shrub layer consists of arrow-wood (Viburnum dentatum) and highbush blueberry (Vaccinium corymbosum). The herbaceous layer is dominated by Canadian mayflower (Maianthemum canadense) in the spring and by cinnamon fern (Osmunda cinnamomea) in the summer. The study site contains a toposequence typical of New England landscapes (Wright and Sautter 1988), with moderately well (MWD), somewhat poorly (SPD), and poorly (PD) drained soils included within the catena. Natural drainage classes are established based on observations and inferences on runoff, permeability, and internal drainage (Wright and Sautter 1988). Moderately well drained soils have internal drainage that is restricted to some degree, with mottles common in the lower part of the subsoil, generally at a depth of 45 to 90 cm. They remain wet and cold later in the spring. Somewhat poorly drained soils remain wet for long periods of time as a result of slow removal of water caused by the presence of a slowly permeable layer within the profile or a high water table, with mottles common in the subsoil at a depth of 20 to 45 cm. Poorly drained soils have high water tables at or near the surface during a considerable part of the year, a thick, dark, surface horizon, and mottles are frequently found within 20 cm of the soil surface. The soil in the upland portion of the catena is mapped as a Hinckley sandy loam (sandy-skeletal, mixed, mesic Typic Udorthent), whereas the soils in the lower portions of the catena are mapped as Walpole sandy loam (sandy, mixed, mesic Aeric Haplaquept) and Scarboro mucky sandy loam (sandy, mixed, mesic Histic Humaquept) (Soil Survey Staff 1981). The mean, range, and coefficient of variation of selected physical and chemical properties of soil from each drainage class are shown in Table 1.
| TABLE 1 Mean, range, and coefficient of variation of soil properties for different drainage classes |
Soil cores (5-cm dia.) were collected below the litter layer (0-5 cm) from three drainage classes (MWD, SPD, and PD) in May 1995. A single plot (24 m × 5 m) was sampled per drainage class within the catena. Each drainage class plot contained five subplots (5 m × 5 m), from each of which 10 samples were taken randomly. Of these 10 points, two were selected for nested sampling, and five nested samples were taken from within a 1 m × 1 m area around each of these two points. In summary, 20 samples were removed from each subplot for a total of 100 samples per drainage class. Samples were placed in sealable plastic bags and stored in the dark at 4°C. Storage of soil samples for 6 to 8 weeks under these conditions has been shown to have no significant effect on phosphatase activity (Speir and Ross 1975; Gerritse and van Dijk 1978).
Phosphatase activity was assayed using fieldmoist soil samples 2 months after collection. The phosphatase assay of Tabatabai and Bremner (1969) was modified as described by Duxbury and Tate (1981) for soils with a high organic matter content. Soil (0.5 g) was dispersed in 2.0 mL of deionized, distilled water and mixed with 0.5 mL of an unbuffered, 2.5 mM p-nitrophenyl phosphate (p-NPP) solution (1 mM in reaction mixture). The assay mixture was incubated in the dark for 3 h at 25°C with continuous shaking at 125 rpm. After incubation, a 1.0-mL aliquot of the assay mixture was centrifuged for 5 min. A known volume (0.5 mL) of supernatant solution was mixed with 3.6 mL 0.5 M NaOH and 0.9 mL 0.5 MCaCl2, and the mixture passed through a Whatman No. 42 filter. The concentration of p-nitrophenol (p-NP) in the filtrate was determined spectrophotometrically at 405 nm using a Shimadzu UV160U dual-beam spectrophotometer. Preliminary experiments showed that p-NP production was linear with time for the duration of the assay (data not shown). The concentration of p-NP found after incubation was corrected for adsorption of the chromophore to soil by determining sorption isotherms for p-NP under conditions identical to the enzyme assay, as described by Gerritse and van Dijk (1978). The absorbance at 405 nm of blanks containing soil in the absence of p-NPP, but otherwise treated in the same manner as analytical samples, was subtracted from all samples. Analytical variability of the method was determined from three replicate assays from five randomly selected sample points per drainage class.
We used field-moist soil because this is believed to be more representative of activity under field conditions (Dick 1994) inasmuch as air-drying has been shown to alter phosphatase activity markedly (Harrison 1979; Malcolm 1983; Speir and Ross 1981; Eivazi and Tabatabai 1977; Gerritse and van Dijk 1978). Similarly, we did not alter the pH of the reaction mixture in order to obtain phosphatase activity measurements representative of the soil as it is found in the field. Although this does not allow us to distinguish between the contribution of acid and alkaline phosphatase activity, others have shown that acid phosphatase predominates in acid soils such as the ones used in the present study (Juma and Tabatabai 1978). The pH of the assay mixture did not change significantly during the 3-h incubation period (data not shown). Soil pH can exhibit considerable spatial variability (e.g., Yang et al. 1995; Riha et al. 1986), and phosphatase activity has been shown to vary with pH (e.g. Dick and Tabatabai 1984). Buffering soil to a common pH would mask variability of phosphatase activity associated with spatial differences in soil acidity.
Soil organic matter content was determined by loss on ignition for 4 h at 550°C (Karam 1993). Soil moisture content was determined gravimetrically at 105°C (Parent and Caron 1993). Soil pH was determined using a 1:2 soil/water (wt:wt) ratio (Hendershot et al. 1993). Oxalate-extractable Fe and Al content was determined by atomic absorption spectroscopy after extraction of 2.0 g airdried, sieved (2-mm mesh) soil with 200 mL of 0.2 M ammonium oxalate (pH 3.0) for 4 h (Soil Conservation Service Staff 1972). Water-extractable inorganic P was determined colorimetrically (Alpkem 1986) after extraction of 1.0 g air-dried, sieved (2-mm mesh) soil with 15 mL of deionized, distilled water for 24 h, followed by filtration through a Whatman No. 42 filter. Water-extractable total P was determined chemically by oxidizing a portion of the water extract to phosphate (Wetzel and Likens 1991), which was subsequently analyzed colorimetrically as described above. Water-extractable organic P was calculated from the difference between water-extractable total P and water-extractable inorganic P.
Three soil cores (5-cm dia, 15-cm deep) were obtained from within each drainage class in March of 1997. Each core was divided into 5-cm segments, and each segment was analyzed independently. The total mass of soil in the core was determined, the soil mixed by hand, and a 5-g (wet weight) sample of soil removed for gravimetric moisture content determination. The remaining soil was spread on a 4.76-mm-mesh brass sieve and the soil washed away from the roots with a gentle water stream. A second sieve (2-mm mesh) was used to catch any roots that were not caught by the larger mesh sieve. After washing, roots retained by both sieves were combined, placed in an aluminum weighing boat, and dried overnight at 70°C. The proportion of total root mass at a particular depth was determined by dividing the mass of roots at a given depth by the sum of root masses from all three depths. The proportion of total mass associated with roots at a particular depth was determined by dividing the mass of dried roots by the mass of dried soil plus roots at that depth.
Normality of the data was determined using the Kolmogorov-Smirnoff test. Differences in phosphatase activity among drainage classes were assessed using a Kruskal-Wallis one-way analysis of variance on ranks and Dunn's method for pairwise multiple comparisons. Differences in proportion of soil mass constituted by roots and in proportion of total root mass in surface soil were assessed using a one-way analysis of variance and a Bonferroni t test for mean separation. Statistical tests were evaluated at the 95% confidence level unless otherwise specified.
Each sampling point was marked after the corresponding sample was taken and was surveyed to a common benchmark using a polar topographic survey method. The resulting data were transformed into Cartesian coordinates, and the coordinates of each point were appended with the corresponding soil property values. These data were then input to GS+ (version 1.21, Gamma Designs Software, Plainwell, MI) to compute semivariograms and maps of soil properties for each drainage class. The semivariogram is described by the equation: Equation (1) where [gamma](h) is the semivariance, E is the expected squared difference between values of samples separated by lag distance h, and z(xi) and z(xi + h) are measured values at points xi and xi + h (Trangmar et al. 1985).
| Equation 1 |
Maps were interpolated using block kriging at a block size of 0.75 m. Semivariogram parameters (lag spacing and maximal lag) were selected to give stable maps. A lag spacing of 0.6 m for all drainage classes gave stable maps if lag spacings were varied by 50%, and this lag was used for all soil properties. Our selection of lag spacing is in agreement with recommendations that the initial lag spacing should be approximately equal to the average spacing of sampling points (Isaaks and Srivastava 1989). We chose the semivariogram model with the highest r2 value to represent the spatial relationship among data points.
Mean phosphatase activity for the soil catena was 1.220 µmol p-NP/g/h (CV = 40.6%; n = 271), with values ranging from 0.152 to 2.544 µmol p-NP/g/h (Table 2). Conditions for soil phosphatase assays vary widely, particularly with respect to pH, substrate concentration, and pretreatment of soil, and these differences make it difficult to compare phosphatase activity values from different studies (Malcolm 1983). Nevertheless, the phosphatase activity values in the present study are within the range of those reported for acidic soils from mixed conifer-deciduous woodlands (Harrison 1979 and 1983), coniferous forests (Pang and Kolenko 1986), and beech forests (Rastin et al., 1988).
| TABLE 2 Mean, median, minimum, maximum, and coefficient of variation for phosphatase activity in soil from different drainage classes |
Phosphatase activity in the catena was not normally distributed. Phosphatase activity was normally distributed in the somewhat poorly and poorly drained soils, but not in the moderately well drained soil. Significant differences (P < 0.01) in phosphatase activity were observed among the three drainage classes tested (Table 2). Activity was highest in the poorly drained soil and decreased as drainage improved. A 10-fold range of phosphatase activity values was observed for the moderately well drained soil, whereas 7-fold and 3-fold ranges of values were observed for somewhat poorly and poorly drained soil, respectively. The coefficient of variation was highest for moderately well drained soil (39.5%), intermediate for somewhat poorly drained soil (34.9%), and lowest for poorly drained soil (25.1%). This is within the range observed by others both for forest soils (Rastin et al. 1988; Harrison 1979) and for grasslands (Bonmati et al. 1991). The mean analytical variability in our study was 4.1% (range: 1.4-7.6%), indicating that the differences among drainage classes were likely not due to analytical uncertainty. The analytical variability in our study is within the range of 11% observed by Speir et al. (1984). By contrast, Bonmati et al. (1991) found that spatial variability of phosphatase in a grassland soil (CV = 36%) was close to the analytical variability (CV = 37%).
The relationship between phosphatase activity and physical and chemical soil properties differed among soil drainage classes (Table 3). Phosphatase activity in moderately well drained soil correlated positively and significantly with soil organic matter, soil moisture, pH, and inorganic P, but not with organic P or extractable Fe or Al. By contrast, in somewhat poorly and poorly drained soils phosphatase activity showed a positive, significant correlation only with soil organic matter and moisture. Multiple linear regression analysis of the phosphatase data from each drainage class failed to identify a combination of soil properties that improved the prediction of phosphatase activity (data not shown).
| TABLE 3 Correlation coefficient (r) of various soil properties with phosphatase activity from moderately well, somewhat poorly, and poorly drained soil |
Juma and Tabatabai (1978) have observed that phosphate inhibits acid phosphatase activity by 21 to 42% at 310 µg P/g soil, but the inhibitory effects are lower (4-21% inhibition) at 31 µg P/g soil. Others (Harrison 1983) have found no significant correlation, positive or negative, between phosphatase activity and inorganic P in soils with inorganic P levels ranging from 50 to 680 µg P/g soil. Values of inorganic P in MWD soil ranged from 0.1 to 47.2 µg P/g soil, with a mean of 7.4 µg P/g soil (Table 2), with 70% of the values lower than 10 µg P/g soil (data not shown). The positive correlation between phosphatase activity and inorganic phosphate levels in MWD soil may simply reflect increased enzymatic release of phosphate from organic phosphorus compounds as phosphatase increases, with the absence of inhibition by inorganic P explained by the its relatively low levels in the soil. Similarly, Harrison (1982) observed that mineralization of organic phosphorus was positively correlated with phosphatase activity in a forest soil with extractable inorganic P levels ranging from 4 to 50 µg P/g soil. The absence of a correlation between inorganic phosphate levels and phosphatase activity in somewhat poorly and poorly drained soils may reflect differences in the sources of phosphatase, e.g., plant communities, known to differ among soil drainage classes in red maple swamps (Allen et al. 1989).
The activity of phosphatase has been shown to increase with soil pH, with maximal activity generally observed over a pH range of 6.5 to 6.9 (Malcolm 1983). The positive correlation between phosphatase activity and pH in MWD soil suggests that this activity is more sensitive to differences in soil pH than in SPD or PD soil, where no correlation was observed.
Numerous previous studies have found a strong, positive, linear correlation between phosphatase activity and soil organic matter and soil moisture in a wide variety of ecosystems (e.g., Bonmati et al. 1991; Harrison 1983; Harrison and Pearce 1979; Juma and Tabatabai 1977; Ross et al. 1975; Speir et al. 1984; Trasar-Cepeda and Gil-Sotres 1987). Soil organic matter is thought to influence phosphatase activity through formation of stable humus-enzyme complexes (e.g., Nannipieri et al. 1988) and by promoting the activities of microorganisms that excrete phosphatases (e.g., Nannipieri et al. 1978). Soil moisture is inextricably tied to organic matter content and may affect phosphatase activity through its influence on soil microbial activity.
Linear regression analysis of the data showed that the change in phosphatase activity with soil organic matter or moisture content (slope, m) increased with increasingly poor drainage (Table 4, Fig. 1), suggesting a nonlinear relationship between phosphatase activity and these two soil properties within the catena. A plot of the pooled data for phosphatase activity versus either soil organic matter or moisture content (Fig. 1) shows that the relationship between these variables is nonlinear, with phosphatase activity increasing disproportionally as organic matter and moisture content increase. Similarly, Trasar-Cepeda and Gil-Sotres (1987) found that acid phosphatase activity increased more rapidly with soil organic matter content in soils with a high organic matter content than in those with a lower organic matter content. Phosphatase activity has also been shown to decrease exponentially as soil moisture decreases to air-dry levels (West et al. 1988).
| TABLE 4 Parameter estimates and standard errors for a linear model describing the relationship between phosphatase activity and soil OM and phosphatase activity and moisture content for moderately well, somewhat poorly, and poorly drained soil |
| Fig. 1. Relationship between phosphatase activity and soil organic matter (A) and water content (B) using data for all three different drainage classes. Lines represent best fit of data to linear model using the parameter estimates shown in Table 4. |
Reasons for nonlinearity in the relationship between phosphatase and organic matter at the landscape scale may be found in the variables that an integrative measurement such as organic matter represents and how these relate to phosphatase activity. Phosphatase is produced both by plant roots and microorganisms, both of which are associated with the presence of organic matter. Root phosphatase activity from different plant species, and from different varieties within a species, has been shown to have different pH maxima and kinetic constants (e.g., Juma and Tabatabai 1988). Differences in plant community composition known to exist among drainage classes in red maple riparian forests (Allen et al. 1989; Golet et al. 1993) may influence the relationship between phosphatase and organic matter. Phosphatase activity has been shown to increase with soil microbial biomass (Frankenberger and Dick 1983) and activity (e.g., Harrison 1982). Although microbial biomass and activity generally increase with organic matter content, this relationship may be nonlinear along the soil moisture and elevation gradient found in a drainage catena, where organic matter content, moisture, and topography interact. The age and composition of soil organic matter are also important determinants of phosphatase activity. For example, Rojo et al. (1990) found that phosphatase activity is poorly associated with stable, well humified organic matter, but fresh organic matter in the larger soil fractions appear to have much greater capacity to associate with phosphatase. The relatively high moisture content and seasonal saturation of poorly drained soil tends to retard microbial degradation of plant debris, which results in higher rates of organic matter accumulation and lower rates of humification than in better drained, upslope areas. Differences in degree of humification along the soil catena may be partially responsible for the relationship observed between phosphatase activity and soil organic matter.
Plant roots may constitute an important source of phosphatases (Martin 1973), particularly acid phosphatase (Juma and Tabatabai 1988) in soil, such that differences in spatial distribution of plant communities may account for some of the differences in phosphatase activity observed among drainage classes. The root zone for most herb layer species in wetland areas is quite shallow (Golet et al. 1993), making this layer a potentially important contributor of root-derived phosphatases to surface soil. We evaluated the relationship between roots and acid phosphatase activity in different soil drainage classes on a limited (n = 3) number of samples for which we determined the distribution of root mass with depth and the proportion of the total soil mass comprised by roots in the top 5 cm. The percent of total root mass in the top 5 cm of a 15-cm-deep core was significantly higher in poorly drained soil, with no significant differences observed between well drained and somewhat poorly drained soil (Fig. 2A). Furthermore, the fraction of the total soil mass constituted by roots in the top 5 cm of the soil was significantly higher in somewhat poorly and poorly drained soil than in well drained soil (Fig. 2B). Mean phosphatase activity for each drainage class increased proportionally with the percent of soil mass constitued by roots (Fig. 3). These results are in agreement with those of Gould et al. (1979), who observed a positive correlation between phosphatase activity and the distribution and root biomass. Furthermore, they suggest that some of the variability in phosphatase activity among drainage classes may be associated with differences in the distribution of root biomass.
| Fig. 2. Proportion of total root mass in surface soil (A) and proportion of soil mass constituted by roots in surface soil (B) in soil from different drainage classes. Values shown are means (n = 3), and bars represent one standard deviation. Bars containing the same letter within a panel were not significantly different (P < 0.10). |
| Fig. 3. Relationship between proportion of soil mass constituted by roots in top 5 cm and mean phosphatase activity in different soil drainage classes. |
Semivariograms were obtained for phosphatase activity, soil organic matter, and moisture content by plotting the autocorrelation function with lag distance. These variables were chosen for spatial analysis because a significant linear relationship was observed among these variables and phosphatase in all soil drainage classes (Table 3). The semivariogram has three important features: nugget, sill, and range (Smith et al. 1994). The nugget (the apparent ordinate intercept) is the amount of variance not explained by the spatial model and is a combination of error terms attributable to inappropriate sampling, analytical error, and random variations. The sill is characterized by a leveling off of the correlogram and indicates the variance between points that are not correlated. The lag value when the correlogram model reaches the sill is the range, and this represents the maximum separation distance within which samples are spatially correlated.
The level of spatial structure at the sampling scale used can be inferred from the ratio, Q, given by Equation (2) where C + Co is the sill and Co is the nugget variance; hence, C represents the variance attributable to spatial dependence. The ratio Q varies between 0 and 1: a ratio of 0 indicates the absence of spatial structure at the sampling and support scale used; as Q approaches 1, a greater proportion of the variability can be explained by the semivariogram model. The spatial structure of phosphatase activity was different among drainage classes (Table 5). Somewhat poorly drained soil had the highest value of Q (0.89), followed by moderately well drained (0.76) and poorly drained soil (0.50). By contrast, the level of spatial structure for soil organic matter was highest for poorly drained soil (0.99), with similar values of Q (0.40) observed for moderately well and somewhat poorly drained soil. Values of Q for soil moisture were 0.45, 0.81, and 1.0 for moderately well, somewhat poorly, and poorly drained soil, respectively.
| Equation 2 |
| TABLE 5 Geostatistical properties of phosphatase, organic matter, and moisture in soil from different drainage classes |
Range values for phosphatase activity were also different among drainage classes (Table 5). The highest range (12.2 m) was observed for moderately well drained soil, followed by somewhat poorly (2.7 m) and poorly drained soil (0.8 m). Range values for phosphatase were relatively close to those observed for organic matter in all three drainage (Table 5). Soil moisture content had range values in the order of those observed for acid phosphatase activity and organic matter for somewhat poorly and poorly drained soil, with much lower range values (0.5 m) found in moderately well drained soil.
Kriged maps of phosphatase activity, organic matter, and moisture in the various soil drainage classes are shown in Fig. 4. If the spatial distribution of phosphatase is controlled by organic matter and/or moisture, co-occurrence of these variables should be apparent in the maps; in those areas where co-occurrence is not observed, additional factors may control phosphatase activity. Map locations are given in (x, y) notation for a point approximately in the center of the area being referred to. In the moderately well drained soil, areas of high phosphatase activity coincided with high organic matter and moisture at (4, 8). A broad area of low phosphatase activity at (0.5, 14) co-occurred with a low organic matter region, but no extremes of moisture were observed in the same area. Co-occurrence of high phosphatase, organic matter, and moisture in somewhat poorly drained soil was observed only at (3.75, 4). Otherwise, phosphatase activity and organic matter content maxima coincided at (4, 16), and minima for these two variables were apparent at (3, 2) and (0.2, 20). Moisture and phosphatase activity maxima coincided at (4, 4) and minima coincided at (0.5, 15). No co-occurrence of extrema was observed between phosphatase and organic matter or moisture in poorly drained soil, but an inverse relationship between phosphatase and moisture content was apparent at (4.5, 4) and (2.5, 12).
| Fig. 4. Maps showing kriged organic matter, volumetric soil moisture, and phosphatase activity for (A) moderately well, (B) somewhat poorly, and (C) poorly drained soil. |
| Fig. 4. (continued) Maps showing kriged organic matter, volumetric soil moisture, and phosphatase activity for (A) moderately well, (B) somewhat poorly, and (C) poorly drained soil. |
| Fig. 4. (continued) Maps showing kriged organic matter, volumetric soil moisture, and phosphatase activity for (A) moderately well, (B) somewhat poorly, and (C) poorly drained soil. |
There are few published studies evaluating the spatial variability of phosphatase, and none have involved forested ecosystems. The available studies have either ignored the possibility of spatial dependence (Bonmati et al. 1991) or performed limited comparisons of variance at two different sampling scales (Speir et al. 1984). The results of Speir et al. (1984), although difficult to evaluate because it is unclear what the distances between samples were, seem to indicate that there was lower variability in phosphatase activity between adjacent cores than between cores taken at an unspecified longer distance apart. Although these results suggest that phosphatase activity is spatially autocorrelated, to the best of our knowledge our study is the first to evaluate the geostatistical properties of soil phosphatase in any ecosystem.
The spatial dependence of phosphatase activity has important implications to sampling of this enzyme activity. Although its association with organic matter and moisture has been known for decades, sampling schemes as well as statistical analyses of soil phosphatase data in previous studies have almost invariably assumed implicitly that the distribution of this activity is spatially independent. That is, phosphatase is treated as if it were randomly distributed in the landscape even though the factors thought to control it are known not to be randomly distributed. Our study shows that, within the riparian area studied, spatial autocorrelation accounts for 30 to 75% of the variability observed in phosphatase activity, depending on soil drainage class (Table 5). Failing to account for this important component of the total variance may, for example, interfere with evaluation of treatment effects, ecosystem comparisons, or temporal differences in landscape-scale studies of phosphatase.
The need for detailed information on the spatial distribution of phosphatase may not be immediately apparent. For example, a comparison of the first-order and kriged means for phosphatase activity from each drainage class (Fig. 5) shows that these values were essentially identical. The similarity in mean values determined by these two different methods is not surprising. Exhaustive sampling, as is the case in the present study, results in identical sample mean and spatial arithmetic average (Isaacs and Srivastava 1989). By contrast, the kriged standard deviation was lower in all cases (Fig. 5). The inclusion of spatial dependence in variance estimates gives rise to the lower uncertainty of mean values estimated by kriging (Isaaks and Srivastava 1989), allowing for greater precision in estimating soil properties at the landscape scale.
| Fig. 5. Comparison of mean (A) and standard deviation (B) for phosphatase activity obtained by first-order statistical analysis and by kriging in different soil dralnage classes. Dashed line represents 1:1 correspondence between values. |
Determination of the spatial distribution of phosphatase activity is invaluable in designing sampling schemes for monitoring purposes. Sampling can be targeted in areas where activity is particularly high or low, which may exhibit a greater response to environmental disturbances. Based on our estimates of the population mean, median, and variance (Table 2), it would require 64, 63, 52, and 30 samples to detect a 20% change in phosphatase activity at the 95% confidence level in the whole catena, the moderately well, somewhat poorly, and poorly drained soils, respectively. Given limited resources, sampling can be targeted in areas where heterogeneity is low, both within a drainage class and within a catena, reducing the number of samples needed to assess differences in activity as a function of time. In more fundamental terms, spatial maps may be used to assess how seasonal changes affect the relative importance of different factors controlling phosphatase activity such as organic matter, moisture, and root distribution.
Our results suggest that the spatial distribution of phosphatase cannot be explained solely by the distribution of organic matter and moisture because spatial coincidence between these variables was not consistent throughout the catena, especially in the poorly drained soil. Other factors must influence the spatial distribution of phosphatase in this drainage class. Differences in composition of soil organic matter may be responsible for a portion of the spatial variability not explained simply by soil organic matter. Soil organic matter determinations are integrative and thus do not distinguish among sources. The distribution of phosphatase in poorly drained soil may be controlled by plant species distribution or, rather, their roots, to a greater extent than in better drained areas because of lower humification in poorly drained soil. Clark and Clark (1981) observed higher phosphatase activity as the plant species diversity increased in an acidic peat soil. The species composition of the herbaceous layer of red maple swamps changes along soil drainage gradients (Allen et al. 1989; Golet et al. 1993). In addition, the relative contribution of roots to soil mass appears to be significantly greater in poorly drained than in somewhat poorly and moderately well drained soil (Fig. 2). This difference may be important in determining the magnitude of phosphatase activity in different soil drainage classes. Roots and their spatial distribution may contribute differently to spatial variability of acid phosphatase depending on landscape position.
Our results show that soil phosphatase activity in a riparian forest varies spatially as a function of position in the landscape (i.e., soil drainage class) and exhibits considerable spatial dependence within a particular drainage class. As in other ecosystems, phosphatase activity is positively correlated with soil organic matter and moisture, but this relationship is different in soil from different drainage classes and is nonlinear when the data for the whole catena are considered, suggesting an interaction between landscape position and level of phosphatase activity. Phosphatase activity exhibits considerable spatial autocorrelation, with 30 to 75% of the total variance within a drainage class explained by spatial dependence. The distance over which phosphatase is spatially autocorrelated increases as drainage improves and in most cases is similar to that for soil organic matter and moisture. Co-occurrence of phosphatase activity with soil moisture and organic matter in space is observed primarily in moderately well and somewhat poorly drained soil, with little spatial coincidence observed between phosphatase, organic matter, and water content in poorly drained soil. We conclude that the phosphatase activity in riparian forest soils exhibits spatial structure and that this spatial distribution is partly controlled by position in the landscape, with the influence of organic matter, moisture, and possibly root biomass distribution as significant factors in determining dependence of spatial variability on soil drainage class.
This research was sponsored by a grant from the U.S. Geological Service/Rhode Island Water Resources Center and the Rhode Island Agricultural Experiment Station. We thank R. Waldron for invaluable assistance with land surveying.
Allen, S.D., F.C. Golet, A. F. Davis, and T. E. Sokoloski. 1989. Soil-vegetation correlations in transition zones of Rhode Island Red Maple swamps. Biol. Rep. 89(8). Fish and Wildlife Service, U.S. Dept. of the Interior, Washington, D.C. [Context Link]
Alpkem Corp. 1986. RFA-300 Operator's manual. Alpkem Corp., Clackamas, OR. [Context Link]
Bonmati, M., B. Ceccanti, and P. Nannipieri. 1991. Spatial variability of phosphatase, urease, protease, organic carbon, and total nitrogen in soil. Soil Biol. Biochem. 23:391-396. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Bonmati, M., M. Pujola, J. Sana, M. Soliva, M. T. Felipo, M. Garau, B. Ceccanti, and P. Nannipieri. 1985. Chemical properties, populations of nitrite oxidizers, urease and phosphatase activities in sewage sludge-amended soils. Plant Soil 84:79-91. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Burgess, T. M., and R. Webster. 1980. Optimal interpolation and isarithmic mapping of soil properties. I. The semi-variogram and punctual kriging. J. Soil Sci. 31:315-331. [Context Link]
Clark, R. K., and S. C. Clark. 1981. Floristic diversity in relation to soil characteristics in a lead mining complex in the Pennines, England. New Phytol. 87:799-815. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Cooper, J. R., J. W. Gilliam, R. B. Daniels, and W. P. Robarge. 1987. Riparian areas as filters for agricultural sediments. Soil Sci. Soc. Am. J. 51:416-420. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Dick, R. P. 1994. Soil enzyme activities as indicators of soil quality. In Defining soil quality for a sustainable environment. J. W. Doran, D. C. Coleman, D. F. Bezdicek, and B. A. Stewart (eds.). SSSA-ASA, Madison, WI, pp 107-124. [Context Link]
Dick R. P., D. D. Myrold, and E. A. Kerle. 1988. Microbial biomass and soil enzyme activities in compacted and rehabilitated skid trail soils. Soil Sci. Soc. Am. J. 52:512-516. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Dick, W. A., and M. A. Tabatabai. 1984. Kinetic parameters of phosphatases in soils and organic waste materials. Soil Sci. 137:7-15. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Dick, W. A., N. G. Juma, and M. A. Tabatabai. 1983. Effects of soils on acid phosphatase and inorganic pyrophosphatase of corn roots. Soil Sci. 136:19-25. ExternalResolverBasic Bibliographic Links Library Holdings
Doran, J. W. 1980. Soil microbial and biochemical changes associated with reduced tillage. Soil Sci. Soc. Am. J. 44:765-771. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Duxbury, J. M., and R. L. Tate III. 1981. The effect of soil depth and crop cover on enzymatic activities in Pahokee muck. Soil Sci. Soc. Am. J. 45:322-328. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Ehrenfeld, J. G. 1987. The role of woody vegetation in preventing ground water pollution by nitrogen from septic tank leachate. Water Res. 21:605-614. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Eivazi, F., and M. A. Tabatabai. 1977. Phosphatases in soils. Soil Biol. Biochem. 9:167-172. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Frankenberger, W. T., and W. A. Dick. 1983. Relation between enzyme activities and microbial growth and activity indices in soil. Soil Sci. Soc. Am. J. 47:945-951. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Gerritse, R. G., and H. van Dijk. 1978. Determination of phosphatase activities of soils and animal wastes. Soil Biol. Biochem. 10:545-551. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Gianfreda, L., and J.-M. Bollag. 1996. Influence of natural and anthropogenic factors on enzyme activity in soil. In Soil biochemistry, vol. 9. G. Stotzky and J.-M. Bollag (eds.). Marcel Dekker, Inc., New York, pp 123-193. [Context Link]
Golet, F. C., A. J. K. Calhoun, W. R. DeRagon, D. J. Lowry, and A. J. Gold. 1993. Ecology of red maple swamps in the glaciated northeast: A community profile. Biological Report 12. U.S. Dept. of the Interior, Fish and Wildlife Service, Washington, DC. [Context Link]
Gould, W. D., R. V. Anderson, J. F. McClennan, D. C. Coleman, and J. L. Gurnsey. 1979. Characterization of a paleosol: Its biological properties and effect on overlying soil horizons. Soil Sci. 128:201-210. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Gregory, S. V., F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem perspective of riparian zones: Focus on links between land and water. Bioscience 41:540-551. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Harrison, A. F. 1979. Variation of four phosphorus properties in woodland soils. Soil Biol. Biochem. 11:393-403. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Harrison, A. F. 1982. Labile organic phosphorus mineralization in relationship to soil properties. Soil Biol. Biochem. 14:343-351. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Harrison, A. F., and T. Pearce. 1979. Seasonal variation of phosphatase activity in woodland soils. Soil Biol. Biochem. 11:405-410. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Harrison, A. F. 1983. Relationship between intensity of phosphatase activity and physico-chemical properties in woodland soils. Soil Biol. Biochem. 15:93-99. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Hendershot, W. H., H. Lalande, and M. Duquette. 1993. Soil reaction and exchangeable acidity. In Soil sampling and methods of analysis. M. R. Carter (ed.). Lewis Publishers, Boca Raton, FL, pp 141-159. [Context Link]
Holloway, J. D., and N. W. Stork. 1991. The dimension of biodiversity. The use of invertebrates as indicators of human impact. In The biodiversity of microorganisms and invertebrates: Its role in sustainable agriculture. D. L. Hawksworth (ed.). CAB International, Wallingford, Oxon, UK, pp 37-63. [Context Link]
Isaaks, E. H., and R. M. Srivastava. 1989. An introduction to applied geostatistics. Oxford Univ. Press, Oxford, UK. [Context Link]
Jacobs, T. C., and J. W. Gilliam. 1985. Riparian losses of nitrate from agricultural drainage waters. J. Environ. Qual. 14:472-478. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Juma, N. G., and M. A. Tabatabai. 1977. Effects of trace elements on phosphatase activity in soils. Soil Sci. Soc. Am. J. 41:343-346. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Juma, N. G., and M. A. Tabatabai. 1978. Distribution of phosphomonoesterases in soils. Soil Sci. 126:101-108. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Juma, N. G., and M. A. Tabatabai. 1988. Comparison of kinetic and thermodynamic parameters of phosphomonoesterases of soils and of corn and soybean roots. Soil Biol. Biochem. 20:533-539. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Karam, A. 1993. Chemical properties of organic soils. In Soil sampling and methods of analysis, M. R. Carter (ed.). Lewis Publishers, Boca Raton, FL, pp 459-471. [Context Link]
Lowrance, R. R., R. L. Todd, J. Fail, Jr., O. Hendrickson, Jr., R. Leonard, and L. Asmussen. 1984. Riparian forests as nutrient filters in agricultural watersheds. Bask. 34:374-377. [Context Link]
Lowrance, R. R., R. Leonard, and J. M. Sheridan. 1985. Managing riparian ecosystems to control nonpoint pollution. J. Soil Water Conserv. 40:87-91. [Context Link]
Lowrance, R., J. K. Sharpe, and J. M. Sheridan. 1986. Long-term sediment deposition in the riparian zone of a coastal plain watershed. J. Soil Water Conserv. 41:266-271. [Context Link]
Malcolm, R. E. 1983. Assesment of phosphatase activity in soils. Soil Biol. Biochem. 15:403-408. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Martin, J. P. 1973. The influence of rhizosphere microflora on the availability of 32P-myoinositol hexaphosphate phosphorus to wheat. Soil Biol. Biochem. 5:473-483. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Mathur, S. P., and A. F. Rayment. 1977. Influence of trace element fertilization on the decomposition rate and phosphatase activity of a mesic fibrisol. Can. J. Soil Sci. 57:397-408. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Mathur, S. P., and R. B. Sanderson. 1978. Relationships between copper content, rates of soil respiration and phosphatase activites of some histosols in an area of Southwestern Quebec in the summer and fall. Can. J. Soil Sci. 58:125-134. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Mitsch, W.J., and J. G. Gosselink, 1993. Wetlands, 2nd Ed. Van Nostrand Reinhold, New York. [Context Link]
Neal, J. L. 1973. Influence of selected grasses and forbs on soil phosphatase activity. Can. J. Soil Sci. 53:119-121. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Nannipieri, P., R. L. Johnson, and E. A. Paul. 1978. Criteria for measurement of microbial growth and activity in soil. Soil Biol. Biochem. 10:223-229. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Nannipieri, P., B. Ceccanti, and D. Bianchi. 1988. Characterization of humus-phosphatase complexes extracted from soil. Soil Biol. Biochem. 20:683-691. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
National Research Council. 1993. Soil and water quality: An agenda for agriculture. National Academy Press, Washington, DC. [Context Link]
Pang, P. C. K., and H. Kolenko. 1986. Phosphomonoesterase activity in forest soils. Soil Biol. Biochem. 18:35-40. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Parent, L. E., and J. Caron. 1993. Physical properties of organic soils. In Soil sampling and methods of analysis. M. R. Carter (ed.). Lewis Publishers, Boca Raton, FL, pp 441-458. [Context Link]
Peterjohn, W. T., and D. L. Correll. 1984. Nutrient dynamics in an agricultural watershed: Observations on the role of a riparian forest. Ecology 65:1466-1475. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Pulford, I. D., and M. A. Tabatabai. 1988. Effect of waterlogging on enzyme activities in soils. Soil Biol. Biochem. 20:215-219. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Ramirez-Martinez, J. R. 1968. Organic phosphorus mineralization and phosphatase activity in soils. Folia Microbiol. 13:161-174. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Rastin, N., K. Rosenplanter, and A. Huttermann. 1988. Seasonal variation of enzyme activities and their dependence on certain soil factors in a beech forest soil. Soil Biol. Biochem. 20:637-642. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Riha, S. J., B. R. James, G. P. Senasac, and E. Pallant. 1986. Spatial variability of soil pH and organic matter in forest plantations. Soil Sci. Soc. Am. J. 50:1347-1352. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Rojo, M. J., S. G. Garcedo, and M. P. Mateos. 1990. Distribution and characterization of phosphatase and organic phosphorus in soil fractions. Soil Biol. Biochem. 22:169-174. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Ross, D. J., T. W. Speir, D. J. Giltrap, B. A. McNeilly, and L. F. Molloy. 1975. A principal components analysis of some biochemical activities in a climosequence of soils. Soil Biol. Biochem. 7:349-355. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Schaffer, A. 1993. Pesticide effects on enzyme activities in the soil ecosystem. In Soil biochemistry, vol. 8. J.-M. Bollag and G. Stotzky (eds.). Marcel Dekker, Inc., New York, pp 273-340. [Context Link]
Skujins, J. 1976. Extracellular enzymes in soil. CRC Crit. Rev. Microbiol. 4:383-421. [Context Link]
Smith, J. L., J. J. Halvorson, and R. I. Papendick. 1994. Multiple variable indicator kriging: A procedure for integrating soil quality indicators. In Defining soil quality for a sustainable environment. J. W. Doran, D. C. Coleman, D. F. Bezdicek, and B. A. Stewart (eds.). SSSA-ASA, Madison, WI, pp 149-157. [Context Link]
Soil Conservation Service Staff. 1972. Soil survey: Laboratory methods and procedures for collecting soil samples. USDA-SCS Soil Survey Investigations Report No. 1. US Govt. Print. Office, Washington, DC. [Context Link]
Soil Survey Staff. 1981. Soil survey of Rhode Island. USDA-Soil Conservation Service, US Govt. Print. Office, Washington, D. C. [Context Link]
Speir, T. W., and D. J. Ross. 1975. Effects of storage on the activities of protease, urease, phosphatase, and sulphatase in three soils under pasture. N Z J. Soil Sci. 18:231-237. [Context Link]
Speir, T. W., and D. J. Ross. 1976. Soil phosphatase and sulphatase. In Soil enzymes, R. G. Burns (ed.). Academic Press, Inc., New York, pp 197-250. [Context Link]
Speir, T. W., and D. J. Ross. 1981. A comparison of the effects of air-drying and acetone dehydration on soil enzyme activities. Soil Biol. Biochem. 13:225-229. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Speir, T. W., D. J. Ross, and V. A. Orchard. 1984. Spatial variability of biochemical properties in a taxonomically-uniform soil under grazed pasture. Soil Biol. Biochem. 16:153-160. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Spiers, G. A., and W. B. McGill. 1979. Effects of phosphorus addition and energy supply on acid phosphatase production and activity in soils. Soil Biol. Biochem. 11:3-8. [Context Link]
Tabatabai, M. A., and J. M. Bremner. 1969. Use of p-nitrophenyl phosphate for assay of soil phosphatase activity. Soil Biol. Biochem. 1:301-307. [Context Link]
Trangmar, B. B., R. S. Yost, and G. Uehara. 1985. Applications of geostatistics to spatial studies of soil properties. Adv. Agron. 38:45-94. [Context Link]
Trasar-Cepeda, M. A., and F. Gil-Sotres. 1987. Phosphatase activity in acid high organic matter soils in Galicia (NW Spain). Soil Biol. Biochem. 19:281-287. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Vought, L. B. M., J. Dahl, C. L. Pedersen, and J. O. Lacoursiere. 1994. Nutrient retention in riparian ecotones. Ambio 23:342-348. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
West, A. W., G. P. Sparling, T. W. Speir, and J. M. Wood. 1988. Dynamics of microbial carbon nitrogen-flush and ATP and enzyme activities of gradually dried soils from a climosequence. Aust. J. Soil Res. 26:519-530. ExternalResolverBasic Bibliographic Links Library Holdings [Context Link]
Wetzel, R. G., and G. E. Likens. 1991. Limnological analyses, 2nd Ed. Springer-Verlag, New York. [Context Link]
Wright, W. R., and E. H. Sautter. 1988. Soils of Rhode Island landscapes. Univ. of Rhode Island Agric. Exp. Sta. Bull. No. 429. [Context Link]
Yang, J., R. D. Hammer, and R. W. Blanchard. 1995. Microscale pH spatial distribution in the Ap horizon of Mexico silt loam. Soil Sci. 160:371-375. [Context Link]
