SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (13)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Agricola
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Related Collections
Right arrow Nutrient Cycling
Soil Science Society of America Journal 66:999-1007 (2002)
© 2002 Soil Science Society of America

DIVISION S-8—NUTRIENT MANAGEMENT & SOIL & PLANT ANALYSIS

Pathways of Soil Phosphorus Transformations after 8 Years of Cultivation under Contrasting Cropping Practices

Zhiming Zhenga, Régis R. Simard*,b, Jean Lafondc and Léon E. Parenta

a Dep. of Soil and Agri-Food Engineering, Laval Univ., Ste-Foy, QC, Canada G1K 7P4
b Dep. of Soil Science, Univ. of Manitoba, Winnipeg, MB, Canada R3T 2N2
c Agriculture and Agri-Food Canada Research Farm, Normandin, QC, Canada G8M 4K3

* Corresponding author (Simardrr{at}umanitoba.ca)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The knowledge of P transformations is essential to understand the P behaviour in soils. Organic (Po) and inorganic (Pi) P fractions were obtained from the 0- to 30- and 30- to 90-cm layers of a Labarre silty clay (fine, mixed, frigid, Humic Cryaquept) by a sequential extraction. Barley (Hordeum vulgare L.) monoculture and 3-yr barley-forage rotation as main plots, and two nutrient sources (mineral fertilizer and liquid dairy manure) as subplots were studied. Path analysis was used to describe soil P transformations between the falls of 1989 and 1997. Nutrient source was the main factor affecting P transformations in the 0- to 30-cm soil layer, whereas crop rotation predominated in the subsoil. Pathways differed more for Po than for Pi pools. In the 0- to 30-cm soil layer, 86 to 97% of the variation in resin-P was the direct effect of added Pi. The NaHCO3-Pi and Po were sensitive to crop rotations and nutrient sources, and were transitory pools of soil P. The NaOH-Pi was a sink for added Pi and a source of NaHCO3-Pi. Labile Pi tended to store as NaOH-Po, which was the sink for added Po and immobilized soil Pi in the mineral fertilizer systems. Conversely, the NaOH-Po was mineralized and contributed to labile Pi in dairy manure systems. The NaOH-Po acted as store of subsoil Po in the monoculture but was the main source of labile P in the rotation systems. The P transformations in this Cryaquept were strongly influenced by nutrient sources and crop rotations.

Abbreviations: LDM, liquid dairy manure • LS, lignosulfate • M, barley monoculture • MIN, mineral fertilizer • Pi, inorganic P • Po, organic P • R, barley-forage rotation


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE PHOSPHORUS TRANSFORMATIONS in soils involve complex mineralogical, chemical, and biological processes. The leaching of bases, the removal of carbonates, and increases in Fe and Al activity that accompany pedogenesis or podzolization cause a shift from primary to secondary Pi forms and also influence the stabilization of organic matter and its associated Po compounds (Walker and Syers, 1976). The abundance and activity of various Pi forms and the turnover of Po in soils control the supply of labile P to soil solution by desorption or dissolution of Pi from the solid phase (Murrman and Peech, 1969), or by the mineralization of Po (McGill and Cole, 1981). Microorganisms deeply affect soil P turnover through mineralization–immobilization processes (Stevenson, 1986). Investigations on the seasonal changes in soil P fractions (Dormaar, 1972; Magid and Nielsen, 1992), and studies using 32P to follow the changes in soil P forms (Tran et al., 1988; Frossard et al., 2000) have shown that the mass transfers among P pools are relatively rapid. However, quantitative assessment of P transformations in soils has lagged behind its conceptual characterization.

The complexity of soil P transformations has led to the use of both conceptual and mechanistic models for their description. Path analysis was introduced to evaluate the mechanisms of P transformations in soils. Using path analysis, Tiessen et al. (1984) found that much of resin-P originated from NaHCO3- and NaOH-Pi in Mollisols, whereas, in more weathered Ultisols, 80% of the variability in labile P was accounted for by Po forms. Xie et al. (1991) showed that, with lignosulfonate (LS) addition to an aqueptic soil, P retention was mainly related to concentrations of P, Ca, organic C, and retained LS-C. Beck and Sanchez (1994) showed that the NaOH-Pi fraction acted as a major sink for fertilizer P in an Ultisol, and that Po was a major primary source of plant-available P in unfertilized soil during 18 yr of crop production in Peru. Zhang and MacKenzie (1997) indicated that, for an Alfisol receiving manure and fertilizer P, Po accumulated as NaOH-Po through resin-P. Resin-P along with NaOH-Pi were the major sinks for added P. When only inorganic fertilizer P was added, most NaHCO3-Pi was directly supplied from fertilizer P through NaOH-Pi. Thus, path analysis can reveal changes in P forms, which depend on soil type, climatic conditions, and cropping practices. It provides not only plausible explanations of observed correlations by constructing models of cause-and-effect relations among variables, but also allows the decomposition of the observed correlation into direct and indirect effects which operate indirectly through subsequent variables (Johnson and Wichern, 1988).

Previous work by Zheng et al. (2001) indicated that the changes in P forms were strongly related to P transformations for the same soil used in the present study, whereas the cause-and-effect relationships between P pools were not identified. In the literature, the distinction of the pathways of soil P transformations under contrasting practices is not well documented. Such information is essential to assess P transformations mechanistically and to understand P behavior in soils. The objective of this study was, therefore, to utilize path analysis, determining the pathways of soil P transformations and clarifying the distinctions of P pools acting as sinks or sources for P in a Humic Cryaquept under contrasting cropping practices.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Design
The study was initiated in the fall of 1989 on a Labarre silty clay (fine, mixed, frigid, Humic Cryaquept) at the Agriculture and Agri-Food Canada Research Farm in Normandin, Qubec, Canada (48°N, 71°W). Four cropping systems consisting of the factorial combinations of two crop rotations and two nutrient sources were studied. The two crop rotations included a continuous spring barley (Hordeum vulgare L., cv. Chapais) monoculture (M) and a 3-yr barley-forage rotation (R). In the rotation, the first-year spring barley was undersown with a mixture of timothy (Phleum pratense L., cv. Champ) and red clover (Trifolium pratense L., cv. Prosper) and was followed by 2 yr of forage production (2 cuts yr-1 and 4.5 Mg ha-1 yr-1 average dry matter yield). Each phase of the rotation was present in each year of the study. Only barley monoculture and the rotation which was in the second year of forage production in 1990 were selected for this study. The effect of nutrient sources was assessed by comparing mineral fertilizer (MIN) to liquid dairy manure (LDM). The mineral fertilizer treatments consisted of 70 kg N ha-1 yr-1 as NH4NO3, 20 kg P ha-1 yr-1 as triple superphosphate, and 58 kg K ha-1 yr-1 as KCl. The LDM was supplied at 51 m3 ha-1 yr-1 from 1990 to 1995 and at near 80 m3 ha-1 yr-1 since 1996 to provide more N in accordance to increased local N recommendations for grass forage (Parent et al., 1996). Nutrients were surface-applied and immediately incorporated to a 7.5-cm depth by one pass of a disk-harrow 2 d prior to barley seeding. The nutrients were not incorporated when applied to established forage. Four treatments were established in a split-plot design with two crop rotations assigned to main plots, and two nutrient sources as subplots of 10 by 5 m each, in four replicates. Herbicides were applied at regionally recommended rates for weed control in spring barley monoculture, but not necessarily in the barley–forage rotation (Stevenson et al., 1998). More detailed accounts of experimental site, climatic characteristics, and field operations were given in Zheng et al. (2001).

Data Collection and Statistical Analysis
Samples were taken from the 0- to 30- and 30- to 90-cm soil layers after crop harvest in the falls of 1989 and 1997. Three composite samples were pushed through 3.5- or 6-mm sieves in the field, and stored at 4°C in plastic containers until analysis. Particle-size separation was done by the hydrometer method (Sheldrick and Wang, 1993). Soil pH was measured in distilled water with a soil/solution ratio of 1:2. Soil water-soluble P was determined in a 1:20 weight/volume ratio of soil/water according to a modification of Sissingh's (1971) procedure (Simard et al., 1995). The ammonium oxalate-extractable P, A1, and Fe were extracted as described by Ross and Wang (1993). The profilewise distribution of the selected chemical and physical properties is given in Table 1. Soil samples were subjected to P fractionation following a modification of the procedure of Hedley et al. (1982), as fully described by Simard et al. (1995). Briefly, a portion of the NaHCO3 and NaOH extracts was acidified to precipitate the extracted organic matter, and the supernatants were analyzed for Pi. Other portions of the NaHCO3 and NaOH extracts, as well as of the soil residue after sequential extraction, were digested in an autoclave (103.4 kPa, 121°C for 1.5 h) by an acidified K2S2O8 oxidation (Environmental Protection Agency, 1971) and were analyzed for total P. The P concentration in the extracts was determined colorimetrically by the molybdate-blue method of Murphy and Riley (1962). The difference between total P and Pi was assumed to be as Po (Tiessen and Moir, 1993). Total soil P was taken as the sum of P in all fractions (O'Halloran, 1993). Labile Pi was obtained by adding resin-P to NaHCO3-Pi. Labile Po was computed as the sum of NaHCO3-Po and NaOH-Po (Beauchemin et al., 1996). Total labile P was obtained as the sum of labile Pi + labile Po. Resilient P was calculated as total P - labile P (Simard et al., 1995). All harvested plant tissues were analyzed for P concentration according to Richards (1993).


View this table:
[in this window]
[in a new window]
  		Table 1. Average values and standard deviations of selected chemical and physical properties in the 0- to 30- and 30- to 90-cm soil layers on a Labarre silty clay at Normandin in 1989.

 
Analysis of variance was conducted independently for four combinations of crop rotations and nutrient sources. Phosphorus inputs from mineral fertilizer and dairy manure and P removal by harvested plant tissues were calculated on an annual treatment basis (Zheng et al., 2001). The calculation of cumulative Pi and Po additions, P removal, and P surplus from 1989 to 1997 are shown in Table 2. The added Pi included the P from inorganic fertilizer and Pi from manure. The added Po was considered as the Po from manure and from crop residues including stover and roots. Dairy manure was assumed to contain 630 g Pi kg-1 and 370 g Po kg-1 (Barnett, 1993). The harvested-crop residues biomass ratios of 2.0 for barley (Bolinder et al., 1997) and of 1.0 for red clover (Bolinder and Angers, 1997) were used to estimate root biomass. The average P content of plant residues was estimated as 0.2%. Since P downward movement was not clearly demonstrated in this study (Zheng, 2001), added Pi and Po were considered as external factors which indirectly affect P transformations in the 30- to 90-cm soil layer.


View this table:
[in this window]
[in a new window]
  		Table 2. Cumulative Pi and Po additions from mineral fertilizer, liquid dairy manure and crop residues, P removal and P surplus from 1989 to 1997 under different crop rotation and nutrient source combinations on a Labarre silty clay at Normandin.

 
Correlation analysis was used to determine the intensity of the relationships between P fractions, and the significance of the relationships was tested at P <= 0.05. Correlation coefficients were taken as the initial coefficients of linear structure equations, which were subjected to optimization in the path analysis. Path analysis was used to evaluate the corresponding changes between soil P fractions, as well as of added P in a conceptual model representing the P transformations in soil. This technique requires development of a conceptual model used to establish relationships among various entities (Li, 1975). It can determine the relative contributions to an endogenous (dependent) variable by several exogenous (independent) variables or other endogenous variables. The PROC CALIS procedure (SAS Institute, 1990) was used to derive the path coefficients between the P pools in the conceptual models. To simplify the modeling procedure, linear relationships between P pools were used, although in some cases curvilinear relationships were slightly more significant. Path analysis is based on the ordinary least square estimate of a set of linear structure equations. The optimized coefficients from these equations are taken as path coefficients (ß). Path coefficients are essentially linear regression coefficients. This statistical procedure also tests how well conceptual models represent graphic interpretations of P transformations in soil. Fit Goodness Index, Chi-square Test, and Akaike's Information Criterion were the criteria selected to determine the fit in the present study (Beck and Sanchez, 1994).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Changes in Soil Phosphorus Fractions
The significance levels of analysis of variance and the mean changes of seven P fractions from 1989 to 1997 in the 0- to 30-cm and 30- to 90-cm soil layers are shown in Table 3. In the 0- to 30-cm soil layer, the increases in all P pools except for resilient P fractions (HCl-P and H2SO4-P) were significantly affected (P <= 0.05) by crop rotations and nutrient sources. The MIN plots showed larger increases in Pi fractions (resin-P, NaHCO3-Pi, and NaOH-Pi) but lower increases in Po fractions (NaHCO3-Po and NaOH-Po) than did LDM plots. The resultant labile Pi and Po pools were mainly affected by nutrient sources, suggesting that the changes in soil labile P forms in the 0- to 30-cm soil layer depended more on nutrient sources than on crop rotations. This observation supports the hypothesis that soil labile P fractions are much more influenced by nutrient additions than by crop rotations (McKenzie et al., 1992a,b). A slight but significant decrease in NaHCO3-Po was found in monoculture plots. The decrease in NaHCO3-Po was related to a decrease in total soil C content in monoculture plots (Zheng et al., 2001). A decline in NaHCO3-Po with cultivation was reported earlier in a mollisolic silt loam (Tiessen et al., 1984) and in a aqueptic sandy clay (Zhang and MacKenzie, 1997).


View this table:
[in this window]
[in a new window]
  		Table 3. Significance level from analysis of variance (ANOVA) and mean change (1989–1997) in P fractions in the 0- to 30-cm and 30- to 90-cm soil layers under different crop rotation and nutrient source combinations on a Labarre silty clay at Normandin.

 
The changes in labile P fractions (resin-P, NaHCO3-Po, NaOH-Po) and resultant total labile Pi and Po in the subsoil (30–90 cm) were only affected by crop rotations. The changes in P fractions in the 30- to 90-cm soil layer were much smaller than those in the 0- to 30-cm layer. This is related to the limited size of P pools except for HCl-P in the subsoil (Table 1), and to a limited mobility of added P (Zheng, 2001).

The fact that total P was not affected by the treatments after 8 yr of cultivation but significant changes happened to a specific P fraction suggest that P transformations between P fractions did occur. Moreover, the changes in NaHCO3-Pi and -Po pools in the soil solum were smaller than for other P pools. The relative stability of NaHCO3-Pi and -Po suggests that they acted as transitory pools rather than as sinks and sources within the soil P cycle (Beck and Sanchez, 1994).

Simple Correlation Analysis
The correlation coefficients among the seven P pools differed between treatments (Table 4). In the 0- to 30-cm soil layer, added Pi was very closely related to the labile and moderately labile Pi pools (resin-P, NaHCO3-Pi, and NaOH-Pi) in all cropping systems. Added Pi and Po were not related to HCl-P in all cropping systems, whereas they were positively related to NaHCO3-Po pool in all cropping systems but M-MIN (data not shown). The absence of correlation between added P and the NaHCO3-Po pool in this system is probably related to lower C inputs than in the other studied cropping systems (Zheng et al., 2001). The NaOH-Po pool was closely related to added Po in all systems, but the relationship between these two variables was better in the MIN systems, in which roots and straw are the sole C inputs, than in the LDM systems, in which the Po also comes from manure. This manure Po is probably transformed to Pi in the soil (Sharpley et al., 1984). The H2SO4-P pool was not related to added Pi and Po except in the R-MIN system, where the relationship was weak (r = 0.55). The correlation between the specific source and recipient P pools were not consistent across the four systems, suggesting that the fate of added P may be specific to the cropping systems under investigation.


View this table:
[in this window]
[in a new window]
  		Table 4. Pearson's correlation coefficients (only coefficients significant at P 0.05 are shown) for P fractions in the 0- to 30- to 90-cm soil layers under selected crop rotation-nutrient source systems{dagger} in a Labarre silty clay at Normandin.

 
In the subsoil, resin-P was strongly related to Po pools in the monoculture systems, but only a weak correlation between resin-P and the NaOH-Po was found in the rotation systems. Resin-P was not as strongly related to other Pi pools as was the case for the 0- to 30-cm soil layer. The NaHCO3-Pi was related to the NaOH-Pi pool in all cropping systems but the relationships were not as close as those found in the 0- to 30-cm layer. Weaker relationships between Pi fractions in subsoil may be attributed to a lower C content and lower microbial activity, as compared to the top 30 cm of soil (Stevenson, 1986; Magid et al., 1996). The NaHCO3-Po and NaOH-Po fractions were only related to each other in the rotation systems but not in the monoculture systems. In the rotation, the NaOH-Pi and NaOH-Po pools were strongly related, and the HCl-P was inversely and weakly related to the H2SO4-P pool (-0.51< r <-0.55). These relationships were not significant in monoculture systems. This may be attributed to the different pathways of transformations between P pools which predominate in the rotation and in the monoculture systems.

Correlation is descriptive and does not necessary indicate cause and effect (Steel and Torrie, 1980). A significant correlation coefficient could be due to noncausal or spurious relationships (Johnson and Wichern, 1988). An observed correlation between added Pi and Po, for instance, cannot be used as proof of a significant causal relationship. Conversely, a nonsignificant correlation coefficient resulting from total associations does not mean that the isolated direct or indirect effect is not significant (Johnson and Wichern, 1988). For instance, the correlation between resin-P and H2SO4-P is generally not significant, but the direct or indirect relationships between these two pools and mid-transitory pools (i.e., NaHCO3-Pi) could be significant. Therefore, to analyze precisely the cause-and-effect relationship and separate this relationship into direct and indirect effects, path analysis is especially recommended (Johnson and Wichern, 1988).

Pathways of Phosphorus Transformations in the 0- to 30-cm Soil Layer
Mineral Fertilizer Systems
The path relationships between the soil P pools in two mineral fertilizer systems (M-MIN and R-MIN) mostly followed the same trend. For example, in the R-MIN (Fig. 1) , path analysis shows that 97% of the variation in resin-P was derived from the addition of inorganic fertilizer P. The NaHCO3-Pi was also mainly the direct result of Pi addition (ß = 0.80) and indirectly through NaOH-Pi (ß = 0.95 x 0.38). In contrast to the correlation analysis, the path coefficient between NaHCO3-Pi and Po was significant (P <= 0.05), and it appeared that the NaHCO3-Pi was immobilized to replenish the decrease of NaHCO3-Po (Table 3), which was further stabilized (outgoing paths) to the NaOH-Po and H2SO4-P pools. Conversely to the correlation analysis, a negative relationship was found between resin-P and NaHCO3-Pi because resin-P and NaHCO3-Pi are part of the same soil P pool which is considered as labile P adsorbed on surfaces of more crystalline P compounds, sesquioxides, or carbonates (Mattingly, 1975). The NaOH-Pi acted as a sink for added Pi, probably through the formation of amorphous and crystalline Al- and Fe- phosphate (Williams et al., 1980). Simard et al. (1995) reported that a significantly larger NaOH-Pi pools in agricultural than in forest Appalachian soils, suggesting that NaOH-Pi is an important sink for added or mineralized P.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 1. Relationships between P pools in the 0- to 30-cm soil layer under mineral fertilizer systems (R-MIN), shown as non-normalized path coefficients on a Labarre silty clay. (The percentage value indicates the partial correlation between added Pi and resin-P. NS, not significant at P <= 0.05).

 
Both NaOH-Po and NaHCO3-Po acted as sinks for added Po from crop residues and root biomass. There was a weak but significant relationship between NaOH-Po and NaHCO3-Po. The NaHCO3-Po tended to reorganize as NaOH-Po, suggesting that the plant availability of Po was reduced; as a consequence, the NaOH-Po constituted the reservoir of soil Po in this system. Therefore, path analysis clearly showed that mineral fertilizer applications tended to transform labile Pi into more stable Po. The same observation was previously reported by Schmidt et al. (1996) and Zhang and MacKenzie (1997) although their soil types were different from that of the present study. This is probably related to the surplus of inorganic fertilizer P addition as compared to plant exports (Table 2). With fertilizer P rates in excess of plant removal, the excess P was supposed to reduce the phosphorylase activity and consequently reduce the mineralization of NaOH-Po (Schmidt et al., 1996). Simard et al. (1995) also reported a larger amount of NaOH-Po in agricultural soils receiving a surplus of P addition in regards to plant needs. However, NaOH-Po is usually the source of plant-available P on less or nonfertilized soils (Beck and Sanchez, 1994).

The H2SO4-P was positively related to the NaHCO3-Po (ß = 0.80) and negatively related to HCl-P (ß = -0.26) pools. An increase in NaHCO3-Po and a decrease in HCl-P accompanied an increase in H2SO4-P in this system, which was inconsistent with the ANOVA data presented in Table 3. The path analysis also showed that H2SO4-P pool was indirectly derived from P addition through transitory pools of the NaHCO3-Pi, NaOH-Pi, and NaHCO3-Po.

Dairy Manure Systems
The pathways of soil P transformations in two dairy manure systems (M-LDM and R-LDM) were comparable. As represented by the M-LDM system (Fig. 2) , the path relationships between Pi pools followed a trend similar to the mineral fertilizer systems (Fig. 1), except for a positive relationship between NaHCO3-Pi and H2SO4-P (ß = 0.46), as also noted by Beck and Sanchez (1994). Eighty-six percent of the variation in resin-P could be explained by the direct effect of Pi addition.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2. Relationships between P pools in the 0- to 30-cm soil layer under dairy manure systems (M-LDM), shown as non-normalized path coefficients on a Labarre silty clay. (The percentage value indicates the partial correlation between added Pi and resin-P. NS, not significant at P <= 0.05).

 
The NaOH-Po and NaHCO3-Po pools acted as sinks for added Po. In contrast with the M-MIN system, a larger impact of added Po was found in the NaHCO3-Po pool (partial R2 = 0.48, model R2 = 0.76) rather than in the NaOH-Po pool (partial R2 = 0.28, model R2 = 0.76). This is attributed to the large bio-availability of added Po from dairy manure. The path coefficients between NaOH-Po and NaHCO3-Po (ß = -1.28) and between NaHCO3-Po and NaHCO3-Pi pools (ß = -0.12) were negative. The moderately labile NaOH-Po was the source of labile NaHCO3-Po and Pi in this system. This clearly indicated that labile Pi pools (resin-P, NaHCO3-Pi) partly resulted from the mineralization of the moderately labile NaOH-Po pool. Such transformations from Po to Pi with manure addition was earlier reported by Campbell et al. (1986) where there was no significant change in Po fraction with manure applied every 3 yr, whereas labile Pi levels were increased, suggesting that transformations of soil Po to Pi fractions had occurred. However in the present study, the extent of the changes from NaOH-Po to NaHCO3-Po was less than the inputs of Po from LDM and crop residues; as a result, NaOH-Po was ultimately accumulating in the soil (Table 3).

The NaHCO3-Po is considered as labile Po sorbed on the soil surface plus a small amount of microbial P (Bowman and Cole, 1978; Magid et al., 1996). The overall effect of added Po was an increase in NaHCO3-Po pool, which was negatively related to H2SO4-P [ß = 1.51 x (-0.34)]. This, once again, suggests that dairy manure application tended to promote the accumulation of NaHCO3-Po. This observation is consistent with the hypothesis that the addition of organic C as manure accelerates the mineralization of soil Po (Sharpley et al., 1984). Lalande et al. (2000) reported an increase in phosphatase activity with the addition of liquid hog manure to a gleysolic soil, and that the increased enzyme activity has contributed to the mineralization of added Po. However, Zhang and MacKenzie (1997) reported that manure addition tended to promote the formation of less labile Po at the expense of more labile Po in a sandy soil under corn (Zea mays L.) monoculture. These contrasting results may be related to the type of residues added. Corn stover may be more difficult to decompose than barley straw (Pfeiffer and Koepf, 1980). Also, corn stover is normally incorporated later in the season than is barley straw, and may not have sufficient time to decompose before the following spring in soils that are frozen for most of the winter.

Pathways of Phosphorus Transformations in the Subsoil
Monoculture Systems
The pathways of P transformations in the subsoil (30–90 cm) in the two monoculture systems (M-MIN and M-LDM) were similar. As an example for the M-MIN system (Fig. 3) , the labile resin-P pool originated mainly from the NaHCO3-Pi and NaOH-Pi pools. The negative path coefficient of -0.51 between NaOH-Pi and resin-P indicated that the accumulation of resin-P tended to reduce the formation of NaOH-Pi (Hedley et al., 1982). The overall effect of NaOH-Pi on resin-P was negative (ß = 1.10 x 0.31 - 0.51). This negative relationship between resin-P and NaOH-Pi was earlier reported by Tiessen et al. (1984) for Mollisols and by Beck and Sanchez (1994) for an acid Ultisol which received no fertilizer. The significant outgoing paths along with no incoming paths for NaOH-Pi and NaHCO3-Po, are consistent with the decreases in the sizes of these two pools in monoculture systems (Table 3). A decrease in the NaOH-Pi pool was accompanied by increases in resin- and HCl-P pools. The NaHCO3-Po was mineralized to NaHCO3-Pi and was further reorganized as moderately labile NaOH-Po. Resin-P significantly contributed to the formation of NaOH-Po, and as a consequence, the NaOH-Po constituted a major reservoir of soil Po. This is also the case for Mollisols rich in organic matter (Tiessen et al., 1984). A negative relationship was found between NaOH-Pi and HCl-P; this supports the hypothesis that the formation of relative stable P occurs at the expense of more labile Pi pools (Hedley et al., 1982). The H2SO4-P was not related significantly to other P fractions, as was indicated by the absence of significant correlation with other P pools (Table 4). The path analysis also showed that a relatively high proportion of the total variation in resin-P (r2 = 0.34) was not accounted for by the P pools and inputs indicated in the path analysis. This may be attributed to some variables, such as soil C content and microbial activity, which are not included in the conceptual model, but are widely acknowledged to be functional factors on the turnover or flux of P in soil (McGill and Cole, 1981; Magid et al., 1996; Turner and Haygarth, 2001).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 3. Relationships between P pools in the 30- to 90-cm soil layer under monoculture systems (M-MIN), shown as non-normalized path coefficients on a Labarre silty clay. Paths with broken lines are not in structural equations for path analysis. NS, not significant at P <= 0.05.

 
Rotation Systems
The pathways of P transformations in the subsoil were similar for two rotation systems (R-MIN and R-LDM), as indicated by the following example for the R-LDM sys-tem (Fig. 4) . Resin-P was again mainly related to the NaOH-Pi and NaHCO3-Pi pools. The overall effect of NaOH-Pi on resin-P was negative by a directly negative effect and an indirectly positive effect through NaHCO3-Pi (ß = 0.93 x 0.85 - 1.94). This suggests that competition existed between the Pi sorbed on to the surfaces of Fe and Al oxyhydroxides and carbonates (resin-P) and that of soil Pi fixed as Fe and Al phosphates (NaOH-Pi, Tiessen and Moir, 1993). The NaHCO3-Po appeared to act as a sink for NaHCO3-Pi and NaOH-Po pools and a source of H2SO4-P. Unlike in the monoculture systems, the H2SO4-P pool in subsoil under rotation systems was significantly related to the NaHCO3-Pi/-Po and HCl-P pools, and was mainly derived from HCl-P (partial R2 = 0.31, model R2 = 0.53). This suggests that the transformations between resilient P pools or through transient P pools (NaHCO3-Pi and -Po) would only occur when C inputs are large enough to support microbial activity.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. Relationships between P pools in the 30- to 90-cm layer under rotation systems (R-LDM), shown as non-normalized path coefficients on a Labarre silty clay. Paths with broken lines are not in structural equations for path analysis. NS, not significant at P <= 0.05.

 
The NaOH-Po acted as a source of the subsoil labile P in the rotation systems. Such observation was reported by Beck and Sanchez (1994), who found that NaOH-Po was the primary source of plant available P in soil which did not receive fertilizer inputs, and explained 44% of the variation in available P, in an Ultisol under a rice-corn-soybean rotation. The observations are also consistent with previous reports which indicated that the NaOH-Po pool plays an important role in the overall P dynamics in soils (Tiessen et al., 1984; Beck and Sanchez, 1994; Zhang and MacKenzie, 1997). Our results stress that NaOH-Po can act either as a source or as a sink in P transformations. The relative contribution of this pool to others clearly depended, however, on crop rotations and nutrient sources.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Changes in P fractions and transformations can be revealed by path analysis. Although the path coefficients for specific source and recipient P pools depended on crop rotation and nutrient source combinations, this technique suggests that the relationships between P pools were more influenced by nutrient sources in the 0- to 30-cm soil layer than by crop rotations. The inverse was true for the subsoil (30–90 cm).

The pathways of transformations in Pi pools were similar for different crop rotation–nutrient source combinations. In the surface soil, 86 to 97% of the variation in resin-P was the direct effect of added Pi. The NaOH-Pi acted a sink for added Pi and a source for NaHCO3-Pi. Path analysis also clarified the different roles of Po pools in the transformations induced by crop rotations and nutrient sources. In the 0- to 30-cm soil layer, labile Pi was transformed to less labile Po pools. The NaOH-Po acted as a sink for added Po and immobilized soil P in the MIN systems. Conversely, the moderately labile NaOH-Po contributed to labile P through mineralization in the LDM systems. In the subsoil, the NaOH-Po pool acted as a reservoir of soil P in the monoculture systems, whereas was the primary source of labile P in the rotation systems. Despite the fact that NaHCO3-Pi and -Po were sensitive to crop rotations and nutrient sources, the magnitude of change in NaHCO3-Pi and Po pools were relatively small, compared to other pools, suggesting that they acted as transitory pools.

In spite of the limitation of representing a dynamic system by a descriptive and static path analysis, it is a more effective method than simple correlation analysis to interpret the changes in soil P fractions and the relationships between P pools. It highlights the differences in the pathways of soil P cycle between crop rotation–nutrient source combinations, as well as how the changes in one P pool may affect other soil P pools.


   ACKNOWLEDGMENTS
 
We gratefully acknowledge Mrs. S. Cote, Mr. J. Martin, and Mr. A. Larouche for their assistance in the laboratory; numerous seasonal workers and the personnel of the Normandin Experimental Farm for their excellent technical assistance. Thanks are extended to Dr. G.M. Barnett for reviewing an early draft of this manuscript. The authors also acknowledge the Agriculture and Agri-Food Canada, Nutrite Inc., and Natural Sciences and Engineering Research Council of Canada (Grant #216328-98) financial support for this study.

Received for publication March 26, 2001.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
Z. He, C. W. Honeycutt, B. J. Cade-Menun, Z. N. Senwo, and I. A. Tazisong
Phosphorus in Poultry Litter and Soil: Enzymatic and Nuclear Magnetic Resonance Characterization
Soil Sci. Soc. Am. J., August 20, 2008; 72(5): 1425 - 1433.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
R. M. Bayley, J. A. Ippolito, M. E. Stromberger, K. A. Barbarick, and M. W. Paschke
Water Treatment Residuals and Biosolids Coapplications Affect Semiarid Rangeland Phosphorus Cycling
Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 711 - 719.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Y. Tang, H. Zhang, J. L. Schroder, M. E. Payton, and D. Zhou
Animal Manure Reduces Aluminum Toxicity in an Acid Soil
Soil Sci. Soc. Am. J., September 28, 2007; 71(6): 1699 - 1707.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Z. He, A.-M. Fortuna, Z. N. Senwo, I. A. Tazisong, C. W. Honeycutt, and T. S. Griffin
Hydrochloric Fractions in Hedley Fractionation May Contain Inorganic and Organic Phosphates
Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 893 - 899.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
H. Zhang, J. L. Schroder, J. K. Fuhrman, N. T. Basta, D. E. Storm, and M. E. Payton
Path and Multiple Regression Analyses of Phosphorus Sorption Capacity
Soil Sci. Soc. Am. J., January 1, 2005; 69(1): 96 - 106.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Z. Zheng, R. R. Simard, and L. E. Parent
Anion Exchange and Mehlich-III Phosphorus in Humaquepts Varying in Clay Content
Soil Sci. Soc. Am. J., July 1, 2003; 67(4): 1287 - 1295.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (13)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Agricola
Right arrow Articles by Zheng, Z.
Right arrow Articles by Parent, L. E.
Related Collections
Right arrow Nutrient Cycling

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome