| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, New Zealand
b Section of Animal Production Systems, School of Veterinary Medicine, Univ. of Pennsylvania, 382 West Street Road, Kennett Square, PA 19348
c Dep. of Geological and Environmental Science, Stanford Univ., Building 320, Room 118, Stanford, CA 94305-2115
d USDA-ARS, Pasture Systems and Watershed Management Research Unit, Curtin Road, University Park, PA 16802-3702
* Corresponding author (richard.mcdowell{at}agresearch.co.nz)
Received for publication February 14, 2007.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Abbreviations: ICP–AES, inductively coupled plasma–atomic emission spectroscopy • MRP, molybdate-reactive P • MUP, molybdate-unreactive P 31 • myo-IHP, myo-inositol hexakisphosphate (phytic acid) • P-NMR, 31P-nuclear magnetic resonance spectroscopy • TA, teichoic acid • TC, total confinement • TMR, total mixed ration
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Potential P loss from dairy farms depends on a number of factors, including hydrologic conditions determined by climate, soil type, and topography; chemical forms of P in manure (stored feces on farms of confinement and subsequently applied to fields) or dung (direct fecal deposits from grazing animals); and the ability of the soil to sorb P (McDowell and Stewart, 2005a). For example, dung is P enriched (about 5–8 g P kg–1 dry matter) compared with most topsoils under pasture (0.5–3.0 g P kg–1 soil) (McDowell and Condron, 2004), which means that soil sorption sites beneath dung is probably saturated with P, leaving much of the dung P available for transport to surface water bodies should flow occur. The same can occur if too much manure is surface spread on fields without incorporation.
Many of the factors affecting P loss in the field are out of our control. However, the use of different dairy feeding systems can alter the supply and use of P within the dairy farm. By changing the quantity and type of P in the feed, P in feces is altered in terms of total P concentration and solubility (Dou et al., 2002; Wu et al., 2000). Controls can also be implemented on the location and amount of spread dairy manure or deposited dung (Sims and Kleinman, 2005). However, to manage manure or dung properly, information is required about the forms of P within feces because some are more bioavailable than others and how these change with feeding regime.
To isolate and determine P species in feed, dung, and manure, some researchers have used solution 31P nuclear magnetic resonance spectroscopy (31P-NMR). The majority have examined NaOH extracts alone or in combination with EDTA to maximize organic P recovery. The recovery of P from manures can range from 40 to >90% (McDowell and Stewart, 2005a; Turner, 2004; Toor et al., 2005a). This technique isolates many different classes of P species, some of which are bioavailable to algae as orthophosphate, by direct uptake or after enzyme hydrolysis. In general, 31P NMR has shown that orthophosphate diesters are more labile than orthophosphate monoesters or phosphonates. Other inorganic species, such as polyphosphates (of which pyrophosphate is the smallest form), can also be created as a result of biological activity (Condron et al., 1985). By examining NaOH-EDTA extracts of feed, feces, and manures, Toor et al. (2005a) showed that myo-inositol hexakisphosphate (myo-IHP) in the diets of dairy cattle in farms from Northeastern and Mid-Atlantic states was partially converted into orthophosphate by the time it was processed into manure. However, compared with corn and other imported grains rich in myo-IHP, pasture forages contain little myo-IHP (Peperzak et al., 1959). Consequently, the question arises of whether changes in the feeding regime affect the speciation, and hence bioavailability, of P in feces.
In addition to changes with feed, several studies have shown that the solubility and loss of P from manure or dung changes as it dries. McDowell et al. (2006) showed that the potential for P loss in overland flow decreased exponentially with time as dung dried and was an order of magnitude less after 2 d due to the formation of a surface crust that prevented interaction with moist dung beneath. Smith et al. (2001) came to a similar conclusion vis-à-vis the spreading of cattle slurry on a gently sloping field. Although the availability of dung-P to flow decreased with drying, McDowell and Stewart (2005a) also showed with 31P NMR of NaOH-EDTA dung extracts that the proportion of labile organic P species such as orthophosphate diesters decreased via degradation. Ajiboye et al. (2004) showed that labile organic P in beef cattle manure could transfer to inorganic P as moisture decreased, although no measurement was made of possible organic P in the acid extraction (He et al., 2006).
One disadvantage of using alkaline extraction to infer the properties of feces and land-applied manure is that it does not inform us about the potential for P loss in overland or subsurface flow. To resolve these issues, Kleinman et al. (2002) presented data showing that P loss in overland flow via artificial rainfall is highly correlated to P extractable from manure by water at a solid to solution ratio of 1:200. More recent work by Dou et al. (2007) indicates that dilute acid extraction can estimate the P in dairy feces irrespective of sample handling (e.g., drying) or Ca and/or Mg content, which were shown to affect P solubility in water (Chapuis-Lardy et al., 2004).
Examination of water extracts of dung or manure have shown radically different distributions of P species compared with the same dung extracted by NaOH-EDTA (Turner and Leytem, 2004; McDowell and Stewart, 2005b), especially in terms of polyphosphate and orthophosphate diester concentrations. The examination of dilute acid extracts has only been done as part of a sequential fractionation protocol (Turner and Leytem, 2004), providing an incomplete measure of the total pool. However, by using water, NaOH-EDTA, and dilute acid as separate extractants rather than as sequential extractants, a full picture of P forms, bioavailability, and potential to be lost in flow can be achieved.
Our objectives were to explore how P form, and hence bioavailability, differs in the feed and feces of dairy cattle found in herds from four different systems in the Northeastern USA: (i) total confinement (TC) with adequate P in feed, (ii) TC with excessive P in feed, (iii) a hybrid of confinement and grazing, and (iv) predominantly grazing with some supplemental grain (low-input New Zealand-style system). Extractants used were NaOH-EDTA to maximize organic P extraction, water to determine P immediately available to runoff, and dilute HCl to estimate the animal's P supply-utilization status as a potential screening tool for the presence of P overfeeding on dairy farms (Dou et al., 2007). As a secondary objective, all fecal samples were air-dried and extracted to estimate changes that occur with drying as would happen if rotational grazing and dung deposition occurred every 2 to 4 wk or if manure were spread when rainfall or overland flow events were unlikely (i.e., during dry periods).
|
Materials and Methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
500 g each) per herd. Information on dairy system, feed composition, and fertilization or supplementation of pasture or feed with P was gathered (Table 1
). Forage samples were obtained 24 h before fecal sampling for pastured cows to ensure that sampled forage represented a true source of fecal P given that it takes up to 24 h for feed to pass through cows grazing pasture. Feed samples were obtained 48 h before fecal sampling for confined cows given that it takes up to 48 h for feed to pass through confined cows. Pastures were sampled from areas that had not been grazed for 20 d before grazing by our cattle. Fecal samples were kept below 4°C and analyzed within a week. Half of each sample was dried at 55°C and ground to <2 mm. Wet and dry composite samples combining 50 g from each of the 10 samples were also made for extraction and 31P NMR analysis.
|
Sample Analysis
The samples collected and analyzed were six feed samples (three TMR [systems 1–3], one grain [system 4], and two forages [systems 3 and 4]), 40 fresh individual and four composite fecal samples, and 40 dried individual and four composite fecal samples. The six feed and 44 dried-ground fecal samples were acid digested with a microwave-assisted digester (Walter et al., 1997); the digests were used for determination of total P, Fe, Mn, Al, Ca, and Mg by inductively coupled plasma–atomic emission spectroscopy (ICP–AES).
Feed samples were extracted in NaOH-EDTA; each of the 44 fresh and 44 dried fecal samples was extracted in water, dilute acid, and NaOH-EDTA. For water extraction, 2 g wet feces were immersed in 98 mL water; 0.3 g dried-ground feces were immersed in 100 mL water; or 1 g dried and ground feed sample were placed in 100 mL water and shaken for 1 h, centrifuged (8000 x g for 20 min), and filtered through a Whatman #42 filter paper. An aliquot of each extract was analyzed for total P, Fe, Mn, Al, and Ca via ICP–AES. For wet and dried fecal composites, bulk extracts were prepared to yield enough lyophilized material for NMR analysis. Fecal samples (20 g wet or 3 g dried) were immersed in 1 L water and shaken for 1 h. Extracts were filtered through Whatman #42 filter paper and immediately frozen and freeze-dried for 31P-NMR. All extracts were analyzed for pH and colorimetric determination of molybdate-reactive P (MRP). For the dilute acid extraction, the same parameters as water extraction were used, except water was replaced by dilute HCl (0.012M HCl). All samples were analyzed for pH; colorimetric determination of MRP; and total P, Fe, Mn, Al, and Ca via ICP–AES. For NaOH-EDTA extraction (Bowman and Moir, 1993), 2 g (dry weight equivalent) of wet or dry feces or dry feed were shaken for 5 h with 40 mL 0.25 M NaOH + 50 mM EDTA, centrifuged at >8000 x g for 20 min, and filtered (Whatman #42). A 1-mL subsample was taken and diluted with deionized water to 10 mL for colorimetric determination of MRP, and for total P, Fe, Mn, Al, and Ca by ICP–AES (feeds and fecal composites only). For feeds and wet and dry fecal composites, the remaining 39 mL were frozen and lyophilized for analysis by 31P-NMR.
Analysis by 31P NMR
The lyophilized NaOH-EDTA extracts were re-dissolved in 0.75 mL D2O, 0.3 mL 10 M NaOH, and 1.5 mL of deionized water. The lyophilized water extracts were re-dissolved in 0.75 mL D2O, 0.4 mL 10 M NaOH, and 1.5 mL of the NaOH-EDTA extracting solution and were allowed to stand for 30 min with occasional vortexing. All of the lyophilized NaOH-EDTA extracts were dissolved, but only about half of the material from the lyophilized water extracts was used, and the pH was confirmed to be >12.5. After dissolving, samples were centrifuged for 20 min at 1500 g, after which the supernatant was transferred to 10-mm NMR tubes and stored at 4°C before analysis within 12 h.
The lyophilized HCl extracts were re-dissolved in 0.8 mL D2O, 0.4 mL 10 M NaOH, and 1.6 mL of the NaOH-EDTA extracting solution and centrifuged for 20 min at 1500 g, and the supernatant was transferred to 10-mm NMR tubes. The remaining material in the centrifuge tube was washed with an additional 0.75 mL of the NaOH-EDTA extracting solution, centrifuged, combined with the prior supernatant in the NMR tube, and stored as described previously.
Solution 31P NMR spectra were obtained using a Bruker ADVANCE 500 MHz spectrometer equipped with a 10-mm broadband probe. The NMR parameters were 90° pulse, 0.68 s acquisition time, 4.32 s pulse delay, 25°C, and 2500 scans (4 h).
Phosphorus compounds were identified by their chemical shifts (ppm) relative to an external orthophosphoric acid standard, and the orthophosphate peak for each sample was standardized to 6 ppm in all spectra to simplify comparisons among samples. Peak areas were calculated by integration on spectra processed with 5 Hz line-broadening (and checked with 1 Hz line-broadening) using NUTS software (2000 edition; Acorn NMR, Livermore, CA). Peaks were grouped into compounds or compound classes if specific identifications could not be made. Inorganic P compounds included orthophosphate (6 ppm), pyrophosphate (–4.3 ppm), and polyphosphates (–3.9 ppm and a number of peaks in the region from –4.7 to –27 ppm, which may include ATP and ADP). Detected organic P compounds included phosphonates in the region from 10.4 to 27.9 ppm, orthophosphate monoesters from 6.4 to 7.7 ppm and between 2.9 and 5.7 ppm, and orthophosphate diesters from –0.4 to 2.5 ppm (Newman and Tate, 1980; Cade-Menun and Preston, 1996; Turner et al., 2003b; McDowell and Stewart 2005b). The orthophosphate monoesters were divided into four regions: myo-IHP was identified by peaks at 5.5, 4.6, 4.2, and 4.1 ppm if they could be clearly identified or by multiplying the peak at 5.5 by 6 (Turner, 2004); monoester 1 from 6.4 to 7.6 ppm, which includes some inositol P compounds (Turner and Richardson, 2004); monoester 2 between 4.0 and 5.7 ppm (excluding peaks for myo-IHP), which includes inositol phosphates and diester degradation products if present (Turner et al., 2003b); and monoester 3, with peaks from 2.9 to 3.7 ppm, which includes sugar phosphates (McDowell and Stewart 2005b), scyllo-inositol phosphate if present (Turner and Richardson, 2004), and other unidentified orthophosphate monoesters. The orthophosphate diesters were also divided: phospholipids and teichoic acid from 2.5 to –0.4 ppm (Makarov et al., 2005); DNA (deoxyribonucleic acid) at –0.7 ppm; and other diesters, which included unidentified orthophosphate diester peaks from –0.8 to –3.0 ppm.
Statistical Analysis
Differences in fecal P pools between herds were evaluated by ANOVA in GenStat 6.0 (GenStat Committee, 2002). Means were distinguished by LSD.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
|
|
As with total P, more P was extracted (MRP and total) from feces sampled from the TC herd fed excessive P (system 2) and the predominantly grazing herd receiving supplemental grain (system 4) compared with the TC herd with adequate P in feed (system 1) and the hybrid confinement/grazing herd (system 3) (Table 3). More P was extracted by NaOH-EDTA from feces than from feed. The total P extracted by any of these solutions ranged from 15% (H2O) to 96% (NaOH-EDTA) of total P in feces. The percentage of MUP in extracts ranged from 6 to 66% and was generally greatest in water extracts (23–66%) and least in dilute HCl extracts (6–31%).
Analysis by 31P NMR
Analysis of extracts at pH >12.5 gave a consistent pH in which to make peak assignments and compare herds from each of the four systems (Cade-Menun and Preston, 1996). However, the alkaline conditions made P speciation in the dilute acid extracts problematic. Orthophosphate should be the dominant form in the dilute acid extracts, judging from the colorimetric results (Table 3), but it accounted for only <5% of the spectra (Table 4). We believe that much of the orthophosphate present in HCl extracts precipitated with Ca on pH adjustment for the 31P NMR analysis. Analysis of Ca concentrations before and after pH adjustment indicated that, in all HCl samples, only about 50% (±6% SEM) of the original Ca extracted remained soluble in the NMR tube, and the remainder had precipitated (for brevity these data are not given in tabular form). This made the relative proportion of organic P forms greater than their true shares in the extracts. Therefore, future attempts of P speciation in acid extracts must first address the issue of precipitation before pH adjustment.
A total of 12 classes of P forms could be distinguished (see Fig. 1 and 2 ). The percentage each class occupied in each spectrum is given in Table 4. Due to cost and analysis time, 31P NMR was done only on individual feed samples and composite fecal samples of each treatment. The concentration of orthophosphate and total P in these bulked fecal samples did not vary more than 10% from the means of replicate samples (Table 3), and hence the bulk samples are considered representative of each treatment.
|
|
For fecal samples, 31P-NMR results varied with the type of extractant and sample handling (Table 4). In NaOH-EDTA extracts, orthophosphate dominated in spectra of all samples (53–80%). This was followed by total orthophosphate monoesters (13–33%), including myo-IHP (generally <10%). Orthophosphate diesters were dominated by phospholipids and teichoic acid (1.2–6.8% of total spectra). Unlike feed samples, phosphonates were detected in all fecal samples despite their relatively small proportion (0.4–2.1%). Condensed inorganic P (pyrophosphate and polyphosphate) ranged from 1.9 to 8.6%. Compared with samples extracted by NaOH-EDTA while wet, a higher percentage of each spectrum was occupied by orthophosphate in dried samples (Table 4). There was a small but consistent increase in myo-IHP associated with the dry samples, accompanied by decreases in other monoesters. The percentage of polyphosphate also decreased in dry samples, whereas changes in orthophosphate diesters were small and inconsistent.
In water extracts of fecal samples, orthophosphate was the dominant P form for systems 1 and 2 (about 50–68%) but not for systems 3 and 4, depending on sample handling (Table 4). Total orthophosphate monoesters comprised 13 to 42% of the spectra, with myo-IHP generally <10%. Orthophosphate diesters (10–31%, mostly as phospholipids and teichoic acid) were several times greater than in NaOH-EDTA extracts. This may suggest possible hydrolysis with the NaOH-EDTA procedure (see Discussion), or it may indicate that orthophosphate diesters are extracted more efficiently in water than other P forms, given the lower recoveries of total P in water than with other extractants (Table 3). Unlike the NaOH-EDTA results, the effects of drying on P forms in water extracts were inconsistent and seemed to be system dependent. For example, system 4 had an increased percentage of orthophosphate (34.6–51.0%), which was accompanied by decreases in condensed orthophosphates and orthophosphate monoesters, including myo-IHP.
In fecal samples extracted by HCl, the spectra were dominated by orthophosphate monoesters followed by orthophosphate diesters (system 1 wet and dry, system 3 wet, system 4 dry, or by orthophosphate diesters—particularly lipids and teichoic acid) followed by orthophosphate monoesters (system 2 wet and dry, system 3 wet, system 4 dry). The greatest relative percentages of polyphosphates and phosphonates were observed in HCl extracts. This is most likely due to the removal of orthophosphate; because 31P NMR determines relative percentages, the large orthophosphate peak in the spectra from other extractants reduces the phosphonate and polyphosphate peaks to very small peaks.
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
The dilute acid extraction was designed to assess the P intake and utilization status of the cow through fecal P testing (Dou et al., 2007). This is based on a recent study by Chapuis-Lardy et al. (2004), who reported that the pH and Ca content of dairy feces and sample handling (wet- vs. dry-based extraction) had significant impacts on water-soluble Pi measurement, although dietary P remained the main factor affecting fecal P concentrations. Dou et al. (2007) replaced deionized water with dilute HCl as the extractant and found that extractable P was independent of pH and Ca changes or sample handling. Furthermore, dilute HCl-extractable Pi was highly correlated to total P in feces (r = 0.92) and dietary P concentrations (r = 0.83, n = 25). Nevertheless, measures must be taken to remove Ca from the dilute HCl extract and to prevent P from precipitation if complete 31P NMR spectra are to be obtained. Examination of organic and inorganic P species among extracts by 31P-NMR showed distinct differences; the greatest was the near absence of orthophosphate (<5% of total P in spectra; Table 4, Fig. 1 and 2) relative to that detectable by colorimetry in HCl extracts (69–94% of total). Some studies have shown that colorimetry tends to overestimate orthophosphate (e.g., McDowell and Sharpley, 2001). However, this has been attributed to liberation of P from colloids and the acid-mediated hydrolysis of labile organic P species, neither of which is likely to account for the up to 90% difference to that we detected by 31P-NMR. In contrast, the extraction of Ca by dilute HCl is far more efficient than water or NaOH-EDTA. Hence, when pH was adjusted to >12.5, it is likely that Ca-P precipitates formed, and therefore P would have been undetectable by solution 31P-NMR. Quantification of species in dilute HCl extracts may be better achieved with other colorimetric methods that include enzymes specific to certain P compounds and less influenced by acid-mediated hydrolysis (e.g., He and Honeycutt, 2005).
If most of the orthophosphate was lost via precipitation, then one would expect the relative proportion of other species, such as phosphonates and monoesters, to increase in HCl extracts. In addition, these species tend to be resistant to changes in pH, whereas other species are susceptible to hydrolysis (McDowell and Stewart 2005b). For instance, Leinweber et al. (1997) found more orthophosphate and orthophosphate monoesters and less orthophosphate diesters in 0.5M NaOH extracts of soils and some manures compared with 0.1M NaOH extracts. Similarly, McDowell and Stewart (2005b) found that adjusting the pH of overland flow from soils and water extracts of dung to pH >12.5 caused increases in orthophosphate in the order of 50–100% at the expense of orthophosphate monoesters and diesters. Analysis of dilute HCl extracts revealed the presence and form of phosphonates and orthophosphate monoesters, consistent with the study of Toor et al. (2005b) who found that HCl extracted myo-IHP in addition to, or included in, Ca-P. However, due to precipitation, the utility of dilute HCl extracts to give an overall picture of P in manure by NMR analysis was limited. Solutions to this problem could include treatment to remove cations before adjusting pH and adjusting the pH <12.5. McDowell and Stewart (2005b) presented equations to show that separation of peaks is possible from pH >7, and Ca-P precipitates only start to form above pH 8.5 (Lindsay, 1959).
Since the development of an indicator of potential P loss in overland flow from manure by Kleinman et al. (2002), Kleinman et al. (2005) found the mean percent of total P extractable by water to be 60 ± 15% SD (mean total P concentration, 6.9 ± 3.2 g kg–1) in 68 dairy manures submitted to the Pennsylvania State University's Agricultural Analytical Service. In comparison, the percentage of total P extracted by water in our samples showed much greater variation (15–70% of total P). This was mirrored by the percentage of organic P extracted by water. The difference between studies likely reflects the role of storage and handling of manures in the Kleinman et al. (2005) study (as opposed to feces in the current study).
Water extracts yielded a different range and proportion of P species than did HCl and NaOH-EDTA extracts. Few other studies of water extracts by 31P-NMR exist. Among the first, Nanny and Minear (1993) isolated various organic P compounds of low-, medium-, and high-molecular-weight from lake waters and found they varied with season and depth. For soil-solution extracts (0.01 M CaCl2), McDowell et al. (1998) and McDowell (2003) discovered that, although the concentration of orthophosphate monoesters increased with increasing soil P fertilization, the proportion of orthophosphate diesters decreased. Toor et al. (2003) similarly found that the soil-solution was enriched with orthophosphate monoesters in a low-P soil, but their analysis included extraction of the freeze-dried soil solution residue with NaOH-EDTA and analysis of this extract after further freeze-drying and reconstitution. This causes uncertainty due to degradation of sensitive P species.
All studies have found a greater range of compounds, and especially orthophosphate diesters, in water extracts compared with NaOH-EDTA extracts. Of particular note in our data was the greater proportion of lipids and teichoic acids in water extracts than in NaOH-EDTA extracts. This may be due to differing solubility in the extracts or alkaline-hydrolysis of these species (Turner et al., 2003b). Irrespective of this, the data indicated that if used as an indicator for the potential P lost in overland flow then the bioavailability is vastly different from the overall proportions of P species present in feces and detected in NaOH-EDTA extracts. For instance, orthophosphate diesters and phosphonates make up a lower proportion of P in the spectra in NaOH-EDTA extracts, but orthophosphate and polyphosphate make up a larger portion of P compared with water extracts. Some of this may be derived from degradation products, but it also suggests that the bioavailability of water-extractable P determined as MRP is underestimated because orthophosphate diesters are not detectable by colorimetry. Several studies have shown through the use of enzymes that these orthophosphate diesters are readily bioavailable to algae via phosphodiesterase (e.g., Fischer and Amrhein, 1974), and NMR studies have suggested that they comprise much of the P that is cycled within some soils (Turner and Newman, 2005).
Variation with Moisture Status
With time, the concentration and form of P in dung changes due to a combination of physical and chemical breakdown, altering the potential for P loss to overland and subsurface flow and thus soil P concentrations (McDowell, 2006a). For example, moist dung breaks down faster than dry dung (MacDiarmid and Watkin, 1972). Rowarth et al. (1985) showed that sheep dung on flat land in the North Island of New Zealand had decomposed within 28 d in winter but lasted for over 75 d in summer. They attributed the rapid breakdown in winter to a combination of higher sustained soil moisture and greater number of days with rain than in summer, thus enhancing earthworm activity and preventing dung from drying and forming a crust. Upon oven drying, Ajiboye et al. (2004) found that labile P, measured as water-extractable P, increased at the expense of more recalcitrant fractions and especially organic P. However, McDowell and Stewart (2005a) found that air-drying cattle dung decreased water-extractable P. This was attributed to a structural rearrangement of humic macromolecules and increased exposure of hydrophobic surfaces to water during extraction.
Our data showed drastic decreases in water-extractable MRP on drying, whereas differences in total P were less dramatic (Table 3). In fact, NaOH-EDTA–extractable MRP was greater in dried fecal samples than in fresh samples. No significant difference was noted with total P; nor were differences noted for HCl extracts. It is likely that the chelating action of EDTA in an alkaline solution negated the effect of humic material on P solubility by preventing transition metal linkages. However, data show that although the ratio of inorganic to organic P may change with drying, this may not translate to the flux of P being lost to the soil or to flow, as indicated by water-extractable P. For instance, when cattle feces are stored in slurry pits, orthophosphate diesters extractable in NaOH-EDTA, such as phospholipids and DNA, degrade to orthophosphate and orthophosphate monoesters (Hinedi et al., 1988; Toor et al., 2005a). Similarly, McDowell and Stewart (2005a) found that the proportion of orthophosphate diesters extractable by NaOH-EDTA decreased with drying. Although our data show a decrease in the proportion some orthophosphate diesters such as DNA extractable by NaOH-EDTA with drying, others, such as phospholipids and teichoic acids in systems 3 and 4, increased; this could be linked to changing hydrophobicity with drying (Table 4). However, a clear increase in orthophosphate with drying was detectable in all of the NaOH-EDTA extracts, indicating a change in solubility of orthophosphate or a change in P species.
Recently, Bol et al. (2006) found that the total orthophosphate diester to total orthophosphate monoester ratio could be used to trace the incorporation of dung organic P into grazed topsoil (0–5 cm depth). The corresponding ratio in NaOH-EDTA–extractable P in system 4 (feces also derived from predominantly grazed pastures) was 3.5 in fresh feces and had decreased by 40% in dry feces. Bol et al. (2006) found that for dung exposed to the elements, the ratio in dung had decreased by 66%. Hence, although much of the change likely results from the drying process, not all of the change can be explained by this mechanism. Specifically, the orthophosphate diester/monoester ratio of water-extractable P in system 4 decreased by about 58%, indicating a much greater change in mobile organic P. This suggests that although the overall changes in organic P in dung can be traced with NaOH-EDTA, the dynamics would be better estimated by water-extractable P.
Comparison among Systems
The addition of P to diets acts as an insurance policy to ensure that the animal is in optimal condition to metabolize energy efficiently. For lactating dairy cattle in New Zealand, the recommended diet P concentration is 3.2 g kg–1 (Familton, 2003); any P supplied in excess of this is mostly excreted in feces. Most dairy pasture forages have P concentrations in excess of 3.5 g kg–1, such that P nutrition is not considered a significant concern in grazed dairy systems (e.g., New Zealand–style system 4). This is because to increase production, more pasture must be grown, which requires more P fertilizer to be added, removing any possibility of P limitation (McLenaghen et al., 2003). In the USA, the commonly advised P content for feed is 3.7 g kg–1. Although most studies have shown no benefit in added milk production beyond this concentration (e.g., Lopez et al., 2004), some have also indicated that no penalty in production occurs at concentrations as low as 3.3 g kg–1 (Wu et al., 2000; 2003). This leaves considerable scope for decreasing P in feces at the farm scale.
Phosphorus in feed varied from 2.8 g kg–1 to 4.5 g kg–1 (Table 2). The concentration of total P in feces reflected differences in feed, with system 2 (TMR + P) greater than system 1 (TMR only) and system 4 (TMR:forage 0.4 and P in feed 4.5 g kg–1) much greater than system 3 (TMR:forage 1.5 and P in feed 2.8 g kg–1), respectively. However, compared with the herd fed system 4 (New Zealand-style grazing system), the system 2 herd (TMR + P) was fed less P but had a greater P concentration in feces (8.55 g kg–1 compared with 7.54 g kg–1) (Table 2). The difference between feces of the herds from systems 1 and 2 can be attributed to the supplemental feeding of mineral P. Table 4 shows that orthophosphate in feed was similar, but more was present in the feces of system 2 than system 4. The difference probably lies with the mineralization of myo-IHP. Decreases in myo-IHP concentration between feed and feces of dairy cattle led Toor et al. (2005a) to conclude that mineralization occurred in the gut. Our data support this hypothesis in that the proportion of myo-IHP in feces was always less than in feed. However, pasture contains a much lower proportion of myo-IHP than TMR (Peperzak et al., 1959; McDowell and Stewart, 2005a), resulting in a greater proportion of orthophosphate diesters and monoesters in water and NaOH-EDTA extracts of systems 3 and 4 compared with systems 1 and 2 (Table 4). These contain fewer phosphate moieties than myo-IHP, which can have up to six, meaning that their mineralization would release less orthophosphate than mineralization of myo-IHP. However, mineralization of organic P varies depending on the time of year or P supply. For instance, Bromfield and Jones (1970) found that organic P in spring pasture was better mineralized than in summer when P concentration and digestibility decreased. Our pastures and feces were sampled in spring; it would be interesting to see if this difference continued during summer and autumn.
The amount of myo-IHP mineralization varied across herds. For instance, in the P-rich feed of the system 4 herd, the percentage of total P as myo-IHP in the NaOH-EDTA extracts decreased 4.5% from feed (13.3%, based on intake) to feces (8.8%). In the relatively P-poor feed of the system 3 herd, the intake was 20.2%, whereas 8.8% was excreted in feces, a decrease of 11.4%. A similar pattern occurred with the P-rich feed of the TC herd in system 2 (a decrease of 10.6%) and the P-poor feed of the TC herd in system 1 (a decrease of 35.5%). Scott et al. (2007) also found that greater mineralization of myo-IHP occurred when comparing sheep dung and pastures in a low-P fertility system (soil Olsen P concentration = 7 mg kg–1) with dung and pastures in a P-rich system (soil Olsen P concentration = 52 mg kg–1). This suggests that a greater proportion of easily digestible P in the diet inhibited myo-IHP mineralization. Further work should be done to examine this issue.
Compared with feed, phosphonates were present in equal or greater proportions in NaOH-EDTA extracts of feces (Table 4). These compounds with a direct C-P bond are thought to be synthesized as primitive energy carriers by protozoa and to a lesser extent by bacteria and fungi (Kononova and Nesmeyanova, 2002). Their presence in feces has been attributed to formation in the rumen by protozoa such as Tetrahymena sp. (Kandatsu and Horiguchi, 1962).
The presence of DNA and phospholipids in feces has been attributed to their formation in the gut (Toor et al., 2005a). However, feces from the herds in systems 1 and 2 showed only minor increases compared with feces from the herds in systems 3 and 4. A profound increase was observed in phospholipids, which increased by an order of magnitude in the feces from the herds from systems 3 and 4 (Table 4). Although some of this could be attributed to the sloughing of cells and bacteria from the inside of the gut, most probably originate from pasture plant cells: Forages are known to be a rich source of phospholipids (Bieleski, 1973). The contribution from plant cells is also evident in the water extracts of feces from the herds in systems 3 and 4, which have about twice the proportion of phospholipids as that observed in feces from the herds in systems 1 and 2 and about 5 times that present in NaOH-EDTA extracts.
Implications for Management
The adoption of grazing systems in the Northeast USA is advocated largely on the basis that income per unit cost of feed can be more than confined animal operations (e.g., White et al., 2002). Coupled to this is the added benefit of less disease and lower bacterial counts in the milk of grazing animals (Goldberg et al., 1992). In the current study, P in feed of grazing herds results in no more P in feces than is found in the feces of herds fed TMR with an equivalent P intake. However, the application of P fertilizers and fresh growth in spring means that the P concentrations and crude protein of pastoral grasses and clover tend to be elevated in spring compared with the rest of the year (Frame et al., 2002). Evidence from this study indicates that the risk of P loss from the feces of grazing animals is similar if not less than that of confined systems but does not take into account the distribution, handling, and volume of feces. In confined systems, feces are commonly stored until it is spread onto fields. Changes in P forms, and hence bioavailability, can occur during storage, commonly increasing with age (Toor et al., 2005a; McGrath et al., 2005). However, the main problem is having sufficient land to dispose of the manure without increasing the P concentration of the soil due to more P coming into the farm in feed than leaving it in produce (Sharpley et al., 2001). Grazing systems, by definition, require much more land on which to graze cattle, and as a consequence dung is distributed around the farm as paddocks are grazed on a rotational basis. A typical pasture-based dairy farm is stocked at between 1.5 and 3.5 cows ha–1. However, a confined system imports feed from about 3 times more land than a grazed system, and hence may not have enough land to apply the manure at a low enough rate such that P concentrations in topsoil and the potential for P loss are not exacerbated (K. Soder, personal communication). Provided good land management is practiced, such as fencing-off streams and not grazing excessively wet paddocks, the losses in erosion and runoff from a pasture-based system can be very low on a per hectare basis (<1 kg P ha–1) (McDowell, 2006b). This, and the fact that the proportion of bioavailable P forms during summer and autumn is probably going to decrease, means that the potential for P losses from dung in grazed systems is probably less than manure-P loss in confined operations.
|
Conclusions |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Z. He, C. W. Honeycutt, T. S. Griffin, B. J. Cade-Menun, P. J. Pellechia, and Z. Dou Phosphorus Forms in Conventional and Organic Dairy Manure Identified by Solution and Solid State P-31 NMR Spectroscopy J. Environ. Qual., July 23, 2009; 38(5): 1909 - 1918. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Dou, C. F. Ramberg, J. D. Toth, Y. Wang, A. N. Sharpley, S. E. Boyd, C. R. Chen, D. Williams, and Z. H. Xu Phosphorus Speciation and Sorption-Desorption Characteristics in Heavily Manured Soils Soil Sci. Soc. Am. J., January 21, 2009; 73(1): 93 - 101. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
