JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 5 July 2005
Published in J Environ Qual 34:1380-1391 (2005)
DOI: 10.2134/jeq2004.0232
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (14)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Toor, G. S.
Right arrow Articles by Sims, J. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Toor, G. S.
Right arrow Articles by Sims, J. T.
Agricola
Right arrow Articles by Toor, G. S.
Right arrow Articles by Sims, J. T.
Related Collections
Right arrow Animal Waste
Right arrow Surface Water Quality
Right arrow Best Management Practices
Right arrow Nutrient Management
Right arrow Water Pollution

TECHNICAL REPORTS

Waste Management

Establishing a Linkage between Phosphorus Forms in Dairy Diets, Feces, and Manures

Gurpal S. Toora,*, Barbara J. Cade-Menunb and J. Thomas Simsa

a Department of Plant and Soil Sciences, University of Delaware, Newark, DE 19716
b Geological and Environmental Sciences, Stanford University, Stanford, CA 94305-2115

* Corresponding author (gurpal{at}udel.edu)

Received for publication June 14, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS AND IMPLICATIONS
 REFERENCES
 
Effective manure management to efficiently utilize organic wastes without causing environmental degradation requires a clear understanding of the transformation of P forms from diet to manure. Thus, the objective of this study was to establish quantitative relationships between P forms in diets, feces, and manures collected from U.S. Northeastern and Mid-Atlantic commercial dairy farms. Total P in diets ranged from 3.6 to 5.3 g kg–1 dry matter, while the feces had higher P than diets (5.7–9.5 g kg–1) and manures had lower P (2.5–8.9 g kg–1) than feces. The farms with total dietary P of 4.8 to 5.3 g P kg–1 had twofold higher concentrations of phytic acid (1647–2300 mg P kg–1) than farms with 3.6 to 4.0 g dietary P kg–1 (844–1100 mg P kg–1). Much of the phytic acid in diets was converted to inorganic orthophosphate in the rumen as indicated by a reduction in phytic acid percentage from diets (32%) to feces (18%). The proportion of orthophosphate diesters (phospholipids, deoxyribonucleic acid [DNA]) was twice as high in feces (6.2–10%) as diets (2.4–5.3%) suggesting the excretion of microbial residues in feces. Phosphonates (aminoethyl phosphonates and phosphonolipids) were not seen in diets but were detected in feces and persisted in manures, which suggests a microbial origin. These organic compounds (phytic acid, phospholipids, DNA) were decomposed on storage of feces in slurry pits, increasing orthophosphate in manures by 9 to 12% of total P. These results suggest that reducing dietary P and typically storing feces in dairy farms will result in manure with similar chemical forms (primarily orthophosphate: 63–77%) that will be land applied. Thus, both the reduction of dietary P and storage of manure on farm are important for controlling solubility and bioavailability of P forms in soils and waters.

Abbreviations: ADF, acid detergent fiber • CP, crude protein • NDF, neutral detergent fiber • NFC, nonfibrous carbohydrates • NMR, nuclear magnetic resonance • TDN, total dissolved nutrients


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS AND IMPLICATIONS
 REFERENCES
 
HEIGHTENED CONCERNS about accelerated eutrophication of lakes, rivers, and coastal waters due to P enrichment have prompted land managers and regulatory authorities to devise strategies that balance P inputs and outputs in regions of intensive animal production (Sharpley and Tunney, 2000; Sharpley et al., 2001, 2003; Toor et al., 2004). Edwards and Withers (1998) reported that the majority of farms in the UK have an annual agricultural P surplus, with values of >20 kg P ha–1 in intensive-livestock areas and lower values (<10 kg P ha–1) in arable farms. Recently, Toor et al. (2004) observed annual P accumulation of 16 to 47 kg P ha–1 in a New Zealand grassland soil that received P inputs in the form of superphosphate and dairy effluent. It is well known that many dairy farms in different parts of the world have P surpluses, with imported feedstuffs the principal P input to these farms. Excess P in animal diets has been documented to increase P excretion in manures, eventually leading to accumulation of P in soils and increasing the risk of P loss to water (De Clerq et al., 2001; Powell et al., 2002; Toor et al., 2005a).

The National Research Council (2001) recommends 0.32 to 0.38% P in the diets of lactating dairy cows. However, overfeeding P is a common practice on many dairy farms. A recent survey of 612 commercial Mid-Atlantic U.S. dairy farms indicated that the mean concentration of P in diets was 0.44%, which is 34% above NRC recommendations (Dou et al., 2003). The elevated concentrations of P in dairy diets are due to (i) concerns of farmers, nutritionists, and veterinarians that grain and forage diets lack sufficient P to meet animal health and reproduction requirements, and (ii) uncertainties in the absorption of P by dairy cows, which can vary from 64 to 92% (Coates and Ternouth, 1992). Inorganic P that is soluble in water or dilute acid is considered to be available to animals. However, the biological availability of P in animal diets depends on the conversion of feed P into inorganic P or more digestible forms of organic P. The efficiency of P absorption depends on a number of factors such as animal age or body weight, physiological state (e.g., lactating vs. nonlactating), intestinal pH, Ca to P ratio of diet, dietary concentrations of Al, Ca, Fe, Mg, Mn, K, and fat, the source of P (e.g., forages, concentrates, mineral supplements, and salivary P), and the amount of dry matter intake (Irving, 1964; Morse et al., 1992a; Peeler, 1972; Soares, 1995).

Over the last two decades, a variety of methods have been employed to investigate P forms in manures; most have involved some form of chemical extraction with dilute acids and/or bases (Dou et al., 2000, 2002; Leinweber et al., 1997; Sharpley and Moyer, 2000). Recently, there has been increased interest in using advanced analytical methods such as 31P nuclear magnetic resonance (NMR) (Hunger et al., 2004; Toor et al., 2005a; Turner, 2004), X-ray absorption near edge structure spectroscopy (XANES) (Peak et al., 2002; Toor et al., 2005b), and enzymatic hydrolysis techniques (He and Honeycutt, 2001) to characterize P forms in manures. However, limited research is available to compare different analytical methods, or to determine the transformation of P along the continuum from animal to soil by examining the P forms in diets, feces, and manures. If we are to understand how dairy diet modification affects the potential for P loss to water, it is important to characterize P in diets, feces, and manures (mixtures of feces, bedding materials, soil, water, etc.). Consequently, our objectives were to determine, by use of chemical analyses and 31P NMR, if different dairy diets affected P speciation in excreted feces and whether storage and handling practices then altered P forms in manures.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS AND IMPLICATIONS
 REFERENCES
 
Sample Collection
Diet, feces, and manure samples were collected in March and April 2003 from six commercial dairy farms in the Northeast and Mid-Atlantic United States in collaboration with researchers from respective states, with one farm each in Pennsylvania and New York, and two farms each in Maryland and Virginia. These farms were part of a larger study undertaken at the University of Pennsylvania, where dietary materials and fecal samples were collected from individual cows to investigate the effect of dietary P modification on P excretion by dairy cows (Dou et al., 2003). The diets used on the six farms in the current study had a range of P concentrations (low- to medium-P-diet farms: 3.6, 3.9, and 4.0 g P kg–1; high-P-diet farms: 4.8, 5.0, and 5.3 g P kg–1 dry matter) that fell within the limits of an earlier study of Dou et al. (2003) in Northeastern and Mid-Atlantic dairy farms (Fig. 1) . The dietary P concentrations in the six farms in our study were on the higher side of the NRC recommendations of 3.2 to 3.8 g P kg–1 for lactating dairy cows (National Research Council, 2001). All diets in our study consisted of a mixture of forages (48–54%) and concentrates (44–50%), together with small amounts of salt, trace minerals, and vitamins.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between total P concentrations in diets and feces of lactating cows. Open circles represents data from Dou et al. (2003) while black data points are from the present study.

 
Diet samples were manually collected in a 20-L bucket from 10 to 15 locations for a particular dietary P group in a farm during feed delivery. Samples were then thoroughly mixed and a representative 4-L sample was taken. Feces were collected from the same dietary P group of dairy cows by holding a bucket behind each cow (20–25 cows per farm; approximately 2 kg per cow). The samples were mixed and a composite 20-L sample was taken. The manure sample was obtained from liquid slurry pits except one (Farm D) where samples were obtained from bedded pack manure. To collect representative manure samples, we threw a bucket attached to a rope at multiple locations in the slurry pits. The buckets were then removed, the manure was mixed, and a representative 20-L sample was removed for subsequent analyses. This method of manure collection is similar to the method used in the United States to obtain manure samples from liquid manure pits, where typically a plastic tube (20–25 cm wide) is attached to a pole and this sampler is dipped in the slurry pits on four to six locations in a farm (Lorimor et al., 2000). It is important to note that all farms had only one slurry pit where manure was stored, but on many farms, there was more than one dietary P group. Thus, the manure sample we used is not from the same dietary P group for which we collected diet and feces samples but reflects the combined sample for a particular farm. The most common bedding material on most farms was sand, although some farms used straw (Farm D), and shredded paper and paper sludge (Farm E). Diet, feces, and manure samples were kept frozen until use, at which time subsamples were dried at 55°C and ground to pass through a 2-mm sieve before analysis.

Wet Chemical Analysis
Diet samples were analyzed by a commercial laboratory (Cumberland Valley Analytical Services, Maugansville, MD) for fibers (neutral detergent fiber [NDF] and acid detergent fiber [ADF]), nonfibrous carbohydrates (NFC), total dissolved nutrients (TDN), and crude protein (CP). Total P, Ca, Mg, K, Fe, and Al concentrations in all samples were determined, in triplicate, by microwave digestion with concentrated HNO3 followed by inductively coupled plasma–optical emission spectroscopy (ICP–OES) (USEPA, 1986). Total C and N were analyzed, in duplicate, by dry combustion (CNS-2000; LECO Corporation, St. Joseph, MI).

Solution Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy
Two grams of each diet, feces, and manure sample were extracted with 30 mL of 0.25 M NaOH + 0.05 M Na2EDTA at room temperature for 6 h on an end-over-end shaker (Cade-Menun and Preston, 1996). After extraction, the samples were centrifuged at approximately 1500 x g for 20 min. A 1-mL aliquot was removed for ICP–OES analysis after dilution to 10 mL, and the remainder of the supernatant was frozen overnight and lyophilized for 24 to 48 h. Freeze-dried extracts were dissolved in 0.4 mL 10 M NaOH and 2.6 mL D2O and allowed to stand for 30 min with occasional vortexing. Samples were then centrifuged for 20 min at approximately 1500 x g, transferred to NMR tubes, and stored at 4°C before analysis within 24 h. Solution 31P NMR spectra were acquired at 202.45 MHz on a Varian (Palo Alto, CA) UnityINOVA 500 MHz spectrometer equipped with a 10-mm broadband probe, using a 90° pulse, 0.68-s acquisition, 4.32-s pulse delay, and 15-Hz spinning. Temperature was regulated at 20°C (Cade-Menun et al., 2002; Turner et al., 2003b). Total acquisition time per sample was 4 to 9 h (2800–6400 scans), with diet samples needing the shortest acquisition time, and manure samples the longest, to obtain the same quality of spectra. Compounds were identified by their chemical shifts (ppm) relative to an external orthophosphate standard. After standardizing the orthophosphate peak in all samples to 6 ppm, peak assignments were based on Tebby and Glonek (1991), Cade-Menun and Preston (1996), and Turner et al. (2003b). Line broadenings of 1 and 7 Hz were used to separate overlapping peaks. The spectra were processed with NUTS software (Acorn NMR, 2000), using automated peak analysis tools for peak-picking and spectral integration, and the percentages were calculated based on total peak area. Peak areas in the orthophosphate monoester region, particularly for phytic acid, were confirmed with spectral deconvolution (Turner et al., 2003c).

Statistical Analyses
Genstat 4.2, 5th ed. (Lawes Agricultural Trust, 2000) was used to calculate means and standard deviations for diet, feces, and manures. The least significant difference (LSD) test using one-way analysis of variance in Genstat 4.2 was performed on the diet, feces, and manures to test for significant differences (P < 0.05) between these materials.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS AND IMPLICATIONS
 REFERENCES
 
Chemical Characteristics
Diets
Diets used on Farms B, C, and D had higher fiber concentrations (NDF: 36–40%, ADF: 25–30%) and lower concentrations of carbohydrates, dissolved nutrients, and protein (NFC: 32–37%, TDN: 63–68%, CP: 16–18%) than those on Farms A, E, and F (NDF: 29–33%, ADF: 19–21%, NFC: 37–40%, TDN: 72–74%, CP: 18–20%) (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected characteristics of dairy diets.

 
Total C in diets was 466 to 506 g kg–1 dry matter (Table 2). Total N was six- to sevenfold higher (24.8–36.6 g kg–1) than P (3.6–5.3 g kg–1), and increased with dietary P. The greatest differences were between Farm A (24.8 g N kg–1) and Farm F (36.6 g N kg–1). There was no pattern of increase of K, Ca, Mg, Fe, and Al concentrations with increasing dietary P.


View this table:
[in this window]
[in a new window]
  		Table 2. Total concentrations of selected elements in diet, manure, and feces.

 
Feces
Concentrations of C and N in feces ranged from 453 to 489, and 25.2 to 34.2 g kg–1, respectively (Table 2). These values were similar to the C and N concentrations of diets. Feces P concentrations ranged from 5.7 to 9.5 g kg–1. Concentrations of P in feces were positively, but weakly, correlated (r2 = 0.47) with dietary P, following the same trend observed in a major study conducted in the Mid-Atlantic region by Dou et al. (2003) (Fig. 1). Overall, the mean concentrations of P were significantly higher in feces (7.8 g kg–1) than diets (4.4 g kg–1). Similarly, Ca and Mg concentrations were significantly higher in feces than diets; however, C, N, K, Fe, and Al were not significantly different between diets and feces.

Manures
Total C in manures varied from 220 to 484 g kg–1 (Table 2). Total N concentrations ranged between 12.9 and 30.8 g kg–1. Unlike the feces samples, there was no pattern of increase in N in manures with an increase in dietary P. The effects of dietary P on the P content in manures were highly variable. For example, Farm B had 3.9 g P kg–1 in diet and 7.1 g kg–1 in feces, but the manure was low in P (2.5 g kg–1). On the other hand, in Farm C, dietary P was 4 g kg–1, but feces and manures had 9.5 and 8.9 g P kg–1, respectively. Similarly, concentrations of Ca varied from 6.3 (Farm B) to 46.8 g kg–1 (Farm E), while Mg ranged between 2.1 and 10.8 g kg–1.

The mean concentrations of P in manures (5.8 g kg–1) were significantly lower than the feces (7.8 g kg–1). On the other hand, mean concentrations of K were significantly higher in manures (29.3 g kg–1) than feces (7.2 g kg–1). Higher concentrations of Fe and Al were observed in some of the manures (0.7–5.1 g kg–1) relative to feces (<2.0 g kg–1) and concentrations of C and N were significantly lower in manures than feces.

NaOH-EDTA Extraction
Extraction of diets, feces, and manures with NaOH-EDTA recovered 84 to 109% of P (Table 3). The near-complete recoveries of P confirm that a single extraction of NaOH-EDTA is an effective method to extract all P from these materials. Similar P recovery rates have been reported for broiler, swine, and beef cattle manure (Turner, 2004), and for forest floor samples (Cade-Menun et al., 2002; Cade-Menun and Preston, 1996), but recoveries from soil are often lower (Turner et al., 2003a).


View this table:
[in this window]
[in a new window]
  		Table 3. Recovery of P, Ca, Fe, and Al with NaOH-EDTA extraction of diet, feces, and manures.

 
Recoveries of Ca were >91% for diets, >96% for feces, and complete recoveries for manures except for Farm E (65%) (Table 3). The recoveries of Fe and Al were higher from diet samples (Fe: 5–39%, Al: 8–59%) than from feces or manures (<5% except for Farm B [<10%]). The high recoveries of P and Ca and low recoveries of Fe and Al suggest that P in these samples is associated more with Ca than with Fe and Al.

Designation of Phosphorus Shifts in NaOH-EDTA Extracts by Phosphorus-31 Nuclear Magnetic Resonance
The original peak shifts for orthophosphate in each sample ranged from 5.47 to 5.71 ppm, with no statistical differences among sample types. These differences reflect subtle variations in solution chemistry and pH. However, to simplify comparisons among samples in this study and to compare our results with those from other studies, processing software was used to set the orthophosphate peak in each sample to 6.0 ppm. This in turn adjusts the chemical shifts of the other peaks, but does not change their relative positions or peak areas.

Typical spectra for diet, feces, and manure for different farms are shown in Fig. 2 . The inorganic P forms detected in these samples included orthophosphate (6.0 ppm) and pyrophosphate (–4.3 ppm). There was an unidentified peak slightly upfield from pyrophosphate (–4.6 ppm). Peaks in this region include the terminal P groups from ATP and long-chain polyphosphates. However, no other peaks for ATP or long-chain polyphosphates were detected in our samples, so the peak at –4.6 ppm was designated as other inorganic P.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 2. Example 31P nuclear magnetic resonance (NMR) spectra of diets, feces, and manures from (a) low-P-diet farm and (b) high-P-diet farm. The main spectra are plotted to show orthophosphate at the same height, to give relative peak heights. Inset peaks show expanded regions of the main spectra.

 
Within the organic P forms, peaks for two phosphonates were detected: aminoethyl phosphonates at 20.1 ppm and phosphonolipids at 18.3 ppm (Cade-Menun, 2004).

In the orthophosphate monoesters region, three of the four distinctive peaks were observed for phytic acid (myo-inositol hexakisphosphate) at 5.5, 4.4, and 4.1 ppm in most samples (Fig. 2) when plotted with 1-Hz line-broadening and analyzed without deconvolution. The fourth peak for phytic acid occurred as a shoulder upfield on the peak at 4.1 ppm and was included in the peak area under the 4.1-ppm peak. These overlapping peaks could be separated with spectral deconvolution (Fig. 3) . The areas under these peaks were used to determine the phytic acid concentration.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. Details showing the peaks for phytic acid in diet, feces, and manure samples, using samples collected from Farm C as an example. A spectrum for pure phytic acid, also extracted with NaOH-EDTA in the same manner as the farm samples, is shown on the bottom, with the P1, P2, P3, P4, and P6 peaks labeled. The areas under the same four peaks in the farm samples were determined both by manual plotting after integration and by spectral deconvolution. For the diet, feces, and manure samples, the solid line indicates the original spectrum, while the dotted lines show the phytic acid peaks as determined by spectral deconvolution. Other peaks determined by spectral deconvolution are not shown. Spectra are plotted with 1-Hz line-broadening.

 
Peaks for other orthophosphate monoesters were seen at 5.2, 5.0, 4.8, 4.7, 4.2, 3.6, 3.6, 3.3, and 2.9 ppm. These compounds may include other inositol phosphates, mononucleotide, and sugar phosphates, and may also include degradation products produced during extraction (Turner et al., 2003b), which were not specifically identified. Orthophosphate diester species were separated into phospholipids (1.7, 1.3, 1.1, 0.0, 0.5, 0.1, and –0.2 ppm), deoxyribonucleic acid (DNA) at –0.7 ppm, and other orthophosphate diesters (–1.0, –1.5 ppm).

Forms of Phosphorus in Diets, Feces, and Manures
Analysis of diets, feces, and manures by 31P NMR spectroscopy showed that the majority of P in diet, feces, and manures occurred as inorganic orthophosphate and phytic acid (Tables 4 and 5). Orthophosphate diesters (phospholipids, DNA) and phosphonates were minor components in these samples. Pyrophosphate and other inorganic P forms were less than 2.1% of extracted P for all samples.


View this table:
[in this window]
[in a new window]
  		Table 4. Functional classes of P determined in NaOH-EDTA extracts of diets, feces, and manures by solution 31P nuclear magnetic resonance (NMR) spectroscopy.

 

View this table:
[in this window]
[in a new window]
  		Table 5. Functional classes of P determined in NaOH-EDTA extracts of diets, feces, and manures by solution 31P nuclear magnetic resonance (NMR) spectroscopy.

 
Diets
The diet with the lowest concentration of P (Farm A: 3.6 g kg–1) had the lowest inorganic orthophosphate (1577 mg kg–1), while the diet with the highest P (Farm F: 5.3 g kg–1) had the highest inorganic orthophosphate (3089 mg kg–1) (Table 4). Intermediate P diets (3.9–4.8 g P kg–1; Farms B, C, D) had very similar and intermediate inorganic orthophosphate concentrations (2429–2746 mg kg–1). The diet with 5 g P kg–1 (Farm E) had the lowest percentage of inorganic orthophosphate (35% of TP) but highest percentage of phytic acid (47% of TP; Table 5). It should be noted that Farm E diet also had the lowest concentrations of fibers, and the highest concentrations of carbohydrates and digestible nutrients (Table 1). For all farms, dietary total P and phytic acid were significantly correlated (r2 = 0.63, significant at the 0.01 probability level); however, the relationship between dietary total P and inorganic orthophosphate was not significant (r2 = 0.11; Fig. 4a, 4b) .



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between diet, feces, and manure total P with orthophosphate and phytic acid.

 
Concentrations of phospholipids were two- to threefold higher than DNA (Table 4). The farms (E and F) having the highest concentrations of P in diets had higher phospholipids (145–173 g kg–1) and DNA (48–89 g kg–1) than other farms (phospholipids: <93, DNA: <37 g kg–1). Pyrophosphates ranged between 33 and 116 mg kg–1. No phosphonates were detected in any diet.

Feces
The P concentrations in diets were related to P forms in feces for all but Farm C: as dietary P increased from 3.6 g kg–1 (Farm A) to 5.0 g kg–1 (Farm F), inorganic orthophosphate in feces increased from 2898 mg kg–1 (51% of TP) to 5749 mg kg–1 (70% of TP; Tables 4 and 5). Farm C appeared to be an exception; where dietary P was intermediate (4.0 g kg–1), inorganic orthophosphate in feces was highest (6824 mg kg–1; 74% of TP), and phytic acid was lowest (1056 mg kg–1; 11% of TP). In contrast to no relationships between fecal total P and inorganic orthophosphate, there was a significant positive correlation between fecal P and inorganic orthophosphate content of the feces (r2 = 0.96, significant at the 0.001 probability level). However, the relationship between feces P and feces phytic acid was weak (r2 = –0.11 [not significant]; Fig. 4c, 4d).

Consistent with trends observed for diets, concentrations and proportions of phospholipids were higher (205–572 mg kg–1; 3.6–7.5%) than DNA (<241 mg kg–1; <2.6%) in feces. Phosphonates, not seen in diets, were detected in feces with concentrations ranging from 46 to 105 mg kg–1 (0.8–1.6%). Concentrations of pyrophosphates were 51 to 157 mg kg–1.

Manures
Concentrations of inorganic orthophosphate were highly variable in manures, ranging from 1916 to 6771 mg kg–1. The relationship between manure total P and inorganic orthophosphate was significantly linear (r2 = 0.92, significant at the 0.001 probability level) but the relationship between manure total P and phytic acid was not as strong (r2 = 0.31; Fig. 4e, 4f). Inorganic orthophosphate concentration in manure was 3299 mg kg–1 on Farm A and increased to 4453 mg kg–1 on Farm F, with an increase in dietary P from 3.6 to 5.3 g kg–1. Total P was 3.9 g kg–1 in Farm B diet, and manure from this farm had an inorganic orthophosphate concentration of 1916 mg kg–1. Farm C manure paralleled trends with feces, having the highest content of inorganic orthophosphate despite an intermediate P concentration in the diet. Despite these variations in absolute concentrations in Farm B and C manures, the relative proportions of inorganic orthophosphate as a percentage of TP were similar across all farms, ranging from 73 to 77% for all but Farm A with 63% (Table 5). It is important to note that despite the lowest percentages of inorganic orthophosphate in the diet from Farm E (35% of P), the proportion of inorganic orthophosphate in manure on this farm (74%) was similar to the other manures.

There was no relationship between dietary P and orthophosphate diesters (phospholipids, DNA) in manures, and the concentrations of diesters were significantly lower in manures than feces. Phosphonates, which were first detected in feces, were also present in manures with concentrations ranging between 21 and 84 mg kg–1. Pyrophosphates were present in concentrations ranging from 41 to 140 mg kg–1.

Overall, the main trends observed were an increase in the relative proportion of inorganic orthophosphate from diet to feces to manures, accompanied by a decrease in the percentage of phytic acid. For example, inorganic orthophosphate was 55% in diets and increased to 62% in feces and 73% in manures. Conversely, the proportion of phytic acid significantly decreased from 32% in diets to 18% in feces and 9% in manures for all farms (Table 5). The percentage of P as phospholipids was significantly lower in diets (2.3%) than feces (5.6%) and manures (3.7%). The percentage of phosphonates also increased, from 0% in diets to about 1% in feces and manures.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS AND IMPLICATIONS
 REFERENCES
 
Phosphorus Concentrations in Diet and Feces
The overall trend, and variability about the trendline, that we found for the relationship between dietary P and feces P was similar to that observed in a large regional study of 75 commercial dairy farms by Dou et al. (2003) (Fig. 1). For four of the six farms in our study (A, B, E, F), P concentrations were numerically higher in feces (8.4–9.3 g kg–1) produced from high-P diets (5.0–5.3 g kg–1) compared with feces (5.7–7.1 g kg–1) produced using low-P diets (3.6–3.9 g kg–1). A similar trend was found by Toor et al. (2005c) who sampled 40 dairy farms in the Northeast and Mid-Atlantic region and reported that P concentrations were 40% higher in feces produced from 21 farms using high-P diets (range = 4.6–5.8, mean = 5.1 g kg–1 dietary P) compared with 19 farms using low-P diets (range = 2.9–3.9, mean = 3.6 g kg–1 dietary P). However, two farms in our study exhibited anomalous behavior: low-P-diet Farm C (4.0 g kg–1) had a high value for feces P (9.5 g kg–1), whereas high-P-diet Farm D (4.8 g kg–1) had a low feces P value (6.9 g kg–1). Notwithstanding these differences for two farms, past research has consistently shown that excess P in diets usually results in higher P excretion in feces, although the linearity of relationship shown in Fig. 1 can be influenced by a variety of factors affecting cow performance and fecal P excretion (Chapuis-Lardy et al., 2004).

Determination of Phytic Acid Concentration with Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy
Phytic acid has been quantified using 31P NMR in food (O'Neill et al., 1980), sewage sludge (Hinedi et al., 1989), animal feed (Kemme et al., 1999), manure (Turner, 2004), and soils (Turner et al., 2003c), but this is the first study to use 31P NMR to quantify phytic acid and other P forms along a continuum from diet to feces to manure. The peak shifts for phytic acid in our study were upfield of those reported in studies using the same NaOH-EDTA extractant (Turner, 2004; Turner et al., 2003c), most likely due to differences in the chemical nature of the samples. For the majority of our samples, the phytic acid peaks (P2, P1/P3, and P4/P6 peaks; Fig. 3) could clearly be identified, with the P5 peak occurring as a shoulder of the P1/P3 peak and for some feces samples the P2 peak occurred as a shoulder of the orthophosphate peak. These could be separately identified using spectral deconvolution (Turner et al., 2003c). There is some variation among samples in terms of peak shape (Fig. 3), most likely due to variations in Fe concentration (Table 3). In the diet sample, peak areas are clearly in the 1:2:2:1 proportion expected for P2 to P4P6 to P1P3 to P5, as seen in the phytic acid standard. This same peak shape can also be seen for the feces sample after spectral deconvolution. However, the manure samples were slightly different: although the chemical shifts of these peaks were the same as those in the diet and feces samples, the peak areas were not in the 1:2:2:1 proportions. This makes the estimation of phytic acid concentration by 31P NMR spectroscopy less certain in the manure samples than the diet and feces samples used in this study, and may have resulted in an overestimation of phytic acid and an underestimation of other orthophosphate monoesters.

Peak areas were determined manually after automatic integration, and were confirmed with spectral deconvolution. As noted by Turner (2004), it is difficult to estimate the error in NMR spectroscopy without acquiring replicate spectra. However, the differences in peak areas calculated manually or by spectral deconvolution were less than 10% (i.e., 0.1% per 1% of calculated relative peak area), which is within the error estimated for 31P NMR in environmental samples (Kemme et al., 1999; Leinweber et al., 1997). This allows us to be confident in the trends observed in our data, because statistically significant differences were greater than a 10% margin of error.

Relationship between Phosphorus Forms in Diets and Feces
The results of this study show that dairy diets with higher concentrations of P do not necessarily contain higher inorganic orthophosphate concentrations than low-P diets. In fact, in the samples used in this study, the higher P in the high-P farm diets (D, E, F: >4.8 g kg–1) was mainly due to phytic acid, which had twice the concentration as farms with low-P diets (A, B, C: <4.0 g kg–1). This indicates that rations on high-P farms, in the present study, contained organic feed ingredients that were rich in phytic acid, although no detailed information on individual feed components was obtained for these farms. Dairy cows hydrolyze dietary phytic acid during digestion due to the presence of microbial phytase in the rumen. This resulted in a linear relationship between feces P and inorganic orthophosphate (Fig. 4c).

We also noted that for Farms D, E, and F where dietary P was greater than 4.8 g P kg–1, phytic acid was relatively high (1590–2300 mg kg–1) and fiber was relatively low in the diets (Table 1). These farms had fecal phytic acid ranging from 1071 to 1341 mg kg–1. On the other hand, Farm B was low in dietary P (<4.0 g P kg–1), but fecal phytic acid was higher (2035 mg kg–1). Perhaps the low-fiber diets on Farms D, E, and F were more digestible, as suggested by Moreira et al. (2003) and Morse et al. (1992b) and thus phytic acid in the feeds was hydrolyzed more efficiently compared with other farms.

Overall, feces had 14% less phytic acid than the diets, which suggests that phytic acid was hydrolyzed by microbial phytase produced by rumen microflora (Karn, 2001). However, the presence of 18% P as phytic acid in the feces suggests that a considerable amount of phytate was not mineralized in the rumen. Perhaps the presence of phytic acid in feces is due to complexation of phytic acid with other polyvalent cations such as Fe, Al, and Ca present in the diets, making the P in phytic acid unavailable for release during digestion. The reduced recoveries of Fe and Al in the feces relative to diets with NaOH-EDTA extraction (Table 3) further suggest that these polyvalent cations may have formed insoluble complexes with phytic acid in rumen during digestion, resulting in reduced hydrolysis of phytic acid (Dao, 2003). Moreover, mineral supplements containing Ca, Mg, Fe, As, Cu, Cr, Zn, and Se are routinely added in dairy diets. These minerals can exist in association with phytic acid as metal salts, thereby reducing the susceptibility of phytic acid to dephosphorylation during digestion.

Concentrations of phospholipids, DNA, phosphonates, and unidentified orthophosphate monoesters increased in feces relative to diets. Phospholipids in feces may originate from forages, which are rich sources of these compounds (Bieleski, 1973). In addition, the presence of phospholipids, and in particular DNA, in the feces may be attributed to microbial debris that is excreted by the animals (National Research Council, 2001). Phosphonates, which contain C–P bonds, were undetected in diets, but were 0.8 to 1.6% of P in feces. There is clear evidence that the phosphonates excreted in the feces were of microbial origin because the phosphonate forms detected in our study are known to occur in the cilia of rumen protozoa such as Tetrahymena sp. (Kandatsu and Horiguchi, 1962), where they are thought to protect protozoa from the hydrolytic enzymes (Kennedy and Thompson, 1970). These transformations reflect microbial mineralization of phytic acid and the subsequent immobilization of some of the released P into microbial biomass, because proteolytic activity in the rumen is due to bacteria, while fungi produce cellulases and hemicellulases to degrade fiber (Akin and Borneman, 1990) and protozoa ingest bacterial cells and particulate matter, excreting simpler forms of C, N, and P (Baldwin, 1995, p. 211–266).

Results from this study show that fecal P was a combination of (i) inorganic orthophosphate, some of which originated from the hydrolysis of dietary phytic acid and some from the inorganic P contained in feed ingredients; (ii) undigested feed material, which contains phytic acid and other orthophosphate monoesters that we did not measure but which may include degradation products of phytic acid such as lower inositol phosphates, mononucleotide, and sugar phosphates; and (iii) excreted microbial material, such as phospholipids, DNA, and phosphonates.

Relationship between Phosphorus Forms in Feces and Manures
Total P concentrations were lower in manure than in feces, perhaps due to dilution of feces with low-P materials such as urine, bedding material, soil, water, and manure from nonlactating animals that were part of the dairy operations. In addition, the concentrations of C and N were significantly lower in manures than feces suggesting the microbial decomposition of C as CO2 and N as NH4 to NH3 (volatilization) and NO3 to N2O and N2 (denitrification) during manure storage. On the other hand, the significantly higher concentrations of K in manures than feces may be due the fact that dairy cattle excrete most of the K in urine (Haynes and Williams, 1993). In addition, the use of sand as a bedding material on most farms may have also resulted in greater K in manures than feces because sand is known to contain K-bearing minerals such as quartz and feldspar (Sposito, 1989). Similarly, the higher concentrations of Fe and Al in manures relative to feces may be attributed to the addition, during manure handling and storage, of foreign materials such as sand or soil that are rich in Fe and Al. Due to these factors, the manures had more variability in total P and other elements (e.g., K, Ca, C) among different farms compared with feces. Therefore, it is difficult to make direct linkages between P forms in feces and manures. However, the variability among different farms shows that on-farm management can have a significant influence on P forms in manures. Despite these variations, there were some interesting trends in regard to P forms; the proportion of inorganic orthophosphate increased by 3 to 19% from feces to manures, while phytic acid decreased by 2 to 15%, which suggests that phytase excreted in feces, or present in the manure storage areas, may have hydrolyzed phytic acid during manure handling and storage. The significantly lower concentrations of phytic acid, phospholipids, and DNA in manures relative to feces show that these compounds were degraded during manure storage resulting in enrichment of manures with inorganic orthophosphate. McGrath (2004), using 31P NMR and chemical fractionation methods, reported that phytic acid was degraded to inorganic orthophosphate during moist storage of broiler litters. This degradation also more than doubled the concentration and percentage of water-soluble P in litter, indicating that best management strategies should focus on the timing of land application of manures to reduce the risk of soluble P losses. The proportion of inorganic orthophosphate observed in the current study is similar to the inorganic orthophosphate reported in beef-cattle feces (67.4%) by Turner (2004) and is lower than the dairy slurry (86%) produced by lactating dairy cows (Toor et al., 2005a).

Using a chemical sequential fractionation method, Barnett (1994) reported that residual type material (measured in the last step of a sequential chemical extraction method and labeled as nucleic acid) in dairy manures comprised as much as 28% of the total P. The higher percentage of nucleic acid type material determined by Barnett (1994) in manures than the current study (<2.4%) may be due to inclusion of some recalcitrant P forms such as Ca-phosphates in the residual pool, because Barnett (1994) measured P in the residues after acid digestion. The concentrations of phospholipids (extracted with chemical reagents) in dairy, poultry, and swine manures have been reported to be less than 2% (Barnett, 1994; McAuliffe and Peech, 1949; Peperzak et al., 1959), which is slightly lower than the current study (3.5–3.8%).

The concentrations of pyrophosphates and phosphonates in manures were not influenced by variations in dietary P, and their similar concentrations in feces and manures suggest that they are relatively unavailable to animals. Pyrophosphate, which is a short-chain polyphosphate, can be rapidly hydrolyzed in soil, and is used as a fertilizer (Hossner and Melton, 1970); however, reaction with Ca and metals in rumen may have produced pyrophosphate forms more resistant to degradation.


   CONCLUSIONS AND IMPLICATIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS AND IMPLICATIONS
REFERENCES
 
Dairy diets are complex, often-changing mixtures that contain a range of feed ingredients, such as silage, forages, concentrates, and mineral supplements. The heterogeneity in diet composition, in turn, results in feed materials that differ in dry matter, fibers, carbohydrates, protein, and nutrients (Tables 1 and 2) and thus can greatly affect the total P content and speciation. Because of the pressure to reduce nonpoint P pollution of surface waters, Northeastern and Mid-Atlantic dairy farmers have been asked to decrease the amount of P fed and thus, presumably, the amount of P excreted. Inherent to this request is the assumption that reducing P in diets will not result in changing speciation of P in feces or manures. Our results showed, for six diets of total mixed rations from typical Northeastern and Mid-Atlantic dairy farms, that orthophosphate ranged from 35 to 68% and phytic acid from 21 to 47%. From diet to feces, phytic acid was reduced by 14%. However, phytic acid as a percentage of total P had a wide range (from 11 to 29%) for the six farms possibly due to complexation of phytic acid in rumen with polyvalent cations (Fe, Al, Ca). Other factors that may affect the hydrolysis of phytic acid in rumen such as feed digestibility need to be investigated. Because each manure sample was obtained from storage that contained feces, urine, feed refusal, bedding material, sand, and/or soil, we cannot draw a direct linkage for the P speciation of fecal samples with the manure samples. Nevertheless, data for the six farms indicated that regardless of differences in P concentrations or speciation in diets or feces, and for typical manure storage practices in this region, the manures samples were more or less uniform: averaging 73% as inorganic P (range: 63–77%), 18% as orthophosphate monoesters (range: 15–27%), and 7% as diesters (range: 3–8%). Thus, feeding management to minimize excessive dietary P in diets is a critical step and a sound management practice to decrease P excretion in feces and to reduce P surplus on the farm that should be recommended to help mitigate nonpoint P pollution on dairy farms.


   ACKNOWLEDGMENTS
 
This work was funded by USDA Initiative for Future Agriculture and Food Systems Grant no. 2001-52103-11334. We are thankful to the dairy farmers located in the Chesapeake Bay watershed for participating in this project. We thank Dr. Zhengxia Dou of the University of Pennsylvania and other collaborators in the animal science departments of the University of Maryland, Virginia Tech, Cornell, and Pennsylvania State University for helping in identifying these producers. Thanks are expressed to James Hyde of the University of Delaware for his help with sample collection. The NMR analyses were performed at the Stanford Magnetic Resonance Laboratory with support funding from the Stanford University School of Medicine and the assistance of Dr. C. Liu.


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




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
S. Hunger, J. T. Sims, and D. L. Sparks
Evidence for Struvite in Poultry Litter: Effect of Storage and Drying
J. Environ. Qual., June 23, 2008; 37(4): 1617 - 1625.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
R. W. McDowell, Z. Dou, J. D. Toth, B. J. Cade-Menun, P. J. A. Kleinman, K. Soder, and L. Saporito
A Comparison of Phosphorus Speciation and Potential Bioavailability in Feed and Feces of Different Dairy Herds Using 31P Nuclear Magnetic Resonance Spectroscopy
J. Environ. Qual., May 1, 2008; 37(3): 741 - 752.
[Abstract] [Full Text] [PDF]

Home page
 J DAIRY SCIHome page
E. Kebreab, N. E. Odongo, B. W. McBride, M. D. Hanigan, and J. France
Phosphorus Utilization and Environmental and Economic Implications of Reducing Phosphorus Pollution from Ontario Dairy Cows
J Dairy Sci, January 1, 2008; 91(1): 241 - 246.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
K. Gungor, A. Jurgensen, and K. G. Karthikeyan
Determination of Phosphorus Speciation in Dairy Manure using XRD and XANES Spectroscopy
J. Environ. Qual., October 24, 2007; 36(6): 1856 - 1863.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
B. Ajiboye, O. O. Akinremi, Y. Hu, and D. N. Flaten
Phosphorus Speciation of Sequential Extracts of Organic Amendments Using Nuclear Magnetic Resonance and X-ray Absorption Near-Edge Structure Spectroscopies
J. Environ. Qual., October 16, 2007; 36(6): 1563 - 1576.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
Z. He, B. J. Cade-Menun, G. S. Toor, A.-M. Fortuna, C. W. Honeycutt, and J. T. Sims
Comparison of Phosphorus Forms in Wet and Dried Animal Manures by Solution Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy and Enzymatic Hydrolysis
J. Environ. Qual., May 25, 2007; 36(4): 1086 - 1095.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
A. L. Shober and J. T. Sims
Integrating Phosphorus Source and Soil Properties into Risk Assessments for Phosphorus Loss
Soil Sci. Soc. Am. J., March 12, 2007; 71(2): 551 - 560.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
A. L. Shober, D. L. Hesterberg, J. T. Sims, and S. Gardner
Characterization of Phosphorus Species in Biosolids and Manures Using XANES Spectroscopy
J. Environ. Qual., October 27, 2006; 35(6): 1983 - 1993.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
S. Bittman, C. G. Kowalenko, D. E. Hunt, T. A. Forge, and X. Wu
Starter Phosphorus and Broadcast Nutrients on Corn with Contrasting Colonization by Mycorrhizae
Agron. J., March 2, 2006; 98(2): 394 - 401.