8
Published: August 26, 2011 r2011 American Chemical Society 8217 dx.doi.org/10.1021/es200517y | Environ. Sci. Technol. 2011, 45, 82178224 ARTICLE pubs.acs.org/est Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach Mustafa N. Shafqat* ,,and Gary M. Pierzynski Department of Environmental Sciences, COMSATS Institute of Information Technology, Abbottabad, Pakistan Department of Agronomy, Kansas State University, Manhattan, Kansas 66502, United States b S Supporting Information INTRODUCTION Animal manure and biosolids are generally applied to soil on the basis of crop nitrogen (N) needs and are considered a valuable resource to agriculture and crop production. However, the pro- blem is that the N:P ratio is narrower in the manures than that taken up by crops, and thus repeated application will eventually build soil P to the levels where most crops might show little yield response to the added P. 1 Recent soil P analysis data from the United States conrmed that the majority of such soils contained P levels in the excessive or high categories and constituted a high risk for osite P movement and may be responsible for the accelerated eutrophication of surface waters. 25 Phosphorus in animal manure and biosolids is inuenced by the digestion system of the animal, diet, and waste processing methods. 68 Monogastric animals, such as pig, broiler, and turkey lack phytase in the digestive system and depend upon feed additives for adequate P nutrition. 9 More than 50% of the ingested P is subject to fecal excretion. 10 The formation of biogenic phosphate minerals (Ca and Mg phosphate) of low solubility in poultry manure was found to control soil solution P concentrations while such minerals were not detected in dairy manure. 11 Most of the organic P (P o ) was tied up in phospho- monoesters in broiler litter as well as in swine manures while sugarphosphomonoesters and phosphodiesters were the domi- nant constituents of cattle manures. 12,13 It is more dicult to hydrolyze phosphomonoesters than phosphodiesters in the soil environment and thus the majority of P o will contribute little toward bioavailable P in soils amended with manure from monogastric animals. The treatment of biosolids with salts of aluminum, iron, or calcium produces less soluble P than in untreated biosolids. 14,15 Similarly, poultry litter treated with calcium carbonates, alum, iron chloride, or iron sulfate showed a decrease in water-soluble P levels from >2000 mg L 1 to <1 mg L 1 . 16 Despite the occurrence of the large variations in the composition, solubility, and bioavailability of P forms in animal manure, current protocols for land application do not dierenti- ate among manure types, although an estimate of water solubility of manure P and in some cases P mineralization potential are included in manure management guidelines. 17 Bioavailable P in soil refers to that fraction of total P that becomes available to crops during the growing season or that potentially becomes available to algae in fresh water systems. In addition to routine soil testing procedures, iron oxide impreg- nated lter paper, 18,19 anion exchange resin, 20 and hydrous iron oxide sinks in dialysis tubing 21,22 have been used for the Received: February 15, 2011 Accepted: August 26, 2011 Revised: July 20, 2011 ABSTRACT: Animal manure amended soils often contain large amounts of bioavailable phosphorus (P) and constitute high risk for the deterioration of surface water quality through eutrophication. Current standards set for the safe disposal of animal manure through soil application are based on the assumption that phosphorus in all P sources would behave similarly. The primary objective of this study was to understand the inuence of P from several manure and mineral fertilizer sources applied at 0, 50, and 150 mg P kg 1 on two measures of bioavailable P to six soils of dierent initial soil test P levels using corn (Zea mays L.) P uptake and an iron oxide strip method for soil analysis (FeO-P). Total net bioavailable P (TNBP) was calculated by subtracting total P uptake by corn after seven consecutive harvests in control treatments that did not receive P from the P uptake from P-amended treatments. Net biovavailable P after the rst harvest (NBP1) was calculated in a similar fashion but only using data from the rst harvest. Signicant dierences in TNBP and NBP1 were found when comparing P sources. The hog (Sus scrofa) manure had the greatest P bioavailability while turkey (Meleagris gallopava) litter had the lowest among the animal P sources across all soils and levels of P application. Signicant dierences were also found between soils with the highest amounts of TNBP and NBP1 found in the Woodson soil and lowest detected in the Crete soil for most P sources. The FeO-P method was useful in predicting TNBP from most P sources.

Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

  • Upload
    gary-m

  • View
    215

  • Download
    2

Embed Size (px)

Citation preview

Page 1: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

Published: August 26, 2011

r 2011 American Chemical Society 8217 dx.doi.org/10.1021/es200517y | Environ. Sci. Technol. 2011, 45, 8217–8224

ARTICLE

pubs.acs.org/est

Bioavailable Phosphorus in Animal Waste Amended Soils: UsingActual Crop Uptake and P Mass Balance ApproachMustafa N. Shafqat*,†,‡ and Gary M. Pierzynski‡

†Department of Environmental Sciences, COMSATS Institute of Information Technology, Abbottabad, Pakistan‡Department of Agronomy, Kansas State University, Manhattan, Kansas 66502, United States

bS Supporting Information

’ INTRODUCTION

Animalmanure and biosolids are generally applied to soil on thebasis of crop nitrogen (N) needs and are considered a valuableresource to agriculture and crop production. However, the pro-blem is that the N:P ratio is narrower in the manures than thattaken up by crops, and thus repeated application will eventuallybuild soil P to the levels where most crops might show little yieldresponse to the added P.1 Recent soil P analysis data from theUnited States confirmed that themajority of such soils contained Plevels in the excessive or high categories and constituted a high riskfor offsite P movement and may be responsible for the acceleratedeutrophication of surface waters.2�5

Phosphorus in animal manure and biosolids is influenced bythe digestion system of the animal, diet, and waste processingmethods.6�8 Monogastric animals, such as pig, broiler, andturkey lack phytase in the digestive system and depend uponfeed additives for adequate P nutrition.9 More than 50% of theingested P is subject to fecal excretion.10 The formation ofbiogenic phosphate minerals (Ca and Mg phosphate) of lowsolubility in poultry manure was found to control soil solution Pconcentrations while such minerals were not detected in dairymanure.11 Most of the organic P (Po) was tied up in phospho-monoesters in broiler litter as well as in swine manures whilesugarphosphomonoesters and phosphodiesters were the domi-nant constituents of cattle manures.12,13 It is more difficult to

hydrolyze phosphomonoesters than phosphodiesters in the soilenvironment and thus the majority of Po will contribute littletoward bioavailable P in soils amended with manure frommonogastric animals. The treatment of biosolids with salts ofaluminum, iron, or calcium produces less soluble P than inuntreated biosolids.14,15 Similarly, poultry litter treated withcalcium carbonates, alum, iron chloride, or iron sulfate showeda decrease in water-soluble P levels from >2000mg L�1 to <1 mgL�1.16 Despite the occurrence of the large variations in thecomposition, solubility, and bioavailability of P forms in animalmanure, current protocols for land application do not differenti-ate among manure types, although an estimate of water solubilityof manure P and in some cases P mineralization potential areincluded in manure management guidelines.17

Bioavailable P in soil refers to that fraction of total P thatbecomes available to crops during the growing season or thatpotentially becomes available to algae in fresh water systems. Inaddition to routine soil testing procedures, iron oxide impreg-nated filter paper,18,19 anion exchange resin,20 and hydrous ironoxide sinks in dialysis tubing21,22 have been used for the

Received: February 15, 2011Accepted: August 26, 2011Revised: July 20, 2011

ABSTRACT: Animal manure amended soils often contain large amounts ofbioavailable phosphorus (P) and constitute high risk for the deterioration ofsurface water quality through eutrophication. Current standards set for the safedisposal of animal manure through soil application are based on the assumptionthat phosphorus in all P sources would behave similarly. The primary objectiveof this study was to understand the influence of P from several manure andmineral fertilizer sources applied at 0, 50, and 150mg P kg�1 on twomeasures ofbioavailable P to six soils of different initial soil test P levels using corn (Zea maysL.) P uptake and an iron oxide strip method for soil analysis (FeO-P). Total netbioavailable P (TNBP) was calculated by subtracting total P uptake by corn afterseven consecutive harvests in control treatments that did not receive P from theP uptake from P-amended treatments. Net biovavailable P after the first harvest(NBP1) was calculated in a similar fashion but only using data from the firstharvest. Significant differences in TNBP and NBP1 were found when comparing P sources. The hog (Sus scrofa) manure had thegreatest P bioavailability while turkey (Meleagris gallopava) litter had the lowest among the animal P sources across all soils and levelsof P application. Significant differences were also found between soils with the highest amounts of TNBP and NBP1 found in theWoodson soil and lowest detected in the Crete soil for most P sources. The FeO-P method was useful in predicting TNBP frommost P sources.

Page 2: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8218 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

quantification of bioavailable P in soils. The method developedby Myers et al.19 also provides measurements of maximum Preleased (Pmax). Manure amended soils had higher Pmax thanuntreated soils and P was released slower in low P soils than inhigh P soils.

Studies are lacking on the assessment of bioavailable P indifferent P sources applied in varying amounts across manydifferent soils. Therefore, we tested the following hypotheses inthis study: (a) P sources (animal manures from monogastric andruminant animals, biosolids, inorganic P fertilizer) will not havedifferential effects on soil P bioavailability, (b) bioavailable P in allP sources tested will increase proportionately as the amountadded increases, (c) all P sources will release similar amounts ofbioavailable P under limited P removal by plant growth, and (d)FeO-impregnated filter paper bioavailable P will show trendssimilar to TNBP and will offer solutions for rapid assessment ofbioavailable P in animal manures and in manure-amended soils.

’EXPERIMENTAL METHODS

Collection of Soils and Waste Materials. Soils used in thisstudy were collected from different parts of Kansas in the UnitedStates. Selected physicochemical properties of soils are presentedin the Table 1. Surface soil samples (0�15 cm depth) werecollected from areas mapped as Eram-Lebo clay loam (fine,mixed, active, thermic Aquic Argiudolls), Crete silty clay loam(fine, smectitic, mesic Pachic Argiustolls),Wabash silty clay (fine,smectitic, mesic Cumulic Vertic Endoaquolls), Woodson siltloam (fine, smectitic, thermic Abruptic Argiaquolls), Ulysses siltloam (fine-silty, mixed, superactive, mesic Aridic Haplustolls),and Harney silt loam (fine, smectitic, mesic Typic Argiustolls).Soils were air-dried and sieved to pass through <2-mm sieve andstored at 25 �C prior to use in the study.Land application of animal manure and other organic bypro-

ducts is a common practice. Therefore, we utilized five animalmanures consisting of two types of stockpiled cattle (Bos taurus)manures (CM1 and CM2), turkey litter (TL), solid hog manure(HM), and an aerobically digested sludge (SS) in this study.CM1 was relatively fresh while CM2 was collected from oldstockpile with considerable mineral matter mixed with it; bothstockpiles were exposed to weather. Select properties of theanimal manures are given in Table 2. Animal manures were air-dried and ground to pass a <2-mm sieve and stored at 25 �C priorto their use in the greenhouse experiment. Triple super phos-phate (TSP) was also included as inorganic P (Pi) source forcomparative purposes.Greenhouse Experiment. Six soils were amended with 50

and 150 mg P kg�1 soil from each of the six P sources andincubated for 24 h at 200 g kg�1 water content. An unamendedcontrol was prepared in a like manner. There were a total of 78treatments (6 soils � 6 P sources � 2 levels of P addition + 6controls) which were arranged in randomized complete blockdesign with three replications. Subsamples of soil were collectedprior to cropping and designated as T0. Two kg of soil was placedinto pots and ten seeds of corn (Zea mays L.) Pioneer hybrids4662 were planted in each pot. Nitrogen and potassium (K) wereapplied at 150 mg N kg�1 and 60 mg K kg�1 soil. Nitrogen wasapplied from urea in three splits (50 mg N kg�1 each); at sowingand at 10 and 20 days after emergence, while all K was added inthe form of potassium sulfate at the time of sowing during eachcrop cycle. One week after emergence, plants were thinned tosix plants per pot. Supplemental light for 16 h per day was

maintained throughout the experiment. Pots were brought touniform soil�water content (200 g kg�1) twice weekly byweighing with a constant volume of water added to each potbetween weighing. A total of seven harvests of corn wereachieved at 35 day intervals each from the date of sowing. Ateach harvest, only the above ground portion of the corn plant washarvested while roots were kept inside pots to simulate fieldconditions. Plant materials were briefly washed in distilled waterand dried to constant weight at 60 �C and dry weight wasrecorded. Soil samples collected after the seventh harvest weredesignated as T7. Soil samples were air-dried and sieved (< 2mm). Plant and soil samples were stored at 25 �C prior tochemical analysis.Soil and Plant Analysis.Ground subsamples (0.25 g) of plant

material were digested with concentrated sulfuric acid andhydrogen peroxide and P was determined in the tissue digestsusing ICP-AES. Total P uptake during each harvest was calcu-lated by multiplying plant dry weight with the concentration of Pin the plant material. The data on corn biomass, tissue Pconcentration, and total P removed for different treatments arepresented in Supporting Information (Tables S8�S25). Soiltexture and other routine soil and animal manure analysis wereperformed by the Soil Testing Laboratory at Kansas StateUniversity using standard protocols.Total Bioavailable and Net Bioavailable Phosphorus in

Soils Amended with Different P Sources. The crop P removalfrom different treatments during successive harvest was pooledinto total P uptake. The total net bioavailable P (TNBP) at both50 and 150 mg P kg�1 application levels in a given soil wascalculated by subtracting total P uptake in the control treatmentsfrom the total P uptake in different P amended treatments across

Table 1. Selected Physicochemical Properties of Soils Usedin the Study

g kg�1

soil pH

Bray-1P

(mg kg�1) OM sand silt clay texture

Eram-Lebo 6.2 4.5 28 160 440 400 silty clay

Crete 6.4 23.0 22 140 500 360 silty clay loam

Wabash 5.4 21.0 33 180 540 280 silty clay loam

Woodson 6.8 27.0 18 560 380 60 sandy loam

Ulysses 7.7 23.0 16 420 390 190 loam

Harney 7.8 45.0 25 280 530 190 silt loam

Table 2. Selected Characteristics of Waste Materials Used inThis Study

g kg�1

animal

manure characteristics

total

P

total

N

total

C

total

K C:N C:P

cattle manure 1

(CM1)

dairy, fresh

stockpile

8.8 19.8 238 22.3 12.0 27.0

cattle manure 2

(CM2)

dairy, old

stockpile

3.8 7.7 77 10.5 10.0 20.0

turkey litter (TL) dry stockpile 21.0 17.1 279 14.0 16.3 13.2

hog manure (HM) moist and fresh 17.0 45.1 400 11.7 8.87 23.5

sewage sludge (SS) aerobically digested 27.0 50.2 311 4.1 6.20 11.5

Page 3: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8219 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

seven harvests. The same procedure was followed to compute netbioavailable P after the first harvest (NBP1). In this paper fromnow onward TNBP and NBP1 will refer to total net bioavailableP across seven harvests and net bioavailable P during first harvestof the corn, respectively.Iron Oxide Impregnated Filter Paper Bioavailable P. The

Eram-Lebo and Ulysses soils were selected for P release kineticstudies. Soil samples amended with 150mg P kg�1 from each of theP sources along with the control treatments from both soils at T0and T7 were weighed (1 g) and placed into wide-mouth glassbottles. In addition, P sources alone without soil amendment wereweighed to contain same amount of total P. An FeO-coated filterpaper strip enclosed in two spectra mesh screens was placed in thebottle along with 80 mL of 0.01 M CaCl2.

19 The samples wereshaken at 125�135 excursions per minute for up to 72 h. Thesamples were destructively removed after 1, 2, 4, 8, 12, 24, 48, and 72h of shaking. The P adsorbed on the FeO paper was dissolved in 0.2M H2SO4 and inorganic P was determined colorimetrically.23 Thedata from the P release study were fit to the exponential equation

Pd ¼ Pmaxð1� e�ktÞ ð1Þwhere Pd is the amount of P desorbed (mg kg

�1) at time (t), Pmax isthemaximumamount of P released (mgPkg�1), k is the desorptionconstant (h�1), and t is time (h).Analysis of variance was done using Proc GLM of SAS.24

Treatment means were separated at P = 0.05 by using leastsignificant difference (LSD). In case of significant interactioneffect among variables, response of each factor at the levels ofother factors was compared and discussed using LSD valuecalculated by using error mean square at correct error degreesof freedom and by using appropriate alpha value.

’RESULTS AND DISCUSSION

Total Net Bioavailable P (TNBP) from Different P SourceTreatments during Seven Harvests.The pots not receiving anyP input (controls) had variable amounts of corn P uptake (Table S1

in Supporting Information (SI)). The highest amount of P taken upby corn was 50 mg P kg�1 in unamended Crete soil which wassignificantly different from all control treatments. The Ulysses andEram-Lebo soils had the lowest amount of P taken up by cornwhich was 20 and 21 mg P kg�1, respectively, and were alsosignificantly different from all other control treatments. Theamounts of corn P uptake in the remaining unamended soils hadvalues between these two extremes with no significant differencesbetween treatments. It is interesting to note that amount of corn Puptake in these soils has nothing to do with initial levels of soil testphosphorus (STP) (Table 1), which is regarded as an index of Pbioavailability determined by using chemical extractants and uti-lized for P fertilizer recommendations. The initial levels of STPwere 4.5 (deficient) and 23 mg P kg�1 (optimum levels ofbioavailable P) for the Eram-Lebo and Ulysses soils respectively,yet they had similar amounts of corn P uptake.Among TL, HM, and SS (monogastric P sources), TL con-

sistently resulted in the lowest amount of TNBP across all soils andat both levels of P application (Table 3). The range of TNBP wasfrom 36 to 52.5 mg P kg�1 at the highest level of TL application inthe Crete andWabash soils, respectively, and consisted of between24 and 35% of the total P. The range was 10.2 to 21.2 mg P kg�1 atthe lower level of P addition for the Crete and Ulysses soils,respectively, and constituted 20 to 42% of the total P added fromthe TL. This was in sharp contrast with HM, where the TNBPranged from 68.4 to 95 mg P kg�1 at the high level of P addition inthe Eram-Lebo and inUlysses soils, and at the lower level it rangedfrom 24.8 to 37.7 mg P kg�1 for the Wabash and Ulysses soils,respectively. Consequently, corn had taken up between 50 and76% of the total P at the lower level and between 46 and 63% at thehigher level of P application across all soils. These values of TNBPwere nearly double in HM amended soils compared to TLirrespective of level of P application.The SS amended soils had TNBP that ranged from 43 to 63%

at lower level and between 30 and 48% of the total P at higherlevel of P application and were intermediate between TL andHM treatments (Table 3). For ruminant animals, both cattle

Table 3. Interactive Effect of P Source and Rates of Application on TNBP in Different Soilsa

mg P kg�1

P source P rates (mg P kg-1) Eram-Lebo Crete Wabash Woodson Ulysses Harney

0b 21.0 50 34 40 20 37

CM-1 50 19.9 y�d* 16 b*�e* 21.2 w�d* 27 s�y 28.3 r�y 24.2 t�b*

150 55.5 h�k 59.5 g�j 55 h�k 63 f�i 65 e�g 58.5 g�j

CM-2 50 20.5 x�d* 15 c*�e* 17.3 z�e* 22.1 w�d* 27.2 s�y 25.8 t�z

150 56 h�k 48 k�m 55 h�k 63 f�i 70.2 d�f 63.4 f�h

TL 50 15.5 c*�e* 10.2 e* 16.5 a*�e* 13.8 d*e* 21.2 w�d* 17.3 z�e*

150 43.5 m�o 36 n�r 52.5 j�l 42.3 m�o 44.2 l�n 39 n�p

HM 50 29 q�x 27.1 s�y 24.8 t�a* 29.8 q�w 37.7 n�q 35.4 o�s

150 68.4 d�f 76 cd 70.6 d�f 91.7 b 95 ab 79.5 c

SS 50 21.5 w�d* 22.3 v�d* 22.7 t�c* 23.4 t�c* 31.3 p�t 27.1 s�y

150 44 l�n 52.7 j�l 55.5 h�k 72 c�e 71 c�f 66.6 e�g

TSP 50 27.3 s�y 30.8 p�v 22.4 u�d* 31 p�u 17.5 z�e* 22.4 u�d*

150 54 jk 54.4 i�k 55 h�k 101.7 a 42.3 m�o 48.4 k�ma Soil � P rate � P source interaction effect on TNBP is significant at P < 0.01. Means with same letter/letters in columns and across rows are notsignificant at P < 0.05. Means with letters having dash line (�) represent all letters present between the two alphabets. Means with letter containing *represent letter starting after z. bTotal P uptake from control treatments during seven harvests are included for comparative purpose. See Table S1,Supporting Information.

Page 4: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8220 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

manures had the lowest amounts of TNBP in the Crete soil (∼15mg P kg�1) and the highest in the Ulysses soil (∼ 28 mg P kg�1)at lower level of P application and were significantly differentfrom each other (Table 3). Therefore, between 30 and 56% of thetotal P was removed by corn across different soils from bothcattle manures when applied at lower level of P. However, CM1had significantly more TNBP (65 mg P kg�1) in the Ulysses soilthan in Wabash and Eram-Lebo soils (∼55 mg P kg�1) at thehighest level, while CM2 had the least in Crete soil (48 mg Pkg�1) and the highest was seen in theUlysses soil (70mg P kg�1)and were significantly different from each other (Table 3). ForTSP, TNBP constituted 35 and 28% of the total P in the Ulyssessoil at lower and higher levels of P applications, respectively,while nearly two-thirds of added P inWoodson soil was TNBP atboth levels of P application. In the remaining soils, the propor-tion of TNBP was ∼50% at the lower level of P additions whichdecreased to∼30% of total P at higher levels of TSP application(Table 3). As for the influence of levels of P application onincreases in TNBP, all P sources except CM1 in the Woodsonsoil, TL for all soils except Ulysses soil, and CM2 amended CreteandWabash soils showed proportionate or even greater increasesas P addition increased (Table 3).The comparison of TNBP at a given level of P application from

different animal P sources with TSP would allow us to understandthe usefulness of P contained in animal manures as a source ofplant available P. All manure P sources, except TL and SS at bothlevels of application resulted in significantly higher TNBP inUlysses soil compared with the TSP treatment (Table 3). In fact,HM treatment had nearly double amounts of TNBP at both levelsof P application in the Ulysses soil. Similar trends were reported inthe Harney soil but only at higher level of P application (Table 3).Most animal P sources in Woodson soil contributed less TNBPthan TSP treatment with HM was the only exception at the lowerlevel of P application. The HM consistently performed better as Psources under most circumstances, while TL proved to be inferioras a P source, while remaining treatments were nearly as effectiveas TSP in most soils (Table 3).On the basis of data presented so far, we rejected our first

hypothesis and found not only significant variations in TNBP inthe animal manures from monogastrics compared with rumi-nants but significant differences were also seen within eachgroup, especially manures from monogastric animals. Overall,HM contributed the greatest amount of TNBP, while TL was atthe other extreme and had the lowest at both levels of Papplication and across all soils. The SS and both cattle manureshad TNBP that would be rated as intermediate between HM andTL. Relatively constant P bioavailability across all soils amendedwith TL at a given level of P suggests that solubility of sparinglysoluble compounds of P present in TL might be exerting adominant influence and limited role of soil properties in con-trolling P release in these soils. This observation was supportedby the fact that TL had significantly more TNBP in Wabash soilamended with the highest level P application suggesting moresolubility of P forms contained in the TL in this acidic soil (pH =5.4). Inclusion of more soils having soil pH between 4 and 5.5warrants a study to prove further that relatively more P wouldbecome available in acidic soils when amended with TL than inneutral and slightly alkaline soils which were used in the presentstudy. This finding was also supported by Cooperband andGood,11 who identified the presence of biogenic P minerals inpoultry manure which were Ca and Mg phosphates minerals oflow solubility. They did not detect suchminerals in dairy manure.

Moreover, they found minimal influence of antecedent P (high Pvs low P soil) on water-soluble P in poultry manure, whilesignificantly higher water-soluble P was resulted in dairy manureamended high P soil than in low P soil.The order of TNBP at a given level of P application from the

animal manures in this study was HM > CM2 = CM1 > TL. Theamounts of P sequentially extracted withmild extractants (0.01MCaCl2 and 0.5 M NaHCO3) and subsequently added together inthis study (Table S2, Supporting Information) resulted in theremoval of identical amounts (52% of total P) in HM and CM2while, slightly lesser amounts in CM1 (47%) and the lowestamount of P (25%) was extracted in TL. This order was furthersupported by the evidence provided by Turner and Leytem,12

who identified 78% of the total P was present in water (55%)and NaHCO3 (23%) extractable labile fraction in HM. Thisproportion was 55% (11% water +43% NaHCO3) in cattlemanure and 33% (29% water + 5%NaHCO3) in TL. Conversely,the proportions of total manure P present in NaOH and HClfractions were 60% (12 + 48) in TL, followed by 25% (19 + 6) incattle manure, and only 15% (6 + 9) in HM in decreasing order.Moreover, broiler litter contained more resistant Po as phosphatemonoesters, ∼50% in NaOH and HCl fractions, which was 14%in cattlemanure, while HM contained just 8% of the total P. In thepresent study, at lower P application levels between 50 and 76%and at higher levels between 47 and 63% of the total P added fromHMwas net removed by corn across all soils. This proportion was∼30% of the total P in all soils amended with TL (Table 3). Thegreater crop P uptake inHMand in both cattlemanure treatmentsshow parallels with more P present in forms that are extractedwith water and mild extractants than in TL. This suggests theutility of water-soluble and NaHCO3 fractions in determiningproportion of total P added in animal manures that might beremoved during crop P removal.We accepted our second hypothesis and there were some P

sources which had shown proportionate increase in TNBP athigher levels than what was seen at lower level of P applicationalthough it was dependent on the P source as well as the soil. Thisbehavior was unique in either TL amended soils or in Woodsonsoil amended with most P sources (Table 3). Less than propor-tionate increase in TL amended Ulysses and Harney soils mightsuggest less solubility of P forms contained in TL in these highpH soils. Previously, Cooperband and Wood11 identified bio-genic P minerals which were Ca and Mg phosphates of lowsolubility in the TL. Therefore, less solubility of such mineralsmight be expected at high soil pH. This was further supported bysignificantly higher amounts of TNBP in TL amended Wabashsoil which is acidic in nature and favors more dissolution ofbiogenic minerals at higher levels of P application. TheWoodsonsoil had the least amount of clay and the highest content of thesand among all soils which would have offered least interactionwith its constituents and allowed more P to remain in forms thatwere taken up by corn. We also rejected our hypothesis that allanimal P sources might contain P that would be as readily availableas fertilizer P. This was again dependent on the interactive effectof the three variables, for example, lower TNBP in more basic(soil pH = 7.7) Ulysses soil with TSP (Table 3) suggests that soilsolution composition might be such that was causing rapidprecipitation of solution P into secondary Ca phosphate mineralsof low solubility. However, no such reductions in TNBP werereported in the Ulysses soil with animal manures suggesting therole ofmanure constituents especially the amount of soluble organiccarbon compounds which might have entered into complexes

Page 5: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8221 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

with excess soil solutionCa and prevented it from forming secondaryP minerals.Net Bioavailable P after First Harvest (NBP1). The amount

of readily bioavailable P taken up by crops during early stages oftheir growth cycle could prove useful in decreasing the amount ofavailable P in soils amended with animal manures so that its offsitemovement could be minimized. Corn was grown for 35 days aftersowing and it took about 10 days to complete germination andthinning, so the remaining 25 days had themost active P uptake forthe first harvest. The highest amount of NBP1 was 57 mg P kg�1

reported in the Woodson soil amended with 150 mg P kg�1 withTSP (Table 4) which nearly constituted 56% of the TNBP(Table 3) and was significantly different from all treatmentsincluded in the study. Contrary to this maximum, the Crete soilmixed at the same level fromCM2 resulted in net P removal of just2.9 mg P kg�1 which only made up 6% of the TNBP (48 mg Pkg�1). Besides CM2, SS (8.1 mg P kg�1) and TL (5.0 mg P kg�1)amended Crete soil also had statistical similar amounts of NBP1that comprised 10 and 15% of the TNBP, respectively (Tables 3and 4). At lower level of P application, the HM amended Ulyssesand Eram-Lebo soils and TSP amended Woodson soil had nearlyidentical amounts of NBP1 (∼ 17.0 mg P kg�1) which repre-sented 44, 59, and 55% of the TNBP, respectively (Table 3). Thelowest amounts at the same level of P application were found inCM2, TL, CM1, TSP, and SS amendedCrete soil and had 0.5, 0.7,2.4, 3.9, and 5.5 mg P kg�1 which made up <25% of the TNBP.Among the animal P sources, HM applied at higher level of P

resulted in 46, 44, 33, 31, 29, 19.6 mg P kg�1 NBP1 in Woodson,Ulysses, Wabash, Eram-Lebo, Harney, and Crete soil in decreasingorder, respectively, and were significantly different from all other Psources at a given level of P application (Table 4). These NBP1constituted 50, 48, 47, 45, 36, and 26% of TNBP in decreasingorder, respectively (Tables 3 and 4). Only TSP amendedWoodsonsoil produced more NBP1 than the aforementioned treatments.Nearly 50% of the TNBP was removed during the first harvest inWoodson, Ulysses, Wabash, and in Eram-Lebo soils amended withHM at higher level of P. Contrary to the HM, TL amended with

same rate had 17.9, 15.9, 14.9, 12.8, 11.7, and 5mg P kg�1 NBP1 inWabash, Eram-Lebo, Ulysses,Woodson, Harney, andCrete soils indecreasing order, respectively (Table 4) and these constituted 34,36, 34, 30, 30, and 14% of the TNBP during the experiment(Table 3 and 4). These contrasting situations have clear implica-tions in understanding the P behavior in different soils. While allthree factors were involved in controlling the release of P in a giventreatment, but these results suggest that HM contained more P informs that were available during the first harvest while in TL,greater proportion of TNBP was removed in subsequent harvests.Again these results are supported by the observation that morebioavailable P was present in 0.01 MCaCl2 and in 0.5 MNaHCO3

fractions in HM than in TL (Table S2, Supporting Information) inthis study and elsewhere.12 Overall, CM2 contributed significantlyless NBP1 at both levels of P application in the Crete soil comparedto all other soils. The Woodson and Wabash soils amended with150 mg P kg�1 from SS had 30 and 21.4 mg P kg�1 (Table 4)which constituted 42 and 39% of the TNBP, respectively, com-pared to 25 and 33% of the TNBP removed in the Eram-Lebo andUlysses soils, respectively (Tables 3 and 4).The Crete was the only soil where all P sources had signifi-

cantly lower amount of NBP1 and more P was taken up duringthe subsequent harvests (Table 4). In all other soils, a consider-able proportion of TNBP was removed during the first harvest(Table 3 and 4). Therefore, soils which either had morebioavailable P in the control treatments (Crete soil) or thosewhich had higher levels of STP like the Harney soil, will initiallylimit the solubility of P forms in animal manures and thus mightpose slightly greater threat of offsite P movement as lesserquantities of P were removed during the first growth.Phosphorus Release Using FeO-Impregnated Filter Paper.

The results regarding P release from the different P sources arepresented in Figure 1A. The parameter coefficients of the equationare given in Table 5. The order of P release (Pmax) was SS >HM>TL>CM1>CM2. Themaximum amount of P release in the timescale of the study was 8.4 g kg�1 and the lowest was 1.31 g kg�1 forSS and CM2, respectively. The Pmax values also suggested that SS,

Table 4. Interactive Effect of P Source and Rate of Application on NBP1 in Different Soilsa

mg P kg�1

P sources P rates (mg P kg-1) Eram-Lebo Crete Wabash Woodson Ulysses Harney

0b 4.1 22 8.0 17.0 8.0 17.0

CM-1 50 10.7 t�d* 2.4 f*�i* 8.9 v�f* 11.6 q�z 12.2 o�w 6.7 x�h*

150 16.7 j�t 11 s�b* 16.2 j�t 17.2 i�r 17 i�r 14.5 l�v

CM-2 50 10.8 t�c* 0.5 i* 4.8 d*�i* 8.3 w�g* 11.3 r�a* 9.1 v�e*

150 19.3 f�m 2.9 g*�i* 15 k�u 19.8 f�l 24 d�f 19.7 f�m

TL 50 8.3 w�g* 0.69 h*i* 5.2 b*�i* 6.1 y�i* 8.1 w�g* 5.7 z�i*

150 15.9 k�u 5 c*�i* 17.9 g�o 12.8 o�w 14.9 l�v 11.7 p�y

HM 50 17.4 h�q 10.8 t�c* 11.5 q�z 13.7 m�w 17.7 g�p 12.8 o�w

150 31 c 19.6 f�m 33 c 46 b 44 b 29 c�e

SS 50 10.9 t�c* 5.5 a*�i* 9.4 v�e* 8.4 w�g* 12 o�y 8 w�g*

150 10.8 t�c* 8.1 w�g* 21.4 f�k 30 c 23.2 e�h 18.9 f�n

TSP 50 13.2 n�w 3.9 e*�i* 10.2 u�d* 17.2 i�r 5.4 a*�i* 10.2 u�d*

150 22 f�j 14.6 l�v 23.5 e�g 57 a 16.9 j�s 22.9 e�ia Soil � P rate � P source interaction effect on NBP1 is significant at P < 0.01. Means with same letter/letters in columns and across rows are notsignificant at P < 0.05. Means with letters having dash line (�) represent all letters present between the two alphabets. Means with letter containing *represent letter starting after z. bAntecedent P during first harvest included for comparative purposes. See tables in data set S1 Supporting Informationfor more details.

Page 6: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8222 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

HM, and TL, released relatively more P than cattle manures. Theorder of release constant (k) was HM > TL > SS > CM1 > CM2indicating that HM released P much faster (k = 0.27 h�1) thanother P sources. However, the nature of P in CM2 was such that kwas the lowest (k = 0.07 h�1) among all P sources included in thestudy. Moreover, HM, SS, and TL not only released more P (highPmax) but they also released that P much faster than the both cattlemanures when tested without soil amendment.The P release from various P sources in the Eram-Lebo and

Ulysses soil amended at 150 mg P kg�1 at T0 is presented inFigure 1B and C, respectively, and P release after seventh harvestis presented in Figures S1 and S2 (Supporting Information). Thecontrol treatments (Table 6) in both soils released only 7.5 mg Pkg�1 which constituted only 1/3 of the actual P uptake (21 mg Pkg�1). The order of Pmax in the Eram-Lebo soil at T0 was TSP >

CM2 > HM > CM1 > TL > SS, while the order of TNBP wasHM > CM2 = CM1 = TSP > SS = TL. The trends in FeO-P inthis soil agree well with the order of TNBP. Because Pmax valuesfor TSP, CM2, HM, and CM1 were statistically at par with eachother in this method, both methods indicated a similar pattern inP release in this low P soil.

Figure 1. (A) Phosphorus release from organic P sources using FeO-strip method on equal P basis. (B) Effect of P source at 150 mg P kg�1 on P releasefrom Eram-Lebo soil atT0 using FeO-strip method. (C) Effect of P source at 150mg P kg�1 on P release fromUlysses soil atT0 using FeO-strip method.(D) Relationship between Pmax (mg P kg�1) determined using FeO-strip method and TNBP for both cattle manures, TL, and SS plotted together forboth Eram-Lebo and Ulysses soils, as well as for individual soils.

Table 5. Phosphorus Release from Different P Sources onEqual P Basis and Parameter Estimates of Equation 1

waste material k (1/h) Pmax (g kg�1)

CM1 0.08 3.6

CM2 0.07 1.31

TL 0.22 5.40

HM 0.27 6.90

SS 0.16 8.40

Table 6. Interactive Effect of P Source, Time, and Soil onPmax (Parameter Estimate of Equation 1) Using FeO-stripMethoda

mg P kg-1

Eram-Lebo soil Ulysses soil

treatment T0 T7 T0 T7

CM1-150 39 ef 9 ijk 57 c 8.6 ijk

CM2-150 43 e 10 ji 65 b 8.4 ijk

TL-150 31 g 7.2 ijk 32 fg 8.0 ijk

HM-150 42 e 11 hi 53 cd 10 ijk

SS-150 30 g 11 hij 46 ed 8.5 ijk

TSP-150 46 ed 18 h 77 a 11 ji

control 7.4 ijk 2.8 k 7.5 ijk 3.8 jkaMeans with same letter/letters across all soils, treatments, and timecombinations are not significant at P < 0.05.

Page 7: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8223 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

In case of theUlysses soil, prior to cropP removal (T0), the orderof Pmax was TSP >CM2>CM1>HM>SS >TLwhile TNBPwasHM > SS > CM2 > CM1 > TL > TSP in decreasing order,respectively (Table 6). Therewere a number of discrepancies in theUlysses soil; TSP had the largest value of Pmax atT0 but this was notsupported by TNBP results. In reality, TSP might undergoreactions with soil constituents and contributed the least amountsof TNBP in the Ulysses soil. This was also evident in NBP1, whereHM, CM2, and SS had contributed significantly more P than theTSP treatment (Table 4). The HM and SS treatments werereleasing significantly lesser amounts than both cattle manures inthis method but in reality, HM released significantly more thanboth cattlemanures while SS also releasedmore but was statisticallyat par with both cattle manures. Despite the discrepancies, thismethod also shared common trends with the TNBP data for someP sources (Table 3). For example, TL released identical amounts inboth soils that also agreed well with the TNBP data though 1/3 lessthan the latter one; second, SS released more P in Ulysses soil thanin Eram-Lebo, and finally, significantly more P released for allanimal P sources in the Ulysses soil than in Eram-Lebo soil. Therelationship between Pmax at T0 and TNBP using both cattlemanures along with TL and SS had a tight fit when both soilswere plotted together, and 78% of the variation in Pmax wasexplained by TNBP (Figure 1D) with slopes very close to unity.Moreover, rate of change was close to unity for both soils. There-fore, calculating Pmax by using FeO-method could prove useful inpredicting TNBP from cattle manures and broiler litter, as well asfrom SS in some soils. The Ulysses soil released P faster (larger kvalues) while all P sources had identical values of k in the Eram-Lebo soil but this parameter seemed to be of limited use based onthe study of using just two soils (Tables S3 and S4, SupportingInformation).In concluding remarks, we saw not only significant different

amounts of TNBP in manures from monogastric and ruminantanimals but significant differences were found within each group ofmanures as well, thus contradicting the conventional wisdom thatassumes that P in allmanures behaves identically. Among the animalP sources, HM had the highest amounts of TNBP and TL had thelowest across all soils and levels of P application. Both cattlemanures and SS contributed intermediate amounts of TNBP inmost soils. This order in the amounts of TNBP from animalmanures was also supported by sum of the P fractions extractedwith mild extractants (0.01 M CaCl2 and 0.5 M NaHCO3). Mostanimal manure treatments had TNBP that was comparable to TSP.In fact, those soils (Ulysses and Harney) where TSP was contribut-ing significantly lesser amounts toward TNBP, most animal man-ures had significantly more amounts of TNBP. More than 50% ofthe total P in theHMwas approximatelymaking up part ofNBP1 inmost soils, while TL had significantly lesser amount during the firstharvest. The Crete soil with its highest crop P uptake in the controltreatment also had the lowest values of NBP1 with all P sources andat both levels of P application. The FeO-P method proved useful inpredicting TNBP from P sources such as cattle manures, TL, andSS. We believe findings from this study will add new informationregarding better utilization of P in animal manures from theperspective of both crop production and protecting environment.

’ASSOCIATED CONTENT

bS Supporting Information. Tables S1�S25 and FiguresF1 and F2. This information is available free of charge via theInternet at http://pubs.acs.org/.

’AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected] or [email protected]; phone:(92) 992-383591; fax: (92) 992-383441.

’ACKNOWLEDGMENT

Partial support of this project was provided by Ministry ofEducation, Government of Pakistan and Department of Agron-omy, Kansas State University, USA. I am extremely thankful toMr. Rustin Kimmel, Kurt Zimmer, and my wife Tahira Khatoonin collection and grinding of soil and plant samples.

’REFERENCES

(1) Toth, J. D.; Dou, Z.; Ferguson, J. D.; Galligan, D. T.; Ramberg,C. F., Jr. Nitrogen- vs. phosphorus-based dairy manure applications tofield crops. J. Environ Qual. 2006, 35, 2302–2312.

(2) U.S. Environmental Protection Agency. 2000 National WaterQuality Inventory; EPA 841-R-02-001; Office of Water, U.S. EPA:Washington, DC, 2002 (http://www.epa.gov/OWOW/305b)

(3) Sims, J. T.; Edwards, A. C.; Schoumans, O. F.; Simard, R. R.Integrating soil phosphorus testing into environmentally-based agricul-tural management practices. J. Environ. Qual. 2000, 29, 60–71.

(4) Ketterings, Q. M.; Kahabka, J. E.; Reid, W. S. Trends inPhosphorus fertility of New York agricultural lands. J. Soil Water Conserv.2005, 60, 10–20.

(5) Carpenter, S. R.; Caraco, N. F.; Correll, D. L.; Howarth, R. W.;Sharpley, A. N.; Smith, V. H. Nonpoint pollution of surface waters withphosphorus and nitrogen. Ecol. Appl. 1998, 8, 559–568.

(6) Barnett, G. M. Phosphorus forms in animal manure. Bioresour.Technol. 1994, 49, 139–147.

(7) Hinedi, Z. R.; Chang, A. C.; Lee, R. W. Mineralization ofphosphorus in sludge amended soils monitored by phosphorus �31nuclear magnetic resonance spectroscopy. Soil Sci. Soc. Am. J. 1988,52, 1593–1596.

(8) Hristov, A. N.; Hanzen, W.; Ellsworth, J. W. Efficiency of the useof nitrogen, phosphorus and potassium and potential for reducingphosphorus imports on Idaho dairy farms. J. Dairy Sci. 2006, 89,3702–3712.

(9) Maguire, R. O.; Sims, J. T.; McGrath, J. M.; Angel, C. R. Effect ofPhytase and Vitamin D Metabolite (25Oh-D3) in Turkey Diets onPhosphorus Solubility in Manure-Amended Soils. Soil Sci. 2003, 168,421–433.

(10) Poulsen, H. D. Phosphorus utilization and excretion in pigproduction. J. Environ. Qual. 2000, 29, 24–27.

(11) Cooperband, L. R.; Good, L. W. Biogenic phosphate mineralsin manure: Implications for phosphorus loss to surface waters. Environ.Sci. Technol. 2002, 36, 5075–5082.

(12) Turner, B. L.; Leytem, A. B. Phosphorus compounds insequential extracts of animal manures: Chemical speciation and a novelfractionation procedure. Environ. Sci. Technol. 2004, 38, 6101–6108.

(13) Shafqat, M. N.; Pierzynski, G. M.; Xia, K. Phosphorus sourceeffects on soil organic phosphorus: A P31 NMR study. Commun. Soil Sci.Plant Anal. 2009, 40, 1722–1746.

(14) Maguire, R. O.; Sims, J. T.; Cole, F. J. Phosphorus solubility inbiosolids-amended farm soils in the Mid-Atlantic region of the USA.J. Environ. Qual. 2000, 29, 1225–1233.

(15) McCoy, J. L.; Sikora, L. J.; Weil, R. R. Plant availability ofphosphorus in sewage compost. J. Environ. Qual. 1986, 15, 404–409.

(16) Moore, P. A.; Miller, D. M. Decreasing phosphorus solubility inpoultry litter with aluminum, calcium, and iron amendments. J. Environ.Qual. 1994, 23, 325–330.

(17) Maguire, R. O.; Mullins, G. L.; Brosius, M. Evaluating long termnitrogen versus phosphorus based nutrient management of poultrylitter. J. Environ. Qual. 2008, 37, 1810–1816.

Page 8: Bioavailable Phosphorus in Animal Waste Amended Soils: Using Actual Crop Uptake and P Mass Balance Approach

8224 dx.doi.org/10.1021/es200517y |Environ. Sci. Technol. 2011, 45, 8217–8224

Environmental Science & Technology ARTICLE

(18) Van der Zee, S.; Fokkink, L.; Van Riemsduk, W. H. A newtechnique for assessment of reversibly adsorbed phosphate. Soil Sci. Soc.Am. J. 1987, 51, 599–604.(19) Myers, R. G.; Pierzynski, G. M.; Thien, S. J. Iron oxide sink

method for extracting soil phosphorus: Paper preparation and use. SoilSci. Soc. Am. J. 1997, 61, 1400–1407.(20) Raven, K. P.; Hossner, L. R. Soil phosphorus desorption

kinetics and its relationship with plant growth. Soil Sci. Soc. Am. J.1994, 58, 416–423.(21) Lookman, R.; Freese, D.; Merckx, R.; Vlassak, K.; Van Riemsdijk,

W. H. Long-term kinetics of phosphate release from soil. Environ. Sci.Technol. 1995, 29, 1559–1575.(22) Maguire, R. O.; Sims, J. T.; Foy, R. H. Long-term kinetics of

phosphorus sorption and desorption by high phosphorus soils fromIreland and the Delmarwa peninsula, USA. Soil Sci. 2001, 166 (8),557–565.(23) Murphy, J.; Riley, J. P. Modified single solution method for the

determination of phosphorus in natural waters. Anal. Chem. Acta. 1962,27, 31–36.(24) SAS Institute Inc. SAS Software: Changes and Enhancements

through Release 9.12; SAS Institute: Cary, NC, 2009.