The Effect of Quicklime (CaO) on Litter Condition and Broiler Performance

V. Ruiz,* D. Ruiz,* A. G. Gernat,*1 J. L. Grimes,† J. G. Murillo,* M. J. Wineland,†K. E. Anderson,† and R. O. Maguire‡

*Escuela Agricola Panamericana, Zamorano, PO Box 93, Tegucigalpa, Honduras; †Department of Poultry Science,North Carolina State University, Raleigh 27695-7608; and ‡Department of Crop and Soil Environmental Science,

Virginia Polytechnic Institute and State University, Blacksburg 24061-0002

ABSTRACT High levels of phosphorus and pathogensin runoff are 2 major concerns following manure applica-tions to fields. Phosphorus losses from fields followingmanure applications have been linked to the solubility ofphosphorus in manure; therefore, by decreasing manurephosphorus solubility, a decrease in phosphorus loss inrunoff should be apparent. The objective of this researchwas to develop a process using quicklime that wouldresult in reduced phosphorus solubility and bacteriacounts in broiler litter. The 4 litter treatments evaluatedwere T1, new wood shavings without the addition ofquicklime; T2, used, untreated broiler litter; T3, used litterwith 10% quicklime (based on the weight of the litter);and T4, used litter with 15% quicklime (based on theweight of the litter). Body weight, cumulative feed con-sumption, and feed conversion (feed:BW) were deter-mined on a weekly basis through 42 d of age. Mortalitywas recorded daily. Carcass weights and percentages ofcarcass yield without giblets were determined prechill.Litter pH, total phosphorus, nitrogen, soluble phospho-rus, litter moisture (%), and total plate counts were mea-

Key words: broiler, litter treatment, quicklime, phosphorus, litter pathogen

2008 Poultry Science 87:823–827doi:10.3382/ps.2007-00101

INTRODUCTION

Changes in state and federal laws in the United Stateshave restricted the amounts and times when poultry ma-nure can be land applied based on N and P standards.Because of these new standards, waste management anddisposal are among the most critical issues confronting USconfined animal feeding operations. Animal industriestoday are faced with a number of issues because of contin-uous growth, increased size, and concentration of opera-tions. For example, land application of poultry manurecan pose a threat to ground and surface water quality ifapplied in areas at risk for storm runoff or if applied above

©2008 Poultry Science Association Inc.Received February 28, 2007.Accepted February 14, 2008.1Corresponding author: agernat@zamorano.edu

823

sured for each litter treatment on d 7 and 42 after birdplacement. No significant differences were found for BW,feed consumption, feed conversion, mortality, carcassweight, or carcass yield. No breast or footpad blisterswere observed. On d 7, 15% quicklime had higher (P <0.001) pH (11.2) when compared with the other treat-ments. Percentages of phosphorus and nitrogen werelower (P < 0.001) for new wood shavings in comparisonwith the used litter treatments. Soluble phosphorus (ppm)was lower (P < 0.001) for 15% quicklime (2.75) whencompared with new wood shavings (42.2), untreatedbroiler litter (439.2), and 10% quicklime (35.0). Althoughnot significant, 15% quicklime had lower total platecounts (cfu/g) in comparison with the other treatmentson d 1 and 10 postmixing and at 7 d after bird placement.Litter conditions on d 42 after bird placement were simi-lar. We concluded that the use of quicklime as a treatmentfor broiler litter would initially reduce nitrogen and solu-ble phosphorus and bacteria counts without negativelyaffecting bird productivity.

established agronomic rates. The potential presence ofpathogens in manure is another concern. The use of lime[CaO and Ca(OH)2] to kill pathogens in biosolids (sewagesludge) is a well-established process. In the United States,the national Environmental Protection Agency Part 503rule regulates the land application of biosolids and re-quires treatment to reduce pathogens before land applica-tion is permitted (US Environmental Protection Agency,1999). When lime is the option used for pathogen reduc-tion, the rule requires that sufficient lime is added to raisethe pH to 12 for 2 h to kill pathogens (US EnvironmentalProtection Agency, 1999). Another effect that can be ob-served with the addition of quicklime (QL) and increaseof pH in the litter is the liberation of ammonia gas. How-ever, before this type of management practice can be putinto widespread use, questions concerning the environ-mental impact of this type of chemical amendment andits safety in broilers on commercial farms must be ad-

at National C

hung Hsing U

niversity Library on A

pril 12, 2014http://ps.oxfordjournals.org/

Dow

nloaded from

RUIZ ET AL.824

Table 1. Composition and calculated nutrient content of the experimen-tal diets

Starter Grower FinisherItem (0 to 21 d) (21 to 35 d) (35 to 42 d)

Ingredient (%, as fed)Corn 51.50 57.41 62.60Soybean meal, dehulled 39.31 33.90 28.70Ground limestone 1.76 1.64 1.57Dicalcium phosphate 1.48 1.38 1.05Salt (NaCl) 0.39 0.39 0.39Vitamin + mineral premix1 0.25 0.25 0.25Oxitet T-412 0.01 0.01 0.01Sacox3 0.05 0.05 0.05Vegetable oil 5.10 4.83 5.26DL-Met 0.16 0.15 0.12

Calculated analysesCP 22.00 20.00 18.00ME (kcal/kg) 3,080 3,131 3,215Calcium 1.07 1.00 0.90Available P 0.46 0.43 0.35Met + cystine 0.91 0.85 0.76Lys 1.30 1.16 1.01Thr 0.93 0.84 0.75Trp 0.31 0.28 0.24

1The vitamin and mineral premix provided the following quantitiesper kilogram of diet: vitamin A, 10,000 IU (all-trans-retinal); vitamin D,2,500 IU (cholecalciferol); vitamin E, 10 IU (DL-α-tocopheryl); vitaminK, 2 mg; riboflavin, 5 mg; niacin, 35 mg; D-calcium pantothenic acid,10 mg; choline chloride, 250 mg; vitamin B12, 12 mg; folic acid, 0.75 mg;manganese, 70 mg; zinc, 50 mg; iron, 30 mg; copper, 10 mg; iodine, 1.5mg; cobalt, 0.15 mg; selenium, 0.10 mg; mold inhibitor, 7 mg; antioxidant,10 mg.

2Oxitet T-41, broad spectrum, 89 g of oxytetracycline/kg (Hoffmann-LaRoche Ltd., Basel, Switzerland).

3Sacox (for prevention of coccidiosis in broilers), 150 g of lasalocidsodium/kg (Intervet, Millsboro, DE).

dressed (Do et al., 2005). There is an urgent need forinnovative methods of collecting, processing, and dispos-ing of manure, mortalities, and by-products such thatthe environmental impact is minimized. In a study thatinvestigated manures, but not poultry performance, Ma-guire et al. (2006) showed that liming of broiler litters andlayer manures could greatly decrease bacterial counts.

We found no data in the scientific literature on poultryperformance and manure condition that directly relate tothe use of QL for manure treatment purposes for thepoultry industry. Quicklime is a highly reactive productthat reacts with water to produce heat and hydrated lime[HL, Ca(OH)2; Budavari, 1996). However, for centuriesHL was used widely as a sanitizing agent to control cer-tain bacterial pathogens and parasites, and for the chemi-

Table 2. Effect of quicklime on broiler litter condition d 1 after mixing1

Variable T1 T2 T3 T4

pH 5.50 8.57 11.85 12.00Moisture (%) 13.45 13.80 36.80 42.80Phosphorus (%) 0.02 0.64 0.75 0.60Nitrogen (%) 0.08 1.14 1.20 0.99Soluble P (ppm) 8.00 429.00 19.50 12.50TPC (cfu/g) 1,400,000 1,600,000 6,000 1,000

1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10%quicklime (based on the weight of the litter); T4 = 15% quicklime (basedon the weight of the litter); TPC = total plate count (aerobic bacteria).

cal treatment of industrial and municipal sewage beforebiological treatment was developed. In the 1800s, farmersknew the benefits of using animal manure, soils, andburnt limestone (QL) to enhance crop yields (Langenbeck,1917; White, 1947). Yushok and Bear (1948) claimed re-markable effects of HL on several poultry pathogenswhen added to poultry manure. Bennett et al. (2003) eval-uated the effect of added HL at the levels of 5, 10, and20% on survival of Salmonella Enteritidis in used broilerlitter and found a significantly reduced Salmonella recov-ery incidence at 24 h. Incorporating HL at levels of 0.2,1, and 5% in turkey poult litter resulted in no reductionof Campylobacter or Salmonella recovery, but a reductionin overall aerobic colony-forming units was seen (Bennettet al., 2005). The objective of this study was to develop aprocess using QL to treat poultry litter so it could bereused as a bedding material and have physical and chem-ical characteristics acceptable for land application.

MATERIALS AND METHODS

Litter Treatment

Before placing birds in the house, wood shavings andused broiler litter were obtained from a local broiler oper-ation (used by 2 previous broiler flocks). Quicklime wasobtained from a local quarry outside Tegucigalpa. The 4treatments were as follows: fresh wood shavings (T1),used broiler litter (T2), used broiler litter mixed with10% QL (calculated on a weight-per-basis; T3), and usedbroiler litter mixed with 15% QL (calculated on a weight-per-basis; T4). Percentage moisture was determined(AOAC, 1990) for the used broiler litter before establish-ing the 4 treatments. Once the percentage moisture wasdetermined, T1 had no water added. Water was addedto T2 to reach 30% moisture in the litter. For T3 and T4,water was added to reach a level of 85% moisture. Beforeadding the water to T3 and T4, the QL was mixed in thelitter manually with shovels. The 4 treatments were leftin individual piles and turned every other day for 10 d.

Eight subsamples were randomly collected from eachlitter treatment (pile) and thoroughly mixed to obtain 1kg of sample for total plate counts (TPC) for aerobicbacteria. For aerobic plate counts, 50 mL of PBS was addedto each 10-g sample. Samples were then serially dilutedand spread-plated onto tryptic soy agar plates. Afterspread plating, tryptic soy agar plates were incubated at

Table 3. Effect of quicklime on broiler litter condition d 10 after mixing1

Variable T1 T2 T3 T4

pH 5.30 8.27 10.00 11.38Moisture (%) 10.75 10.35 31.95 31.50Phosphorus (%) 0.02 1.22 1.03 1.04Nitrogen (%) 0.09 1.18 0.82 0.91Soluble P (ppm) 8.50 433.00 19.50 14.00TPC (cfu/g) 1,000,000 700 100 100

1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10%quicklime (based on the weight of the litter); T4 = 15% quicklime (basedon the weight of the litter); TPC = total plate count (aerobic bacteria).

at National C

hung Hsing U

niversity Library on A

pril 12, 2014http://ps.oxfordjournals.org/

Dow

nloaded from

QUICKLIME EFFECT ON LITTER CONDITION AND BROILER PERFORMANCE 825

Table 4. Effect of quicklime-treated broiler litter on BW, feed consumption, feed conversion, and mortality1

Parameter T1 T2 T3 T4 SEM

BW (g)Day 7 132.6 134.0 135.0 134.4 2.4Day 14 356.3 355.2 357.8 347.6 4.3Day 21 731.2 737.5 739.2 709.9 10.6Day 28 1,243.0 1,257.4 1,265.7 1,227.3 16.3Day 35 1,760.9 1,775.3 1,815.5 1,777.6 28.2Day 42 2,110.1 2,176.2 2,194.2 2,143.0 30.9

Feed consumption (g/bird)Day 7 132.0 129.0 128.3 127.4 1.5Day 14 500.8 483.9 488.0 473.8 6.0Day 21 1,101.0 1,083.7 1,105.9 1,065.9 12.7Day 28 2,030.4 2,027.9 2,050.0 1,994.2 28.3Day 35 3,040.1 3,077.9 3,113.7 3,028.0 39.6Day 42 4,008.1 4,068.2 4,168.6 4,033.4 47.6

Feed conversion2 (g of feed:g of BW)Day 7 0.99 0.96 0.95 0.94 0.01Day 14 1.40 1.36 1.36 1.36 0.01Day 21 1.50 1.46 1.49 1.50 0.02Day 28 1.63 1.61 1.61 1.62 0.01Day 35 1.72 1.73 1.71 1.70 0.01Day 42 1.85 1.86 1.89 1.88 0.01

Mortality (%)Day 7 0.68 0.34 0.34 1.25 0.02Day 14 0.86 0.51 0.69 1.88 0.03Day 21 1.03 0.86 1.20 1.88 0.03Day 28 1.88 1.71 2.05 2.05 0.03Day 35 3.42 2.74 4.11 4.45 0.04Day 42 4.97 5.65 5.99 5.99 0.19

1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10% quicklime (based on the weight of thelitter); T4 = 15% quicklime (based on the weight of the litter).

2Adjusted feed conversion.

37°C for 24 h. Plates were then examined, and total aerobiccolony-forming units were enumerated and recorded (Se-cretaria de Agricultura y Ganaderıa, 2006). The pH (1:1litter per water extract), total P (determined by usingacid digestion with H2SO4 and H2O2 and analyzed byspectrophotometer), N (Kjeldahl method), soluble P (SP;1:10 litter per water extract and analyzed by spectropho-tometer), and percentage moisture were determined at d1, and a second collection was repeated at d 10. Aftercompleting the 10-d waiting period, the 4 litter treatmentswere divided and allocated in a randomized completeblock design to the 16 experimental pens, giving 4 repli-cates for each treatment. The litters were provided at adepth of approximately 12.5 cm over concrete flooring.

Bird Placement

One-day-old broiler male chicks were received from acommercial hatchery and placed in an open-sided natu-

Table 5. Effect of quicklime-treated broiler litter on carcass weight and yield1

Variable T1 T2 T3 T4 SEM

Carcass weight (g) 1,529.2 1,562.7 1,577.5 1,525.4 15.9Carcass yield2 (%) 72.1 71.6 72.3 71.3 0.003

1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10% quicklime (based on the weight of thelitter); T4 = 15% quicklime (based on the weight of the litter).

2Without giblets.

rally ventilated broiler house receiving a daily photope-riod of 24 h of light. Each of the 16 pens (2 × 3 m) housed72 chicks, placed at a density of 12 birds per square meter.Before placing the chicks, the used, untreated and treatedlitters were top-dressed with approximately 2.5 cm offresh wood shavings. Each pen was heated by a gasbrooder and provided with nipple waterers and tubefeeders. Commercial mash diets (Table 1) and water wereprovided ad libitum. Body weight, cumulative feed con-sumption, and feed conversion (feed:BW) were deter-mined by each pen at 7, 14, 21, 28, 35, and 42 d of age.Litter TPC for aerobic bacteria and pH, total P, N, SP,and moisture (%) were measured for each experimentalpen when birds were 7 and 42 d of age (following proce-dures described previously). Mortalities were recordeddaily. Birds’ footpads and breasts were observed on aweekly basis for the presence of blisters. Carcass weightsand carcass yields (%) without giblets were determined

at National C

hung Hsing U

niversity Library on A

pril 12, 2014http://ps.oxfordjournals.org/

Dow

nloaded from

RUIZ ET AL.826

Table 6. Effect of quicklime on broiler litter condition 7 d after bird placement1

Variable T1 T2 T3 T4 SEM

pH 5.60d 7.65c 8.85b 11.12a 0.29Moisture (%) 5.69c 5.80c 18.01b 22.20a 0.02Phosphorus (%) 0.11b 1.13a 1.14a 1.04a 0.002Nitrogen (%) 0.44c 0.89b 0.95b 1.20a 0.003Soluble P (ppm) 137.05b 445.25a 44.00c 16.75c 27.80TPC (cfu/g) 4,433,000 4,333,000 4,860,000 1,814,000 62,000

a–dMeans within rows without a common superscript are different (P < 0.001).1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10% quicklime (based on the weight of the

litter); T4 = 15% quicklime (based on the weight of the litter); TPC = total plate count (aerobic bacteria).

prechill. A second trial was conducted to evaluate thesame treatment following the same procedures.

Statistical Analysis

Data from each trial were evaluated by ANOVA withGLM procedures (SAS Institute, 1991). There was no sig-nificant trial effect (P > 0.05); therefore, the data from the2 trials were pooled. Percentage data were subjected toarcsine square root of the percentage transformation, andtreatment means were separated by the test of least sig-nificant difference. A probability of P ≤ 0 0.05 was re-quired for statements of significance.

RESULTS AND DISCUSSION

Litter Condition Postmixing

Treating used litter with QL substantially increased pHon d 1 after mixing for the 10 and 15% addition of QL,and that increase was maintained 10 d postmixing (notstatistically analyzed; Tables 2 and 3). A decrease in solu-ble P was also observed. Maguire et al. (2006) showedthat adding at least 10% QL to broiler litter and layermanure decreased soluble P by >90%. Liming has beenshown to reduce the solubility of P in biosolids, probablybecause of the formation of calcium phosphates (Maguireet al., 2001; Penn and Sims, 2002). A similar decrease wasalso observed for TPC in the same treatments. The mixingof QL with the additional moisture added to the littercaused an exothermic reaction, increasing litter tempera-ture to 65°C and elevating pH (Table 2). Yushok and Bear(1948) reported similar effects of HL on several poultry

Table 7. Effect of quicklime on broiler litter condition 42 d after bird placement1

Variable T1 T2 T3 T4 SEM

pH 7.65c 8.12b 8.38ab 8.75a 0.10Moisture (%) 13.92 11.75 12.31 12.96 0.01Phosphorus (%) 0.95c 1.27a 1.17ab 1.14b 0.001Nitrogen (%) 3.01a 2.13b 1.92b 1.85b 0.005Soluble P (ppm) 584.00a 438.75 84.40b 35.00b 82.77TPC (cfu/g) TNTC TNTC TNTC TNTC

a–cMeans within rows without a common superscript are different (P < 0.01).1T1 = wood shavings; T2 = used, untreated broiler litter; T3 = 10% quicklime (based on the weight of the

litter); T4 = 15% quicklime (based on the weight of the litter); TPC = total plate count (aerobic bacteria); TNTC =too numerous to count.

pathogens when added to poultry manure, although theactual data provided were not sufficient for independentevaluation. Maguire et al. (2006) showed that adding QLto broiler litter and layer manure could decrease bacterialcounts by >99%, depending on moisture and QL rate. Adecrease in litter moisture was observed from d 1 to 10and was most probably caused by the heating processoccurring in the litter through composting and chemicalreaction with QL and additional moisture that was addedto the litter. A slight decrease in N content was observedwith the treated litters because of the volatilization ofammonia. An increase in total P was also observed, whichwas probably due to the decreasing moisture level oc-curring from d 1 to 10 (Tables 2 and 3). Similar decreasesthat were observed for TPC 1 d after mixing were alsoobserved for TPC in the same treatments 10 d after mixing(Table 3).

Bird Performance

There were no significant differences in BW, feed con-sumption, feed conversion, mortality, carcass weight, orcarcass yield by treatment (Tables 4 and 5), nor wereblisters observed on the breasts or footpads (data notshown). Contrary to preliminary early poult performancetrials conducted by Bennett et al. (2005), concentrationsof lime greater than 5% added to the litter resulted inocular and respiratory irritation during the first 48 h fol-lowing placement. However, in his earlier study in 2003,concentrations of 0.2, 1, and 5% lime improved poultperformance, apparently associated with the lime treat-ments. This suggests that the treatment of poultry littermay have beneficial effects on growth during the brood-

at National C

hung Hsing U

niversity Library on A

pril 12, 2014http://ps.oxfordjournals.org/

Dow

nloaded from

QUICKLIME EFFECT ON LITTER CONDITION AND BROILER PERFORMANCE 827

ing phase. The inconsistent lime-associated changes inbacterial recovery from the environment in this studywould not explain an improvement in poult performance;however, Bennett et al. (2005) suggested it was possiblethat the lime treatment may have affected a populationof opportunistic pathogens not evaluated in the study. Inour present study, no ocular or respiratory abnormalitieswere observed.

Litter Treatment After Bird Placement

The effect of the liming was still evident 7 d after place-ment (Table 6). Treatment 4 had a significantly (P < 0.001)higher pH (11.12) and percentage moisture (22.2%) whencompared with the other treatments (Table 6). Percent-ages of P and N were significantly (P < 0.001) lower forT1 in comparison with the used litter treatments. Thiswas because the used litter maintained its residual N andP from previous flocks as compared with fresh materialthat had never been used. Soluble P (ppm) was lower (P< 0.001) for T4 (16.7) when compared with T1 (137.0), T2(445.2), and T3 (44.0). This was caused by the effect ofliming in reducing SP in biosolids (Maguire et al., 2001;Penn and Sims, 2002). Treatment effects on litter condition42 d after bird placement (Table 7) showed significantdifferences. Litter pH was higher (P < 0.005) for T3 andT4, with pH of 8.38 and 8.75, respectively. No significantdifferences were observed for percentage moistureamong treatments. Percentage P continued to remainlower (P < 0.01) for T1. Nitrogen content (%) was higher(P < 0.001) for T1 (3.01) than for T2 (2.13), T3 (1.92), andT4 (1.85). Soluble P (ppm) remained lower (P < 0.01) forT4 (35.0) and inclusively for T3 (84.4) when comparedwith T1 (584.0) and T2 (438.7). No significant differenceswere found for TPC among treatments (Tables 6 and 7).As expected, the litters had an increase in bacterial countsafter the birds were placed.

We demonstrated in this study that the use of CaO asa treatment for broiler litter initially reduced N and solu-ble P. Placing birds on CaO-treated litter at levels of 10and 15% (based on the weight of the litter) did not nega-tively affect bird performance or carcass characteristics.As regulations are developed to reduce P losses fromlitter-amended soils, the development of processes to re-duce the solubility of P will become crucial for the eco-

nomic survival of intensive poultry production. Our re-sults suggest that lime treatment can successfully de-crease P solubility in broiler litter. Overall, the limingprocess was able to reduce SP in the litter by more than90%, which should greatly reduce concerns about P lossesin runoff following land application of these materials.

REFERENCES

AOAC. 1990. Official Methods of Analysis. 15th ed. Assoc. Off.Anal. Chem., Alexandria, VA.

Bennett, D. D., S. E. Higgins, R. W. Moore, R. Beltran, D. J.Caldwell, J. A. Byrd II, and B. M. Hargis. 2003. Effects oflime on Salmonella enteritidis in vitro. J. Appl. Poult. Res.12:65–68.

Bennett, D. D., S. E. Higgins, R. W. Moore, J. A. Byrd II, R.Beltran, C. Corsiglia, D. J. Caldwell, and B. M. Hargis. 2005.Effects of addition of hydrated lime to litter on recovery ofselected bacteria and poult performance. J. Appl. Poult. Res.14:721–727.

Budavari, S. 1996. Page 1735 in the Merck Index. 12th ed. Merckand Co. Inc., Whitehouse Station, NJ.

Do, J. C., I. H. Choi, and K. H. Nahm. 2005. Effects of chemicallyamended litter on broiler performance, atmospheric ammo-nia concentration, and phosphorus solubility in litter. Poult.Sci. 84:679–686.

Langenbeck, K. 1917. The practical advantage of burned limeover ground limestone. The Agricultural Lime Bureau,Washington, DC.

Maguire, R. O., D. Hesterberg, A. G. Gernat, K. Anderson, M.Wineland, and J. Grimes. 2006. Liming poultry manures todecrease soluble phosphorus and suppress the bacteria popu-lation. J. Environ. Qual. 35:849–857.

Maguire, R. O., J. T. Sims, S. K. Dentel, F. J. Coale, and J. T.Mah. 2001. Relationship between biosolids treatment processand soil phosphorus availability. J. Environ. Qual.30:1023–1033.

Penn, C. J., and J. T. Sims. 2002. Phosphorus forms in biosolids-amended soils and losses in runoff: Effects of wastewatertreatment process. J. Environ. Qual. 31:1349–1361.

SAS Institute. 1991. SAS User’s Guide: Statistics. Version 5 ed.SAS Inst. Inc., Cary, NC.

Secretaria de Agricultura y Ganaderıa. 2006. Secretaria de Agri-cultura y Ganaderıa, Direccion de Ciencia y Tecnologıa, Co-lonia Lomas Linda Norte, Tegucigalpa, Honduras.

US Environmental Protection Agency. 1999. Biosolids genera-tion, use, and disposal in the United States. US EPA, Officeof Wastewater Management, Washington, DC.

White, J. W. 1947. The use of burned lime products in soilimprovement. Pit Quarry (May).

Yushok, W., and F. E. Bear. 1948. Poultry manure: Its preserva-tion, deodorization, and disinfection. N. J. Agric. Exp. Stn.Bull. 707:3–11.

at National C

hung Hsing U

niversity Library on A

pril 12, 2014http://ps.oxfordjournals.org/

Dow

nloaded from