Bf

VU

a

ARR1AA

KHSZCRSV

1

tgdadrtn

s

gaapnate

w

0d

Industrial Crops and Products 36 (2012) 555–559

Contents lists available at SciVerse ScienceDirect

Industrial Crops and Products

journa l homepage: www.e lsev ier .com/ locate / indcrop

iomass, extracted liquid yields, sugar content or seed yields of biofueleedstocks as affected by fertilizer�

.M. Russo ∗, W.W. Fish1

nited States Department of Agriculture, Agricultural Research Service, Wes Watkins Agricultural Research Laboratory, 911 Hwy. 3W, POB 159, Lane, OK 74555, United States

r t i c l e i n f o

rticle history:eceived 13 October 2011eceived in revised form6 November 2011ccepted 18 November 2011vailable online 14 December 2011

eywords:

a b s t r a c t

Harvesting products from plants for conversion into renewable resources is increasing in importance.Determination of nutrition requirements for the applicable crops is necessary, especially in regionswhere the biofuel feedstock crops have not been historically grown. Sunflower (Helianthus annuus L.), twohybrids and one variety; sweet and grain (milo) sorghums (both Sorghum bicolor L.), one variety each, andsweet corn (Zea mays var. rugosa Bonaf.), four cultivars, were provided the recommended and twice therecommended rate of fertilizer. Biomass, expressed liquid volumes and sugar contents of sweet sorghumand sweet corn were determined. Grain yields of milo and sunflower and oil content of sunflower were

elianthusorghumeaultivarateweet corn

determined. Sweet corn stalk sugar levels were below what is expected from field corn (maize), and werenot affected by fertilizer rate. Sweet sorghum biomass and sugar content were within expected rangesand not affected by fertilizer rate. Milo grain yields were higher with increased fertilizer. Seed yield inSunflower, which was below expected levels, was inconsistently affected by fertilizer rate, years or vari-eties. Overall crops year and cultivar/variety had more effect on results than did fertilizer. There does notappear to be a reason to provide fertilizer above recommended rates in production of these crops.

ariety

. Introduction

Biofuel feedstock production is increasing and crops are some-imes established in areas where they historically have not beenrown. Nutrition is one of the most important inputs of a crop pro-uction system. Fertilizer level affects establishment in any cropnd is of importance in annual crops that require seed be placed inirect contact with soil. It is important that fertilizer levels be cor-ect for early and continued plant development. It is also importanthat fertilizer requirements be determined for crops established in

ew production areas.

Sunflower (Helianthus annuus L.) and sweet sorghum and grainorghum, milo, (both Sorghum bicolor L.) are row crops considered

� Mention of a trademark, vendor, or proprietary product does not constitute auarantee or warranty of the product by the U.S. Department of Agriculture (USDA)nd does not imply its approval to the exclusion of other products that may be suit-ble. USDA employees prepared the article as part of their official duties. Copyrightrotection under U.S. copyright law is not available for such works, and there iso copyright to transfer. The fact that the private publication in which the articleppears is itself copyrighted does not affect the material that is a work product ofhe U.S. government, which can be freely reproduced by the public. The USDA is anqual opportunity employer.∗ Corresponding author. Tel.: +1 580 889 7395; fax: +1 580 889 5783.

E-mail addresses: vincent.russo@ars.usda.gov (V.M. Russo),ayne.fish@ars.usda.gov (W.W. Fish).1 Tel.: +1 580 889 7395; fax: +1 580 889 5783.

926-6690/$ – see front matter. Published by Elsevier B.V.oi:10.1016/j.indcrop.2011.11.019

Published by Elsevier B.V.

to have use as biofuel feedstocks which respond to fertilizer levelin traditional production regions (Dahnke et al., 1992; Whitney,1998; Almondares et al., 2009). Sweet sorghum is unique as a bio-fuel feedstock in that it contains simple sugars through much of itsdevelopment that can be readily converted to ethanol through fer-mentation. In addition the remaining biomass present after stalksare crushed has the potential to be treated so complex sugars andother carbohydrates can be converted to ethanol when the technol-ogy becomes available on a commercial scale. Milo has the potentialas a biofuel feedstock due to the starch stored in seed. The seed pro-vide a biological package that allows the transport and long-termstorage of material that can be converted to bioethanol.

Another possible candidate as a biofuel feedstock is sweet corn(Zea mays var. rugosa Bonaf.) grown for harvest of stalks, beforeears are formed, from which sugars can be extracted. Fertilizerrequirements for sweet corn harvested for the ears are understood(McCraw et al., 1987), but it is not clear if these requirements affectsugar content in stalks harvested for extracted sugars. Russo et al.(1998) determined that about 85 mg mL−1 of sugars was foundin stalks of a “supersweet” cultivar at the silking (R1) stage asdescribed by Ritchie and Hanway (1982), and values changed withplant development and senescence stage. Presumably senescenceaffects levels of simple and more complex sugars in other mono-

cots including sweet sorghum. However, unlike for field and sweetcorn senescence patterns in sorghum have not been extensivelystudied.

556 V.M. Russo, W.W. Fish / Industrial Crops and Products 36 (2012) 555–559

Table 1Typical analysis for pH, nutrientsa on a dry basis, and electrical conductivity (EC) of chicken litter manure used.

pH C N P2O5 K Ca Cu Fe Mg Mn Na S Zn EC

8.2 679.3 51.4 31.9 52.2 113.9 0.1 52.9 18.8 1.2 3.1 13.1 0.3 2487

, analL

rCaewcfufac

2

ltetnfsvaowtls

sPcbnKstara(

tseMiss2vA

wtc

a C, N, P2O5, K, Ca, Cu, Fe, Mg, Mn, Na, S, and Zn are in kg Mg−1 and EC is in �Saboratory, Stillwater, OK.

In some instances the crops may be grown where manureesources, which can be used to supply nutrition, are available.rops use soil nutrients at various efficiencies. Crop residues canlso supply nutrients when returned to the soil. Amount of nutri-nts supplied can affect both activities. It is necessary to determinehether fertilizer recommendations are sufficient for candidate

rops so that biofuel feedstock production can be maximized andertilizer assets used in an economical manner. This research wasndertaken to clarify fertilizer requirements of traditional biofueleedstock crops where the crops have not been traditionally grownnd for crops that may potentially be used as biofuel feedstockrops.

. Materials and methods

Experiments were conducted on a Bernow (fine-loamy, sil-iceous, thermic, Glossic Paleudalf) soil at Lane, Okla., from late-Mayo mid-August 2009 and late-April to early-August 2010. Differ-nces in planting dates were dictated by rainfall occurring prioro planting in 2009. The experiments were conducted on threeon-contiguous fields of the same soil type about 300 m apart

rom each other. One field contained Sunflower (previous cropquash, Cucurbita spp.); the second field contained sweet corn (pre-ious crop Mustard, Brassica spp.) and a third field contained sweetnd grain sorghum (milo), which followed sorghum. In all fieldsverwintering wheat (Triticum aestivum L. subsp. aestivum), whichas established between the previous crops and the experimen-

al crops, was mowed in mid-March of each year and the residueeft on the soil for two weeks before being incorporated into theoil.

Fertilizer rates were determined based on soil test results andtate or regional recommendations (McCraw et al., 1987; Izekor andorter, 2004; Zhang et al., 2009). Synthetic fertilizers were broad-ast applied to soil in which sunflower and sweet corn were toe sown preplant at 95 or 190 kg ha−1 of synthetic N from ammo-ium nitrate. All P (80 kg ha−1) from single super-phosphate and(185 kg ha−1) from muriate of potash, were applied preplant. To

oil supporting sweet and grain sorghum two rates of chicken lit-er (3.36 or 6.72 Mg ha−1; nutrient content Table 1) were broadcastpplied preplant. In all fields the fertilizer material was incorpo-ated using a Liliston rolling cultivator (Bush Hog Corp., Selma, Ala.)nd final, flat, seedbeds formed using a multi-facet finishing toolDo-All, Forrest City, AR).

All crops were established from seed in rows on 0.9 m cen-ers. Plot sizes for crops were: sweet corn (each cv.), 4 m × 80 m;orghum (sweet and grain), 9 m × 106 m, and sunflower (each vari-ty), 8 m × 76 m. Sweet corn cvs. Florida Staysweet, Incredible, anderit, and the forage sweet corn, cv. TC1101, were sown with an

n-row spacing of 20 cm on 21 May 2009 and 26 April 2010. Grainorghum, a field run mix, and the sweet sorghum, var. Della, wereown with an in-row spacing of 7 cm on 20 May 2009 and 27 April010. Sunflower varieties 657, 820OH, and a Peredovik (high oil)ariety were sown 16 cm between plants on 22 May 2009 and 22pril 2010.

For sunflower and sweet corn the herbicide Dual (1.2 L ha−1)as applied preplant in accordance with label directions. No syn-

hetic herbicides were applied to sweet and grain sorghum. Sweetorn received 5 cm per week of water through precipitation and

ysis performed by the Oklahoma State University Soil, Water & Forage Analytical

irrigation; sweet and grain sorghum were irrigated based on aminimum soil moisture reading of 35–40 kPa obtained with a soilmoisture meter (model TEMP-200, Aquaterra, Fremont, CA). Sun-flower was generally rain-fed with a single irrigation occurringonce on 3 June 2010, the need for which was surmised from plantappearance.

Samples of sweet corn and sweet sorghum were harvested forbiomass from 3.5 linear meters in a row from each replication.Sweet corn was harvested on 5 August 2009 and 21 July 2010;sweet sorghum was harvested on 18 August 2009 and 4 August2010. Stalks were weighed and passed through a press manufac-tured at this station to express liquids. The volume and weight ofexpressed liquid was determined. Subsamples of expressed liquidswere processed for quantification of sugars by high performanceliquid chromatography (HPLC) by first centrifuging the liquid at16,000 × g for 15 min at room temperature, diluting the liquid withdistilled H2O to the desired level, and filtering the diluted solu-tion through a 0.45 �m cellulose filter. HPLC was carried out on aVarian ProStar (Vista, CA) ternary solvent system equipped withan autosampler and RI detector. Quantitative sugar profiles of thediluted subsamples were obtained with a 250 mm × 4 mm aminocolumn (5 �m, LunaTM, Phenomenex, Torrance, CA). Glucose, fruc-tose, and sucrose, were eluted with an isocratic system of 80%acetonitrile/20% H2O at a flow rate of 1 mL min−1 and a columntemperature of 35 ◦C.

Grain sorghum and sunflower were harvested for seed yieldwith a Gleaner combine (Agro Corporation, Duluth, GA) over theentire plot excluding a guard row on either side. The grain sorghumwas harvested on 24 August 2009 and 9 August 2010, and the sun-flower on 21 September 2009 and 12 August 2010 to determineresponse to fertilizer level as it affected seed yield. Sunflower seedwere fed into a screw press (model CLB-300, Cropland Biodiesel,Lynden, Wash.) in 2010 using heat at 93 ◦C to extract oil; themachine was not available in 2009.

Due to differences in production practices, plant habit, har-vesting method, product of harvest, and time of harvest therewere no direct comparisons between crops. All experiments werearranged in randomized complete block designs with three repli-cations. The data were subjected to analysis of variance in SAS (ver.9.1, SAS, Inc., Cary, NC). If interactions were significant they wereused to explain results. If interactions were not significant meanswere separated with the Ryan–Einot–Gabriel–Welsch multipleF test.

3. Results

Average minimum and maximum temperatures and precipita-tion were in some instances different for the crops in 2009 and 2010(Table 2). For sweet corn minimum and maximum air tempera-tures were lower in 2010 but precipitation was higher. For sorghumminimum temperature was slightly lower in 2010, but maximumtemperature and precipitation were similar in both years. For milominimum and maximum temperatures were similar in both yearsbut precipitation was slightly higher in 2010. For sunflower mini-

mum temperature was lower in 2010, maximum temperature wassimilar between years and precipitation lower in 2010. Due toweather plants were sown about 3 weeks later and harvested about2 weeks later in 2009 than in 2010.

V.M. Russo, W.W. Fish / Industrial Crops and Products 36 (2012) 555–559 557

Table 2Average minimum and maximum air temperatures and precipitation for the duration of growth of the various crops in 2009 and 2010.

Crop Year Date Average temperature (◦C) Precipitation (cm)

Plant Harvest Min. Max.

Sweet corn 2009 21 May 20 July 18.6 31.4 15.52010 26 April 6 July 17.6 28.7 23.7

Sorghum 2009 20 May 18 August 19.2 31.5 31.82010 27 April 4 August 18.7 31.3 32.8

Milo 2009 20 May 24 August 19.1 31.4 32.12010 27 April 9 August 19.2 31.5 34.5

Sunflower 2009 22 May 21 Septemb2010 22 April 12 August

Table 3ANOVA results for effect of year, cultivar/variety and fertilizer rate on biomass andamount of extracted liquids in sweet corn and sweet sorghum feedstock crops.

Source Biomass Extractedliquid

Percent sugar

Frua Glu Suc Total

Sweet cornYear (Y) ** ** NS NS NS NSCultivar (C) ** ** NS NS NS NSFertilizer rate (F)b NS NS NS NS NS NS

Interactionc

Y × C NS ** NS NS NS NSSweet sorghum

Year (Y) ** * * NS ** **

Fertilizer rate (F)a NS NS ** ** NS NS

NS, non-significant.a Fru = fructose, Glu = glucose and Suc = sucrose.b For sweet corn fertilizer refers to level of N applied; for sorghum fertilizer refers

to level of chicken litter applied.c Shown is the only significant interaction.

3

baya(wvas

TI

NId

3.4. Sunflower

* Significant at P ≤ 0.05, ANOVA.** Significant at P ≤ 0.01, ANOVA.

.1. Sweet corn

Year and cultivar, but not fertilizer rate, affected sweet corniomass and extracted liquids (Table 3). The year × cultivar inter-ction did not affect biomass and extracted liquids. The biomassield over N-rate averaged 25.25 Mt ha−1 and extractable liquidsveraged 531.4 L ha−1. There was more biomass produced in 201028.0 Mt ha−1) than in 2009 (22.5 Mt ha−1). The highest biomassas produced by ‘TC1101’ (29.7 Mt ha−1), and the other culti-

ars averaged 23.8 Mt ha−1. Except for ‘Incredible’, which was not

ffected, there was less extracted liquid produced by the otherweet corn cultivars in 2009 than 2010 (Table 4).

able 4nteraction of year and cultivar/variety on amount of extracted liquids in sweet corn.

Source Extracted liquid (L ha−1)

Cultivar Year

TC1101 2009 517.42010 1170.3**

Florida Staysweet 2009 82.42010 318.9*

Incredible 2009 444.12010 514.3NS

Merit 2009 240.72010 962.7**

S, non-significant.n this type of analysis the first value in the group is not assigned a significanceesignation; value in a combination is compared to the value directly above it.

* Significant at P ≤ 0.05, Least squares means analysis.** Significant at P ≤ 0.01, Least squares means analysis.

er 23.8 30.6 46.418.8 30.8 34.5

Percents of fructose, glucose, sucrose and total sugars were notaffected by treatment (Table 3). Average fructose, glucose, sucroseand total sugars were 2.07%, 2.53%, 0.37% and 4.97%, respectively.

3.2. Sweet sorghum

Year, but not litter rate, affected sweet sorghum biomassand amount of extracted liquid; year affected percent fruc-tose, sucrose and total sugars and litter rate affected percentfructose and glucose (Table 3). Over fertilizer rate biomass aver-aged 44.5 Mt ha−1 and extractable liquids averaged 3932.4 L ha−1,respectively. There was less biomass and extractable liquidsproduced in 2009 (31.7 Mt ha−1; 3517.2 L ha−1) than in 2010(57.4 Mt ha−1; 4287.4 L ha−1).

There was higher fructose in 2009 (3.75%) than in 2010 (3.23%);glucose averaged 4.36% over the years; there was no sucrose foundin 2009 but sucrose level was 4.54% in 2010, and total sugar waslower in 2009 (7.85%) than in 2010 (12.39%). The recommendedlitter rate produced more fructose (3.84%) and glucose (4.78%) thandid the higher litter rate, 3.15% and 3.93%, respectively. Litter ratedid not affect sucrose (avg. 2.27%) and total sugars (avg. 10.12%).

3.3. Grain sorghum

Year did not, but litter rate did, affect grain sorghum yield(Table 5). Averaged over year seed yield was 4.12 Mt ha−1. Plantstreated with the recommended litter rate produced 3.85 Mt ha−1

which was less than that produced with twice the recommendedlitter rate 4.39 Mt ha−1.

Year and variety affected seed yield but fertilizer rate did not(Table 5); fertilizer rate interacted with the other variables to affect

Table 5ANOVA results for effect of year, fertilizer rate on grain sorghum (milo) and sun-flower seed yield and sunflower oil yield.

Source Seed yield Sunflower hotpress oil yield

Grain sorghum Sunflower

Year (Y) NS ** –a

Variety (V) –b ** **

Fertilizer rate (F) * NS NSInteractiona

Y × F NS ** –a

V × F –b ** NSV × Y × F –b ** NS

NS, non-significant.a Oil yields obtained only in 2010.b Only a single source of grain sorghum (milo) tested.* Significant at P ≤ 0.05, ANOVA, shown are the only significant interactions.

** Significant at P ≤ 0.01, ANOVA, shown are the only significant interactions.

558 V.M. Russo, W.W. Fish / Industrial Crops

Table 6Interaction effects of year by fertilizer by variety on sunflower seed yield.

Source Seed yield(kg ha−1)

Variety Year Fertilizer rate

657 2009 Recommend 674.6Twice recommended 222.2**

2010 Recommend 230.0Twice recommended 270.1NS

820OH 2009 Recommend 424.2Twice recommended 182.4**

2010 Recommend 265.1Twice recommended 369.8**

Peredovik type 2009 Recommend 196.2Twice recommended 396.5*

2010 Recommend 136.7Twice recommended 236.6NS

NS, non-significant.In this type of analysis the first value in the group is not assigned a significanced

yswd(mvtNINFNNa

4

gtieiy

ryhydbrid

Pstslmsla

esignation; value in a combination is compared to the value directly above it.* Significant at P ≤ 0.05, Least Squares Means analysis.

** Significant at P ≤ 0.01, Least Squares Means analysis.

ield. Only variety affected oil yield (Table 5). Hybrid 657 had aignificantly higher oil yield (50.1%) than did var. 820OH (39.2%)hich was greater than for the Peredovik type (31.5%). There wereifferences in amounts of seed produced by sunflower in the yearsTable 6). In 2009, for plants of hybrid 657 receiving the recom-

ended N-rate, yield was higher than for the elevated N-rate. Thesealues were not different than in 2010 regardless of N-rate. Whenhe hybrid 820OH was used, and plants received the recommended-rate, yields were higher than when the elevated N-rate was used.

n 2010 yield increased when plants received the recommended-rate and increased again when the elevated N-rate was applied.or the Peredovik type in 2009 yield was higher when the elevated-rate was used. Yield was less in 2010 when the recommended-rate was used and not different when the elevated N-rate waspplied.

. Discussion

In spring, there is typically a residual of 2–4 kg ha−1 of nitro-en if synthetic nitrogen fertilizer was applied the previous yearo the sandy loam soil used for these experiments. Soil test resultsndicated this was the case for all crops at the beginning of thexperiment. Residual nitrogen was not considered a factor affect-ng results. The same amounts of chicken litter were used in bothears.

Fertilizer rate was not responsible for all of the differencesecorded on crops. Increasing fertilizer rate did not affect biomassield or sugar concentration in sweet sorghum and sweet corn. Theigher fertilizer rate increased milo seed yield, and sunflower seedield was inconsistently affected by higher fertilizer. Vegetativeevelopment will likely not require higher nutrition rates that maye needed for seed development and maturation. Where fertilizerate did have an affect, it appears results were moderated, by grow-ng conditions in the years, and were somewhat inconsistent andid not totally drive the outcomes.

Pappelis and Williams (1966) in field corn and Russo andappelis (1994) in sweet corn reported the correlation of internodeenescence rating with growth stage. It was further determinedhat sugar levels in tissues at previous stages of development, andenescence levels, could predict sugar levels in various tissues atater stages of development (Russo et al., 2004). Russo et al. (1998)

easured sugars in sweet corn internode tissues at various growthtages for a “supersweet” cultivar and determined that sucroseevels were highest at the R1 (silking) stage as described by Ritchiend Hanway (1982). Percent sucrose in the sweet corn cultivars

and Products 36 (2012) 555–559

sampled indicated that the plants had not reached the stage wherelarger portions of sugars were sucrose. Sweet corn stalk harvestwas based on visual markers. It may be that these markers are notprecise enough to determine when to harvest sweet corn stalksfor optimum sucrose content. However, it may be preferable tohave higher levels of simple sugars than higher levels of sucrose atharvest since simple sugars are more easily converted to ethanol.Although not determined it may be that senescence in sweetsorghum stalks may be able to be correlated with sugar levels andused to maximize harvest for extractable sugar.

Sweet sorghum yields were similar to other locations; the per-cent of soluble sugars was on the low end of the expected range, andthese and the total volume of extracted liquids were not affected byfertilizer rate. Grain sorghum (milo) yields were as good as otherproduction areas, and increased by 12% with increased fertilizer.Based on biomass and seed yields the relatively high pH of thechicken litter apparently had little effect on crops. When appliedto the soil the pH was likely reduced since the soil used is consis-tently below pH 6.5 and the alkaline nature of the chicken litterreduced. Sunflower seed yields were below what is expected forother production areas; oil percent was within expected rangesand increased fertilizer rate did not improve yield or oil.

There is little information on use of sweet corn as a biofuel feed-stock. For field corn (maize) total soluble carbohydrates was about10.5% which is higher than the average for the sweet corn culti-vars tested (Almondares et al., 2009). Other factors are involved inyield and/or oil percent of the crops tested and require additionalresearch. However, additional fertilizer did not provide a benefit tobiomass yield and/or oil production in crops.

4.1. Conclusions

As the need for renewable plant based sources of fuel becomemore critical it becomes necessary for producers to understand themost efficient level of fertilizer to use to deliver feedstock yieldcomponents. Also, use of crops for end products for which they werenot intended, i.e., sweet corn as a potential source of carbohydratesfor conversion to biofuels, or for crops where they have not tradi-tionally been grown, i.e., sunflower in southeastern Oklahoma, cancreate challenges for growers. In addition alternate sources of fertil-ization, i.e., manure, needs study in production of biofuel feedstockcrops. For milo the higher rate of chicken litter increased seed yield.In most instances the use of the higher fertilizer rate, synthetic orchicken litter, had no effect or reduced yields of the specific compo-nent. The inclination to use more than the recommended amountof fertilizer may result in increased cost without a correspondingbenefit to yield of the components specific to the use to which thecrop is being put.

Acknowledgement

The authors thank Northeast Texas Community College, Mt.Pleasant, Texas, for access to the screw press used to extract oilsfrom sunflower seed.

References

Almondares, A., Jafarinia, M., Hadi, M.R., 2009. The effects of nitrogen fertilizer onchemical compositions in corn and sweet Sorghum. Am. Eurasian J. Agric. Envi-ron. Sci. 6, 441–446.

Dahnke, W.C., Fanning, C., Cattanach, A., 1992. Fertilizing Sunflower. SF-713. NorthDakota State Univ., Fargo, ND.

Izekor, S.E., Porter, O.A., 2004. Sunflower Production in Arkansas. Cooperative Exten-

sion Program, FSA-2018. Univ. Arkansas-Pine Bluff, Pine Bluff, AR.

McCraw, D., Motes, J.E., Criswell, J.T., McCollum, G., 1987. Sweet Corn Production.OSU Extension Facts No. 6021. Oklahoma State University, Stillwater, OK.

Pappelis, A.J., Williams, J.R., 1966. Patterns of cell death in elongating corn stalks.Trans. Illinois State Acad. Sci. 59, 495–498.

Crops

R

R

R

V.M. Russo, W.W. Fish / Industrial

itchie, S., Hanway, J.J., 1982. How a Corn Plant Develops. Spec. Rpt. 48. Iowa StateUniv., Ames, IA.

usso, V.M., Pappelis, A.J., 1994. Cell death patterns in stalks of a supersweet sweetcorn cultivar. HortScience 29, 919.

usso, V.M., Roberts, K., Wright, J.R., Maness, N., 1998. 13C-nmr spectroscopy tomonitor sugars in pith of internodes of a sh2 corn at developmental stages.HortScience 33, 980–983.

and Products 36 (2012) 555–559 559

Russo, V.M., Collins, J.K., Perkins-Veazie, P., 2004. Carbohydrate distribution in asweet maize with the sugar enhanced endosperm genotype. Cereal Res. Com-

mun. 32, 347–353.

Whitney, D., 1998. Grain Sorghum Production Handbook. Kansas State Univ., Man-hattan, KS.

Zhang, H., Raun, B., Arnall, B., 2009. OSU Soil Test Interpretations. Okla. State Univ.,Stillwater, OK, Extension Facts No. PSS-2225.