Calculation of threshold Olsen P values for fertilizer response from soil properties

RESEARCH ARTICLE Calcula ion o h eshold Olsen P alues o e ilize esponse om soil p ope ies Rami o Recena 1 &Isabel Díaz 2 &Ma ía Ca men del Campillo 3 &José To en 3 & An onio Delgado 1 Accep ed: 28 July 2016 /Published online: 23 Sep embe 2016 #INRA and Sp inge -Ve lag F ance 2016 Abs ac Phospho us (P), a non- enewable esou ce, needs o be used mo e e icien ly in ag icul u e. This e- qui es using soil P es s. Howe e , he P es h eshold alues o e ilize esponse depend on many soil p op- e ies, some o which may be use ul o es ima e hese h eshold alues, o he s no . The e o e, we sea ched he e which soil p ope ies a e use ul o es ima e P h eshold alues. We calcula ed he h eshold alues o Olsen P and 0.01 M CaCl 2 ex ac able P o 18 ep esen a i e ag- icul u al soils o he Medi e anean egion o Spain. Fo ha , we pe o med a P s a a ion expe imen in which whea and sun lowe we e al e na i ely po -c opped. Resul s show ha Olsen P h eshold alues a e nega i ely co ela ed o P bu e capaci y ( o −0.74, P lowe han 0.001), clay con en (−0.82, 0.001), pH (−0.76, 0.001), and Fe oxide con en (−0.55, 0.05). Mul iple eg ession models in ol ing clay, pH o soil o ganic C, and phos- pha ase ac i i y o o ganic hyd olysable P accoun ed o as much as 87 % o he a iance in calcula ed Olsen P h eshold alues. In pa icula , he e is a majo e ec o o ganic P on Olsen P h eshold alues. Single models based on ou inely measu ed soil p ope ies such as clay con en and pH made accu a e p edic ions o Olsen P h eshold alues wi h 2 o 0.81 and P lowe han 0.001. Keywo ds Olsen P .A ailabili y index .Th eshold alues . Bu e capaci y .O ganic P 1 In oduc ion Ag icul u e depends on he use o phospho us (P) e ilize s ob ained om phospha e ock, which is a non- enewable and s a egic esou ce (Keyze 2010; Van Vuu en e al. 2010). The demand o P e ilize s is expec ed o g ow as he logical esul o he g owing ood needs o an inc easing popula ion, whe e- as he p oduc ion o PR is expec ed o peak in he nex decades (Sch öde e al. 2011;Ryane al.2012). Expec ed u u e P sca ci y is hus eme ging as a global challenge o humankind (Co dell and Nese 2014). In addi ion, he use o P in ag icul- u e is globally e y ine icien , in such a way ha only 15 % o P applied o ag icul u al soil goes in o he ood chain (Co dell e al. 2009; Wi he s e al. 2014). The o me conside a ions poin o he need o a mo e e icien and sus ainable use o P in ag icul u e. In pa icu- la , e iliza ion schemes should ely, among o he ac o s, on accu a e dis inc ion o P- esponsi e si es (Recena e al. 2015) in o de o concen a e his esou ce on soils in which he highes e u n o P e iliza ion is o be expec ed. This dis inc ion is usually assessed wi h he use o an “a ailable P es ”o a “soil P es ” ha is co ela ed wi h plan P up ake. The e is no a uni e sal soil P es alid o e y di e en soils, and only o Eu ope, 16 soil P es s ha e been p oposed (Ney oud and Lische 2003; Delgado and Scalenghe 2008). Commonly used soil P es s a e based on ex ac ion wi h a chemical eagen , as is he case o wo ldwide used Olsen P es (Olsen e al. 1954). The key issue o hei p ac ical use is he de ini ion o h eshold alues below which yield esponse o e ilize applica ion can be expec ed (Delgado and Scalenghe 2008;Tange al. *An onio Delgado [email protected] 1 Depa amen o de Ciencias Ag o o es ales, ETSIA, Uni e sidad de Se illa, C a. U e a km 1, 41013 Se illa, Spain 2 Depa amen o de Mecánica de Fluidos e Ingenie ía Ae oespacial, ETSIA, Uni e sidad de Se illa, C a. U e a km 1, 41013 Se illa, Spain 3 Depa amen o de Ag onomía, Uni e sidad de Có doba, Edi icio C4, Campus de Rabanales, 14071 Có doba, Spain Ag on. Sus ain. De . (2016) 36: 54 DOI 10.1007/s13593-016-0387-5 2009). Howe e , al hough soil P es s a e deemed o be use ul o assessing P e ilize needs, hey can be inaccu- a e because he de ini ion o hei h eshold alues is no always p ecise. In ac , ac ual h eshold alues o a gi en es can ange widely among hose soils o which i s use is ecommended, e.g., o Olsen P (Delgado e al. 2010; Sánchez-Alcalá e al. 2015). Thus, Olsen P is no always well ela ed o plan P up ake (Delgado and To en 1997; Kulhánek e al. 2007; Tandy e al. 2011). In soils wi h e y simila p ope ies, Delgado e al. (2010) obse ed a wide ange in h eshold alues o Olsen P, which clea ly depended on he soil P bu e capaci y. Sánchez-Alcalá e al. (2014) epo ed ha he h eshold alues o Olsen P we e a ec ed by p ope ies ela ed o he P so p ion capac- i y o he soil such as ca bona e and Fe oxides con en s. These h eshold alues we e es ima ed conside ing ha plan s can abso b su icien P abo e a gi en P concen a ion in solu ion. Howe e , ecen s udies ha e e ealed ha P concen a ion in soil solu ion o a su icien P supply may di e om one soil o ano he (Sánchez-Alcalá e al. 2015). The e is no much in o ma ion on soil p ope ies con olling his h eshold P concen a ion in solu ion. In o he plan g ow h expe imen s, ca bona e con en and pH we e he soil p ope ies a ec ing Olsen P h eshold alues (Sánchez-Alcalá e al. 2015). Recena e al. (2015)demon- s a ed ha he Olsen P p edic i e alue o plan P up ake is a ec ed no only by hose soil p ope ies in luencing he equilib ium be ween ino ganic phospha e in solu ion and ha in solid phase bu also by ac o s ela ed o o ganic P dynamics. Howe e , hey did no p o ide e idence o how h eshold alues o Olsen P we e a ec ed by all hese soil p ope ies. In spi e o all hese con ibu ions, he e is no a whole iew o all po en ial soil ac o s a ec ing h eshold alues o a gi en soil P es . I can be hypo hesized ha he e is a need o mo e p ecise me hods o es ima e P- esponsi e si es. One possibili y is he de elopmen o es s based on mo e complex P ex ac ion me hods such as hose using esins (Delgado e al. 2010) o di usi e g adien s hin ilms (Tandy e al. 2011; San ne e al. 2015), which a e mo e sensi i e o soil P bu e capaci y. Ano he op ion is he de elopmen o single models ha a e based on adi- ional soil P es s such as Olsen P in combina ion wi h soil p ope ies ela ed o he dynamics o soil P. This would mean ha h eshold alues could be accu a ely es ima ed on he basis o ou inely de e mined soil p ope ies. These models should ely on a deepe knowledge on hose soil p ope ies a ec ing he ela ionship be ween plan P up ake and chemically ex ac able soil P. In his ega d, he objec- i es o he p esen s udy we e as ollows: (i) iden i ica ion o soil p ope ies a ec ing h eshold Olsen P alues o e ilize esponse in a se o ep esen a i e Medi e anean soils whe e his es is usually ecommended and (ii) o de elop simple models based on ou inely de e mined soil p ope ies o es ima e accu a e h eshold alues o use o Olsen P as soil P es in hese soils. 2 Ma e ials and me hods 2.1 Soils The se o soils used in his s udy was he same one as used by Recena e al. (2015), wi h one addi ional soil included (18 soils in o al). The soils we e classed as Mollisols, En isols, Incep isols, Al isols, and Ve isols acco ding o soil axonomy (Soil Su ey S a 2010). The ag icul u al use o hese soils wasde ailedinRecenae al.(2015), and he addi ional soil conside ed was used o he ypical biannual ain- ed c op o a ion in Spain, i.e., whea –sun lowe . The main p ope ies o he 0–20-cm soil laye , om which samples we e aken, a e desc ibed in Table 1. Me hods o soil sampling, p ocessing, analysis o basic p ope ies, s udy o P and Fe o ms, and P so p ion capaci y om P so p ion iso he ms a e desc ibed in Recena e al. (2015). I on o ms included ci a e-asco ba e ex ac able Fe (Fe ca ), mainly ela ed o poo ly c ys alline Fe oxides, and ci a e-bica bona e-di hioni e ex ac able Fe (Fe d ), mainly e- la ed o c ys alline Fe oxides. In addi ion, he concen a ion o phy ase hyd olysable P in he alkali (NaOH) and ci a e- bica bona e ex ac s o he sequen ial ac iona ion scheme desc ibed by Recena e al. (2015) o s udy P o ms was also de e mined. These ex ac s in he sequen ial ac iona ion scheme a e supposed o con ain ino ganic and o ganic P mos - ly adso bed o p ecipi a ed as soluble me al phospha es (Saa ed a e al. 2007). The phy ase hyd olysable P was es i- ma ed as he inc ease in molybda e eac i e P (Mu phy and Riley 1962) a e 30 min o incuba ion o he ex ac wi h phy ase (Bio-Feed phy ase L, No ozimes, Bags ae d, Denma k) a 37 °C. To his end, he enzyme was added o he soil ex ac in 0.2 ml o bu e solu ion p epa ed wi h 0.4 M C 2 H 4 O 2 and 10 mM e hylenediamine e aace ic acid a pH 5.5 o a inal ac i i y o 10 nKa ml −1 and he eac ion was s opped by adding 15 % ichlo oace ic acid. Phospho us concen a ion in 0.01 M CaCl 2 ex ac s (P CaCl2 ) was used as a p oxy o he concen a ion o P in he soil solu ion. This ex ac ion was pe o med in dupli- ca eanda asoil:ex ac an a ioo 1:10using2go soilin 20 ml o ex ac an in polye hylene alcon ubes ha we e end-o e -end shaken a 2.5 s −1 o 30 min. The ex ac was cen i uged a 1000g o 10 min, and hen a po ion o 2 ml cen i uged in Eppendo ubes a 19,000g o emo e soil pa icles wi h a diame e >0.05 μm (Sánchez-Alcalá e al. 2015). In he emaining solu ion, molybda e eac i e P was de e mined acco ding o Mu phy and Riley (1962). Olsen P was de e mined as he molybda e eac i e P in bica bon- a e ex ac s (Olsen e al. 1954). 54 Page 2 o 8 Ag on. Sus ain. De . (2016) 36: 54 As desc ibed by Recena e al. (2015), om a collec ion o soil si es in which he indi idual samples anged widely in Olsen P, hose soil samples wi h he wo ex eme Olsen P alues we e selec ed om each si e and named “low-P”and “high-P”samples acco ding o hei P s a us. The high-P and low-P samples we e examined in o de o check ha hey di e ed by less han 5 % in he main soil p ope ies. Fo he expe imen desc ibed below, soil was g ound o <6 mm. Olsen P le els in high-P samples we e all well abo e h eshold alues o e ilize esponse acco ding o he cu en s age o knowledge, e en in samples wi h he lowes alues (9 mg kg −1 ) which we e samples wi h a e y high P bu e capaci y and hus wi h expec ed h eshold alues below 5mgkg −1 (Delgado e al. 2010). Fo mos low-P samples, Olsen P was assumed o be close o h eshold alues (da a no shown; means and ange in Recena e al. 2015). 2.2 Expe imen al se up An expe imen in ended o deple e soil P was conduc ed in a g ow h chambe (Fig. 1) whe e du um whea and sun lowe we e g own un il an hesis wice, whe eby simula ing a ypical ain- ed c op o a ion, o bo h high-P and low-P samples. Thus, o each ype o sample o each soil, ou successi e c ops we e g own. Polys y ene po s (5.5-cm diame e , 15-cm high) illed wi h 300 g o soil we e used. Seeds we e p e- ge mina ed on a mois ened Pe i pla e o 14 days and, a e ha , ge mina ed in ays wi h pe li e as subs a e and i iga ed wi h deionized wa e . A e 16 days, seedlings we e ansplan ed. Fo each soil and sample, he numbe o po s pe deple ion s age dec eased om i e in he i s s age o wo in he las one because 300 g we e emo ed a e each s age o u he labo a o y analysis. A e ansplan ing, po s we e wa e ed daily wi h 20 ml o a Hoagland- ype nu ien solu- ion wi hou phospho us a pH 6–6.5, sandwiching be ween h ee i iga ions and one wi h he same olume o deionized wa e o a oid saliniza ion o he soil. Plan s we e g own wi h a 14-h pho ope iod, a ela i e humidi y o 45 % (day pe iod) and 60 % (nigh pe iod), and an ac i e pho osyn he ic adia- ion o 22 W m −2 . The composi ion o he solu ion applied was he ollowing (all concen a ions in mM l −1 ): MgSO 4 (2), Ca(NO 3 ) 2 (5), KNO 3 (5), KCl (0.05), Fe-EDDHA (0.01), H 3 BO 3 (0.009), MnCl 2 (0.0023), CuSO 4 (0.0005), ZnSO 4 (0.002), and H 2 MoO 4 (0.0005). 2.3 Plan and soil analysis a e c opping A e each c opping, plan shoo s and oo s we e ca e ully sepa a ed, washed, and inally d ied in a o ced ai o en un il cons an weigh a 65 °C o a leas 48 h. Roo and shoo d y ma e was hen measu ed. A e each c opping s ep, soil om all he eplica ions o each soil and ype o sample was mixed and p ocessed by d ying and g ounding Table 1 Mean, s anda d de ia ion (SD), and ange o he p ope ies o s udied soils (n= 18), including ini ial Olsen P in high-P samples, ci a e-asco ba e ex ac able Fe (Fe ca ), ci a e-bica bona e- di hioni e ex ac able Fe (Fe d ), phy ase hyd olysable P in he NaOH (NaOH h ), and ci a e-bica bona e (CB h ) ex ac s o high-P samples; mean o he phospha ase ac i i y in he ou c ops in high-P samples; cons an s A and b in he F eundlich equa ion (Y + Yo = A •X 1/b ) in high-P samples; and P bu e capaci y a h ee concen a ions in solu ion equilib ium (e), 0.2 mg P l −1 ,and1mgPl −1 Gene al soil p ope ies Clay Soil o ganic C Ca ca bona e equi alen Ac i e Ca ca bona e equi alen Fe ca Fe d pH Olsen P gkg −1 mg kg −1 Mean ± SD 288 ± 155 10.6 ± 3.8 270 ± 234 81 ± 67 1.02 ± 0.74 5.52 ± 3.05 7.8 ± 0.5 24 ± 10 Range 48–640 5.5–20 0–723 0–178 0.17–2.44 2.04–13.24 6.5–8.3 9.1–49.1 O ganic P- ela ed p ope ies NaOH h CB h Phospha ase mg kg −1 mg PNP kg −1 h −1 Mean ± SD 5.21 ± 5.25 7.0 ± 4.6 142 ± 18 Range 0.48–24.02 0–13.5 105–172 P ope ies ela ed o P adso p ion AbBu e capaci y e Bu e capaci y 0.2 Bu e capaci y 1 Mean ± SD 72 ± 32 2.22 ± 0.31 348 ± 413 78 ± 35 32 ± 14 Range 16–140 1.52–2.64 9–1480 18–151 8.5–67 Th eshold alues o Olsen P and P concen a ion in CaCl2 ex ac s (P CaCl2 )andP alues and R 2 o he Ca e-Nelson model P CaCl2 P alues R 2 Olsen P P alues R 2 mg l −1 mg kg −1 Mean ± SD 0.025 ± 0.043 0.0203 ± 0.0240 0.72 ± 0.14 9.2 ± 3.2 0.0326 ± 0.038 0.68 ± 0.19 Range 0.002–0.175 0–0.0868 0.47–0.96 4.9–16.5 0–0.1065 0.38–0.96 Median 0.006 0.0117 0.74 8.9 0.0164 0.69 Subsc ip s o bu e capaci y: e, es ima ed a equilib ium concen a ion; 0.2, es ima ed a 0.2 mg l-1 P in he solu ion; and 1, es ima ed a 1 mg l-1 P in he solu ion Ag on. Sus ain. De . (2016) 36: 54 Page 3 o 8 54 o 6 mm; 300 g o soil was wi hd awn o chemical anal- ysis and g ound o pass a 2-mm sie e. In hese samples, molybda e eac i e P in bica bona e and 0.01 M CaCl 2 ex ac s was de e mined as desc ibed p e iously. 2.4 Da a analysis Da a om high- and low-P samples we e analyzed as a unique se o da a o each soil. D y ma e yield o whea and sun- lowe was exp essed on a ela i e basis conside ing as he maximum yield he a e age d y ma e ob ained in high-P samples o whea and sun lowe in hei i s c op. Because Olsen P alues we e gene ally abo e 12 mg kg −1 be o e he i s sun lowe c op (da a no shown), and sun lowe is e y e icien in ex ac ing P (Delgado and To en 1997), sun low- e d y ma e yield was no es ic ed; only wo soils showed lowe Olsen P alues be o e sun lowe (6 and 9) which we e conside ed non-limi ing due o he high P bu e capaci y o he soil (Delgado e al. 2010). Th eshold alues o Olsen P and P CaCl2 we e es ima ed acco ding o he Ca e and Nelson me hod (1971). This me hod p o ides h eshold alues o each P ex ac ion below which he yield signi ican ly de- c eased om he maximum ela i e yield de ined o each soil and seems mo e accu a e o de ine h eshold alues o soil P es han o he me hods such as he linea o quad a ic pla eau i ings o he Mi sche lich- ype equa ions (Malla ino and Blackme 1992). The c i ical le el o he Ca e-Nelson model was es ima ed as ha alue o he soil P es (P CaCl2 o Olsen P) ha maximized he sum o squa es be ween wo popula- ions o soil P es alues, i.e., abo e and below h eshold alues (Geng e al. 2014), using he ANOVA op ion o S a g aphics 5.1. 3 Resul s and discussion The se o soils used in his s udy anged widely in p ope ies ela ed o P dynamics (Table 1). Rega ding phy ase hyd olys- able P, his can be assumed o be mainly monoes e s which can be po en ially hyd olyzed by he ac ion o hizosphe e enzymes and can po en ially con ibu e o P up ake by plan s. On a e age, amoun s o hese monoes e s adso bed o p ecip- i a ed as soluble me al phospha es we e no negligible since hey we e equi alen o hal o he bica bona e ex ac able ino ganic P in soil (Olsen P, Table 1). As expec ed, he soil P bu e capaci y was posi i ely co ela ed o clay con en (Shi ani e al. 2005) and he con en o Fe ela ed o Fe oxides in soil (Fe ca +Fe d ). On he con a y, soil P bu e capaci y was nega i ely co ela ed wi h he a io o Fe in poo ly c ys alline oxides o ha in c ys alline oxides (Fe ca /Fe d )(Table2shows P bu e capaci y a 1 mg l −1 ). This nega i e co ela ion is ex- plained by he usual lowe a ini y o P o poo ly ela i e o highly c ys alline Fe oxides (Colombo e al. 1994). I on in Fe oxides and he Fe ca /Fe d a io we e co ela ed wi h clay con en (Table 2) hus e ealing ha oxides a e pa o he clay ac ion bu also ha he a io o poo ly c ys alline o c ys alline Fe oxides inc eased wi h dec eased clay con en in soil. On he o he hand, he Fe in Fe oxides was nega i ely co ela ed wi h he ac i e calcium ca bona e equi alen , e idencing ha min- e al wea he ing and subsequen p ecipi a ion o oxides du ing soil genesis is inhibi ed when soil pH is bu e ed by ca bona e. O e all, he Ca e-Nelson model p o ided easonable es i- ma es o h eshold alues o bo h soil P es s: median alues o he po ion o a iance explained was 69 % o Olsen P and 74 % o P CaCl2 (Table 1). Th eshold alues o Olsen P we e in he ange usually obse ed in ield expe imen s (Colomb e al. 2007; Bai e al. 2013) and we e simila o hose desc ibed by Delgado e al. (2010) in po expe imen s using a simila amoun o soil. Obse ed h eshold alues o his soil P es anged widely be ween soils (Table 1). Such a deg ee o a - iabili y in h eshold alues o Olsen P was obse ed by Delgado e al. (2010) in a g oup o clay soils wi h a na ow ange o a ia ion in soil p ope ies. Sánchez-Alcalá e al. (2015) also obse ed a wide ange o Olsen P and P CaCl2 h eshold alues in a g oup o soils which, in ha case, anged widely in hei p ope ies. Th eshold alues o bo h indices we e posi i ely co ela - ed, and o bo h es s, h eshold alues we e nega i ely co e- la ed wi h clay (Table 2). In addi ion o clay, soil p ope ies wi h which Olsen P h eshold alues we e co ela ed we e he ollowing: pH, P bu e capaci y, and Fe in Fe oxides; hese co ela ions being also nega i e (Table 2). Nega i e co ela- ion wi h clay and Fe in Fe oxides e eals ha h eshold alues dec eased wi h inc eased P so p ion capaci y o he soil. Since clay and Fe oxide con en we e posi i ely co ela ed in he g oup o soils s udied (Table 2), he ela i e con ibu ion o clay and Fe oxides o he P dynamics is di icul o assess. Fo he same P concen a ion in solu ion, P bu e capaci y usually inc eases wi h inc easing P so p ion capaci y. This con ibu es o explain he co ela ion o clay and Fe in oxides wi h soil P bu e capaci y (Table 2), which may explain a leas in pa he obse ed nega i e co ela ion o h eshold alues wi h clay and Fe oxides. As epo ed by Delgado e al. (2010), P bu e capaci y seems o accoun o mo e a ia ion o he Olsen P h eshold alues when i is measu ed a ela i ely high P concen a ion in solu ion (1 mg l −1 ) a he han a low concen a ion (da a no shown). Al hough he co ela ion o pH wi h P bu e capaci y and clay makes i di icul o ex ac clea -cu con- clusions, he e ec o pH can be also asc ibed no only o i s in luence on soil P dynamics bu also by i s co ela ion wi h o he mine alogical p ope ies, such as so ben su ace ype, also a ec ing a ailabili y o plan s (Delgado and To en 1997). Th eshold alues o Olsen P can be accu a ely p e- dic ed in ou soils om mul iple eg essions in ol ing di e - en soil p ope ies. Clay con en , pH, and phospha ase ac i i y 54 Page 4 o 8 Ag on. Sus ain. De . (2016) 36: 54 in he hizosphe e explained nea 90 % o a ia ion in Olsen P h eshold alues (Fig. 2a). In addi ion, clay, soil o ganic ca bon, and he phy ase hyd olysable P in he ci a e-bica bona e ac ion explained 75 % o a ia ion in h eshold alues o Olsen P (Fig. 2b). These ela ionships e eal ha h eshold alues o his soil P es was in luenced no only by hose ac o s a ec ing he equilib ium o ino ganic phospha e be ween he solid and liquid phases o soil, such as P bu e capaci y, bu also by ac o s ela ed o o ganic P dynamics, such as he amoun o hyd olysable o ganic P o he phospha ase ac i i y in he hizo- sphe e. O e all, lowe h eshold alues can be expec ed wi h inc eased con en s o hyd olysable o ganic P (Fig. 2b) due o he con ibu ion o he o ganic P o plan supply. This con ibu- ion is no aken in o accoun in usual soil P es which a e based on he de e mina ion o ino ganic P (usually molybda e eac i e) in soil ex ac s. The wide ange o P CaCl2 h eshold alues obse ed suppo s he con en ion ha he P concen a ion in soil solu ion abo e which he e is a su icien P up ake by plan s a ies widely de- pending on soil p ope ies, consis en wi h p e ious esul s (Sánchez-Alcalá e al. 2015). These alues we e a ec ed nega- i ely by ac i e calcium ca bona e, and posi i ely by Fe ca /Fe d , and hyd olysable o ganic P in NaOH and ci a e-bica bona e ex ac s (Table 2;Fig.3). This e eals he po en ial con ibu ion o o ganic P o P supply o plan s. The e ec o he a io Fe ca /Fe d on h eshold P CaCl2 can be explained by he lowe a ini y o P o poo ly ela i e o highly c ys alline Fe oxides. This implies ha a highe P concen a ion in solu ion can be in equilib ium wi h he Table 2 Co ela ion coe icien s be ween di e en p ope ies o s udied soils: h eshold alues (TV) o Olsen P and P in 0.01 M CaCl 2 ex ac s (P CaCl2 ), clay, pH, mean o he phospha ase ac i i y in he ou c ops in high-P samples, o ganic C, ac i e Ca ca bona e equi alen (ACCE), Fe in Fe oxides (Feca + Fed), a io o ci a e-asco ba e o ci a e-bica bona e- di hioni e ex ac able Fe (Fe ca /Fe d ), P bu e capaci y a 1 mg P l −1 (PBC 1 ), phy ase hyd olysable P in ci a e-bica bona e ex ac s (CB h ), and he sum o phy ase hyd olysable P in NaOH and ci a e-bica bona e ex ac s ( NaOH h +CB h )inhigh-Psamples P CaCl2 TV Clay pH Phospha ase O ganic C ACCE Feca + Fed Feca/Fed PBC 1 CB h NaOH h +CB h Olsen P TV 0.63*** −0.82*** −0.76*** 0.26 ns 0.11 ns −0.31 ns −0.55* 0.42 ns −0.74*** 0.11 ns 0.32 ns P CaCl2 TV −0.80*** −0.61** −0.33 ns −0.18 ns −0.54* 0.28 ns 0.68** −0.67** 0.25 ns 0.77*** Clay −0.60** 0.52* −0.31 ns −0.44 ns 0.52* −0.66** 0.85*** −0.25 ns 0.69** pH 0.41 ns 0.17 ns 0.71** 0.19 ns −0.73** 0.57* 0.27 ns 0.2 ns Phospha ase 0.39 ns 0.28 ns 0.23 ns 0.29 ns 0.52* −0.36 ns −0.48* O ganic C −0.14 ns 0.1 ns 0.12 ns −0.23 ns 0.23 ns 0.17 ns ACCE −0.59** −0.70** 0.34 ns −0.02 ns −0.35 ns Feca + Fed −0.35 ns 0.50* 0.23 ns 0.44 ns Feca/Fed −0.46* −0.24 ns 0.55* PBC 1 0.13 ns −0.42 ns CB h 0.77*** ns no signi ican *P < 0.05; **P < 0.01; ***P < 0.001 Fig. 1 Po s a a ion expe imen : asun lowe in g owing chambe wi h bincipien P de iciency symp oms and cclea de iciency symp oms and dwhea wi h eP de iciency symp oms Ag on. Sus ain. De . (2016) 36: 54 Page 5 o 8 54 solid phase a he same amoun o so bed P wi h inc easing a io o poo ly o highly c ys alline Fe oxides. Thus, he ype o so - ben su aces was also a ele an p ope y a ec ing h eshold alues. Phospho us bu e capaci y es ima ed a 1 mg l −1 seems a c ucial p ope y explaining a iabili y in h eshold alues o OlsenPandP CaCl2 . This is in ag eemen wi h he cu en s age o knowledge (Hol o d 1980; Bolland e al. 1994;Ehle e al. 2003). Signi icance o soil P bu e capaci y in ou long- e m deple ion expe imen o explaining h eshold alues o Olsen P was much g ea e han ha obse ed by Recena e al. (2015) o explaining he a io o plan P up ake o Olsen P when P a ailabili y in g owing media was sca ce. This e eals ha P bu e capaci y may be c ucial o poin o he le el o P in soil a which he e is no limi a ion in P up ake by plan s, bu no so ele an in explaining he P up ake by plan s below his le el. Phy ase hyd olysable P in NaOH and ci a e- bica bona e ex ac s was signi ican in explaining he Th eshold alues o Olsen P (mg kg-1) 2 4 6 8 10 12 14 16 18 Es ima ed h eshold alues o Olsen P (m g k g -1) 2 4 6 8 10 12 14 16 18 246810121416 Th eshold alues o Olsen P (mg kg-1) 2 4 6 8 10 12 14 16 18 b a Fig. 2 Es ima ion h eshold Olsen P alues o e ilize esponse as a unc ion o aclay, pH, and mean phospha ase ac i i y in he hizosphe e (mean o he ou c ops in high-P samples); Y = 32 −0.015 clay −3.26 pH + 0.05 phospha ase; R 2 =0.87;P< 0.001; and coe icien s o each explica i e a iable signi ican a P< 0.05 in he leas signi ican case and bclay, soil o ganic ca bon (SOC), and phy ase hyd olysable P in CB ex ac s o high-P samples (CB h ); Y = 14 + 2.05 SOC −0.22 CBh −0.02 clay; R 2 =0.75;P< 0.001; and coe icien s o each explica i e a iable signi ican a P< 0.1 in he leas signi ican case. Th eshold alue can be es ima ed also as a unc ion o clay and pH 36– 0.012 clay—3.02 pH; R 2 = 0.81; P < 0.001; and coe icien s o each explica i e a iable signi ican a P< 0.01 in he leas signi ican case and mean absolu e e o = 1.12 Th eshold alues o PCaCl2 (mg L-1) -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 Es ima ed h eshold alue o PCaCl2 (mg L-1) -2,5 -2,0 -1,5 -1,0 -2,5 -2,0 -1,5 -1,0 Th eshold alue o PCaCl2 (mg L-1) -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 b a Fig. 3 Es ima ion o he loga i hm o he h eshold P CaCl2 alues o e ilize esponse as a unc ion o aclay and he a io o Fe ca o Fe d , Y=−2.02 −0.0018 clay + 2.92 Fe ca /Fe d ,R 2 =0.57,P< 0.001, and coe icien s o each explica i e a iable signi ican a P<0.05inbo h cases and bpH and he sum o phy ase hyd olysable in NaOH and CB ex ac s, Y = 2.38 −0.62 pH + 0.034 (NaOH h +CB h ), R 2 = 0.54, P< 0.01, and coe icien s o each explica i e a iable signi ican a P< 0.05 in bo h cases. Th eshold alue can be also es ima ed as a unc ion o clay and ac i e Ca ca bona e equi alen (ACCE), Y=−1.13 −0.039 ACCE −0.002 clay. R 2 = 0.56, P< 0.01, and coe icien s o each explica i e a iable signi ican a P<0.05 54 Page 6 o 8 Ag on. Sus ain. De . (2016) 36: 54 h eshold alues o bo h soil P es s when he alue conside ed was ha a he beginning o he deple ion expe imen o he high-P samples (Fig. 2b and Fig. 3b). In addi ion, phospha ase was signi ican explaining Olsen P h eshold alues when he mean ac i i y o he ou deple ion s eps in ol ing high-P sam- ples was aken in o accoun . This con ibu ion o o ganic P o P supply o plan s was p e iously e idenced by Recena e al. (2015) when he a ailable P in he g owing media was e y low in a sho - e m P deple ion expe imen . Howe e , ou long- e m P deple ion expe imen e ealed ha o ganic P also a ec s he soil P es alue abo e which plan s can ake up P wi hou dec ease in yield. The p esen esul s e eal ha he signi ican in luence o o ganic P- ela ed ac o s on h eshold alues o p ac ical use o Olsen P as soil P es in assessing P e iliza ion. Mo eo e , he con ibu ion o o ganic P o P up ake by plan s was no es ic ed o s a a ion in ino ganic a ailable P as p e iously obse ed (Recena e al. 2015); i seemed o be mo e ela ed o o ganic P o ms a ailable o enzyma ic hyd olysis in he hizosphe e. O ganic ca bon in soil also seems o con ibu e o explain Olsen P h eshold alues (Fig. 2b). I is well known ha dissol ed o ganic ma e a ec s he equilib- ium o P in soil h ough compe i ion wi h phospha e o he so ben su aces and inhibi ion o me al phos- pha e p ecipi a ion (Delgado e al. 2002; Saa ed a e al. 2007). In addi ion, inc eased o ganic ca bon in soils esul ing om o ganic ma e addi ion can p omo e a dec ease in soil P bu e capaci y (Sui and Thompson 2000) which also con ibu es o explain po en ial e ec s o soil o ganic ca bon on soil P es h eshold alues. When compa ed wi h he cu en s age o knowledge, he p esen wo k p o ides a mo e comple e iew o soil ac o s a ec ing he de ini ion o h eshold alues o Olsen P and P in soil solu ion, as es ima ed, e.g., by ex- ac ion wi h 0.01 M CaCl 2 . This was achie ed by a deep analysis o he soil p ope ies in ol ed in P dynamics. F om hese esul s, new single models o es ima e h esh- old alues based on ou inely de e mined soil p ope ies can be p oposed. In he case o Olsen P, a model based on clay and pH explains 81 % o he a ia ion wi h an es i- ma ed mean absolu e e o o 1.12. This in p ac ical e ms does no ha e a ele an economical impac in assessing P e iliza ion. De elopmen o hese models should be pe - o med, o p ac ical ecommenda ions, a ield scale and o di e en c ops because ac o s a ec ing oo de elop- men and he po en ial con ibu ion o subsu ace ho izons a e ele an o he supply o P o c ops. Ou esul s go beyond he p e ious ones o Recena e al. (2015), who desc ibed he ele an ole o p ope ies ela ed o o ganic P on P up ake by plan s. In ou case, o ganic P- ela ed p ope ies a ec he le el o P in soil abo e which he supply o P o he plan is assu ed, con ibu ing hus o de ine h eshold alues o bo h soil P es s. 4Conclusions Fac o s ela ed wi h he soil P bu e capaci y s and as he mos ele an o explain he h eshold alues o bo h soil P es s. Howe e , ac o s ela ed o o ganic P dynamics a e also in lu- en ial on he alue o soil P es in soil abo e which he e is a su icien P supply o plan s hus con ibu ing o de ine h esh- old alues o bo h soil P es s. In p ac ice, de ini ion o accu- a e h eshold alues o Olsen P in soils o Medi e anean egions should p o i om wo basic soil p ope ies, namely, clay con en and pH. In a simple manne , his could be achie ed by de ining di e en anges o h eshold alues de- pending on he clay con en and pH o soil. Acknowledgmen s This s udy was unded by he Spanish Minis y o Science and Inno a ion and he Eu opean Regional De elopmen Fund o he Eu opean Union h ough he Na ional Resea ch, De elopmen and Inno a ion P og am (Plan Nacional I + d + i, AGL2011-29893-CO2-01, and AGL2011-29893-CO2-02) and by he Regional Go e nmen o Andalusia (P ojec AGR 63-85). M . Rami o Recena was a Ph.D. s uden wi h a g an om he Regional Go e nmen . Fe ibe ia S.A. and OCA Leb ija collabo a ed in he collec ion o soil samples o he p esen s udy. Re e ences Bai Z, Li H, Yang X, Zhou B, Shi X, Wang B, Li D, Shen J, Chen Q, Qin W, Oenema O, Zhang F (2013) The c i ical soil P le els o c op yield, soil e ili y, and en i onmen al sa e y in di e en soil ypes. Plan Soil 372:27–37. doi:10.1007/s11104-013-1696-y Bolland MDA, Wilson IR, Allen DG (1994) E ec o P bu e capaci y and P e en ion index o soils on soil es P, soil es P calib a ions and yield esponse cu a u e. Aus J Soil Res 32:503–517. doi:10.1071/SR9940503 Ca e RB, Nelson LA (1971) A simple s a is ical p ocedu e o pa i ioning soil es co ela ion da a in o wo classes. Soil Sci Soc Am P oc 35: 658–660. doi:10.2136/sssaj1971.03615995003500040048x Colombo C, Ba ón V, To en J (1994) Phospha e adso p ion and deso p- ion in ela ion o mo phology and c ys al p ope ies o syn he ic hema i es. Geochim Cosmochim Ac a 58:1261–1269. doi:10.1016 /0016-7037(94)90380-8 Colomb B, Debaeke P, Jouany C, Nolo JM (2007) Phospho us manage- men in low inpu s ockless c opping sys ems: c op and soil e- sponses o con as ing P egimes in a 36-yea expe imen in sou he n F ance. Eu J Ag on 26:154–165. doi:10.1016/j.eja.2006.09.004 Co dell D, Nese TS (2014) Phospho us ulne abili y: a quali a i e amewo k o assessing he ulne abili y o na ional and egional ood sys ems o he mul i-dimensional s esso s o phospho us sca - ci y. Glob En i on Chang 24:108–122. doi:10.1016/j. gloen cha.2013.11.005 Co dell D, D ange JO, Whi e S (2009) The s o y o phospho us: global ood secu i y and ood o hough . Glob En i on Chang 19:292– 305. doi:10.1016/j.gloen cha.2008.10.009 Delgado A, To en J (1997) Phospha e ich soils in he Eu opean Union: es ima ing o al plan -a ailable phospho us. Eu J Ag on 6:205–214. doi:10.1016/S1161-0301(96)02048-5 Delgado A, Mad id A, Kasem S, And eu L, del Campillo MC (2002) Phospho us e ilize eco e y om calca eous soils amended wi h humic and ul ic acids. Plan Soil 245:277–286. doi:10.1023 /A:1020445710584 Ag on. Sus ain. De . (2016) 36: 54 Page 7 o 8 54 Delgado A, Scalenghe R (2008) Aspec s o phospho us ans e om soils in Eu ope. J Plan Nu Soil Sci 171(4):552–575. doi:10.1002 /jpln.200625052 Delgado A, del Campillo MC, To en J (2010) Limi a ions o he Olsen me hod o assess a ailable phospho us in eclaimed ma sh soils. Soil Use Manage 26:133–140. doi:10.1111/j.1475-2743.2010.00264.x Ehle P, Mo el C, Fo yma M, Des ain JP (2003) Po en ial ole o phos- pha e bu e ing capaci y o soils in e ilize managemen s a egies i ed o en i onmen al goals. J Plan Nu Soil Sci 166:409–415. doi:10.1002/jpln.200321182 Geng X, Guilla d K, Mo is TF (2014) Rela ing u g ass g ow h and quali y o equen ly measu ed soil ni a e. C op Sci 54:366–382. doi:10.2135/c opsci2013.03.0145 Hol o d ICR (1980) E ec s o phospha e bu e capaci y on c i ical le els and ela ionships be ween soil es s and labile phospha e in whea g owing soils. Aus J Soil Res 18:405–414. doi:10.1071 /SR9800405 Keyze M (2010) Towa ds a closed phospho us cycle. De Economis 158: 411–425. doi:10.1007/s10645-010-9150-5 Kulhánek M, Balík J, Če ný J, Ned ěd KB (2007) The in luence o di e en in ensi ies o phospho us e ilizing on a ailable phospho- us con en s in soils and up ake by plan s. Plan Soil En i on 53: 382–387 Malla ino AP, Blackme AM (1992) Compa ison o me hods o de e min- ing c i ical concen a ions o soil es phospho us o co n. Ag on J 84:850–856. doi:10.2134/ag onj1992.00021962008400050017x Mu phy J, Riley JP (1962) A modi ied single solu ion me hod o he de e mina ion o phospha e in na u al wa e s. Anal Chim Ac a 27: 31–36. doi:10.1016/S0003-2670(00)88444-5 Ney oud JA, Lische P (2003) Do di e en me hods used o es ima e soil phospho us a ailabili y ac oss Eu ope gi e compa able esul s? J Plan Nu Soil Sci 266:422–431. doi:10.1002/jpln.200321152 Olsen SR, Cole CV, Wa anabe FS, Dean LA (1954) Es ima ion o a ail- able phospho us in soils by ex ac ion wi h sodium bica bona e. USDA ci cula 939. US Go e nmen P in ing O ice, Washing on DC Ryan J, Ib ikci H, Delgado A, To en J, Somme R, Rashid A (2012) Signi icance o phospho us o ag icul u e and he en i onmen in he Wes Asia and No h A ica egion. Ad Ag on 114:91–153. doi:10.1016/B978-0-12-394275-3.00004-3 Recena R, To en J, del Campillo MC, Delgado A (2015) Accu acy o Olsen P o assess plan P up ake in ela ion o soil p ope ies and P o ms. Ag on Sus ain De 35:1571–1579. doi:10.1007/s13593-015- 0332-z Saa ed a C, Velasco J, Pajuelo P, Pe ea F, Delgado A (2007) E ec s o illage on phospho us elease po en ial in a SpanishVe isol. Soil Sci Soc Am J 71:56–63. doi:10.2136/sssaj2006.0162 Sánchez-Alcalá I, del Campillo MC, Ba ón V, To en J (2014) The Olsen P/solu ion P ela ionship as a ec ed by soil p ope ies. Soil Use Manage 30:454–462. doi:10.1111/sum.12141 Sánchez-Alcalá I, Campillo MC, To en J (2015) C i ical Olsen P and CaCl 2 -P le els ela ed o soil p ope ies: esul s om mic opo ex- pe imen s. Soil Use Manage. doi:10.1111/sum.12184 San ne J, Mannel M, Bu ell LD, Hoe e C, K euzede A, Wenzel WW (2015) Phospho us up ake by Zea mays L. Is quan i a i ely p edic - ed by in ini e sink ex ac ion o soil P. Plan Soil 386:371–383. doi:10.1007/s11104-014-2271-x Sch öde JJ, Smi AL, Co dell D, Rosema in A (2011) Imp o ed phos- pho us use e iciency in ag icul u e: a key equi emen o i s sus- ainable use. Chemosphe e 84:822–831. doi:10.1016/j. chemosphe e.2011.01.065 Shi ani M, Sha ia mada i H, Kalbasi M (2005) Phospho us bu e ing capaci y indices as ela ed o soil p ope ies and plan up ake. J. Plan Nu 28:537–550. doi:10.1081/PLN-200049235 Soil Su ey S a (2010) Keys o soil axonomy, 11 h edn. USDA-Na u al Resou ces Conse a ion Se ice, Washing on, DC Sui Y, Thompson ML (2000) Phospho us so p ion, deso p ion, and bu - e ing capaci y in a biosolids-amended mollisol. Soil Sci Soc Am J 64:164–169. doi:10.2136/sssaj2000.641164x Tandy S, Mundus S, Yng esson J, de Bang TC, Lombi E, Schjoe ing JK, Hus ed S (2011) The use o DGT o p edic ion o plan a ailable coppe , zinc and phospho us in ag icul u al soils. Plan Soil 346: 167–180. doi:10.1007/s11104-011-0806-y Tang X, Ma Y, Hao X, Li X, Li J, Huang S, Yang X (2009) De e mining c i ical alues o soil Olsen-P o maize and win e whea om long- e m expe imen s in China. Plan Soil 323:143–151. doi:10.1007 /s11104-009-9919-y Van Vuu en DP, Bouwman AF, Beusen AHW (2010) Phospho us de- mand o he 1970–2100 pe iod: a scena io analysis o esou ce deple ion. Glob En i on Chang 20:428–439. doi:10.1016/j. gloen cha.2010.04.004 Wi he s PJ, Neal C, Ja ie HP, Doody DG (2014) Ag icul u e and eu o- phica ion: whe e do we go om he e? Sus ainabili y 6:5853–5875. doi:10.3390/su6095853 54 Page 8 o 8 Ag on. Sus ain. De . (2016) 36: 54