scieee Open visual document viewer

Rapid assessment of selected free amino acids during Edam cheese ripening by near infrared spectroscopy

Mlček, Jiří,Šustová, Květoslava,Rop, Otakar,Juríková, Tünde,Humpolíček, Petr,Balla, Štefan

Abstract

The study focuses on rapid determination of free amino acids produced during the ripening of cheese, by using near infrared spectroscopy. Analyses of 96 samples of Edam cheese (30% and 45% of fat in dry matter) were performed at monthly intervals up to the ripening age of 6 months. In total, 19 amino acids were analysed with infrared spectrometer using two different methods, either in the regime of reflectance in the integrating sphere of the apparatus or using a fibre optic apparatus with the fibre optic probe. Reference data based on high-performance liquid chromatography were used for calibration of the spectrophotometer. Calibration models were developed using a partial least square algorithm and tested by means of cross-validation. When measured with the integrating sphere and with the probe, the values of correlation coefficients ranged from 0.835 to 0.993 and from 0.739 to 0.995, respectively. Paired t-test did not show significant differences between the reference and predicted values (P < 0.05). The results of this new calibration method showed the possibility of near infrared technology for fast determination of free amino acids, which occur during the ripening of Edam cheese. The content of free amino acids allow us to prepare Edam cheese quickly and efficiently for sale or to prepare the material for processed cheese.

Full text

Rapid assessmen o selec ed ee amino acids du ing Edam cheese ipening by nea in a ed spec oscopy Jiří Mlček1, K ě osla a Šus o á2, O aka Rop3, Tünde Ju íko á4, Pe Humpolíček5 , Š e an Balla4 1Tomas Ba a Uni e si y, Facul y o Technology, Depa men o Food Technology, Zlín, Czech Republic 2Mendel Uni e si y, Facul y o Ag onomy, Depa men o Food Technology, B no, Czech Republic 3College o Business and Ho el Managemen , Depa men o Gas onomy, B no, Czech Republic 4Cons an ine he Philosophe Uni e si y, Facul y o Cen al Eu opean S udies, Ins i u e o Na u al and In o ma ics Sciences, Ni a, Slo akia 5Tomas Ba a Uni e si y, Cen e o Polyme Sys ems, Polyme Cen e, Zlín, Czech Republic Recei ed July 2, 2012 Accep ed Ma ch 19, 2013 Abs ac The s udy ocuses on apid de e mina ion o ee amino acids p oduced du ing he ipening o cheese, by using nea in a ed spec oscopy. Analyses o 96 samples o Edam cheese (30% and 45% o a in d y ma e ) we e pe o med a mon hly in e als up o he ipening age o 6 mon hs. In o al, 19 amino acids we e analysed wi h in a ed spec ome e using wo di e en me hods, ei he in he egime o e lec ance in he in eg a ing sphe e o he appa a us o using a ib e op ic appa a us wi h he ib e op ic p obe. Re e ence da a based on high-pe o mance liquid ch oma og aphy we e used o calib a ion o he spec opho ome e . Calib a ion models we e de eloped using a pa ial leas squa e algo i hm and es ed by means o c oss- alida ion. When measu ed wi h he in eg a ing sphe e and wi h he p obe, he alues o co ela ion coe icien s anged om 0.835 o 0.993 and om 0.739 o 0.995, espec i ely. Pai ed - es did no show signi ican di e ences be ween he e e ence and p edic ed alues (P < 0.05). The esul s o his new calib a ion me hod showed he possibili y o nea in a ed echnology o as de e mina ion o ee amino acids, which occu du ing he ipening o Edam cheese. The con en o ee amino acids allow us o p epa e Edam cheese quickly and e icien ly o sale o o p epa e he ma e ial o p ocessed cheese. NIR, ipeness o cheese, me hod o pa ial leas squa es, p o eolysis The key ac o o ageing in cheese is i s chemical composi ion. The wa e con en is one o he mos impo an ac o s in egula ion o cu d hyd olysis and mic obial g ow h in manu ac u ed cheese (Janš o á e al. 2010). Ne e heless, changes in and analysis o majo compounds p esen ed in young cheese a e no good indica o s o he cou se o ipening. This is because hese cheese componen s a e s able and mo e o less unchanging (He ing e al. 2008). On he o he hand, howe e , changes in con en s o pep ides and ee amino acids a e much be e indica o s o cheese ageing. Cheese acqui es i s ypical as e, smell, consis ency and appea ance only in he cou se o ageing, due o e men a ion p ocesses ha change all h ee basic milk componen s, i.e. lac ose, p o eins and a . P o eolysis is he mos complex and in he majo i y o cheese ypes also one o he mos impo an biochemical p ocesses ha ake place in he cou se o ageing. The concen a ion o amino acids, he inal p oduc s o p o eolysis, is dependen on he ype o cheese and s age o cheese ageing. In Edam cheese, he ollowing amino acids a e he mos abundan : glu amic acid, leucine, a ginine, lysine, phenylalanine and se ine. The ype and quan i y o ee amino acids p esen in cheese in luences i s as e and indica es he deg ee o ageing and i s p og ess. The composi ion and spec um o ee amino acids change depends on he ype and age o cheese. Du ing cheese ageing, he concen a ions o amino acids a e inc easing wi h he excep ion o a ginine because i s concen a ion dec eases in he inal s age o ageing (McSweeney 2004). F ee amino acids hemsel es can subs an ially ACTA VET. BRNO 2013, 82: 191–196; doi:10.2754/a b201382020191 Add ess o co espondence: Ing. Jiří Mlček, Ph.D. Depa men o Food Technology Facul y o Technology, Tomas Ba a Uni e si y in Zlin nam. T.G. Masa yka 275, 762 72 Zlin, Czech Republic Phone: +420 576 033 030 Fax: +420 576 031 111 E-mail: [email p o ec ed] h p://ac a e . u.cz/ con ibu e o he as e o cheese because hey a e one o i s senso y componen s (Lane e al. 1997). Good knowledge o he s age o cheese ageing is he e o e an impo an ac o o i s u he de elopmen and o he p oduc ion o p ocessed cheese (Bunka e al. 2004). The aim o he s udy was o e alua e he possibili y o using nea in a ed spec ome y o de e mina ion o ee amino acids which a ise du ing he ipening o Edam cheese as an impo an indica o o he p oduc ion o p ocessed cheese. The esul s ob ained by measu emen s pe o med in he in eg a ing sphe e and by p obe we e compa ed. Ma e ials and Me hods Analyses we e pe o med on Edam cheese samples o igina ing om wo dai y ac o ies (ma ked as A and B) in he Czech Republic. Each ac o y supplied ou di e en p oduc s. Two o hem con ained 45% and 30% o a in d y ma e (FDM), espec i ely, and wo we e so-called s a ing cul u es (Lac ococcus lac is subsp. lac is, Lac ococcus lac is subsp. c emo is and Lac ococcus lac is subsp. c emo is, Lac ococcus lac is subsp. lac is, Leuconos oc mesen e oides subsp. c emo is, Lac ococcus lac is subsp. diace ylac is, S ep ococcus he mophilus, Lac obacillus hel e icus). Each o he samplings in ol ed h ee samples. Analyses o hese samples we e pe o med a mon hly in e als up o he ipening age o 6 mon hs. The cen al pa (co e) and ind o cheese we e e alua ed sepa a ely and he e o e he o al numbe o cheese samples was 96. E e y mon h he cheese samples we e analysed o he p esence o ollowing 19 aminoacids: soluble y osine, yp ophan, a ginine, o ni hine, lysine, his idine, glu amine, se ine, aspa agine, glycine, h eonine, alanine, p oline, aspa ic acid, aline, glu amic acid, leucine, isoleucine and phenylalanine wi h he aim o calib a e he Fou ie ans o m in a ed (FTIR) appa a us o indi idual componen s which change in dependence on he deg ee o cheese ipeness. Concen a ions o ee amino acids we e es ima ed in he liquid ch oma og aph (HPLC) Agilen 1100 Se ies wi h he mass de ec o Agilen MSD 1456 VL (Agilen Technologies, USA) based on he me hod o A do and Polych onidau (1999). All eagen s including s anda ds we e p o ided in he EZ: aas ™ amino acid analysis sample es ing ki by Phenomenex Inc. (To ance, CA, USA). Da a ob ained om HPLC we e used o calib a ion o he nea in a ed spec opho ome e . Edam cheeses we e analysed wi h he Nicole An a is nea in a ed spec ome e ollowing he p ocedu e acco ding o Mlček e al. (2011). Two di e en me hods we e used: in eg a ing sphe e (IS) and ib e op ic appa a us wi h a ib e op ic p obe. Calib a ion models we e c ea ed wi h he TQ Analys e sion 6.2.1.509 (The mo Elec on Co po a ion, Madison, USA); s a is ical me hods used we e simila as desc ibed by Mlček e al. (2006). Resul s Al oge he 96 samples o Edam cheese we e used o e alua e he amino acids. The esul s ob ained by measu emen s pe o med in he in eg a ing sphe e and using he p obe we e compa ed. All de eloped calib a ion models showed high co ela ion coe icien s (Table 1). The highes co ela ion coe icien s we e ound in models o phenylalanine (0.993) and yp ophan (0.992) in samples o Edam cheese measu ed in he in eg a ing sphe e and o phenylalanine (0.995) and o ni hine (0.993) measu ed by he p obe. When using he in eg a ing sphe e and p obe, he lowes alues o co ela ion coe icien s we e ound in he calib a ion models o glycine (0.835) and p oline (0.739), espec i ely. In gene al, i can be concluded ha he e we e no ma ked di e ences in alues o co ela ion coe icien s when using bo h a o emen ioned me hods. Co ela ion coe icien s o alida ion anged om 0.733 o 0.983 and om 0.558 o 0.960 when using he in eg a ing sphe e and p obe, espec i ely. Simila ly as wi h calib a ion esul s also in case o he co ela ion coe icien o alida ion he bes esul s we e ob ained o phenylalanine (R = 0.983 and 0.960 when measu ing in he in eg a ing sphe e and using he p obe, espec i ely). The wo s esul s we e eco ded o glycine (R = 0.733) and p oline (R = 0.558) when measu ing wi h he in eg a ing sphe e and using he p obe, espec i ely. Table 2 indica es ha in case o he p obe, be e s anda d de ia ions o calib a ion alues we e ob ained o en amino acids (o ni hine, lysine, his idine, glu amine, glycine, h eonine, alanine, aspa ic acid, leucine, phenylalanine) whe eas be e s anda d de ia ions o alida ion alues we e ob ained only o i e amino acids (his idine, y osine, glu amine, glycine and alanine). Excep o he amino acids men ioned abo e, o he amino acids 192 showed be e esul s ob ained by measu ing in he in eg a ing sphe e. This means ha he in eg a ing sphe e can be ecommended o he es ima ion o amino acids in cheese by means o NIR spec oscopy because lowe e o alues a e ob ained. Fo all amino acids he alues o calib a ion coe icien s o a ia ion (CCV) we e highe han 5% and he alues o p edic ion coe icien o a ia ion (PCV) exceeded 10%. The inc eased a iabili y esul ed om complex dependences and also om a mino p esence o amino acids in cheese. The lowes alues o CCV we e eco ded o phenylalanine when measu ing bo h in he in eg a ing sphe e and using he p obe (CCV = 6.48% and CCV = 5.84%, espec i ely). As a as he alue o PCV was de ec ed, he lowes alues we e eco ded also in case o phenylalanine (PCV = 10.12% and PCV = 15.81%, espec i ely). Discussion NIR spec oscopy shows good abili y o se alues o a s, wa e con en and p o eins in na u al and p ocessed cheese and he ob ained esul s co ela e well wi h o he e e ence me hods (Rod iguez-O e o e al. 1995; McQueen e al. 1995; Adamopoulos e al. 2001; Cu da e al. 2001; McKenna 2001). These me hods a e o in e es especially when moni o ing he p ocess o cheese ageing because hey enable analysis o changes in hei composi ion. The composi ion o young cheeses is a key ac o o hei u he ageing and he mois u e con en is one o he mos impo an ac o s o he egula ion o cu d hyd olysis and o cheese mic obial g ow h. 193 Table 1. Values o ee amino acids in Edam cheese, es ablished by e e ence me hods (n = 96), calib a ion and alida ion esul s o ee amino acids in Edam cheese es ablished by in eg a ing sphe e In eg a ing sphe e x ± SD (nmol/g) Calib a ion Valida ion R SEC (nmol/g) CCV (%) R SEP (nmol/g) PCV (%) A ginine 10.57 ± 7.68 ns 0.970 1.87 17.69 0.837 4.22 39.92 O ni hine 230.06 ± 146.96 ns 0.977 31.6 13.74 0.966 38.1 16.56 Lysine 258.89 ± 149.34 ns 0.958 42.7 16.49 0.930 55 21.25 His idine 66.40 ± 46.23 ns 0.929 17.1 25.75 0.874 22.6 34.03 T yp ophane 17.43 ± 9.03 ns 0.992 11.6 6.65 0.949 28.6 16.41 Ty osine 95.85 ± 45.43 ns 0.949 80.9 9.60 0.884 213.0 22.22 Glu amine 166.61 ± 118.54 ns 0.983 22.1 13.27 0.915 47.9 28.75 Se ine 162.23 ± 90.15 ns 0.912 36.9 22.75 0.881 42.7 26.32 Aspa agine 282.07 ± 177.73 ns 0.890 81 28.72 0.849 94.1 33.36 Glycine 126.75 ± 85.32 ns 0.835 47 37.08 0.733 58.6 46.23 Th eonine 110.59 ± 71.19 ns 0.941 24 21.70 0.918 28.2 25.50 Alanine 238.52 ± 160.57 ns 0.848 85.2 35.72 0.788 99.4 41.67 P oline 200.80 ± 179.42 ns 0.935 63.8 31.77 0.821 102 50.80 Aspa ic acid 46.75 ± 29.45 ns 0.975 6.5 13.9 0.961 8.16 17.46 Valine 292.14 ± 232.48 ns 0.967 59.5 20.37 0.940 79.5 27.21 Glu amic acid 610.20 ± 453.28 ns 0.959 129 21.14 0.919 179 29.34 Leucine 497.40 ± 380.19 ns 0.973 88.2 17.73 0.953 115 23.12 Isoleucine 92.57 ± 85.88 ns 0.969 21.2 22.90 0.940 29.3 31.65 Phenylalanine 274.69 ± 150.39 ns 0.993 17.8 6.48 0.983 27.8 10.12 x - a e age alues, SD - s anda d de ia ion, R - co ela ion coe icien , SEC - s anda d e o o calib a ion, CCV - calib a ion coe icien o a ia ion, SEP - s anda d e o o p edic ion, PCV - p edic ion coe icien o a ia ion, ns - non-signi ican di e ence (P < 0.05) In he cou se o cheese ageing, he cu d is hyd olysed by enzymes o igina ing om enne , by na i e enzymes o milk, and by mic oo ganisms p esen in he cheese ma ix. Casein is deg aded o pep ides and hese a e u he deg aded o indi idual amino acids ha can become p ecu so s o a bad as e, which is caused by a ious acids, alcohols, aldehydes, and ke ones (McSweeney e al. 2000). Types and amoun s o ee amino acids p esen in cheese in luence i s as e and indica e o cha ac e ise he s age o ageing. The spec um o ee amino acids in cheese changes in dependence on i s ype and age. The au ho s ound ha he e we e di e ences in he con en o ee amino acids no only be ween indi idual ypes bu also be ween indi idual ba ches o cheese; hese di e ences we e caused by a ying con en s o d y ma e , sal , amoun s o added lac ic acid bac e ia, and o so-called non-s a e lac ic acid bac e ia (NSLAB) which occu in cheese unde so a unexplained condi ions (Øs lie e al. 2004; Øs lie e al. 2005). The deg ada ion o some speci ic ee amino acids is impo an o he de elopmen o senso y subs ances con ained in cheese (Y on e al. 2001). Me hionine is deg aded o me hane hiol, which is he cause o unpleasan sulphu ous s ench o cheese. The deg ada ion o b anched-chain amino acids (leucine, aline and isoleucine) and o a oma ic amino acids (phenylalanine, y osine a yp ophan) esul s in o ma ion o a ious aldehydes and alcohols (i.e. ue a oma ic cheese componen s) and o a ious ca boxylic acids (i.e. p ecu so s o a oma ic compounds) (Y on e al. 2001). Recen ly i was ound ha also deg ada ion o se ine (Se ) esul ed in p oduc ion o s ong a oma ic compounds (Liu e al. 2003). 194 Table 2. Values o ee amino acids in Edam cheese, es ablished by e e ence me hods (n = 96), calib a ion and alida ion esul s o ee amino acids in Edam cheese, es ablished by p obe P obe x ± SD (nmol/g) Calib a ion Valida ion R SEC (nmol/g) CCV (%) R SEP (nmol/g) PCV (%) A ginine 10.7 ± 7.57 ns 0.925 2.87 27.15 0.821 4.35 41.15 O ni hine 242.64 ± 152.7 ns 0.993 18 7.42 0.946 49.5 20.40 Lysine 285.73 ± 186.02 ns 0.976 40.2 14.07 0.927 70.1 24.53 His idine 71.56 ± 53.91 ns 0.970 13.1 18.31 0.911 22.2 31.02 T yp ophane 16.89 ± 7.96 ns 0.951 24.7 14.62 0.819 46 27.23 Ty osine 93.74 ± 44.04 ns 0.957 128.0 13.66 0.929 163.0 17.39 Glu amine 171.09 ± 118.48 ns 0.991 16.1 9.41 0.931 43.4 25.37 Se ine 160.39 ± 86.18 ns 0.900 37.6 23.44 0.844 46.3 28.87 Aspa agine 316.46 ± 234.94 ns 0.937 82.3 26.01 0.901 102 32.23 Glycine 155.73 ± 119.56 ns 0.940 40.7 26.14 0.888 55 35.32 Th eonine 117.98 ± 76.06 ns 0.986 12.8 10.85 0.889 35.2 29.84 Alanine 264.84 ± 193.46 ns 0.912 79.3 29.94 0.873 94.8 35.80 P oline 188.34 ± 144.10 ns 0.739 97 51.50 0.558 122 64.78 Aspa ic acid 51.68 ± 29.00 ns 0.981 5.62 10.88 0.934 10.4 20.12 Valine 306.08 ± 244.40 ns 0.946 79.4 25.94 0.877 118 38.55 Glu amic acid 656.52 ± 471.24 ns 0.955 140 21.32 0.915 191 29.09 Leucine 577.19 ± 416.78 ns 0.979 84.5 14.64 0.935 148 25.64 Isoleucine 121.70 ± 120.81 ns 0.975 26.7 21.94 0.916 48.7 40.02 Phenylalanine 282.67 ± 160.03 ns 0.995 16.5 5.84 0.960 44.7 15.81 x - a e age alues, SD - s anda d de ia ion, R - co ela ion coe icien , SEC - s anda d e o o calib a ion, CCV - calib a ion coe icien o a ia ion, SEP - s anda d e o o p edic ion, PCV - p edic ion coe icien o a ia ion, ns - non-signi ican di e ence (P < 0.05) Good knowledge o he s age o ipeness o na u al cheese is impo an in o ma ion o he subsequen manu ac u ing o p ocessed cheese (Fo man 1996; Fox 2004). Based on he highes alues o co ela ion coe icien s o bo h calib a ion and alida ion expe imen s, he model de eloped o he es ima ion o phenylalanine using bo h a o emen ioned me hods can be ecommended as he mos eliable. Es ima ion o p oline, alanine, aspa agine and glycine a e he leas eliable because hey showed no only he highes di e ences be ween coe icien s o co ela ion bu also he highes a iabili y. Pa ame ic - es co obo a ed ha he e alua ed e e ence and ins umen al me hods did no gi e di e en esul s ( s a < k i ). Fo all models used o he es ima ion o ee amino acids, NIR spec oscopy seems o be sui able o o ien a ion pu poses because only in he model o phenylalanine he PCV alue measu ed in he in eg a ion sphe e was a he le el o 10% (PCV = 10.12%); whe eas using he p obe he CCV alue was equal o 5.84%. Fo he measu emen o his amino acid we can e alua e his model as eliable. Howe e , compa ing bo h me hods o measu emen , measu ing wi h he in eg a ing sphe e seems o be mo e sui able. Al hough he es ima ion o ee amino acids by means o NIR spec oscopy ga e, as expec ed, a li le wo se esul s (mainly due o he ac ha compa ed o he basic componen s o cheese, he accu acy o es ima ion o some less equen compounds was lowe ), i is possible o ecommend he use o NIR spec oscopy o an o ien a ion es and es ima ion o hese subs ances mainly due o i s speed. I could con ibu e no only o he e alua ion o he deg ee o ipeness o Edam cheese bu also when selec ing op imum aw ma e ial o making p ocessed cheeses. Compa ed o adi ional me hods o de e mina ion o he chemical composi ion and quali y o cheese, he NIRS me hod enables a as , simple and simul aneous e alua ion o mo e componen s a once and p ope ies o cheese wi hou he need o any chemicals. I is also a non-des uc i e me hod sui able o ood e alua ion e en in in e -ope a ional analysis. Acknowledgemen This s udy was suppo ed by p ojec s NAZV KUS QJ1230044 which we e inanced by he Minis y o Ag icul u e o he Czech Republic. Re e ences Adamopoulos KG, Goula AM, Pe opakis HJ 2001: Quali y con ol du ing p ocessing o e a cheese – NIR applica ion. J Food Compos Anal 14: 431-440 A do Y, Polych onidau A 1999: Labo a o y manual o chemical analysis o cheese. COST 95, Luxembou g. Ed by Luxembou g: O ice o O icial Publica ions o he Eu opean Communi ies Buňka F, H abě J, K áčma S 2004: The e ec o s e ilisa ion on amino acid con en s in p ocessed cheese. In Dai y J 14: 829-831 Ču da L, Kukačko á O, No o ná M 2001: Nea -in a ed spec oscopy and i s applica ion o milk and dai y p oduc s analysis. Chem Lis y 96: 1-18 Fo man L 1996: Dai y echnology II. Ed. VSCHT, P aha, 228 p. Fox P 2004: Cheese: chemis y, physics, and mic obiology. Else ie , Ams e dam, 617 p. He ing P, Hanuš O, F elich J, Py loun J, Macek A, Janů L, Kopecký J 2008: Rela ionships be ween he esul s o a ious me hods o u ea analysis in na i e and en iched milk. Czech J Anim Sci 53: 64-76 Janš o á B, D ačko á M, Cupáko á S, Přidalo á H, Pospíšilo á M, Ka píško á R, Vo lo á L 2010: Sa e y and quali y o a m esh goa ’s cheese in he Czech Republic. Czech J Food Sci 28: 1-8 Lane CN, Fox PF, Johns on DE, McSweeney PLH 1997: Con ibu ion o coagulan o p o eolysis and ex u al changes in Chedda cheese du ing ipening. In Dai y J 7: 453-464 Liu SQ, Holland R, McJa ow P, C ow VL 2003: Se ine me abolism in Lac obacillus plan a um. In J Food Mic obiol 89: 265-273 McKenna D 2001: Measu ing mois u e in cheese by nea in a ed abso p ion spec oscopy. J AOAC In 84: 623- 628 McQueen DH, Wilson R, Kinnunen A, Jensen EP 1995: Compa ison o wo in a ed spec oscopic me hods o cheese analysis. Talan a 42: 2007-2015 195 McSweeney PLH 2004: Biochemis y o cheese ipening: in oduc ion and o e iew. In J Dai y Technol 57: 347–357 McSweeney PLH, Sousa MJ 2000: Biochemical pa hways o he p oduc ion o la ou compounds in cheeses du ing ipening: A e iew. Lai 80: 293-324 Mlček J, Rop O, Dohnal V, Šus o á K 2011: Applica ion o nea in a ed spec oscopy o es ima e selec ed ee amino acids and soluble ni ogen du ing cheese ipening. Ac a Ve B no 80: 293-297 Mlček J, Šus o á K, Simeono o á J. 2006: Applica ion o FT NIR spec oscopy in he de e mina ion o basic chemical composi ion o po k and bee . Czech J Anim Sci 51: 361-368 Øs lie HM, Eliassen L, Flo aag A, Skeie S 2004: Pheno ypic and PCR-based cha ac e iza ion o he mic o lo a in No egia cheese du ing ipening. In J Food Mic obiol 94: 287-299 Øs lie HM, Eliassen L, Skeie S 2005: Pheno ypic and PCR based cha ac e iza ion o he mic o lo a in P as cheese du ing ipening. In Dai y J 15: 911-920 Rod iguez-O e o JL, He mida M, Cepeda A 1995: De e mina ion o a , p o ein, and o al solids in cheese by nea -in a ed e lec ance spec oscopy. J AOAC In 78: 802-806 Y on M, Rijnen L 2001: Cheese la ou o ma ion by amino acid ca abolism. In Dai y J 11: 185-201 196