. n 1.. 0 V. . V... “V. .. ..0.0 d 1.. C ,0. V. ..V ..u... ..0 .0. 37v ~ 7 u . V V. .. r . - 1 V: ......V 0 1. v. . ... 0.~ 9:“. 70.. I. .0.~..T.....L..F1wb0vu ’30.. ....u£i.~0.h 0.34Y‘0m80m0 L”: $...J~ “I To, 0 .AV? . A .1. . .1 1 030.0. .. .4011.- 0 .0. V 00. 1 ...0 010.00. . V 0 .05 \ V . t. . V0 .0. .0. .00. v0 .0. 74. .0 52.0 . 11.0.. .40... . I ... 1. 0..: ..N. .m... . o r .. . .flNVflS . 0.. 3.01 ..V... w o .. .01.. . 0 is“ .03 . . a? .0.. .0 0.3“} .u 1V“ .0 0 .0 V. . .0.0..0 000. 0...0I.~00.0 00. 03.0000 0..}. V .00.. 0. 0. 1 . 000.0 0000 :0- . . .V. 0 A. ..040 0:. .0... u 03003.0 1 . . .0..0000.4..0.V ..a’. . a. . 000 au‘0oo00 .0. 1.00 .00000 -0... .. .-. 000....00’ 0 00.... . . V . .- .0.00 .03. ‘7. . .0. 0..-. ... .20 .‘0 \ . . . C. 0..!0.0..0’..J V V 0 . .V 9 . ..A V .. . .0... 0..-0.0 Z... .7 0.0. . . 0 00. u, .0 o. 0 ...V .23.... En .u... b...».u.Vr .56 . 3...... 0.1.00.0: . ~ 00“ ..0 .206. 0.0.. 0 0 . ..0 .. .,0 . n v.“ V». ..uuu§§...n. .V 0 000.0 0. 04.00 O . .V . 0 0a. '0 .0. 0.. 0M} 00 50.0 .0... “0 ~ 03...... .734... 0.3%“ .3 “\f&.%. V V . .V 0 I. . V L .104. .2... .0000..f..800. . . u .00 I...» 0. .1. 3 . .0 . 0034 .. .2u3001 401......000H. 00.; .V... :30! .00 .0... 0 0 3.0.3.0.... .. . .t. n V... .. 1 ..V .. . :.:.V 2...... . V :9... 3.9a .1. .V. . $0.03!)..ch . V 8.0%.. .V... 2.3%» n .: 1. ..000 V . . 0 .. L0 00- 0.0 “0... V . . I. m02...flufi0h~ dfuvcirrm". 0 00.00 h who-hm . t 0. . .V...I. 0 2....00 0. . 0 V V. .. 5 ..m. 0. V. _ 9...... 3.. aux... . .. A... V .3 .. .a. 4.. .0“: 3.3.0..." . . u. 1. V. . . o. .4»... V. :2. 03:00.. no.“ .. . . . a.» 0.. . 13000.10: 0 IV 00....a.00.0. 0 .0030. .0 .V ..MWOHUL . “on"... .«00..00A ...0.4uo~.v0V400"0.0 0.0 v 00. 0....00 . . 0 '0 V 0 . . 00.. .. S.....J..0O.“$.’-ci 3.3:..de . 0“ 'd...‘ 0"; 03.33... 0 .000 ‘0 0.0.0 0. 0‘ . 0“ "flw. II 0.. $0.03.. Lfluhfldn-‘dkmawocwmfiuhmu-‘D. tn.“ ‘0 090. ”0 1.0.. 0 .00!".0.r 00 .0 [00.0000 )“ ‘8 V 0‘ 00. .0 00.000 ..nfimmwuaunm 0“ 000“” 2.031.“.0 “dw‘az. 0! an .- 000000 . 01.0. 0.50... 1.4.... 00 0.. V. . . 00.. V . .. . .033. 0 ..0 0.. 00. .0 .2 V. .0 0. 0..V0.ln...0 . 00-. . 0.. 0 0 0.0 :. . 0 V V «0.0000. $0000.03 0003 . . . 0. 7... .0 0...0!..0v.A..00 000.. . . 0.; 00 .0. 0 . . .0 0 :0. 00.0 0.. 0 .0v . V .3... f0 . 0 V .0. 0 . .. 0. 0.....- 00.0 0. r...... 300....- “a 0:0. .. '0. . . ~. ..0 . 0.00. 10: . Ix 00‘ v“, .A 0.0.01. .. o .3. V “M. .0 0w. .0. V V V .0.0 0.0.0. 00V. 0.0 u 0. 0. .90... 00 a. .. . ~00“ .00000‘fl‘00011040f‘. 0. r0 0..: J 0 "3+“? .008 .V. 3.3.0 . ..0003-20. V. .X V ..0. .000. . :09. 902.01.. 0 003050.00. V... 0.: ‘00}. 00 0. , - 00000 00 . V . it. v. 0 u 3. x . b . i t; . .L r . .. 0.: 3....000... all 000. . n . .V. 0L0. V . V . ..00000... :0 'éai. 0 ..03 4.“ 0.000.. 0. V 3.. .00 . . .4CV ..Vw .. .- mhi.flw.u (.VVHV. VrAV. .3...— . 0. V: 0.! . . V. .11 .0 .008 .1 1.0-0.0...23 Z... :0 . V V V V V V. 0. . 0 30 00. .0 2 .V .0. : h .A a k. V . A . VV . - .00....0 0V. . V .000. .0000.’ v . 0.....u00n001010 00qu 00. ..0..0.0 .. 0v. ... . . V . .. . . 0 0 fags}. .0 0" 0nd. in”: “.0"; 4 0 4 0 0‘ 7 “Cs. 'rnst.vw~fuxy‘0}1~uu.bflxq ..0. .10.. 3.3.... 3.: 3.1.42.0}..avfio0. . V V . V 1.30000. .. .00 V 0.0.3.020... . V V V V 3...: £0 . 000. u 0. V . . _ . fif....:.u:r.s1tzb: 1.00... .V V V... 00 .V .A .000 . I... 0 ..0. . 0000.000 :00 0... . . 0 . 0.. 0 V V . 0 0 0 1 0%“... IV .0 .1131 V t ..V... .0. .0... 0..... 0.8.0. . . V . 30' 4 0 .l.....0..u£3.0..0000000 0 V ..0 0‘ . IV. 0 A.0 1.0. .IYHLVBV)0(0.VL. 1&0... .0 .0 .V 000130 0 0.00 V . . V V . u .000. 03.00000 0. - .00. 00k- 1 00 ’00, 0r.0.1.rl\ .VV); VL... 01$. 0.?‘000 . 000‘... ..d. 0700.00 . VV 0.00.1.0. 0.000 .0 . .000“! 0 kr-Ln‘h .0‘0‘ .r 01.4..» .A r . . . V . A A .000 40.: 7 . ..v. 0 . . .V. :1... . . 0..-0.0.Vufl0 20““ .0 "9.. V V .-0.....v|r. ... ..V.... 00 . .v . V00 «00 0 1......0..._..0 . 000A. V 5.0060 0 .3 .0 0. '10. 0 $00 .&.0000'00W0-0 a 00 .V .0 . V ..0 .. _ .00... V 0 . V 0 0 .0. A ..0000.000. 000...: 0 .3200. r0 V .. ..V .0 0 .0 2. . . ..0 2.3.... 3.0 00.00.. ..0 00.0. ..v. .30). 001.00. . 0 n .V... V . V V .0 0000. 00.0.0003 .. ..A . . .0. I V. ... .. 30.0.0502..- 0 0. .0 0 .0. . .o . 0. .0 0‘ V 10.00 :23! V. 0. .020 a. 0.. .V V . V 1. . f. 0, 0 V V. . VV ..A... 00.0 2.00.0211 .Ca. V .7 00.00.00 .00 V 3 ..0 . .10 . 0.. V . V .0. V . V V V V . 00 0002.09... VV .0 . . V. a . . . f _ 0 V ... A . : . . 0.00.00 00000 . 000. 00.0000 000“ :0... .0. V .0... 00 00... . 0 V . V ’0 .0 0.00400 00 00 ...0.0 .3001". V0. .0 00.00 .0059 000100 1.0 o .0 ..00. . 0.. V .00..0.~..0.00. 0.0 00000 .0 .0J’V.4.0~ . V0 3. . . .00. . .30 . . . . .33.... 00.03 9.00. . . .. .00 . V 00 000.. .00 v 0 . . 0. . a . . c 0.0 . 0 0. 0. . r V . . . . 0.00 00 0000. .0. .00 ‘- 0...00-000 . V 0.4 . .90.. 00.09000 ... . 00". 000‘ 000 000- ’. . ... V V 0.0 .. L. 00 V. ..V.0 .. .0 .V 0 ,’00.lv <0. ‘0. V 00.0.0.0. . 0 00b . 0V. 0.00000 . 0.00.10. 0 0 .4... V0. ’. ..0 00.; V00. 4 00 10 .. 0. 5. 7|. . 00.. . A 0. .. V 0 . o . £00: a 0 .00 0" .00 0V. .. K 0.0 .0 . . 0.0. i o 00 00 0 00.1,: 00. u 0 V 2.0 0:250... .. .0000 000200. . .0VV . . V0 .00 000 000 03.0.03...) . . ii ..0 .0 1 0.. V. .. I3 3 000000. .00 0.0...0 0000.0. 0.0.10. 0... 0..-.00 02.". V- . '0. i 000 000000. 000000 .0 0 .0 0, 0 nzt’tn .U 0'10; 0‘ II .. 0.. 0 . 00.0000v-‘00 .00.0.30000 ”3305?”... (.IVV 00 .. . . c...” t . YV‘ 00.00000 .3 0 .0 000.... 2.. 3.3.0.;300500HJ0Q 0......000...'.l. o0 . 0 - A0. V V .I . . 0.100 0' 3’. 03.00.00! 0 .- .. . I. 0. 1.00. 000. . . 0 30.00 ’33001 .'. 000 #100: ..00000 0.. V 3.30.00 .0. o 3.. V. .0 05V“...0‘50.0~3 .. 0V... 0.... .0 . V.... i .. .0005 30.03. 0 0 V! .0 000. 1.00.. 0.0 08V . . 000 0 .2. ' . 3... ..JV .. 3.. 3.32.... it. ”V.Vtufiflr V ..3; 3 . 0 ‘I0 0. 000... V100. 000:00 - .0 0.10.0.0: 1.00.00 0.0 h... 00.00”."{0 V V -114: 0.. w. .’V.00a.031.. 1 .01... .00....‘31. 003.01.... . 0.. r. .0520...” fl VWV. ..H0 ytki.) T. .03 0..1000...0.0..I.00. .000 . 0.. .00»0..00 00.00. .000 000004.010...- 0 0 . V 4' V VI~0>‘.. .frr . $2.30... ..A. :1. a..- t V» :0. . . “0.....“ "tour-642%.“... “moan"... {03“. . 6:30.“): :00! $35.... . In. .t.t.;182t€ . . 0.000. . . 0 0 l I O L . V . . 8.02.... 8..."..b000..%31.3.0l 032...}... no»: on 214-. 30.0. 3 In! .0 Lu... 0.. 0?. V 1!“.m 130.. I 0 ..0 0. .0”...00050.V.0, .I 000.0 0.“ .000.- 000. .‘c. .".u0 §..‘t’r”fip 9‘ V Qfil 1. a 4... rt 1. .\ .0. .3... 0.. 1. 2.0.0.2? . 2.... .00.. 0.0 00 {coon-0.0.0 3.333.330. Sis-J0... 0.00. 1.... 00““. 5.0.... 000 {303“ J. 0 . J0 V. 1 1%.... V . . . 0 00 o . 000.00. . 0.9.00 0.07 .0 . 0 ...0 . :0. V0.0 000 .l 0 0.000 . 0‘T .51“? 0.00 0:00....00000000 :00. 0‘ 00.0! 0. 0 .0 d V .V.0.V..10.!3..0V V V a V. .2. .V V V ._ :83 3.). 30.001 .. x. w. 2.00.00 0.00000300‘ 5.00.00 3.... 80L}; ‘I.0'.00.0200 0 . . V . .V. VV. 0 V . 13A . 0 4.00030 .10 . . 000000. . 0.. . .0..\00‘0C0. .00-0'.00.40.)‘ .60 0 :0. . . 0.)}... . V 0.0.0.0 3’000: . ..o V. d..a.00...0OVAVfi. . 0..0 V V .0 . V . . V.0.0 . 3.000 4-0A . V . 100000 0’. . ”i .04. 0‘ a . . .03.: .. .. 0.020.. 0.2.02030100 00.00 50.3.). 3 ‘00..." 0 ogwmaa .0010 V V 6.32.00. .0.'.00000000..,:02~ 9%.. ..I .I.0-000 £00.00. 0000.. 0a. ‘ 3. V, . I .00.. . .000.” 30.3. 0......- AV. III . .. 00. . 3 .il 1......» ......02m. 00......au53. 00:0 . .unnV "30......0wbn0‘. .0850»: I 00. 0000.0 0 0..... .0. a 0 I. . V 4| 0.000 V . “v.03..." 0..0.u|.0 .0... 000 ..00. .1"... float.“ 0 . . 1.00 . .00 .00}: 0.00 0.00....L0 1013.0 .2. V0 . .V ..00 .00000 .0. .0 . 1' 1‘ 0 00 . 0. 0 V «Hr-.3..- zhflJV . 00105.... gin?! ”1 $2.00 0.0... . "IS'IQ'Ogt‘It10 H a.” .....V.... v... . 9...). J ... .10....00 I... .n a a .3808 3 .10.. 00.2.0...- .5 . .. a o .8“. 00. .0 .0 03.0.:0H30. 0§L0l0.u0..00\0«u—09I .0. Dantiiiy..oigl.2.§10. 00 V .. .0903... .3028... l 0 30.00.20 .3330. 2.0.3.0085]... . 500005! 0.0”:«0 . ‘0 «00.0 1000. V. 30.... 0.0. V200. {- 000‘... 00.2.0.- 03“}. $0....- 3, ..3 fl 000 00? 00 n . 00. 1. .0 .0... 0]....00. .0 0306.000. .00. 0.00.. .0..o03..0«,,3.......00~ 5.00.0 :0...) 0.003....)0-00 0020‘. 0000 .0 0 A.- 0 .0 00.0.00 0.... .7. .00 00000 00...... 0.0.0 30.! 0 0 . 0100.00.00 0....000 a... J 00.!00. 00000.00 .00.. Jon-‘03.". 0000].. .00 000 all... 3.1.... 9......- 0.....V.0.....V. . 3...... so... .. 0.1.43.3... . . 00.331.18.27. ....0.... V ......V V1.5.VSI..S..9.....0....:0.,.00. >00 000.100.? 09".}. . V . . . . .00 . a. . 0 .2 0. 0 . . Va... . . r .0. . 0 . ‘00 0.0.00... ..0 .3400 -0 ‘....0. 140.000.4300“? .0 . 0000."... 310.50300000...00.0.00:J0’00000 30000.00. V V. V ... .. ..0. 0.. . .15 0.01 c . . .0 .2833. . 30 ..J 10.0.. I 00 000 0..-0.0!! 0 0... '...0. .20... ‘2‘ .0 0.3 .‘3 000 .00 I000. «a 0.0 0 0...0..00. . . . V V. . 0 . . 0. .0000. n . 000 ..00 . 000 0..-0 0...! . 0. 0 0 0000.0 0‘ 0.4.0 0 0fi'h 00; 0&030? _V. 00’ “00000.000’0000“ 0.: au?_ru03: 0.0.: .0) .— .0 .00: . .0091; . . 30.3. .24.. ..n. V. . 00 .0000'3o0‘v0..0....0 . 0.0.0.000... . .A 00.00 0.:.V 4. .. 700000 0.00: 000.7900“ - 10.0.50 ’1V003...’3_0000 01‘3‘300’00.-.u’0.50.}...‘0! 0.20.0. V . .C r. . . ..V .01 1.. ... 71.. 00.00.35. . .. . .. . 3.0 0.0.x...00. 0 .V 0.. V 000 1.0. .. 020.042.000.000.:000 .0... 0.30.00.00.00.- 00000-‘0‘. (050.)...0.“”.. 0. 0. .9. _ 1.2.. O .0 2.3.0310 . V V.. .10 V. V . . 00.00:... 0.... 0.. . 00.11 00.000.331.02203. .10... o0.....00 In"... .20.... 4000.... 0 .00 0.0.0. 00.000000..N00.’0. .00 2 . . . V . . . . t V; V .V V. x .. . . V . V . V . . V 3 .V. . .. lo... ... o 2...... ...0 .4... . 40.0...0 t..u0.0.0..!10.o:.9 «2.1.0.151: 0.. ..10000.0..30.l.0....)9 .0 .shlo... 03.0.. n 2. . . v . ..0 V. 00 3...... . L V. 3. 1.0... V V V .01 .. 00. 0! .63.... ..0.000o.0.0....3. .0 . .0!...0...0...0.0.0...0 012.... 30000.1...) 3.4V. 0..-000......0‘. 0. 9.00w30300t1000900z515000V .0 1"? . V.“ v .V .0. 0. 0. 0 V >- V . 0 0. . .0 . V 000.....00 0 000000100 .0005. O. 00... V 0A. .0Vo\0’.0.0.. "00%.... .0 5.000 0.0.0500 0 1...! .0 000000- 00000. 00090.0000...00..\0.’0"“ 0;; 00", VV . . . . .0. . . .. V 0.0 .. 0. 0. .. ..n. . 7 0. I. L o 900 0.40.9. V. 00 ..0 . 0 V 04V '0 0 ‘0 000.00. 000.0000.‘0.01.... .30300: 00100 000.00.. 100.110.... 4. 00.00 40.0000 J . " .f.". .“00. A‘J‘. .- _ _ .0 ..V _ . V V .0 .3. 2.0-0.0...0'. ..3. .- .00 .00} 0VAV .00.. o... . . ..00 00000 ..0 .J 0 000.05....) 0.100 0000 I} ..0.’.¢.0l..00 «0.4V -0...000§000000. .02 ‘ 0000‘ .V. 0 0. .. . .. 0.. 1...... . .3... :2 D V..- o. . 00.0...SI: 0 ..0 . 2... 22.30.: ..V. V 0;. 3.. 8.. ...uv80: 39.030... . 08......05’7 ...\0 ..(00....3008 0!. “ .003W.l.0.300' 0..-:75 0.00 33%;. ..h‘o. . V. .. . -. V. . . ...V V V . V. I: ..5. ..V. . 9.. 5.0.3....) -23.. r 0...... 3... .. ... . . . 2. . . . .331 0...... 1. 1. 03.0 2...... .V. . .. 01. .2. :0...- 3: )..|.. :02! 00:... .01 .0 .320... If... .. 2 .1 . . ..J . ..V V... ........_.. 0.. 000.3 . .V! ..i 00. . .3 3.000 .V.. . 0. . .0. 0.000.000 V V 00. 0.20:... .021... 3700. 0 0.0.“)...00J ..00 4.30.0.3; ‘00. 301.000.53.055}, .00 10.00000... ..0..0...90.\00$0. 0.0.Qu0V0.$.0 0. 6.. g!“ 0.30000 . 0 .000. V V t V ...A 3 .0 . 0... .00 . 3 . . V0 .. . .u . 0. 00‘1“...” .0000. 03.! 0 n 0 c . u 0. 00000000 000,0"0 0.00- 0 0..-4.0 3.....- ... .IV.0-’00.v0.0000.000 ’0’.)0.0.‘.0 000 . .00 000.00 0'0 0 .V ..V. . .V . . _ V: 0.0 ... 2... .01.. V00 00 ..0. V... .A .. 0..-030000.11 .. .. 02’ .0 . .30.... 2.0 .0. 0 I. 000.0 0 0' 0.«.0»30..V-0...7.. 0.... 00.0.0. .00 u..000...000.0.0.0000 a...00.~0.«a0.0.0.00”0000| .0 .. . . V 3...: .0... . . ‘ 0V0 .13....‘0 .. V 00 r0. 0 .. 00014.3: .0! 0 00.00 - 0 9.009 . V I. . 0V . 70......- 0 0.00.0:«0'.0. \00 030.01.230.00... .0 .0000 (0“00‘00 0. .V V. .3: .- .7 . .V .83.... n . I V n. .0: .l . . V ......V0._V.V..a. ..0 0n. 1 .00.. 900 0.073... ....lgld0..d0..03..03: .00..000300000‘0..0.0.00000 . .. .. .0. . . .. .. 0.. ..A. . . 03.0 3.0!: ... .V .020; 70.. 300.00... 0..! 10“.- .230... 0001 . 1 .... 0 . .V .0. 2.... 00. .9 .. .00 20.01....I0. .. . . 0.... .V .500... ..2 .0...0I40A0V00.. . 0.. I .0 I. . 0.. .0 :0.- . V . V . 1‘... .0023 . . 00. ..0 0..-0.0 . VI, 0 .|0.00 00...} . 0 . 0.0.. . 0.. n . .0000...V000 I 0; 0. .. 00 0.0 00.00 V V . . ..uvl 0000.00d‘ V. V1.33... 3.... 0. .V . .A.V..0 .. .u... . 0.0.0.! 0 . 9.0.0.0.... 000...} I ... 0.. .. ..n 00. . 0 n v o .0 V. 00. . . .. V . a0 . .. . 0 ...V . V .00 .100... v . . ..0 . .30 . 0 0 . 5.00 .0. .. 0. 00 0 . V .. Z 0. 0:0. .0 V‘... .0 ... 0......0.0 0 . I130: .0 0000.30.00 V '0 .2. V . _ . . . . u I. 0...... ...0..000.I.V .. .0400 01 0.0.01.0 v"0..0V. 0. ..c 0000 0 0 .000... .0000}. 1.00.0000 0.:- . . .V I. I 0 V: . 0.0. 0.. .0 47.031300: .0 . n .V .0....A.A .000... V. .03.). 0.000 v.0:- . 0.... .030... 00?... 3.0—. '00 00...] 0 000.. 0. . . 0 0.0 .0.. 0 .. .100 . 0 0. . . 0 0 n . 0000. 0.0 0.. 00000 I 0.0 0.000 . .2. I 0 . 0 . ’00 0 o’d 0 0.00.. 0 .0 0 0.0.0 0 .. x .. . ..V V. l.‘ 0 0 0.. . 0 . . 0000‘ V 00.0.. .0. 00001.00. .5. 0.0.0.0 030 0.0.0 0.0.} 4,000. .1. . .. ._. , 0 0 V. . . .V .N‘ O V 0 JIO 20000.- .0 0001,0000 ’.!0 . v 0. 0 .00 I V ...V . 0' 90 V V0. V. . . V 0. 4.010 3.... V 0 V00 .3 ..0.V0 v 03.0003 00. 0.400. 00... «’00. 00. . A I 0 :3 V 0. O o. ..0. . .0 ....VA . .. .. . . . ..3 . .. . . . . V . V ‘2'. . .0: ..0 0.0. .2 0.. n . . r . ..V... .0. . V . . .. 30.0.040000.5.5.0...0200u0l “V . V b .000. . .000- 0 0 . 0 0 .V 0 . - 3. 0 00.. . V . . V 1 . o... 00.0. .0103? 00.0000 0.0.0 . V V . . V 00.0 00.00.!“00’... . . ..4 00.71.... V . .1): V .V V. V V . 0. . .o . V . V V . . . . V V... 06......0 .: a. £2.00). . n . .. 0.1:... :0. .0. : V.... V V... .920. ..o. . 3... . . r... 0.00.... 00 .22 0 0 .V.S........!0.V .01. ....l..0.0 V. V V 04.040.33.00... 5.1203000. . ... .0 7... . . . .. . .0 2 .140 .v 00.0. .. 0.. . V .1. . V . ..a. 3.0... . . V .00. . 300.....9’. 1.0 v . o .. J o .010 .V0.0 ....o0 ... .5 V ..0 . V0 .050..- I‘. 3.00.. . V 00. .00.0 V ..0’ 05,00 .0 . . . n . . V V .. . o 00 2 .. 0 . V . 0 .0 . . . 0 ...V 01.1000 . . 0 .0 00‘1’18 0. .0 0 .00300040. 0.000‘... c. . 0 . . V0: 1.0 . 000...... . u 0 . 0 0 ..V. I. V . 0.0. u... . ..A ..0 v 00... . 030003.... . . 0... . Al I. . . 0 . V . u 00 0 20..- 0'50.‘ 0. . .0V. .0‘.000000..V0.".l .7 ’ .00! 00. .0 0.. . .V . . V 00 . .0 o . V V V.V. .0 .. .. .VV. . . . .v . 8.0. .0 0 .. V V. 0.0 u . . n 0 0 . 00 O . 0 .0 V... V A ..0 00. 0..... 000...... 000.00. .9.- . 0.1...000.‘ .0000000 I 0000070 .1. . .V .: V 3.. .3 .3000... ..0 I... 3:... .10.. 0 . o. . 12.110 3.... 1:18.... 3...... . ..0V. ..00...61.......l...0V 2.... .. ..0. .0... 00.310... .0. , 30010.1...000...» . $0.... D 030.)..00‘.t0.\110.:030“8.o 0.003.; r .. V . 0 . . V . V . V... . . .. .0 . .V .0 A0. .0 n . 00v . . o o. .0}... .0 .I...0\ 1.-., . a 0:32.03... . ..0 . . . V 5... 3000.. 20.000040... 0.030000930030303 0.338.033: 0.. .06“. 300100.00’3 30.,00‘ 00'. 2 V0: 3.0 V . V 0.. . V V... . . . .0 V 00.50.... 7.3...0 0.0.1.0 .00. 0 a. V.V... V0... . 0 .... . 1.30.400. 2 0.0.. . . 0 .1000: 00 .380 .00.00...0‘1.0..0.’ 2000.31.00.20.‘ dz: 0.0.50.0...3,...‘130000.00.‘zx.0.? . 3000300000., . . . . . .V0 0 0 0. 0 V......... A 0-V V V 0.. V . . .0 .V. 0. V. 000. 0 0 3 u 0 I0 .0 V 400 0 u 0.0 0. .. . u I 0.. 20000.0 0 0 . I . . 0.100 0.. . .300,V0’00-00Q"O. 00V 00.0000)"‘0- 00 350.} l ‘0 0.0 I. V 0...... .0 . -00 3.0. .V . .v. V.. I 30.05:}. 0 0. V.. V 0 . 1.0.0 0.00 A . u; 0-V0000.V.0-00.00.0:000 .0. 0 .00.... 90.. .00’0.\.000‘001000‘l00. 00‘ 0‘00_’. .0‘ I) n n l“ ,‘00 0 .03 . . 0 - . V . 0 0 . 0. 0 0 ... . 1.- :00. .30... 0 .0511}: O 0 0.0 . 0 A00 . ..0L . 0. 0 0:000:50 00 . ’00' 2000., 5 I) a .0 ‘ 0...“ . . . . h- 0 . . i V 0:0. V . 0. . V... 4 0.00.000 3.0%. 0. .0 0 V0 0 .. 0.. .0 020. v 0 . 0....0 0.0000 00. . V ..0... )0. A V V V. .0- 000 I. . . V . . 00. .. n-l. 0 . .. V n 0. . .0 .00 3V0. . 0 . .0... 00 . . . ..l 0. s0 ’0...’ 0 .0. .0.0 ... 0 V‘- . . . 10’00’00: 0. V . V V . . . . 000.}‘0' V V 0. 00» . . . 1. u . . 0 ..V . . . :20 . 0 _ .1 . .V :1... a ..V... ..V... . 0 0...!3... I... .2 1...; ..... . 0..). 0013030000020! .000... . .0 0 .0 . V . .00.. 0..0.it.lui0.3‘. V. 0.0. 0. .. 0. . ..A 0.0 ..0 000‘ .00 r . . .0 0. . 0 0.. S..- .V ._ V. 0.2 .11. . 0 . 001V... .000 .00 1.0.00 0 .. 0 0. 00 V V . . V . . .. V V . 000,00 . . A . _ . . . . 2.... . 0.1.5.073. . V. .0 .. . . . 0 1.53.311... .050 0. . .V. .30.... 21.03.0500: . .V . . . V . V V V. o .0 V1.0 .0 V. . 0.. . A A V .. .000... 0... . l . I .01, V .0.0 V 0. . . . . V . . V V .. . . w 0 . . . . . .. 0.. .000 V0... .. . .l . . V . . V. V V 0000..0...'.0 .03 .000'0’00. .0.0.0.0 0.0.3....00 . . . V. . . . 0. 01 . . . . .0V V... V V V . V V. .65.! .. .. u. .0 V. V . . V . . . V V .0. 0 _ 0 V V. ... V 01 .0 . 0V 0 0 ..V .. V V V . . V 0.. . .. 0 .000. V 0.0. . 0.. 0.0000 .00. 00000.0. . . A V .r . 0 .1. ..-V V. ..f... I. .. t ... Y. I. . V. . . V V .00.. . ..n . .‘0<. 0..-0 . . 0.. . . V0 . ..V 0.0 0.. 3.0.. .. . V V . V V V V. . .. .. . 0.. ..0. V0.00: V . 00 u .0 4 | V 0 3.. . . . ...0 . 0 .v . .0v00. 0..0. '..0.0.y.0 00.0. 0. V... .‘00’0130 ‘0."00.a’ 00 . ..0000 . 70. V . V. .00 .0o. .4 3. 0.... 0 .3 V. .2 .3 0.0.0 0 .. 0001 00.0.0 z... . 0..-5,1001 0,000.0.” . .0 . . .0 .‘I ..V . V.V. V . 2......0 . . 0 .t . . V V . A V . 0.)}... 030.0000 000 ' .0V.. A . 0.. 0 . 0 .a . .. v. . 0.3.00. V . . V . . .. . . V 000000.. 03.1-00.0: 0‘01 .00 . . 0. V .0 . . .1... 0 Y..- 0.0. . . V. V . V 324 000000030. 00. . A .V....... .0! V . .......1 .......0.... .V. V .01.. . V . V .73113030010 ..0. V V 0 . ..0 .. VV. 0 0 V. V V .V . 3...)...- 0.0.. .00) 0. 0..!- V .» V0 V . 0 0. . V V .0 . 0.0.00- . 3.0. .0 . . V .. 0000 3.0.. . .000\0.....0000 .u. 1.201~.~!.9‘ g£'\0.fi . 0.0.0.0 03300.00 . . a: n0 ...0 . I l .V 0 o. 0:... 0. .3 u.0.0. 0.. . 0.0 0 0 . 000 u. 0 0 .1301 0:00.000 . 0.3.0000 . v a a. ..V V: 0 o 9 .V 0 . . . 0 0 . V... .1 0..; . 7 . 0 . . . . 0 7.. 0 .2... 00 0.00. .000. 00 ..A . 00.....0... .V.0v ’.0V . . .. . 0111003,... 0‘0; 0.3V . .L. . .. ..0 a 0 . .. 0 ..0 . ... . . . . _ . . . . . 00. . . 0 0.... .33... 11.0. . . .00 0.9 0 .00 0‘ . V I .70 0....3 0 . .' 00,000‘0015'I0.‘500‘~V‘i.}34uug\4 V . V .A V 0 ..0 a . c 0 0 0 . . 0.. c 0 . 0.0 0 0 . - o 0 .. 00 0 0 0 . 3|... 0." 0'. tzy.0.0.0.“"Oa‘lV .‘000‘0007‘.§u00‘¢1“0r'000“0 . 0 . 0. . .00 .0 ..A .. . 0 .20. .V . . 0 . . d. V 0 . . o . 0 .. .0 0. ._ . . 00 . 000. .. . .00 .000. .00., . 0 00... 1,50. 0!.0’00100‘000'0.-V‘.00 .050... ‘00:.nl 0:.5‘l0‘.‘ . \ 0.0 c .2... . 0 . .5. . . . . . 0.. o u 0 . . 0 . 0 f... .10 0 .0 0 .1 V. . 0.. V .0...00001. 0.0.0. ..000000...‘ .’0.0 0.. (,0...‘ 0. 0V'|\ .0 00. . $.100l0‘07 90.00.00 310....004000000‘0q03009100.‘ 2... V . . . .. . .. in . 0 .. . u 70.... . V V. ..V. . .V. V: . I. .1 .. V .010 . 3.. o . V . 0. 3.9.00... 0.... I0 . 0 100.030.2017.. 0 500. .12.. .0... l. .43....001h000 ‘..\¢!..-¢...;000..0.0.514.. 1t‘03:~.0.1~0.32lo‘. A . .9 : . . .V. V. v . . .. {7.0.700 i... . 2. u . V0.4 . 0 ... .o . . V. . .V. V V V V .0? 02.0%....0... ... V... . T... . . 00. 030. ..O0....l.!0 1:03. 53.300.50.00314-12000100.‘ ). 001,010.0‘.‘ 00.1%... 09.". ._ 0 a 7.0 . .17 V3 .40 . :0... .J. . . . . . .. V. V. . . . .V...V . . V V 1.3.0.... V V . V V . V V . V . o 0.. 39.1.9.1...- 0.. 0.40. 1.0 (to .Voo...0.‘\.0l. ..3 0.1.5.30 .10. .600 9.0.0 1.3311... 10.30.: .. 0 .V. V 0 0 0 0.0.0 0 . r . 0\ .1 0| . 0.0 00 . . 0 . . u . .. u u . . 0 V0! . .00 ..I. V. V V . 0 . a ..c .00; 0. A V V V 1.000.000," 0‘. 0 010. 00:5. 9. .0’0‘3’000'.\‘000II$V000 "0. 50.3 . a... . 0 . . . . ..V . V... A . 0 V . 0 .1 . . . V.V . . . V. V .. . . V V .V. . A ..V... 0. 1 v. {3000.05.03 ..V..0..7 0. 00’.00...0 0.000.000: “II..¢.\! 0 ..0. V .. .. . .V . 0 L 0 .. .- 0 V. - L. V .0 0 . c. .0 .0 .‘0 ..0 u .. u . . 0.0. .. . V 90‘.0000‘00.V0V\.V . «00.0.... .0 .V . ....1 . 0. . . . . .. V . ... .0; . ..... V 00.2.0... V‘.V..0 .00. 0V.- .33 0 0 0 v.0 . . 0.. . ...V ..0 ..0. V .0 0 0 . 0 0000.. 0 0.0 00 .0 0n. 3.... 0.. .I. .0 .. 0 .00.!000; 0. . V . 1 .V . 0 . 0 .0 . . . . v .0 . . 0 u 0 . 0. . .04. .0... . .. .03? :0 .002. . V .. a0 0 . . 02:00.. :0. 0. . 2.00 V. ... V0. .. 0 V n .5 V... .10... . .. 0 .. u . .010 V .. V ..010..... V .....-..0. 0 . V . . . ... .1 . .0. . 0 V V .A? . 0 . . V. ..0 0 . ..0. . .a 4... .. 0. . 0 .. . 0. .V0. . a 0 V 3‘00. . . . . f. . o. . ..0 0 V. ..V .00.. . 0 0 0.0... . o 0. u .0 u . . O 00 V... . V. .0 . 0.0.1.0.) V 0. 33.0 \0 00 3 .2 . .V . . 0..-0.. .00..- 1 .. .0' . . . . ..0 .. 00 V.“- ..,0‘..u0 ’00.; . '0 .V. . . . .V... . . o . I .0 . .....0VV0.0 . .0v 0.. 0 ._ .V . 0. .00 ...... .0 V . V0; . X .. 0 0 0.. .. . g . .0. V V . (I. ..000 . . 0..... 0. .1 . ..4 .0 . .uv . . 0 ...01.0.00. .- . . . ..00A ... 0.0l0.. 00.. 0...0. . V . 0.. .0..0 .. ...0.000 0.00: .01.... .V V . '3‘ 000 .. V .. .1. V. V; . :0... 0. 0. .V . V . .0 . V V . .. 0. V c .. V.. . . . .2 . .. . Y 0.7 V V. . 0... . . .. .0 . ...V.... .0..0 V . . V V \g‘V-..-0.(00 . 0 ..1’ . V.V _ . o 0. I. . . . 0 0... . .V . 0 000- ....0 V .00. .0..0..'....0000.000.. 00- .6.. 0.0.0 0.0... . V 0 .020..0 .00.0v.09..V . . . V . ’00. {.‘0‘ . Vno 0.3.00 . . A V .. . . . A V1.0 . ..00 . . .V 00 . .0000V3.0 .00 ..3 0. ...0 ..0. . 00.0.00. .: ..V V . . z0tt‘002' 0 . Y. . 0. V... A 00. 0 0 . .00 0 000‘ 0. 0.0V 0 v.0... 0 0. . 0 0.0.10 ...000..... V. . . u . 00V . .V .0. .0. 0 . o 0.. ..uuV. . 0.- . :1. 0..'00 . . . 000000 '0 . . . 0. . I _ 0 .0. . 0 . . .0 00 0 0 ..0 v0 000 n l 1 0.0V. .0 . 3.. 0 00. .0 3.0 0 00' .0 .000... ... 0 0 I 00 .. 4 .0. . 00 .V 0400 .30’0‘00000’. 0.0 0 .2 . 0 . V .. . V . . I 0 0. .0. . o o n - .. . . 0 A 0 0 a . . . .a 00’. $0 :0. . v 0 . . . 0 . V 0 .. .0! V.. AA... .00. . .20 o . .V .00... .0. 0 .V0 . ..000.0.. A 000VL.‘0. ‘I“ . V .. V... ..00: . V.0V.-0.000.0...0~ ..oV...V..A.00|.0...0..00..V..._ 00 .al...0.0-0V. 20000.03: 0 . V . . 4 . V V . . 0 v 0 0 o 0 . . .. V .0 0 a. . 0 V V _ 0 p 0 . 0A 00 . 0 0;. 0. 0 .v. . 0 .. (I0 -0 0A {00.000'0“00 0. I V 0. 0 .V 1 vv ..0 .30 0. V .. A. . V . . 0.1.. C v . Q. 010.. 0 u o .0... n . . .1. 0:. 0 0‘0... 0.0 0 . . ‘0 0 0. 0.0 0 .0.0 0.00.30.100“. . . V . ‘0‘£.lo0. 0 0 0 I V V 0 . ..0 0V .00. . V.. I. .. 0. V.. . 2.. .V 0.0 . . . 0 . .000 V .02 . IV-cA V 1.00... 0. 0.0. .15. V0 . o...0... . . 0. .V- 0 . ... ... V V . . 000. h 0 ! V ..0.. .. . . . . .0 .V. .13.. 0.0000 . 00 4 ..... I . 0 0 .0 .. V..n0. 00. 100... V . on... 700... o .00 . . V A .00. 200003034" ['0‘ .V .V. . .0 . .0 . 0. V . V .V . . . ..v V V.. c. . . V. . . . .0. .. . . VV .0 30...... . A . . . . .. . V .4000 .- 5‘ 0010100.: 0. 00..¢.0¢0..l0.00.00§ . .. ..0 .. V0 . 0.3.70.0- 0 . .0 .. u 0 ..0 .V .0. 3.. . .. 0.10 .0 w .2. . . .00 4 Z... 0 0 ..0V...00.\..V. . V .V. 0.0. 1...: 0 p0..§'vl.l¢10.l 100.00.00.01 0’ 0000.00.00; 00. . A 0. c . .. 0 . 0.... . . ..A... y .4. 31.00 . V L . s .V V V..V.V 00. a... ..Jr .00 10.5 0000.0.\ .0 .20.! 0|. 0. 0C0 .u.. 00,0.00 0o 0‘00 .V 0..) . ..A. .V. 0 p . c . . . V A. . 0. 0. . 0 . V 4.1.0055. 1900.10.30‘? 000. .. V . 0 . .. 0 0 . v0.. . .V V . V.-.. .013’ .0. VI. ..0 a. A . . 0 ._ . O .. ..0 V 0 . 0 ..0. 0.0.. . V V . 000200 .0} O. 0. I 0.... . .0 $000 0 . .0 0.300 ..0 . . . 0 V I .. . 0 00 0 .. A ....V....VV . .000 u... a. V . V 00.0010 1 . . I... .0. 0.. . .1 . . . . . .0 0 . .. . 0 o 0.. 0 J . .. . . 0 0 . 0 u . 00001.0. 3.0.0.; 3.0000021... 0.. . 00V. . ...VV. .. . . V. 0. . . V0 . V. .. .... . I. . . V 3.. . .. . ...V.., . . . . ..fi ._. V V... V.... I 0.. .0 v V a A. ... . 0 . 0 0V. .0. 0 000 0. . 0.0 0. . ... . . V. . 0 0 0.. 0. .0. . . 0 . . 0.. u o. . 1000 0.... 00.10.20.1'0. 0000. 0 . 00 0000) . .. 0.0 u .0. 2... V... V... .. . 0 0 V 0 .0 0....... 00.00.. .0 . .V 0... .0 V .V... ..V 0 . ... . . V . . V 30.03.01. .00!0 0.. Q VV..V.0 . V . V0.0. 0 ..v. 0...V 0 0 a... VA. 0 .. . .. 1.05... .V. c0000.... . V . .0 090.513.. 00.0.). 0.0 0- a... . .V . V V . V . .. .7901). V 0 V V. A 0 V . .0 _ v . 0 0 0. . .o 0 . 0 . .0 . V . 0. . . n . .00. .V V 0 0.. . . u 0 000 . V . . 0o. .0. V.. 0 . 0 V V . . . . . V. V . . VV . . 1... . u. .00 Va. . V.-. 0 0 0. V V. 0 . 00000 00 .9 0 0.0 .V. .01. 0 .0 V. .0. Z. 00.00.30. 0.30 ..V.0. ..0 ..0. 30-0—3003 .00....00. 0.3.00.0 2200. . V . V . . V V . . V . 01 .0 0 . .. . . 0 0 . V 0 2 V n . . . . 0 0 . I. . . .. .. 0. . V 0. V. ..V V . 0 - 0 0.. . 0. .. . . .V .510... . V V . . . . . . . V V a“ ." 0.0.10. 0.00 A\ .0! 0V. .2 .V0 . . . .- 0. . 0 . . V .0 ... 0... V0 0 . . .1 . . V V . a 0.0 .0 V V 10.00 0. .V . . 0. . V. .V. 0 .10: 1.... - .V. 0. .0 000 A. .0) (0..... V... .00 ..0 00.0.. 31.0... ..8. . . .V . . V V . V .0.I0..2~...‘ ’0V0.0000- ‘30 .. ‘. 3|. 0 1001.3... .000- 00 0 0 1. l V 0 A 0 3 .00 0 . . .. V . . n» 0. 00 . 0 V0 .. 4 .V .0. .. 0 n .V V V 0 0 0 . . . V . . 0. a.) .0070}.- u 1...... 001. 1.0010. . 3‘1. 3‘ 0... ‘00- 0.000‘9010‘1000. 1'000 00 V 0 . V .I. . . V 30 4 0 . 0 VA. 3.: ..0 . .. . . V . . u a V V A 2.0} .20. V0... . V 0.30..» ... .33.. A. 0 1.... .31.... .000. u. . .o. . .0. V . . . 0 110...... . .0. 1.1.3.0. 1.00.0 . V000 10.33:...DEO .0. 0. .10) . 3.301.131.0030... 0|- . . . .0 0 . . o . . . . . . . _. . .. V V. .V V. V.. 0 0 . . .. V0 .2 0 . . . V V o. . 0 . V V . . .V .V. . . 0 V1... .. V 4 .. 0 0. V 0:02.. 0004 3.00.00‘ .01.; ’00.. 0100.A00v0( 5:0,. 0.000“ o ..0 0 V ...0 . .V. . 0 . V . 0 . 3.... .2 33.0.0 I. . 0. J . V. ... .0000. .32 00... ...V . V. I V. . . .20.... 3.200000 0.01.... .0. .o A . .00. .0 0 o . 03.110.10.00 ..1V 3.1.20.9. AO.....¢\...0.0 0 .. 00"... 0 0010.00.00 V %y .. . . V V . .. . V. 0. .V0 . A 0 1.0 0.0 .0.- . .. . n .. Q . .v 4 . V. .0. V . 0 .0... . 0.. 00. ~ ’0 . .3 0A". V. ..00 A. .0. IQ... 10.. .10 0. 0.0 .V .1 .. . 0.. 9... V. V .. ......V. .0 5..-. .AV {0;}... to-.. .. 7.0.. ..0 0..-03.01.30 1.0 .£.0...0.. 0 0 0.} V V .0 0.. 0 0 . . ..0 0 .00. . V. 0 0. . . u . . .. L V0. . .. 0 .V .00 .00 ..... . . V . 0 ~00. o. o . 0 .V . 0 . . .0 . . 0 .. . .0. . . . .0. V0. ..0. 0O . 0.. ... ..00 . .. 0... 0. v 00 . 0..J0.0 A V 0 .V. 0. 0.0. .. 000 0V . ..0V ..u . V 0 0. O r V 0. V .. .0. 0 0 . .0 V . . ..V 1 0 00 t .g 0 .. .n. ...l . 500.00 . 0 0.. .0 4 . V .4 V .. . 0 .. V .. . ....~...V 2.0.0.3.. 010.....00: V0. .0 I .0. . 001.0l..V...V0.. 2...... .3 00V .0. )0... V V V . V 0...... .3. 30.0.. 0. 8.3.3.0. . .2... 0.300;... V V\ .30 £330.". 0010.0. . . n. V . . 0 v 0. n A 0 - 0 . V V . V . V. . V .. ..4... u... ‘0 0.0 A . V .A 0 v :3... ..V.. . V . 1.. 00 0. 0 . . V . I .. .40.. 00.1.0.000!‘ . ’.VV.V000.0.0.. I. 0. . 0 . . 0. u . v . V.V . .000. a 0 001.. 0.0.0 o . . .0 V... .. V V ..0 . v . 0». 0. 3.0 ... . 3.0.. 0.. .00.... 00. V 0.0. 1000.0 i. 0 00‘. 0. .V.0V. .J. V. I0 V, V VV .1 .1000‘0n0- 0. ‘V..0.|"0.00\00.0..u0 I00..l\ t . - . 0 . 0 . 0 V n . V. 0V.VV é . .2 . V 0 V. . 03.. V . u ..v . .n 1.00. a .0 .. V . ...0 0 .V .0.0 .0 ..I . .0 .. . 0 4 . .. .l 0 . 0 . V.V . .000.) .a 0 0. 0... ....V... l.0~.‘.l.‘. QVV .'. 0 0 ... $.0 . .. . 4 . T .0 . ..u. . . .V .. V. 1 0.100.. 0 V....... 0 .. . .0 0. 0V. . . A 0. .. 1 0 . V ..V: . . . .. .2 . 000 A. 00. 00 .. .0. . . . . .AV-’0 .... ... 30.. 20.0 I 00.0. V. 3 )0... . . .a 7..-}... 0.0 0.000 410 .l000‘0 50-00.. 0: .. . . 00 . 04 .. . ..V . .. .0 .V . . . . . V . 0 . 0 . . 0 V. 0 1;! V I V. V _ . 0.00. . . . .0. . 0. ... . VV . V ... ..00.0 . . I- i... .. V. V r .0 0.0!...V0‘Q001 .0’ .. 00.0.0; v0.0.1- . ... 0.0 0 . 0.. .000. . r . . a ..n a . .V V .0. . 0 . 0V 0. . . 000.0 .0. 000.0.~: ... 0 ..0 1 . . u. .1 ..V V 00.. .00. 0.0.1.3. . 1500:“. Q. . .0 5 .0801. . I... 00 0 0 V 0. . . . 0‘00_00-000(,.u 0. 00!...” A . .u . . .. . .. . .. n V . 000 . V. .v . . V . .0 . 0.. . . . A .V. . 0. . 0 . . 00 .A .V . l. V V. . ..l .100. . l0 00000.0...0...’0.000.0‘0 1.0 - $0. 0 .V.. .. A .2 V 3’0... .9 o . 4 . . . .. . :3 0 . .03... .0. . V1.00 .. ...Vn0 . V n. .. -.. 2. - V .. 1...... . A. .V. ....0..V‘. . ..01.0.0...‘..0.01""0| 0 0.V .0 . .0 0 . .. ..V «I . o 3...}; ..0 ..0.... .0.. .0. 00. 00 .0 V ..V ..V.A0 . V 0.. V .0 . . .n .0. 00 .0. . . 2. 0 . V. . V 4. 0.10 .... ..V 0011.0: .12“ )1 .01, . .l. . .V. .. u 0 . I. :0: . . 0 V 20 . .. . 0 .\ u . . n s . 000 0 0.2.0.. 0. -.. . 0 0 .0 0 . .y .. .04‘. ..I 0 .V . 040A 0.900 . V. . V . . ..V q . 4.. 0.-v . .'- . . . ~ou 00,110 .1 ‘4‘ . .- V. .0. 0 . . . 0 A . r O . 0 0... . O I?! u .0. 0. t 0 00 V . . .. .0: . .. ..0. 1000 . 0‘. V.00'00..; . .V... V .A.. . 0‘00. .5000 0. . \00‘0‘0 I0u00’“0 \I 0 $00.00.. . 0 0 v :0 . . u 0 . 0.0.0.0 0. I 0 . \ t 0 . 0 0 .0 . . q 0. . L0... . v. . .V . .V .. E.- .. 000 00... V. 7.1 .\.. .1 I0 D0oVO00‘r'. '0‘ 0.00. I'C V .0. . . . u u 0 . I . 0 0 . .. .0. .. . . V. .2 . . . .... ..0. ..0 V .3 . .0 .0: . . V 0.. . 0 .. .0. V. V a .. t .0 .- -.- .. .0 .0 \ w 0.00 1100 ... 00 0 ...00\|r01 . . 00' 0.0.0 1.... .. 0. 0 u .. Q . ..0. . . 0 .0 o . . . 0 . .0 .. . V. . . . A. (0210.0... 2.. . 10.. . .Vl .. . . L .0. 0 .1... .. V t . . .n u. a .V! 00.0.. o p #0.. ‘00.. r ... V .. I . . . 5 . .V . . 00 00. V... . .0. 0. V. . . . a . n. ...0 0. \ .V.. 0. . 10 I40. . 0 .v‘lilov’pn..\..\. 10-0 . VV. . 3.0 0A.AV0 ..... 0.00 o . 0 O . . _ 0 0 . v. 0 V. . ...00~0.V0 V. 0 I. 0. . 0 0 . a ..V 0 .V . ..V a. . I.» .00 0 . 0 0 l 0 ’0. .00 0 .0 0.... ..o. 0 . '00.. . V911. 0.. 0.. 000.1 I . 00V 0.. . V 0 ..V V 0. . V V . . . 0. . .0... 0 0 . .... o. . .. 0 .. . 0 0 ..0 ..0 . I .t' 0 V V V015. 0 00‘.\.0\O‘0 ' . .. V 0 . i. . 0 V . V .0. .V ..0.00 0. ..A. V . .0 V. V . . .0 a.‘ . t 0 A . ... 0.. _ I. 0. .10 . 1... 1.... . :- . .0 .. ... . 0 . x 1.0. V. . I 000 .10. . . . l _ .. 0.. 0 V . u . V 0 . 0 0 . v V- ..0 . .. .. 000 0. .. t .0 .0 . . V . . _ I 0;!L 1 u .50. .0. .0. Volt .‘.0.000.0.0.l 0-0.0. . . . I .0. 0. 13.00.07. ..0 0.00 a ... . . .0V0. . u... 00 . .....0.0.. 0. ..V u... 0. o . 0. . . . .... . v. . 0 0 0.0.0.. . t 1-! .040. .V V . . V . . 0 . .v . V. a :0... I . . .. .. S A . 0 . o . V.V . . . . . 1.... - 0 . .0 ..V. o... . . . \ « a 7.0». . 0.. . .n s 0.70. 0. 0.. 0. . 0 .V . . . .0. .. o 0.00.. . 03-. . 0 0 o .. y . . . .. ... . 0 0.0 V. c I l .0 0 . 000.0 \. . V 0 V .. V .V .. A. V . n . 0 . n .0 A. I 0 ..0 0V I 0.. 0 0 . . . . .. . .. s . .0. V . . . 0.. .10. 0 0.0 o 0 . ..V 0 0 u . . 0. .0. 0. A .0... q . V . n. 0 .0. 0 . .V 0‘. 0 0 0'. I V 0 ..0 .. 0 . 0V0 . .01 .n u 0. I. . .I _ . .. . . V I. .. 3. .. L ..V . . . V . 0 . . . . . . c . 4v... . .0 . . V 0 . .V . .. . . 1 0 . V 0. L L . . 0. 0 . 0 . n 2. . ..l O _ 0. a . V 0. . . . I 1 4 .0: .0 . . 0.0. V1... 0 . . 0 131.....- 1.3.3.: V V. V.V.VVV V.V.V .? V0 *0 V . .r 0.0 V VV V . L .. VV . . . .. n V . . V .V 00.05.-. ..V V 9.0!! . 0 0. )0 f V .4... ..«t 1.9.; 57...! .VrhfifiaV V T V .V . V V.. »_.__.VP+V.0..r.....wV . V ....... 0.735.-..V1r590. 01.1 I00}. 9 . it! ‘0 "a. *0 V .37.»? ‘0'. v. V, \“AHA.~V ..z . . 7V: .V.V.~V.?.V..re.yv V. .VV ”:2...- . V“ .V V . 3.. V. 5.... . I .V . . . V. . A. . 0.... . A. .. 2. .1); .. V n . . . a b. . . . .V 0 VV. V.V... . . V 0 at... VrVV. V... . .2... r; 5.31; (V.V r? .V V 1. V V V I V V .00... .I .00. . a .. A Y 0 Lt. .. . J 0 \V b ..V . V .0 0.0 0.0.0. In. .00 .0003! .1.|.o.000‘.. .0 .000 y 0.. .nn 0 .~ «a K '1 L0 . 0.: ... V.V. ‘ . 0 r 7‘ .L— .0. 7 l V _..00. L. 0.91. V: .54 V....A...0 ..3 1 . .13.... 50.... . .03.. V _.. Vt. 00......Sq’u 3.....0 .0. . SI 0000 u 0. “1.0 0 .. 00 * V 0* . . . . . V V . . Z. ..0 . . 0 4.2.00. .5: 0. I... 0. .c 00 . 0 ..0. .A. V.. .. A...” .0 . .. .V .3 0.. 0“ c . . 1.00001.) 0. 0. .r: A .4. . V .053 31.0.03; 0...‘...00 1 00 0. V .. .001 00 . 0. . 00.. r 2. 0 . T... . ..J \..0 0 0. . . ”0.0.0.... 00. .10. 00 2. 0 .. . V . 01...... 30 20.4.8.0 0.0 .V0 . ...l.......l..00.0 0.0... . . .0 0V. 0 . .0 3.1.0.0 0. 000. .0. 0.30-5001). 0.0.0 1’. 00 .. 000 . V I. 9. V. . . V 0.1.60... .4 . . . .32.... . . . 1.2.1.0 $00.00.. . . 1 ...0..V.1I....... '20..- .. o 3.00.0 ..00. 3.3.8 . . . .1 0.0.30: ..0 . ..0 . 0.. .. .. . . 3...... . 5.. .. c... 1.0 .Q.. ... 0... ...V... .......0V0..0 V.V...IVV .0... 5.70.... .V. .V. .V... V..._ A I... . V.. .0. V . V... V . V 0 . l .. 1. . V 2 . \ 9 . 0 0 V ' ‘0 I V‘I’ ' . 0230:0100... 0 ..n..\ lrl V V. V 31.01. 3..) 0.0. ....V. V V I .n 0 0 .3... c. 0.... . . . V . II This is to certify that the dissertation entitled EVALUATING THE IMPACT OF GLIRICIDLA SEPI UM ON SOIL ORGANIC MATTER IN MAIZE-BASED CROPPING SYSTEMS IN SOUTHERN MALAWI presented by Tracy L. Beedy has been accepted towards fulfillment of the requirements for the Doctoral degree in Crop & Soil Science /4/ Ma Profissor’s Signature 7' December 16, 2008 Date MSU is an Afiinnative Action/Equal Opportunity Employer LIBRARY Michigan State University _A_._._._._.-.--_.-.-.-.-._.-.-._ PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 KzlProleccaPres/CIRCIDateDuo.indd EVALUATING THE IMPACT OF GLIRICIDIA SEPIUM ON SOIL ORGANIC MATTER IN MAIZE-BASED CROPPING SYSTEMS IN SOUTHERN MALAWI By Tracy L. Beedy A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Crop and Soil Sciences 2009 SIL oft ht) U; ABSTRACT EVALUATING THE IMPACT OF GLIRICIDIA SEPIUM ON SOIL ORGANIC MATTER IN MAIZE-BASED CROPPING SYSTEMS IN SOUTHERN MALAWI By Tracy L. Beedy There is considerable interest in soil organic matter technologies and fertility sources in intensified maize-based cropping systems in Southern Africa. This study investigated the effect of the Gliriciclia sepium (Jacq.) Walp and maize intercrop on soil organic matter (SOM) fractions and soil fertility indices in southern Malawi. Part I investigated the response of SOM fractions to the intercrop, inorganic nitrogen (N), and phosphorus (P) on a long-term trial established in 1991 on a Ferric Lixisol in southern Malawi. Soil was sampled to a 20 cm depth in July of 2006. SOM, available P, exchangeable 1C, and CBC were determined on whole soil samples. Particulate organic matter (POM) was separated fi‘om the soil by size and density fractionation and analyzed for C and N. The intercrop had a positive effect on all SOM fractions. With the intercrop, POM was increased by 37%, carbon in POM by 60% and nitrogen in POM by 78% compared to sole maize. C/N ratio of POM decreased from 17.9 with sole maize to 15.5 with the intercrop. After 14 years, predictors of soil fertility and SOM fractions were significantly greater under the intercrop than under sole maize. In part 2, an on-farm study in the same district described characteristics of intercrop users and their placement of the intercrop on different soils. Thirty-seven households using the intercrop and 28 households not using the intercrop were interviewed in June and July of 2006. S} Cl l S 0 gm- field Questionnaires were completed with each household concerning demographic statistics and the soil types found and cropping systems used in their fields. In female-headed households use of the intercrop was 66% compared to 49% in households headed by males. lntercrop use was 50% among households in the lower half of the socioeconomic scale and 63 % in the upper half. Households using the intercrop had an average of 0.4 hectares of land, while those not using the intercrop had an average of 0.83 hectares. Placement of the gliricidia/maize intercrop on sandy M’chenga soils was less common than on Katondo and ‘Makande soils. In part 3, the SOM and soil nutrient effects of the intercrop were compared to prevailing soil management practices in smallholder’s fields. Soil samples were taken from the 2005/6 planting rows from the fields with and without the gliricidia intercrop. Soil samples were analyzed for SOM, sand content, available P, exchangeable K+ and CEC. Soil organic matter values associated with the intercrop were not significantly different from those for other cropping systems. When the gliricidia/maize intercrop fields were analyzed separately, soil type, elevation, and their interaction affected trends in soil organic matter. A positive relationship between soil organic matterand elevation was evident in the two finer-textured soil types, Makande and Katondo. The addition of pigeonpea, Cajanus cajan (L.), to the maize cropping system at elevations below 877 meters was associated with an increasing trend in SOM compared to the sole maize cropping system. The emergence of increasing trends in SOM correlated with increasing elevation and fineness of soil texture in fields with the gliricidia/maize intercrop may indicate that a SOM benefit will develop over time in fields at relatively higher elevations with finer soil textures. ACKNOWLEDGEMENTS I would like to thank my major professor, Dr. Sieglinde Snapp and the Michigan State Crop & Soil Science Department for the opportunity and support to conduct this research. I am appreciative of the help and guidance of my committee members, Dr. Darryl Warnke, Dr. Scott Swinton, Dr. Festus Akinnifesi, through this process. I am grateful to Dr. Anne Ferguson and the office of (then) Women in Development at Michigan State University for administrative support and to USAID for funding that made this project possible. The ICRAF research program for southern Africa provided key logistical and administrative support without which this program would not have been possible. I would like to thank the sixty- four cooperator households in Zomba District, Malawi, for their generosity in sharing with us their farming practices, and to Mr. Ladson Chirwa and Miss Martha Chiwaya for soil sampling and information gathering. 4 Laboratory analyses by Miss Andrea Posigian are gratefirlly acknowledged. iv PREFACE Chapter 1 in this dissertation was written in the style required for publication in A griculture, Ecosystems, and Environment. Chapters 2 and 3 were written in the style required by A grqforestrjv Systems. CH4; Chafac Vang U TABLE OF CONTENTS LIST OF TABLES ........................................................................... .ix LIST OF FIGURES ........................................................................ xi KEY TO SYMBOLS OR ABREVIATIONS ........................................ xiii CHAPTER ONE Long-term Impact of Gliricidia septum Intercropping and Inorganic Fertilizer on Soil Carbon and Nitrogen in a Maize-based Cropping System in Southern Malawi Abstract ................................................................................. 1 1. Introduction ........................................................................ 3 2. Materials And Methods ........................................................... 7 2.1. Site description and management ..................................... 7 2.2. Establishment and management of gliricidia ........................ 7 2.3. Soil sampling and analysis .............................................. 8 2.4. Data analysis ........................................................... 10 3. Results ............................................................................. 12 3.1. Soil organic matter .................................................... 12 3.2. Particulate organic matter ........................................ ' ...... 13 3.3. Available nitrogen ...................................................... l4 ' 3.4. Soil fertility indices ................................................... 14 3.5. Soil Texture and Soil Fertility ....................................... 15 4. Discussion ......................................................................... 16 4.1. Site soil characteristics and production history ................... 16 4.2. Soil organic matter .................................................... 16 4.3. Particulate organic matter ............................................ 17 4.4. Soil texture and soil organic matter ................................. 20 4.5. Available nitrogen ..................................................... 21 4.6. Soil fertility ............................................................ 19 5. Conclusions ............. - ........................................................... 24 Tables .................................................................................. 25 Figures ......................................................... . ........................ 28 References ............................................................................. 34 CHAPTER TWO Characteristics of Smallholders and Their Placement of Gltricidia septum on Various Soil Types in Southern Malawi Abstract ............................................................. . ................. 40 Introduction ........................................................................... 42 Materials and Methods .............................................................. 45 Survey Methods ............................................................. 45 vi Placement of the intercrop ................................................. 46 Data analysis ................................................................. 47 Results .................................................................................. 49 Demographic data ........................................................... 49 Placement of the intercrop .................................................. 50 Discussion ............................................................................. 52 Demographic data ........................................................... 53 Placement of the intercrop ................................................. 54 Conclusions ............................................................................. 56 Tables .................................................................................. 57 Figures ................................................................................. 59 References ............................................................................. 63 CHAPTER THREE Evaluating the Impact of Gliricidia Sepium on Soil Organic Matter in Smallholder Cropping Systems in Southern Malawi Abstract ................................................................................ 70 Introduction ........................................................................... 72 Materials and Methods ............................................................... 77 Agroecology .................................................................. 77 Gliricidia intercrop practices ............................................... 78 Survey methods .............................................................. 78 Soil sampling and analysis ................................................. 79 Field measurements and calculations ..................................... 80 Statistical analysis methods ................................................ 81 Results ................................................................................. 83 Environmental variables ................................................... 83 Management variables ...................................................... 84 Response variables .......................................................... 84 Multiple soil fertility enhancing practices ............................... 85 Soil organic matter .......................................................... 86 Variability of soil organic matter among villages ....................... 86 Soil organic matter and explanatory variables ........................... 87 Soil organic matter in fields with the gliricidia/maize intercrop. . 88 Soil organic matter in fields without the gliricidia/maize intercrop. 89 Discussion ............................................................................. 90 Soil fertility enhancing practices .......................................... 91 Soil organic matter and explanatory variables ........................... 93 Soil organic matter in fields with the gliricidia/maize intercrop. . 95 Soil organic matter in fields without the gliricidia/maize intercrop. 97 Conclusions ............................................................................. 98 Tables .................................................................................. 99 Figures ................................................................................ 1 04 References ........................................................................... 1 08 vii APPENDIX ................................................................................... 115 Household questionnaire ............................................................ 116 Farm Field questionnaire ............................................................. 119 viii LIST OF TABLES CHAPTER ONE Table 1.1 Mean, maximum, minimum, and standard deviation for response variables ........................................................................................ 25 _ Table 1.2 Significance of cropping system, N rate, P rate, and clay and sand content on soil organic matter, particulate organic matter, and carbon in particulate organic matter of whole soil .................................................... 26 Table 1.3 Significance of cropping system, N rate, P rate, and clay and sand content on available N, Bray P, cation exchange capacity, and potassium (1C) values of soil ................................................................................... 27 Table 1.4 The effect of cropping system and added N or added P on concentration of plant nutrients in study soils, with standard errors of treatment means in parentheses following concentrations ...................................................... 27 CHAPTER TWO Table 2.1 Sampling frame for the on-farm study .......................................... 57 Table 2.2 Mean and standard deviation for continuous variables and frequencies for categorical variables among the 65 households surveyed ................................. 57 Table 2.3 Chi-Square test of independence for resource availability and gender and for the use of the gliricidia/maize intercrop and available land and labor ............... 58 Table 2.4 Two-way table of fiequencies for Chi-square test of cropping system placement on farmer-defined soil types ...................................................... 58 CHAPTER THREE Table 3.1 Sampling frame for the on-far'm study ......................................... 99 Table 3.2 Mean and standard deviation of environmental, management, and response variables, divided by extension planning area ................................. 100 Table 3.3 Mean and standard deviation of soil organic matter above and below ix the mean elevation for the sub-group of fields, divided by soil type and crepping system .......................................................................................... 101 Table 3.4 The effect of cropping system, sand content, soil type, elevation, fertilizer N, rainfall, village, household nested within village and residual variance on soil organic matter and CEC in all fields .............................................. 101 Table 3.5 The effect of sand content, soil type, number of soil fertility methods, elevation, rainfall, fertilizer N, gliricidia population, and age of gliricidia on soil organic matter in fields with the gliricidia intercrop .................................... 102 Table 3.6 The effect of sand content, soil type, elevation, rainfall, fertilizer N, pigeonpea population and age on SOM in fields without gliricidia ..................... 103 LIST OF FIGURES CHAPTER ONE Fig. 1.1 Monthly precipitation, Makoka Research Station, 2003-2006 ............... 28 Fig. 1.2 The effect of fertilizer N rate and gliricidia on soil organic matter and particulate organic matter. Error bars represent standard error of treatment means. Bars with the same letter within each cropping system are not significantly different ......................................................................... 29 Fig. 1.3 The effect of clay content and gliricidia on soil organic matter, estimated at endpoints and quartiles of clay content. Error bars represent standard error of treatment means ........................................................... 30 Fig. 1.4 The effect of gliricidia and fertilizer N on carbon and nitrogen in particulate organic matter. Error bars represent standard error of treatment means. Bars with the same letter within each cropping system are not significantly different ......................................................................... 31 Fig. 1.5 The effect of gliricidia, fertilizer N and fertilizer P on CfN ratios of particulate organic matter. (a) Sole maize (b) Gliricidia/maize intercrop Error bars represent standard error of treatment means. Bars with the same letter within each P rate are not significantly drfferent 32 Fig. 1.6 The effect of gliricidia and sand content on available N pre-incubation and post incubation, estimated at the quartile boundaries of sand content. Error bars represent standard error of treatment means ......................................... 33 Fig. 1.7 The effect of gliricidia and clay content on cation exchange capacity, estimated at the quartile boundaries of clay content. Error bars represent standard erroroftreatment means33 CHAPTER TWO Figure 2.1 Map of Zomba district, with the study fields and villages ................. 59 Figure 2.2 Comparative prosperity of households and use of the gliricidia/maize, as estimated by the number of improvements made to the family dwelling among tin roof, glassed windows, and fired bricks ................................................ 60 xi Figure 2.3 Gender of household head/s and use of the gliricidia/maize intercrop... 60 Figure 2.4 Relationship of available resources and gender. .' ........................... 61 Figure 2.5 Landholding size and number of people per household ................... 61 Figure 2.6 Frequency of cropping systems on soil types ............................... 62 CHAPTER THREE Figure 3.1 Three-year average annual rainfall from the 2003 to 2006 cropping seasons in the four extension planning areas ............................................. 104 Figure 3.2 Planting pattern for the maize gliricidia intercrop given a 75cm spacing between rows and single-seeded planting pattern within rows ......................... 104 Figure 3.3 Soil organic matter as a function of sand content, by cropping system, including all fields ............................................................................ 105 Figure 3.4 Soil organic matter as a function of sand content, by soil type, including all fields ........................................................................................ 105 Figure 3.5 The effect of soil type and elevation on soil organic matter in fields containing the gliricidia/maize intercrop .................................................. 106 Figure 3.6 The effect of age of gliricidia planting on gliricidia population .......... 106 Figure 3.7 Soil organic matter as a function of elevation in fields without the gliricidia/maize intercrop, by cropping practice .......................................... 107 xii C/N CEC ICRAF POM POM-C POM-N POM-C/N SOM WAC KEY TO SYMBOLS OR ABREVIATIONS C arbon/nitrogen ratio, ratio of carbon to nitrogen Cation exchange capacity lntemational Centre for Research in Agro forestry (see WAC) Particulate organic matter Carbon in particulate organic matter Nitrogen in particulate organic matter Carbon/nitrogen ratio in particulate organic matter Soil organic matter World Agroforestry Centre xiii CHAPTER ONE LONG-TERM IMPACT OF GLIRICIDIA SEPIUM INTERCROPPING AND INORGANIC FERTILIZER ON SOIL CARBON AND NITROGEN IN A MAIZE- BASED CROPPING SYSTEM IN SOUTHERN MALAWI ABSTRACT There is considerable interest in the role of soil organic matter technologies and fertility sources in intensified maize-based cropping systems in Southern Afiica. Yield gains and seasonal N dynamics for the Gliricidia septum-maize intercropping system have been studied, but the long-term effects of the intercrop on soil organic matter fiactions and N pools have not been documented. Soil organic matter (SOM), particulate organic matter (POM) and soil nutrient responses were quantified in response to a gliricidia intercrop at three rates of added inorganic nitrogen (N) and phosphorus (P), compared to sole maize. This trial was established in 1991 on a Ferric Lixisol in southern Malawi. Gliricidia was established at a 150 cm x 90 cm spacing. Calcium ammonium nitrate was applied at 0, 46, and 92 kg N ha", and triple superphosphateat 0, 20 and 40 kg P ha'l annually. Soil was sampled to a 20 cm depth in July of 2006. SOM, POM, carbon and nitrogen in POM (POM-C, POM-N) inorganic N, Nitrogen mineralization potential, available P, exchangeable K, CBC and base saturation were determined. The gliricidia/maize intercrop had a significant and positive effect on SOM fractions: SOM was increased by 12.5% and POM was increased by 37% under the intercrop versus sole maize. Fertilizer additions of N and P did not result in significant changes in SOM in l both production systems, although N fertilizer increased POM by 16% in the intercrop and 24% in sole maize. The gliricidia/maize intercrop was associated with marked increases in POM-C (60%) and POM-N (78%). A decrease was observed in C/N ratio of POM from 17.9 with sole maize to 15.5 with the intercrop indicating N enrichment although depletion of N due at the end of the cropping season might have been expected. The gliricidia/maize intercrop and increasing soil clay content were associated with significantly increased soil CEC. After 14 years, predictors of soil fertility and supporting SOM fractions were significantly greater under the gliricidia/maize cropping system than under sole maize. 1. INTRODUCTION Soil organic matter is often the single largest source of nutrients for smallholder farming systems in Southern Africa according to Mtambanengwe et a1. . Because maize is the staple crop in Malawi (Chirwa et al., 2003), as in the rest of southern Africa, food security in the region depends heavily on enhancing the performance of this crop. How to build up and maintain soil fertility is a central issue in increasing agricultural productivity in Africa (Mafongoya et al., 2006a). The use of nitrogen fertilizers to improve maize production in the landlocked African countries is constrained by the high fertilizer costs, shipping costs and difficult logistics in terms of transport and timeliness. Price ratios of nitrogen to maize in Malawi may be almost double those in Mozambique, which borders Malawi on the east, and is able to bring fertilizer cheaply by sea to the northern maize production region (Mafongoya et al., 2006a). Compounding this fertilizer availability problem is the low level of fertility and depleted soil organic matter on smallholder farms due to continuous cultivation without adequate soil replenishment. Thus, examining soil organic matter dynamics is crucial to understanding soil fertility issues in southern Africa. Annual and perennial intercrops and rotations have been developed in this region to enhance soil fertility (Snapp et al., 2002a). A wide range of physical, chemical, and biological interactions occur in both agroforestry systems which affect plant nutrient I cycling (Schroth, 1998). Several leguminous trees and shrubs such as Sesbam'a sesban (L.) M6112, Tephrosia vogelii Hook, f., Gliricidia sepium (Jacq.) Walp. And Leucaena leucocephala (Lam) have been developed and introduced in southern Africa as renewable soil nitrogen (N) and soil organic matter (SOM) sources in maize-based 3 cropping systems (Kwesiga et al., 2003). Sesbania, tephrosia, and gliricidia have been investigated in southern Malawi for soil fertility replenishment (Harawa et al., 2006). Sesbania and tephrosia are used as improved fallows and relay intercrops, while gliricidia is managed as a long-term intercrop. Gliricidia is one of a group of species which tolerates periodic coppicing and removal of leaves for use as green manure or mulch of crops. The quality of leaves from these species varies widely. Gliricidia produces a very high quality green manure (Makumba, 2003). When managed as an intercrop, Makumba et al. (2003) found that the primary rooting zone of gliricidia was below that of maize, and Harawa et al. (2006) found that gliricidia reduced leaching of N below the maize rooting zone compared to sole maize. A long-term trial established on a Ferric Lixisol at the Makoka research station in southern Malawi gliricidia has been shown to increase maize yield (Akinnifesi et al., 2007). In a meta-analysis of the performance of woody and herbaceous legumes, Sileshi at al. (2008) found coppicing legumes to be associated with greater yield increases in maize than non-coppicing legumes. This trial allows investigation of fertilizer interactions with organic matter technologies, at three rates of application for both N and P. This interaction of inorganic and organic nutrient sources is important in developing feasible production systems in Africa (Sanchez et al., 2007). Long—term research station trials, such as the one in this study, play an important role in providing detailed information on biological variables over multiple cropping seasons (Richter et al., 2007). Oorts et al. (2000) evaluated the effect of several agroforestry species on soil organic matter fi'actions and the effect of these fractions on the cation-exchange capacity of the soil. They found that the organic inputs fiom the agroforestry species were associated with an increase in silt-sized particles with the highest charge density of the organic fractions, and noted the promise of these systems for increasing the CEC of weathered soils. Release of nutrients such as nitrogen fi'om soil organic matter depends on mineralization of its biologically active (or labile) fractions (Barrios et al., 1996). Labile soil organic matter (SOM) can be assessed effectively by characterizing the particulate organic matter (POM) fiaction. The POM fraction of SOM is typically low-density (usually 1.4-2.2 g cm'3) and/or coarse (53-250um), and commonly derived fi‘om biomass additions from the previous year (Wander, 2004). Szott et al. (1999) found this fraction to be larger after ‘ short-duration fallows with species having high-quality foliage such as sesbania and gliricidia. I Waddington et a1. (2004) found N mineralization potential to be dependent on annual organic inputs from in crop residues or manure, where residues were recycled through animals. Because N is the most common limiting nutrient to maize growth, and N is mineralized from POM it is important to study the interaction between POM and the mineralization of nitrogen. In an earlier study of N dynamics in the gliricidia/maize intercrop trial in southern Malawi, Makumba et a1. (2003) found that the yield increase associated with N additions from gliricidia and added inorganic N was significant, but this yield increase was not associated with increased total soil organic matter. Akinnifesi et al. (2007) also found maize yield to be related to first and second order N in the monoculture and to the first and second order of N and rainfall in the intercrop. Soil texture is also an important variable in soil nutrient cycling (Giller et al., 1997; Hao et al., 2006) however the relationships between soil carbon, soil texture and microstructure are complex (Plante et al., 2006). Because soil texture is heterogencus, spatial variability of soil can also be an important factor in SOM characterization. Kravchenko et al. (2006) found that assessment of soil carbon based on both treatment effects and spatially sampled data such as topography and texture improved the acCuracy of analysis in a trial comparing tillage regimes. The current study investigates SOM and POM response to the gliricidia intercrop and three rates of added inorganic nitrogen (N) and phosphorus (P), compared to sole maize. The first objective was to determine whether the gliricidia intercrop significantly increased SOM, POM, and labile N pools, and whether the effect of gliricidia interacted with the effect of inorganic N and P fertilizer. The second objective was to determine whether the gliricidia intercrop increased soil’s ability to supply plant nutrients beyond those added as inorganic N and P. 2. MATERIALS AND METHODS 2.1. Site description and management This trial was established in 1991 as a randomized complete plot design with three replications to test the factorial combination of two cropping systems, three N fertilizer rates (0, 46, and 92 kg N ha") and 3 P fertilizer rates (0, 20 and 40 kg P ha") with three replications (Akinnifesi et al., 2007). The soil at the trial site is classified as a Ferric Lixisol (Ikerra et al., 1999), with an initial SOM level of 15.2 g kg", based at the Makoka Research Station (latitude 15° 30' S, longitude 35° 15' E). The rainfall pattern at the station is unimodal with most rain falling between November and March. The mean annual rainfall between 1997 and 2006 was 964 mm. Monthly rainfall during the growing seasons of 2003-2006 ranged fi‘om 11.5 mm to 363.9 mm (Figure 1.1) and varied widely between cropping seasons. Nitrogen was applied annually, four weeks after planting. Phosphorus was applied annually as triple super phosphate except for 1993- 2002, because no P yield effects were demonstrated in the initial years of the trial. Phosphorus applications were resumed in 2002 at the rates listed above (Akinnifesi et al., 2007), which corresponded to 0, 50%, and 100% of the recommended rates (Malawi Government Ministry of Agriculture, 1996). 2. 2. Establishment and management of gliricidia The 6.7 x 6.0 m plots were separated by iron sheets inserted to a 1m depth to minimize root encroachment. The three replicate blocks were separated by a l m path. Planting ridges were established at 75 cm spacing. Gliricidia was planted in alternate ridges (150 cm spacing) and with a 90 cm spacing within the ridge, which results in a 7 gliricidia population of 7,400 trees ha". The trees were pruned to a height of 30 cm, in October, December, and February of each year, and the leaves and tender twigs incorporated into the maize ridges, which were rebuilt in October of each year. Biomass was recycled within each treatment plot each year. After the onset of rains, usually in November or December, both sole maize and gliricidia/maize plots were planted 'with maize hybrid NSCM 41. The maize was seeded at 30 cm intervals in all of the ridges, resulting in a plant population of 44,400 plants ha". The maize grain was harvested after drying down in the field, usually in May. Sub-samples were taken for moisture determination, and yield and biomass data were expressed as kilograms per hectare on a dry weight basis. A more detailed description of site management may be found in Makumba et al. (2006b) and Akinnifesi et al.(2007). 2. 3. Soil sampling and analysis Soil samples (0-20cm) were taken in July, 2006, composited from six samples per plot using a straight-walled soil probe with 2cm diameter. The samples were mixed, air- dried, passed through a 2 mm sieve, and stored. A sub-sample of the soil was dried overnight at 40°C, crushed with an Agvise flail-type grinder and analyzed for organic matter, pH, available phosphorus (P), exchangeable potassium (K), magnesium (Mg), and calcium (Ca) (A & L Great Lakes Laboratories, Fort Wayne, IN). Organic matter was determined by loss on ignition at 360 ° C and the data correlated with and reported as Walkley-Black titration. Soil pH was determined in a 1:1 soil to water slurry. Available phosphorus and exchangeable cations were extracted according to Mehlich III (Mehlich, 1984), and analyzed by inductively-coupled plasma spectrometry (ICP). The P data was correlated to and reported as Bray P-l (Bray and Kurtz., 1945). The data for 8 exchangeable cations correlated to and reported as a l N ammonium acetate extraction (McIntosh, 1969). Percent base saturation and CEC were calculated from the results for exchangeable cations. Particle size distribution, nitrogen mineralization potential and particulate organic matter were determined on subsamples of soil that had not been ground. Particle size distribution of the samples was determined by gravimetric sedimentation after removal of the organic fraction with 30% H202 (Gee and Bauder, 1986). Nitrogen mineralization was determined using an aerobic 30-day incubation of rewetted soils after Haney, et al. (2004). One 10 g sample was dried at 105° C overnight for moisture determination. Four sub—samples (40 g each) were weighed into half-pint glass jars. Water was added to 50% of field capacity (Saxton and Rawls, 2006), and the containers were covered with parafilm in which four 1 cm holes had been made for atmospheric exchange. The quadruplicate samples were incubated at 25° C in an incubator with extra jars of water to maintain high humidity levels. Soil moisture was returned to 50% of field capacity, measured gravimetrically, on a weekly basis. After 30 days of incubation, each sample was extracted with 200 ml of 2 N KCl. Initial inorganic N was determined from duplicate 10 g samples of non-incubated soil extracted with 2N KCl. The extracts fiom the non-incubated and incubated soils were analyzed with a Westco Smartchem analyzer for nitrogen from NH4+, NOz', and N031 Nitrate and nitrite were determined by hydrazine reduction and ammonium was determined with the Berthelot reaction (WestcoScientific, 2007). POM was separated from the soil samples by size and density fractionation (Cambardella and Elliott, 1993; Wander et al., 1994) and the values presented as a fraction of total soil weight. Quadruplicate 25 g samples of soil were dispersed in 125 mL 0.5 g L" sodium hexametaphosphate for 17 hours on aireciprocating shaker at 180 rpm. The dispersed fraction was passed through a 53 um sieve. The fraction remaining on the screen, containing both sand and POM, was well washed with deionized water, and dried 48 hours at 60° C. The dry sample was carefully transferred into 50 ml plastic conical centrifuge tubes, and 35 mL of sodium polytungstate of density 1.85 Mg/m3 added. The tubes were capped, and inverted slowly several times to ensure that the suspended material was completely wetted and mixed with the sodium polytungstate. The tubes were then placed in a vertical position and centrifirged at 1000 rpm for 30 minutes. The POM plus sodium polytungstate was decanted onto a 20 um mesh nylon filter connected to a vacuum filtration system. The POM remaining on the mesh was rinsed thoroughly with deionized water to remove traces of sodium polytungstate, and dried overnight at 60° C. Sodium polytugnstate was recycled according to Six et a1. (1999). The weight of POM was recorded for each sample and the quadruplicate samples were combined and ground for 2 minutes each in a Shatterbox 8515 mill (SPEX SamplePrep, Metuchen, NJ). Carbon and nitrogen were determined on the ground samples by dry combustion in a CHNS analyzer (Costech ECS 4010, Costech Analytical Technologies, Valencia, CA). 2. 4. Data analysis Analysis of variance was conducted to determine the effect of the gliricidia intercrop and inorganic N and P rates, as well as the continuous variables of sand and clay content on response variables related to the organic fraction: SOM, POM, POM carbon and POM nitrogen. Response variables relating soil fertility to the treatment factors were available 10 at pl W: cm. inorganic N, N mineralization potential, Bray P1 phosphorus, exchangeable 1C, CBC, and percent base saturation,. Ancovas were conducted using the PROC MIXED procedure in SAS (Littell et al., 2006) to determine if soil textural variables influenced the relationship between the treatment factors and the response variables. Where soil texture had no significant influence, the models were three-way Anovas using SAS Type 3 analysis, which uses expected mean squares to estimate variance components. In the PROC MIXED analysis, standard errors are calculated considering both the fixed effects and random effect components of the model. Planned contrasts were included in the Anovas for treatments which had similar N additions from inorganic and organic sources. Because variances of exchangeable 1C were unequal among gliricidia and sole maize, this analysis was conducted using the restricted maximum likelihood method, and different variances specified for the two . groups. I Nitrogen mineralization was calculated by subtracting inorganic nitrogen (NH4+ plus N02" plus N03) values in non-incubated samples from the corresponding values in the incubated samples, and expressed as mg N per kg dry soil. POM values were expressed as g per kg of dry soil mass. Carbon and nitrogen in POM were expressed as mg per kg dry soil, and as a CM ratio. Because SOM is often spatially correlated (Kravchenko et al., 2006), Soil organic carbon values, and soil sand and clay content were analyzed for spatial correlation using the PROC MIXED procedure in SAS (Littell et al., 2006). Spatial modeling did not improve the fit of the soil organic carbon model, and was dropped from the modeling strategy. 11 3. RESULTS Southern Malawi lies in the southernmost portion of the eastern Rift Valley. . Elevations range from less than 100m above sea level (asl) in the southern Shire valley to more than 2000m atop Mount Mulange. Zomba district lies mostly between 300m and 1200m (Benson et al., 2002). Rainfall is unimodal (Figurel. l), with the majority of rains falling between December and April. Population density for Malawi averages 105 persons km'z. In southern Malawi, population densities range from 50 to 400 persons km' 2 outside of urban centers. The population density of Zomba district ranges from 100 to 200 persons krn’2 (Benson et al., 2002). Rainfall in southern Malawi ranges from .600mm to 1600mm annually. Most of Zomba district receives 800mm to 1200mm of rainfall annually (Reynolds, 2000). 3. 1. Soil organic matter Soil organic matter ranged from 18 to 33 g kg" of total soil mass across all gliricidia, N and P treatments. The presence of gliricidia increased SOM 3.4 g kg" beyond the SOM maintained in the sole maize cropping system (Figure 1.2) averaged across N and P treatments. At the median clay content, the gliricidia Imaize system contained 26.2 g kg" SOM compared to 22.8 g kg" in the sole maize system. The increase of SOM was constant across the range of clay content (Figure 1.3). The clay content of the soil ranged fi'om 18.4 to 44.6 percent, the average base saturation was 67% and the average pH was 5.6 (Table 1.1). The presence of clay increased SOM 0.19 g kg" for each 1% increase in clay content. 12 3.2. Particulate organic matter Both the presence ofgliricidia and fertilizer N positively influenced POM values (Table 1. 2). The presence of gliricidia increased POM by 4.7 g kg" (3 7%) at each rate of N addition (Figure1.2). With 48 kg ha fertilizer N added to the system, POM increased from 12.6 g kg" in the sole maize plots to 17.3 g kg" in the intercrop. Fertilizer N addition increased POM by 2.4 g kg" over no N additions, which amounts to 18% with the intercrop and 33% in the sole maize plots. Both gliricidia and fertilizer N had positive effects (Table 1.2) on carbon in the POM fraction (POM-C) and nitrogen in the POM fiaction (POM-N). The presence of gliricidia was associated with increased POM-C by 0.76 g kg" or 60% (Figure 1.4), and POM-N by 0.06 g kg" or 78% (Figure 1.4). With 48 kg ha N fertilizer addition, POM-C was increased from 1.27 g kg" with sole maize to 2.03 g kg" with the intercrop. At the same rate of fertilizer N addition, POM-N was increased fi'om 0.071 g kg" with sole maize to 0.126 g kg" with the intercrop (Figure 1.4). The first increment of N fertilizer was associated with increases in POM—C and POM-N, 0.32 g kg'l and 0.015 g kg" respectively. POM-C and POM-N values with similar levels of inorganic N + organic N were also contrasted. The treatment combination gliricidia + 0 added N is associated with a POM- C value of l .71 g kg", contrasted to sole maize + 48 kg ha" added N, with to 1.27 g kg" of POM-C (p = 0.0025). The same contrast is associated with POM-N values of 0.1 1 1 and 0.071 g kg" respectively, (p=<.0001). The ratio of C to N in POM varied from 14.7 to 18.6 (Figure 1.5). The gliricidia, N and P treatments resulted in statistically significant decreases in the UN ratio of POM l3 (Table 1.2). The presence of gliricidia was associated with a UN ratio of 15.6 compared to 17.9 for sole maize, averaged across N and P treatments (Figure 1.5). The first and second increments of N fertilizer were associated with UN ratios of 17.2 and 16.3 compared to 16.8 for no N fertilizer, averaged across cropping and P treatments. The first and second increments of P fertilizer were associated with C/N ratios of 16.4 and 16.9, compared to 16.9 with no P fertilizer, averaged across cropping and N treatments. 3.3. Available nitrogen The effect of gliricidia and fertilizer N on inorganic N values (ammonium+nitrate+nitrite) prior to incubation (Table 1.4) was positive (Table 1.3). Eighty-seven percent of inorganic N was in ammonium form This ratio was relatively uniform among the three N treatments. Gliricidia increased inorganic N from 7.93 g kg" in sole maize to 12.8 in the intercrop. The effect of increasing P addition was negative. Nitrogen values after incubation were negative, and were 22% to 32% of pre-incubation values, indicating immobilization (Figure 1.6). Initial inorganic N values were greater in the gliricidia treatment and immobilization values were correspondingly greater. The effects ofgliricidia and fertilizer N treatments on N mineralization potential were significant. 3.4. Soil fertility indices The gliricidia intercrop, fertilizer N, and increasing clay content were associated with decreasing soil phosphorus values (Table 1.4) while soil phosphorus values were increased by increasing P additions (Table 1.3). Decreasing soil potassium values were associated increasing fertilizer N and P additions in the sole maize system (Table 1.3, 14 Table 1.4), while the opposite trend was observed in the intercrop. Increasing soil potassium was also associated with increasing clay content in the gliricidia cropping system, but not in sole maize (Table 1.3). Gliricidia, the first increment of N fertilizer, clay content and the interaction of gliricidia and clay content significantly increased cation exchange capacity (CEC) (Table 1.3 and Table 1.4). Base saturation was significantly increased only by N addition. 3. 5. Soil Texture and Soil Fertility Sand content modified the effect of gliricidia on inorganic available N, which was increased by 0.074 g kg" for each 1% increase in sand content in the sole maize cropping system, and decreased by 0.29 g kg" for each 1% increase in sand content in the intercrop (Figure 1.6). The interaction between gliricidia and sand content was also evident in N mineralization values, and similar in proportion (Figure 1.6). Soils in the gliricidia cropping system also maintained CEC between 8.5 and 9 cmol kg“I of soil across the range of clay content (Figure 1.7), while CEC varied significantly with clay content in the sole maize treatments. 15 4. DISCUSSION 4.]. Site soil characteristics and production history Though the base saturation of Lixisols is comparatively high, the CBC and absolute leveliof plant nutrients is low, which makes recurrent inputs of plant nutrients a pre- condition for continuous cultivation (FAO, 1999a). Given the limited natural fertility of these soils, the gliricidia/maize intercrop could play an important role in replenishing labile organic matter pools, an important plant nutrient source for smallholders continuously cropping Lixisols. Organic matter technology effects on soil properties is often not discemable in studies with a two to three-year time flame (Lehmann et al., 2001 ). This trial is an opportunity to study treatment effects on soil organic matter levels that have been built over a period of fifteen years. Makumba et al. (2006b) gives maize yield ranges in this trial from 1.5 to 4 Mg ha" for the sole maize treatments and flom 4 to 6 Mg ha" between 1993 and 2003. One of their conclusions is that the maize yield in gliricidia treatments averaged 1.9 times that in the sole maize treatments over the decade. The average N content in the leaves is given as 29mg N per g of leaf, and the average annual gliricidia biomass addition as 4.8 Mg ha’ I yr". From these figures, N added through gliricidia biomass averaged approximately 139 kg ha" yr" between 1993 and 2003. 4. 2. Soil organic matter The gliricidia cropping system had a positive effect on all SOM flactions. Soil organic matter ranges across all treatments were consistent with soil organic carbon values of 1 1.4 to 29.1 g kg" reported by Akinnifesi (2007) in a previous study in the 16 same trial. The effect of gliricidia on both SOM and POM (Table 1. 2) in soils may be due to both the direct effect of decomposition of gliricidia biomass and an indirect effect as plant nutrients released in gliricidia decomposition increase maize biomass production, which is then recycled within each treatment plot. Between 1998 and 2002, biomass additions averaged near 1.1 Mg ha" in the sole maize treatments and near 4.5 Mg ha" in the gliricidia treatments (Makumba, 2003). Sherrod et a1. (2005) found that differences in stover production accounted for 80% of the variability in SOC associated with increased cropping intensity. Akinnefesi et a1. (2007) reported a significant first-order , interaction between N rate and gliricidia with production of stover biomass between 2002 and 2006 in this trial. However, this interaction was not carried forward into the SOM reported in this study. After 14 years of biomass additions, SOM values were 15% higher in the gliricidia/maize intercrop than in the sole maize cropping system, however fertilizer N and P treatments did not have a significant effect on SOM. 4. 3. Particulate organic matter In addition to the significant effect associated with gliricidia in maintaining SOM and POM (Table 1.2), increases in fertilizer N were associated with significant increases in .POM values (Figure 1.2). Moran et a1. (2005) found that addition of mineral N increased the rate of soil C respiration and SOM decomposition. Our study followed a similar pattern, in which the 48 kg ha" rate of fertilizer N was related to an increase (Pr>F=0.0169) in POM values flom 10.2 to 12.6 g kg" (Figure 1.2). Use of the gliricidia intercrop and addition of fertilizer N to the cropping system resulted in increased POM-C and POM—N compared to sole maize and no addition of 17 fertilizer N. Vanlauwe et a1. (2000) found that N in the POM flaction was significantly related to N uptake by maize in coarse savannah soils. Moran (2005) found enrichment of POM-N with N fertilizer to be greater than with other organic matter flactions in a greenhouse study of rice residue decomposition. Figure 1.4 demonstrates that POM-C and POM-N values with the gliricidia intercrop but no N fertilizer were greater than POM-C and POM-N in the sole maize cropping system. The increase in POM-C and POM-N amounted to 35% and 56% respectively. A planned contrast showed these differences to be statistically significant (Table 1.3). Marriot et a1. (2006) reported similar results in a comparison of organic and conventionally fertilized trials in the US. Makumba et al. (2006a) demonstrated annual N additions flom gliricidia that averaged near 100 kg ha", but produced grain yield increases commensurate with 50 kg N ha" applied as calcium ammonium nitrate. This indicates that gliricidia is associated with a lower yield efficiency than fertilizer N. The enrichment of POM-C and POM-N, however, demonstrates the potential of gliricidia-derived N to become available to maize in subsequent years. The lessening of N efficiency in a given year may be less important than the hedge against the risk of a shortfall in availability of fertilizer N in the firture, where POM-N provides that buffer. POM C/N ratios were decreased by the gliricidia cropping system, and N and P additions (Table 1.2 and Figure 1.5). This may have important irrrpacts on early-season maize nutrition and yield. Maroko et a1. (1998) found a similar effect with a sesbania fallow on both an alfisol and an oxisol. The effect of the gliricidia was far greater than the effect of the N and P additions. Since the C/N ratios were reduced flom 17.9 to 15.5, decomposition of POM should release N upon decomposition, rather than immobilizing l8 N. This result is contradicted by the data in Figure 1.6, which shows greater immobilization with the gliricidia intercrop. This effect, however, may be associated with the dry season, when soil was sampled. This immobilized N may be remobilized with the onset of seasonal rains and addition of gliricidia green manure and added mineral N. Makumba et al. (2006a) found that gliricidia green manure added in October increased yields of maize planted in December more than green manure added later in the cropping season, so that re-mobilization would be expected to occur after the 30 days of interval used in the incubation. 4. 4. Soil texture and soil organic matter Giller et a1. (1997) state that temperature and moisture availability control the initial rate of decomposition of SOM, while texture controls the extent to which decomposition eventually proceeds. The tendency of finer-textures soils to maintain higher SOM levels has been widely reported. The tendency of clay colloids to slow decomposition of organic matter may result both flom adherence between colloid and organic molecules (Giller et al., 1997), and increased aggregation with increasing clay content (Vanlauwe et al., 1998). In our study, SOM increased with clay contents increasing flom 18.7% to 44.6% equally in gliricidia and sole maize treatments (Figure 1.3), and the additive effect of the gliricidia and clay raised SOM to 29.4 g kg" in the gliricidia treatments with the finest texture, compared to 21.1 g kg" SOM in the sole maize treatments with the coarsest soil texture. 19 4. 5. Available nitrogen Inorganic N and N mineralization potential (N MP) under these treatments was characterized to elucidate more clearly the N-supplying power of POM in the maize system when intensified by addition of the gliricidia intercrop and/or by fertilizer N. Giller (1997) listed the three principal sources of N for crop production in Aflica as biologically fixed N, organic biomass recycled in the field and mineral N fertilizers. Each of these sources was included in this study, as all maize biomass was recycled in each treatment in both cropping systems and the biological and inorganic sources of N were systematically varied in the research design. The increase in inorganic N observed in the gliricidia cropping system was presumably derived flom both biological and fertilizer N (except in N=0 plots). In a field study of seasonal inorganic N in smallholder maize production on sandy soils in Zimbabwe, Chikowo et a1. (2004) found that mucuna (Mucuna pruriens), and soybean (Glycine max) rotations were associated with 50% increases in soil inorganic nitrogen over the season compared to unfertilized maize, which was similar to the increases found in our study. The high proportion of ammonium-N in the pre- incubated N analysis was typical of soils in climates with a pronounced dry season. Because mineralization proceeds under slightly drier soil conditions than nitrification, ammonium—N tends to accumulate during the dry season (Giller et al., 1997). The materials incorporated in this trial, gliricidia prunings and maize stover, have been described as high and low quality materials with C:N ratios of 16 and 86, respectively, by Makumba (2003). Though the maize stover had not been incorporated at the time of sampling, maize root fragments 20 SIT cor the the WM sum imn‘ Crop. deco clay Cllicj mmh “'0 Lllt (Ham 46.5 CEC. . 3.830“; offini smaller than the 2mm sieve size were observed in the soil samples, and might have contributed to immobilization of N. Vanlauwe et al. (2005) observed complete immobilizatiOn of N of in soil with added maize stover in a 28 day incubation, as compared to incubation of soil with gliricidia leaves added, which mineralized 30% of the N present in the leaves. Makumba (2003) also noted initial N immobilization with the gliricidia/maize residue combination and related increased immobilization to years with lower rainfall totals and increased remineralization of N to higher rainfall totals in some study years. In a temperate field study, Gentry et a1. (1998) also linked N immobilization to crop root residues and low rainfall cropping seasons. The effect of increasing soil sand content on available N under the gliricidia cropping system (Figure 1.6) may derive flom protection of organic matter flom decomposition with increasing clay content. This is supported by the significant effect of clay on total soil organic matter (Figure 1.3). However, leaching of N can also limit the efficiency of N derived flom annual legumes in coarse soils under heavy rainfall in southern Malawi (Mwato et al., 1999). This leaching of N from coarse surface horizons would occur equally in agroforestry species, but may also be limited by tree uptake of N (Harawa et al., 2006). 4. 6. Soil fertility - Overall, gliricidia had positive effects on available N and P in the soil, and on CBC, although not on exchangeable K. Nitrogen fertilizer treatments were positively associated with all soil fertility indices in the intercrop, while the shorter term application of fertilizer P had a substantial effect only on soil available P. Vanlauwe et a1. (2000) found Olsen P levels under unfertilized alley cropping and sole maize in West African 21 savannah soils did not differ, indicating that the woody species were not associated with increased availability of P to the cropping system. The decrease in Bray-1 P values in our study under gliricidia may have been the result of an export of soil P in increased grain and biomass yields (Table 1.4). Akinnifesi et al. (2007) found that gliricidia and the first increment of N and P additions had a synergistic effect on grain yields. Mweta et al. (2007) has described maize uptake of P in excess of the soil’s ability to supply P in this trial. ‘ Both designations of the soil, a Ferric Lixisol (F A0), or Oxic Haplusalf (USDA) (Ikerra et al., 1999), highlight the presence of Fe and A1 oxides in the clay fraction which typically fix P, and lessen the ability of the soil to supply sufficient P under high production conditions. Mweta et al. (2007) reported that the Fe-P flaction was the dominant form of inorganic P in soils in this trial. The fixation capacity of the soil in our study may be altered by the extent to which SOM binds with Fe and Al oxides, lessening their fixation capacity, however, not sufficiently to maintain soil P in maize cropping systems. Gliricidia significantly increased available P in all soils, but available P more than doubled as clay content declined flom 36% to 24%. Decrease in soil potassium (K) values associated with increasing N additions in the sole maize system (Table 1.3, Table 1.4) may have been an export effect of higher yields similar to that observed with soil P. Soil K+ is rarely deficient in Malawi agricultural soils (Snapp, 1998), and the Ministry of Agriculture has not recommended K+ additions (Malawi Government Ministry of Agriculture, 1996). The significant effect of gliricidia and N addition and their interactions on CBC (Table 1.4) could be related to the significant increase of SOM flom these treatments. 22 Though the increase was only 0.2 cmol kg", it is operationally important because CEC serves as a conduit for the supply of nutrients for maize in a given year, and lessens the leaching of monovalent cations such as K+ and N114+ in coarse-textured alfisols. The significant effect of clay and its interactions on CBC should be due to clay’s character as a source for much of the CEC at neutral or near neutral pH in most soils. The highly significant interaction of clay with gliricidia in CEC (Figure 1.7) demonstrates the ability of gliricidia to maintain soil CEC through SOM additions in soils with low clay contents. The ability of SOM to maintain CBC and soil fertility in sandy soils is an important factor in maize-based cropping systems in alfisols. 23 5. CONCLUSIONS The results flom this study are consistent with the maize/gliricidia intercrop increasing SOM levels compared to sole maize. As production systems are intensified, tree intercrops may play an important role in preventing or slowing soil degradation and maintaining the supply of plant nutrients to maize in intensified cropping systems in the coarse soils typical in the region. Particulate organic matter pools, which act as a reservoir of crop nutrients, were enhanced by the gliricidia intercrop, which was also associated with a significant decrease in POM C/N ratio flom 17.9 to 15.5, potentially increasing N availability to crops. However, because of immobilization during the dry season timing of application of gliricidia very early in the cropping season is important so that immobilization may be reversed (Makumba et al., 2006a). That immobilization is eventually reversed is supported by findings that maize yields in the gliricidia cropping system averaged four times those in the sole maize system (Akinnifesi et al., 2007). 24 TABLES Table 1.1 Mean, maximum, minimum, and standard deviation for response variables. Mean Max. Mln. St. Dev. SOM (g kg") 24.6 33.0 18.0 3.41 POM (g kg") 14.0 23.0 6.6 3.80 POM-C (g kg") 1.58 2.49 0.63 0.504 POM-N (g kg") 0.09 0.16 0.03 0.035 POM cm 17.1 19.6 14.6 1.39 Avail. N (mg kg") 10.8 28.4 4.49 4.64 NMP (mg kg") -7.72 -1.10 -222 4.00 Bray P (mg kg") 61.0 123 17.0 24.4 Exch. K“ (mg kg") 176 373 44.0 75.6 CEC (cmol kg") 8.49 ‘ 10.5 5.00 1.11 Clay content (%) 28.1 44.6 18.4 6.32 Sand content (%) 56.0 66.2 41.7 5.76 Base saturation (%) 67.2 86.3 50.9 8.55 pH 5.62 6.20 5.00 0.25 25 Table 1.2 Means and significance Of cropping system, N rate, P rate, and clay and sand content on soil organic matter, particulate organic matter, and carbon in particulate organic matter of whole soil. Standard errors are in parentheses following the means. SOM POM POM-C POM-N POMC/N Statistical sigificance (Pr>F) Cropping System (CS) <.0001 <.0001 <.0001 <.0001 <.0001 N fertilizer (N) NS 0.0169 0.0058 0.0026 0.0013 P fertilizer (P) NS NS NS NS 0.0399 Clay 0.0161 NS NS NS NS Sand NS NS NS NS 1 NS N*P NS 0.0471 NS NS NS CS. x N NS NS NS NS NS Flamed contrast:(intercrop + N fertilizer=0) versus ‘ (sole m_aiize + N fertilizer=48 kgha") 0.0025 <.0001 ' hdeans Sole Maize (8 kg") (8 kg") (8 kg") (8 kg") (8 kg") N =0 20.7(0.37) 10.2(2.6) 0.95(0.31) 0.056(0.018) 17.9(0.85) N=48 21.9(0.38) 12.6(2.6) 1.27(0.29) 0.071(0.018) 18.3(0.82) N=96 22.2(0.20) 12.3(1.7) 1.31(0.20) 0.080(0.012) 17.5(0.68) Maize/Gliricidia N =0 23.3(0.29) l4.9(1.8) 1.71(0.23) 0.111(0.013) 15.6(0.66) N=48 24.6(0.30) 17.3(2.6) 2.03(0.21) 0.126(0.013) 16.0(0.62) N=96 24.8(0.12) 17.0(1.7) 2.07(0.11) 0.135(0.006) 15.1(0.49) Sole Maize 21 .6(0.32) 11.7(2.3) l.18(0.27) 0.069(0.016) 17.9(0.78) Maize/Gliricidia 24.3(0.24) 16.4(2.0) 1.94(0.18) 0.124(0.011) 15.5(0.59) N=0 22.0(0.33) 12.6(2.2) 1.33(0.27) 0.084(0.016) 16.8(0.76) N=48 23.3(0.34) 15.0(2.6) 1.65(0.10) 0.098(0.016) 17.2(0.72) N=96 23.5(0.16) 16.0(1.7) 1.69(0.16) 0.108(0.009) 16.3(0.59) Interactions not listed were non-significant at 0.05 for all response variables. *NS: non-significant (Pr > F) >0.05 26 Table 1.3 Significance of cropping system, N rate, P rate, and clay and sand content on available N, Bray E, cation exchange capacity, and potassium (K+) values of soil. Inorg. N Brag P K+ CEC Statistical significance (Pr>F) Cropping System 0.0054 0.0014 NS 0.0002 N <.0001 0.0495 <.0001 0.01 15 P 0.0334 <.0001 0.0147 NS Clay NS <.0001 NS 0.0005 Sand NS NS NS NS Cropping System x N NS NS NS 0.0304 Cropping System x clay NS NS 0.0010 0.0009 N x clay NS NS NS 0.0104 P x clay NS NS NS 0.0291 Cropping System x sand 0.0232 NS NS NS P x sand 0.0416 NS NS NS Interactions not listed were non-significant at 0.05 for all response variables. NS: non-significant (Pr > F) >0.05 Table 1.4 The effect Of cropping system and added N or added P on concentration of plant nutrients in study soils, with standard errors of treatment means in parentheses following concentrations. Inorg N NMP Exch. K+ CEC pH Bray P 6118 kg") (mg kg") (mg kg") (cmol kg") (mg kg") Sole Maize Sole Maize N=0 6.2(3.1) 3.4(1.7) 166(39.8) 7.75(0.99) 5.8(0.09) P=0 40.4(12.5) N=48 6.6 (3.1) 3.0(l.8) 132(401) 8.82(0.99) 5.6(0.08) P=20 68.6(12.5) N=96 11.0(2.1) 5.6(0.8) 81.0(24.9) 8.22(0.65) 5.3(0.16) P=40 87.3 (8.2) Gliricidia! Maize ‘ Gliricidia/Maize N = 0 II.I(2.3) -8.0(l.7) 96.4(27.5) 8.01(0.7l) 5.9(0.09) P = O 31.7 (9.1) N=48 11.5(2.3) -7.6(1.8) 130(27.8) 9.08(0.71) 5.7(0.16) P=20 61.6 (9.0) N=96 15.8(1.2) -10.2(0.8)181(12.6) 8.48(0.36) 5.5(026) P=40 78.8 (4.7)_ 27 FIGURES E1 2003/4 I 2004/5 I 2005/6 Precipitation (mm) Fig. 1.1 Monthly precipitation, Makoka Research Station, 2003-2006. 28 20 3:” 3.9 .15" Q) I: a. E U "510‘ at 1°." 0 3 a . 3 5 .0 E a 9" a 0 30- Soil Organic Matter (g kg") Sole Maize Gliricidia N Rates (kg ha") E10 I48 I 96 Fig. 1.2 The effect of fertilizer N rate and gliricidia on soil organic matter and particulate organic matter. Error bars represent standard error of treatment means. Bars with the same letter within each cropping system are not significantly different. 29 35 30 25 20 15 10 Soil Organic Matter (g kg-‘) 0 I I I T 0 10 20 30 40 50 Clay Content (%) +Sole Maize +Gliricidia Fig. 1.3 The effect of clay content and gliricidia on soil organic matter, estimated at endpoints and quartiles of clay content. Error bars represent standard error of treatment means. 30 9 .° .° 9 O O A .3 Ci ‘0 N 0'! l l I J Nitrogen in POM (g kg") 0 "o w Carbon in POM (g kg") Sole Maize Gliricidia Nrate (kg ha") E10 I48 I96 Fig. 1 .4 The effect Of gliricidia and fertilizer N on carbon and nitrogen in particulate organic matter. Error bars represent standard error of treatment means. Bars with the same letter within each cropping system are not significantly different. 31 (a) 20 .L O .3 on C/N Ratio :1 —l m .3 0| .1. h (b) 17 - 16.5 ,. 16 - 15.5 - 15 » 14.5 - CIN Ratio 0 N rate (kg ha") E10 I48 I96 Fig. 1.5 The effect ofgliricidia, fertilizer N and fertilizer P on UN ratios Of particulate organic matter. (a) Sole maize (b) Gliricidia/maize intercrop Error bars represent standard error of treatment means. Bars with the same letter within each P rate are not significantly different. 32 15 1] o’ 10 p .E if; 5 E” 0 ._ Z 2 5 I: _ - g :3 a -$ < a '10 ' o 9- -15 52 56.25 60.7 Sand Content (%) DPre-inc. Gliricidia I Pre-inc. Sole Maize I Post-inc. Gliricidia I Post-inc. Sole Maize Fig. 1.6 The effect of gliricidia and sand content on available N pre-incubation and post- incubation, estimated at the quartile boundaries of sand content. Error bars represent standard error Of treatment means. 10 + Sole maize CEC (cmole kg“) .5 2 -IMGfifickfia o . . . . . . 20 22 24 26 ' 28 30 32 Clay content (%) Fig. 1.7 The effect of gliricidia and clay content on cation exchange capacity, estimated at the quartile boundaries of clay content. Error bars represent standard error of treatment means. 33 REFERENCES Ajayi, O.C., S. Franzel, E. Kuntashula, and F. Kwesiga. 2003. Adoption of improved fallow technology for soil fertility management in Zambia: Empirical studies and emerging issues. Agroforestry Systems 59:317-326. Akinnifesi, F ., W. Makumba, G. Sileshi, O. Ajayi, and D. Mweta. 2007. Synergistic effect of inorganic N and P fertilizers and organic inputs flom Gliricidia sepium on productivity of intercropped maize in Southern Malawi. Plant and Soil 294:203-217. Barrios, E., and M.T. Trejo. 2003. Implications Of local soil knowledge for integrated soil management in Latin America. Geoderma 111:217-231. Barrios, E., R]. Buresh, and J .1. Sprent. 1996. Organic matter in soil particle size and density flactions flom maize and legume cropping systems. Soil Biology and Biochemistry 28:185-193. Benson, T., J. Kaphuka, S. Kanyanda, and R. Chinula. 2002. Malawi, An Atlas Of Social Statistics National Statistical Office, Government of Malawi and the lntemational Food Policy Research Institute. Bray, RH, and LT. Kurtz. 1945. Determination of total, organic, and available forms Of phosphorus in soil. . Soil Sci. 59:39—45. Cambardella, C.A., and E.T. Elliott. 1993. Carbon and Nitrogen Distribution in Aggregates flom Cultivated and Native Grassland Soils. Soil Sci Soc Am J 57:1071-1076. Chikowo, R., P. Mapfumo, a. P. Nyamugafata, and KB. Giller. 2004. Maize productivity and mineral N dynamics following different soil fertility management practices on a depleted sandy soil in Zimbabwe 102:119-131. Chinangua, L.L.R. 2006. Adoption of soil fertility improvement technologies among smallholder farmers in southern Malawi, Norwegian University Of Life Sciences, Oslo, Norway. Chirwa, T.S., P.L. Mafongoya, and R. Chintu. 2003. Mixed planted-fallows using coppicing and non-coppicing tree species for degraded Acrisols in eastern Zambia. Agroforestry Systems 59:243-251. Cochran, W.G. 1954. Some methods for strengthening the common x2 test Biometrics 10:417-451. Ellis, F., M. Kutengule, and A. Nyasulu. 2003. Livelihoods and Rural Poverty Reduction in Malawi. World Development 31:1495-1510. 34 Ettema, CH. 1994. Indigenous Soil Classification Systems Introduction to Ethnopedology. University of Georgia, Athens,.GA. FAO. 1999. Lecture Notes on the Major Soils Of the World [Online] http://www.fao.org/DOCREP/003/Y l 899%] 899e08a.htm. FEWSNET. 2007. Malawi Food Security Update Food Security Update. Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis, p. 383-411, In A. Klute, ed. Methods of Soil Analysis, Pt. 1. Physical and Mineralogical Methods Agronomy Monographs 9 - 2nd edition. American Society of Agronomy, Madison, WI. Gentry, L., M. David, K. Smith, and D. Kovacic. 1998. Nitrogen Cycling and tile drainage nitrate loss in a corn/soybean watershed. Agriculture, Ecosystems & Environment 68:85-97. Giller, K.E., W.D. Sakala, and PL. Mafongoya. 1997. Building Soil Nitrogen Capital in Aflica, In R. J. Buresh, ed. Replenishirrg Soil Fertility in Aflica. Soil Science Society of America, Madison, Wisconsin. Haney, R.L., A.J. Franzluebbers, E.B. Porter, F.M. Hons, and DA. Zuberer. 2004. Soil Carbon and Nitrogen Mineralization: Influence of Drying Temperature. Soil Sci Soc Am J 68:489-492. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania, USA. Harawa, R., J. Lehmann, F. Akinnifesi, E. Femandes, and G. Kanyama—Phiri. 2006. Nitrogen dynamics in maize-based agroforestry systems as affected by landscape position in southern Malawi. Nutrient Cycling in Agroecosystems 75:271-284. Ikerra, S.T., J.A. Maghembe, P.C. Smithson, and R.J. Buresh. 1999. Soil nitrogen dynamics and relationships with maize yields in a gliricidia-maize intercrop in Malawi. Plant and Soil 211:155-164. Kamanga, B. 2002. Understanding the farmer's agricultural environment in Malawi International Maize and Wheat Improvement Center (CIMMYT). Kravchenko, A.N., G. P. Robertson, X. Hao, and D.G. Bullock. 2006. Management Practice Effects on Surface Total Carbon: Differences in Spatial Variability Patterns 98:1559-1568. Kwesiga, F., F.K. Akinnifesi, P.L. Mafongoya, M.H. McDermott, and A. Agumya. 2003. Agroforestry research and development in southern Aflica during the 19905: Review and challenges ahead. Agroforestry Systems 59:173-186. Lehmann, J., M. da Silva Cravo, and W. Zech. 2001. Organic matter stabilization in a Xanthic Ferralsol of the central Amazon as affected by single trees: chemical 35 charicterization of density, aggregate, and particle size flact ions. Geoderma 99:147-168. , Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. Wolfinger, and O. Schabenberger. 2006. SAS for Mixed Models, Second Edition SAS Institute Inc., Cary, NC, USA. Mafongoya, P., A. Bationo, J. Kihara, and B. Waswa. 2006. Appropriate technologies to replenish soil fertility in southern Aflica. Nutrient Cycling in Agroecosystems 76:137-151. Makumba, W. 2003. Nitrogen Use Efficiency and Carbon Sequestration in Legume Tree- based Agroforestry Systems. A Case Study in Malawi, Wageningen University and Research Center, Wageningen, The Netherlands. Makumba, W., B. Janssen, O. Oenema, and F. Akinnifesi. 2006a. Influence of time of application on the performance of gliricidia prunings as a source of N for maize. Experimental Agriculture 42:51-63. Makumba, W., B. Janssen, O. Oenema, F.K Akinnifesi, D. Mweta, and F. Kwesiga. 2006b. The long-term effects of a gliricidia-maize intercropping system in Southern Malawi, on gliricidia and maize yields, and soil properties. Agriculture, Ecosystems & Environment 1 16:85-92. Malawi Government Ministry of Agriculture, 1., and Livestock Development. 1996. Guide to Agricultural Production, Lilongwe, Malawi. Maroko, J.B., R.J. Buresh, and RC. Smithson. 1998. Soil Nitrogen availability as Affected by Fallow-maize Systems on Two Soils in Kenya. Biology and Fertility of Soils 26:229-234. Marriott, E., and M. Wander. 2006. Total and Labile Soil Organic Matter in Organic and Conventional Farming Systems. Soil Science Society of America Journal 70:950- 959. McIntosh, J.L. 1969. Bray and Morgan soil test extractants modified for testing acid soils flom different parent materials. Agron. J. 61:259-265. Mehlich, A. 1984. Mehlich no.3 extractant: A modification of Mehlich no.2 extractant. Commun. Soil Sci. Plant Anal. 15:1409—1416. Moran, K.K., J. Six, W.R. Horwath, and C. vanKessel. 2005. Role of Mineral-Nitrogen in Residue Decomposition and Stable Soil Organic Matter Formation. Soil Science Society of America Journal 69: 1 739-1 736. Mwato, I.L., A.B.C. Mkandawire, and SK Mughogho. 1999. Combined inputs of crop residues and fertilizer for smallholder maize production in southern Malawi Aflican Crop Science Journal 7:365-373. 36 Mweta, D.E., F.K Akinnifesi, J.D.K. Saka, W. Makumba, and N. Chokotho. 2007. Use Of pruning and mineral fertilizer affects soil phosphorus availability and flactionation in a gliricidia/maize intercropping system. Aflican Journal of Agricultural Research 2:521 -527. Nkamleu, GB, and V.M. Manyong. 2005. Factors Affecting the Adoption of Agroforestry Practices by Farmers in Cameroon. Small-scale Forest Economics, Management and Policy 42135-148. Oorts, K., B. Vanlauwe, O. Cofie, N. Sanginga, and R. Merckx. 2000. Charge characteristics of soil organic matter flactions in a Ferric Lixisol under some multipurpose trees. Agroforestry Systems 48: 169-1 88. Pattanayak, S., D. Evan Mercer, E. Sills, and J.-C. Yang. 2003. Taking stock of agroforestry adoption studies. Agroforestry Systems 57: 1 73-1 86. Phiri, A., G. Kanyama-Phiri, and S. Snapp. 1999. Maize and sesbania production in relay cropping at three landscape positions in Malawi. Agroforestry Systems 47: 1 53- 162. Plante, A.F., R.T. Conant, C.E. Stewart, K. Paustian, and J. Six. 2006. Impact of Soil Texture on the Distribution of Soil Organic Matter in Physical and Chemical Fractions. Soil Sci Soc Am J 70:287-296. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania. Reynolds, L. 2000. Malawi: Country Pasture/Forage Resource Profiles Grassland and Pasture Crops. FAO. Richter, D.d., Jr., M. HOfmockel, M.A. Callaham, Jr., D.S. Powlson, and P. Snrith. 2007. Long-Term Soil Experiments: Keys to Managing Earth's Rapidly Changing . Ecosystems 10.2136/sssaj2006.0181. Soil Sci Soc Am J 71 :266-279. Saxton, K.E., and W.J. Rawls. 2006. Soil Water Characteristic Estimates by Texture and Organic Matter for Hydrologic Solutions 70: 1569-1578. Schroth, G. 1998. A review ofbelowground interactions in agroforestry, focussing on mechanisms and management options. Agroforestry Systems 4325-34. Sherrod, L.A., G.A. Peterson, D.G. Westfall, and LR. Ahuja. 2005. Soil Organic Carbon Pools after 12 Years in No-Till Dryland Agroecosystems. Soil Science Society Of America Journal 69:1600-1608. ' 37 Sileshi, G., F. Akinnifesi, O. Ajayi, and F. Place. 2008. Meta-analysis Of maize yield response to woody and herbaceous legumes in sub-Saharan Aflica. Plant and Soil 307:1-19. Six, J., P.A. Schultz, J.D. Jastrow, and R. Merckx. 1999. Recycling Of sodium polytungstate used in soil organic matter studies. Soil Biology and Biochemistry 31:1193-1196. Snapp, S. 1998. Soil Nutrient status of smallholder farms in Malawi. Communications in Soil Science and Plant Analysis 29:2571-2588. Snapp, S. 2008. Agricultural Systems: Agroecology and Rural Innovation for Development Academic Press, Elsevier. Snapp, 8., G. Kanyama—Phiri, B. Kamanga, R. Gilbert, and K. Wellard. 2002a. Farmer and Researcher Partnerships in Malawi: Developing Soil Fertility Technologies for the Near-tenn and Far-term. Experimental Agriculture 38:411-431. Snapp, S.S., D.D. Rohrbach, F. Simtowe, and HA. Freeman. 2002b. Sustainable soil management options for Malawi: Can smallholder farmers grow more legumes? Agriculture, Ecosystems & Environment 91:159-174. Szott, L., C. Palm, and R. Buresh. 1999. Ecosystem fertility and fallow firnction in the humid and subhumid tropics. Agroforestry Systems 47:163-196. Tripp, R. 2008. Agricultural Change and Low-Input Technology, In S. Snapp and B. Pound, eds. Agricultural Systems: Agroecology and Rural Innovation for Development. Elsevier. Vanlauwe, B., C. Gachengo, K. Shepherd, E. Barrios, G. Cadisch, and CA. Palm. 2005. Laboratory Validation of a Resource Quality-Based Conceptual Framework for Organic Matter Management 10.2136/sssaj2004.0089. Soil Sci Soc AmJ 69:1 135-1145. Vanlauwe, B., S. Aman, K. Aihou, B. Tossah, V. Adebiyi, N. Sanginga, O. Lyasse, J. Diels, and R. Merckx. 1998. Alley cropping in the moist savanna of West-Aflica: III. Soil organic matter flactionation and soil productivity. Agroforestry Systems 42:245-264. Vanlauwe, B., K. Aihou, S. Aman, B.K. Tossah, J. Diels, O. Lyasse, S. Hauser, N. Sanginga, and R. Merckx. 2000. Nitrogen and phosphorus uptake by maize as affected by particulate organic matter quality, soil characteristics, and land-use history for soils flom the West Aflican moist savanna zone 30:440-449. Waddington, S., W.D. Sakala, and M. Mulugetta. 2004. Progress in Lifting Soil Fertility in Southern Aflica. Fourth lntemational Crop Science Congress. 38 Wander, M. 2004. Soil Organic Matter Fractions and Their Relevance to Soil Function, p. 67-102, In F. Magdoff and R. R. Weil, eds. Soil Organic Matter and Sustainable Agriculture. CRC Press. Wander, M.M., S.J. Traina, B.R. Stinner, and SE. Peters. 1994. Organic and Conventional Management Effects on Biologically Active Soil Organic Matter Pools 58:1130-1139. WestcoScientific. 2007. Nitrate-Nitrite in Water, Waste Water, Soil Extracts and other aqueous samples SmartChem Method 1421-1006, Brookfield, Connecticut. Whiteside, M., and S. Carr. 1997. Services and policies needed to support sustainable smallholder agriculture in Malawi [Online]. Available by Agricultural Services Reform in Southern Aflica http://wwwcldis.org/fi111text/ma1awi.pdf(verified May). 39 CHAPTER TWO CHARACTERISTICS OF SMALLHOLDERS AND THEIR PLACEMENT OF GLIRICIDIA SEPIUM ON VARIOUS SOIL TYPES IN SOUTHERN MALAWI ABSTRACT Gliricidia sepium (Jacq.) Walp has been promoted as an intercrop in maize-based cropping systems in southern Malawi to increase yields and improve soil quality. However, smallholders use and placement of the cropping system by soil types has not been investigated. Thirty-seven households using the gliricida/maize intercrop were interviewed in June and July of 2006. Two questionnaires were completed with each household concerning specific cropping practices, and demographic statistics in the household. Socioeconomic status, the number and gender of the people in the household, total land area available to each household, soil types found and cropping systems used on the land were documented so that household characteristics could also be related to use of the gliricidia/maize intercrop. The total area of land used by each household ranged flom 0.75 hectares to 2.8 hectares with an average of 0.58 hectares per household. Households using the intercrop had an average of 0.4 hectares of land, while those not using the intercrop had an average of 0.83 hectares. A chi-square test of this relationship indicated dependence of use of the intercrop and landholding size. Female-headed households were more heavily represented in use of the intercrop than households headed by males with 66% using gliricidia, versus 49% intercrop use among households headed 40 by males. lntercrop use was 50% among households in the lower half of the socioeconomic scale and 63 % in the upper half. Use of the intercrop was found not to be dependent on the number of adults between 15 and 49 years of age in the household. Placement of the gliricidia/maize intercrop on sandy M’chenga soils was less common than on finer-textured Katondo and Makande soils. Nevertheless, a chi-square test showed placement of the intercrop to be independent of soil type. Among the households surveyed, gender, income, and the number of adults per houshold appeared to present little barrier to use of the intercrop. Households with lower than average landholdings were more likely to use the intercrop. The intercrop was placed on all soil types, though there was a trend toward placement on finer-textured soils (Figure 2.5). 41 INTRODUCTION Declining yields in sub-Saharan Aflica have been attributed to soils degraded flom continuous cropping with few inputs. Whiteside and Carr (1997) have described 60% of Malawian smallholders as net buyers of maize. They must work in other’s fields to buy or earn food. Their own fields are often cultivated and weeded poorly or late, reducing yields and forcing households further down the poverty spiral with each cropping season. Maize is the staple crop among most of Malawi’s population. Malawi’s maize production fell short of the food requirement in 4 of 10 years flom 1997 to 2007 (FEWSNET, 2007). Maize production depends heavily on rainfall, soil fertility, and availability of nitrogen (N) inputs (Giller et al., 1997). The use of nitrogen fertilizers to improve maize production in the landlocked countries of southern Aflica is constrained by the high fertilizer prices and difficult logistics in terms of transport and timeliness. Smallholders in southern Malawi have listed high prices as the major limitation to expansion of fertilizer N use (Chinangua, 2006). Both annual and perennial legumes have been rotated and irrtercropped with maize to increase soil N in circumstances where addition of nitrogen fertilizer is limited (Phiri et al., 1999; Snapp et al., 2002b). Several agroforestry species such as Sesbania sesban (L.) Merr., T ephrosia vogelii Hook, f., Gliricidia sepium (Jacq.) Walp. And Leucaena leucocephala (Lam) have been developed and introduced in southern Aflica as renewable soil nitrogen (N) and soil organic matter (SOM) sources in maize-based cropping systems (Kwesiga et al., 2003) 42 Gliricidia was introduced into southern Malawi to increase biological N fixation in areas where high demographic pressures required the use of intercropping strategies rather than rotations (Akinnifesi et al., 2007). Gliricidia may be intercropped with maize without reducing maize populations if it is regularly coppiced (Makumba, 2003). Tripp (Tripp, 2008) lists labor requirement as a fundamental determinant of the acceptability of a technology to farmers. Labor availability has been found to be negatively related to adoption of the gliricidia/maize intercrop in southern Malawi (Chinangua, 2006). Because gliricidia must be regularly coppiced, preferably in October and again in December or January (Makumba, 2003), it adds a labor requirement to the cropping cycle. For some smallholders, the timing of the second pruning conflicts with labor needed for tobacco production (Harawa et al., 2006). Snapp (2008) has described crop species as adapted to particular features of the landscape. Barrios et al. (2003) have described smallholders classification of soil types in Honduras as heavily dependent on slope position and plant species descriptors. However, in the same study, soils in Columbia were categorized by color and soils in Venezuela were categorized by vegetation and texture. Ettema (1994) has compiled a list of 20 published soil ethnographies in which indigenous peoples describe soil color and texture to categorize soil types. In Nigeria, Raji et al. (2006) found smallholders using landscape position and soil fertility to describe primary soil categories, then adding more detail by using color and texture. Kamanga has studied smallholder’s categories for soils in central and southern Malawi, facilitating smallholder’s descriptions of the use of different intercrops on specific soil types in our study. 43 In moving technologies that show biological promise for increasing soil fertility flom research stations to farms, it is important to study the technology under the variety of growing conditions imposed by variation in soil types, elevation and temperature, rainfall and other biophysical factors. It is also important to verify that technologies such as the gliricidia intercrop may be integrated into local farming systems not only by well- to-do innovators, but also by households such as those described by Whiteside and Carr, which are more subject to the cycle of poverty and land degradation. The current study investigates whether the gliricidia/maize intercrop has been successfully used by households which are relatively impoverished, female-headed, or have restricted access to land or labor. It also describes smallholder use of the gliricidia/maize intercrop as one of a number of intercrops typical of the region, and investigates whether their placement on various soil types is independent. 44 MATERIALS AND METHODS The households and fields in this study were located between 15° 12' and 15° 37' S latitude and 35° 12' E and 35° 23' E longitude, in Zomba district, in southern Malawi. Zomba district was chosen as the location for this study because of the high concentration of fields with the gliricidia/maize intercrop within the district and within individual villages. Three other districts were considered, but had lower incidence of use of the gliricidia/maize intercrop. Survey methods Thirty-seven of the households interviewed were chosen purposively chosen based on ICRAF records for their use of the gliricidia/maize intercrop. In addition, twenty-eight fields were chosen flom those near and similar in soil type to the gliricidia/maize fields, flom households that did not adopt the maize/gliricidia intercrop (Table 2.1). The non-adopter households chosen represented nineteen villages scattered across 35% of the area of the district. Because the sampling was purposive rather than random it is possible that the sample membership and the properties of interest are not entirely independent. For example, those with higher income often have more education, and thus more access to information on innovations such as agroforestry practices. Because they have more information the probability that they will adopt agroforestry practices may be higher than the probability for a low income household. To the extent that the probabilities diverge, the two groups are not directly comparable. The information presented in the results section assumes that any divergent probabilities have less influence than the variables listed, for example that the necessity of intensifying a 45 smaller than average size landholding influences the decision to adopt the gliricidia intercrop more than the difference in information access. The two questionnaires (Appendix 1) used in the survey were developed by the author at Michigan State University, and pre-tested and refined with the assistance of the study enumerator and three users of the gliricidia/maize intercrop near Lilongwe. In the I study, carried out in the eastern half of Zomba district, two questionnaires were completed with each household, one concerning specific cropping practices used in the fields that were sampled, and another concerning demographic statistics of the household. Socioeconomic status, the number and gender of the people in the household, total land area available to each household, soil types found and the cropping systems used on the land were taken so that household characteristics could also be related to use of the gliricidia/maize intercrop. The total land area available to each household, the soil types found and the cropping systems used on the land were noted so that use of the gliricidia/maize intercrop could analyzed in context of typical cropping systems. A rough gauge of comparative prosperity the household was taken by noting improvements in the quality of the dwelling. A score of zero represented a house without a metal roof, glassed windows, or fired bricks, while a score of three indicated addition of all of these elements. Placement of the intercrop Brief information was taken on the aggregate land available to the household, including the size of each field, the soil type predominant there and the cropping system typically used in the field. The area of the gliricidia intercrop and comparison fields was measured with a Garmin Etrex Legend GPS. This GPS unit has an accuracy ofi 15m, 46 which was adequate while measuring fields greater than one-tenth hectare. Other fields areas were described by the farmer in relation to the size of the gliricidia intercrop field. Four soil types were included in the questionnaire: Katondo, Makande, M’chenga, and Mtsilo. These classifications were among seven listed by Kamanga (2002), as used by smallholders to describe soil types in central and southern Malawi. The cropping systems identified by the farmers included sole maize, maize/pigeonpea intercrop, maize/gliricidia/pigeonpea intercrop, maize/gliricidia intercrop, tobacco, and vegetables. Data analysis Means and standard deviation are presented for continuous variables such as landholdings and number of people per household (Table 2.2). Frequencies are presented fOr categorical variables such as farmer-defined soil types. A Pearson’s Chi-square test was used to test independence of four relationships: gender and prosperity, landholding size and gliricidia use, number of adults and gliricidia use, and cropping system and soil type. Pearson’s Chi-square test is a simple method for examining statistical independence of categorical data (Schabenberger and Pierce, 2002, p. 385). A two-way table was constructed for each relationship with the observed flequency each possible outcome. Expected flequencies were calculated for each cell in the table using the flequencies of occurrence of each of the categories (Table 2.3). Pearson’s Chi-square test was used to test whether the occurrence of the possible outcomes indicated independence or dependence of the two categories. Four fields typically used for tobacco production and one field with the mtsilo soil type were excluded flom the intercrop placement 47 analysis, and the male household was left out of the gender and prosperity analysis as the presence of categories with very few entries makes the Pearson’s chi-square test less reliable (Cochran, 1954). Two households were excluded flom the analyses of gender and prosperity because of missing data. 48 I. acumen-.9“ w RESULTS Demographic data Households using the gliricidia/maize intercrop were found in all the prosperity groups studied (Table 22, Figure 2.2). Increased use of the intercrop was seen in the more prosperous groups, with gliricidia use at 60% in the poorest group, 43% and 60% of the second and third groups, and 67% of those householdswho were the most well off (e. g., who could afford three improvements to their dwelling). The group was structured with 37 households using the intercrop out of 65 households, or 57%, so that the two poorer groups of households were reduced below the average in use of the intercrop, and the two more prosperous households were increased above the average in use of the intercrop. Household size ranged flom one to fifteen people, with an average of 5.5 people per household. The average number of people between the ages of 15 and 49 across all households is 2.3, and the standard deviation is 1.4. A chi-square test of independence indicated that use of the gliricidia intercrop was independent of the number of prime-age adults resident in the household (Table 2.3). The gliricidia/maize intercrop was used by households headed by male and female individuals (Table 2.1, Figure 2.3). Female- headed households were more heavily represented in use of the intercrop than households headed by couples with 66% using gliricidia, versus 49% gliricidia intercrop use among households headed by couples. Because gender and socioeconomic status may be linked, i.e. households without an adult male may have less access to resources, this relationship 49 was tested. Among the households in this study, gender status was found to be independent of socioeconomic status (Table 2.3). Land available to the household ranged flom 0.75 hectares to 2.8 hectares with an average of0.58 hectares per household (Table 2.2). Households using the intercrop had an average of 0.4 hectares of land available for their use, while those not using the intercrop had an average of 0.83 hectares of land available (Figure 2.4). The chi-square test indicated that landholding size and use of gliricidia were not independent (Table 2.3). Although land is sometimes allotted based on size of households, the number of people in the household and land available were not strongly correlated (Figure 2.4). Placement of the intercrop The households described three major categories of soil types available to them: Katondo, Makande, and M’chenga. These soil types occurred in similar proportions among respondents fields, with 59 fields in which the Katondo soil type predominated, 59 in Makande, and 58 in M’chenga (Table1.2). Of the four cropping systems considered, maize/pigeonpea was the most common with 107 fields. There were 37 Maize/gliricidia/pigeonpea and 22 sole maize fields. The gliricidia/maize intercrop was the least common with 10 fields. The most common combination of cropping system and soil was the maize/pigeonpea intercrop on M’chenga soils (Figure 2.5). The two intercrops which contained gliricidia were more commonly placed on Katondo and Makande soils, and less common on M’chenga soils. The chi-square test of independence indicated that cropping system placement was statistically independent of soil type, as the test statistic of 5.7 did not exceed the 50 In- J.‘ "n- S.- critical value of 12.6 (Table 2.3). Thus, the trends seen in the present data set toward placing gliricidia intercrops on finer-textured soils may or may indicate widespread farmer preferences. When the chi-square test was done with the two gliricidia cropping categories combined, the test statistic of 4.6 also did not exceed the critical value of 9.5. 51 DISCUSSION Demographic setting Southern Malawi is a densely populated landscape (Peters, 1996). Maize harvest provides rural households in southern Malawi with the bulk of their food supply, though the majority of househo lds are net purchasers of maize (Peters, 1996). Most landholdings in southern Malawi are less than 0.5ha per household (Chirwa et al., 2003), on which maize, pigeOnpea (Cajanus cajan L.), beans (Phaseolus vulgaris L.), pumpkin (Cucurbita pepo L.), and COWpea (Vigna unguiculata L.) are commonly cropped. Maize and pigeonpea are the most common intercrop (Chirwa et al., 2003). Ellis et al. (2003) also found that 40% of household landholdings in Malawi were less than 0.5ha. Benson et al. (2002) gave a population density of near 200 persons per km2 for Zomba district, versus 105 persons km2 for Malawi. Benson et al.(2002) in the Malawi, An Atlas of Social Statistics, classify households according to the use of traditional versus permanent housing, where traditional corresponds to our prosperity rating of 0, and permanent housing corresponds with our rating of 3. Benson et al.(2002) give 66% of Malawian houses as traditional versus 15% permanent, with similar proportions for Zomba district. They also designate 31% of households in Malawi as female-headed in 2002 (Benson et al., 2002), while more than 45% were female headed in Zomba district. In our sample, 41% ofhouseholds were female-headed, which is near the district average. Therefore the sample of households in our study were comparable in landholdings to others in southern Malawi, with 0.58ha of land. In our sample, 41% of households 52 were female-headed, which is near the district average of 45 %. The households in our sample, however, used less often in traditional housing, which could indicate higher prosperity and ability to invest in agricultural improvements. The households in our sample were similar in the number of female-headed households to the district average, which was well above the national average. Demographic data Pattanayak et a1. (2003) have found measures of income and assets to be positively correlated with agroforestry adoption, except when these assets were flom non- farm sources. The finding flom the survey were generally consistent with Pattanyak. Households in the upper half of the prosperity scale in our study used gliricidia 10% more often than the average for the whole group, while households in the lower half had 7% less gliricidia use than the average. Whiteside and Carr (1997) find that 60% of the smallholders in Malawi are net buyers of maize, and in many cases labor in neighbor’s fields to buy or earn maize. As a result, many have insufficient time to apply soil improvement practices on their own land. Households in the lower half of the socioeconomic scale in our study used the gliricidia/maize intercrop less than the average of all households. This reduction in intercrop use may stem flom the circumstances described above; however even in these most irrrpoverished groups, adoption of the technology did occur. Pattanayak et al. (2003) have found that households with higher proportions of males were more likely to adopt agroforestry practices. Among the smallholders in our study, the oppositecase was true. Sixty-six percent of female-headed households used the gliricidia/maize intercrop, while 49% of households headed by a couple used the 53 intercrop. The somewhat greater use of the gliricidia/maize intercrop by female-headed households in Zomba district implies that the intercrop is manageable and useful under - the conditions faced by female-headed smallholder households. Given that gender-based measurements often reflect both economic and cultural realities, it may be that this group of female-headed households are no more constrained flom adopting the intercrop by these economic realities than households headed by couples. The lack of dependence between gender and socioeconomic scale in this sample supports this hypothesis. Ajayi et al.(2003) found farm size to be positively correlated with adoption of improved fallows in five of seven studies reviewed. The difference in these findings and ours may be in the cropping systems referenced. The improved fallows studied by Ajayi require land to be taken out of maize production, while the gliricidia/maize does not, making it more acceptable to those with smaller plots of land. Nkamleu et al. ( 2005) found different factors to affect adoption of different agro forestry practices in Cameroon, and warned against generalization. Placement of the intercrop Kamanga (2002) lists seven soil types by indigenous descriptors as common agricultural soils in Malawi. Of these, he describes M’chenga and Katondo as the two most common agricultural soils in Malawi. These two were listed by smallholders in our study, as well as Makande soils. Kamanga (2002) equates Katondo soils with ferric rhodustalfs, Makande soils with vertisols, and M’chenga soils with sandy ferralitic soils. Though the three soil types were found in almost equal measure among the smallholders (Table 2.2), the gliricidia/maize intercrop was used less on sandy M’chenga soils than on the finer-textured Katondo and Makande soils. However, the placement of cropping 54 systems was found to be statistically independent of soil type. The trend toward establishing gliricidia on finer-textured soils did not reach statistical significance, but it should not be ignored. Focus group discussions conducted in April of 2006 in Kasungu district indicated that gliricidia was less drought resistant in sandy soils than in fine- textured soils. The issue Of the drought resistance of gliricidia in different soil types deserves firrther attention. E- u. 55 CONCLUSIONS Household with limited resources were associated with a moderate reduction in the use of the gliricidia/maize intercrop. However, many of these resource-poor households were able to use of the technology. The fact that gliricidia intercrop users had half the average landholdings as non-users may indicate that it is especially useful in high population density regions such as southern Malawi. Since female-headed households were more heavily represented in the use of the intercrop than households headed by a couple, the constraints faced by female-headed households do not seem to limit the use of the intercrop, an important issue, given the loss ofworking-age adults to HIV/AIDS in the current generation. Since cropping system placement was found to be independent of soil type, it might be assumed that the gliricidia/maize cropping system could be promoted for all soil types. However, the issue of drought resistance of gliricidia in differing soil types should be more carefully investigated to avoid burdening impoverished households with unnecessary production risk. 56 TABLES Table 2.1 Sampling flame for the on-farm study. EPA Cropping fistem No. villages No. households No. fields Thondwe gliricidia 2 4 4 paired maize 4 non-gliricidia 3 3 Total 7 1 I Dzaone gliricidia 7 14 14 paired maize 14 non-gliricidia 1 1 1 1 Total 25 39 Mpokwa gliricidia 4 8 8 paired maize 8 non-gliricidia 7 7 Total I 5 23 ‘ Malosa gliricidia 6 l l 11 1 paired maize 11 non-gliricidia 7 7 Total 18 29 §r_and total 1 9 65 102 Table 2.2 Mean and standard deviation for continuous variables and flequencies for categorical variables among the 65 households surveyed. Mean Minimum Maximum St. Dev. Land available (ha) 0.58 0.075 2.75 0.55 Number of people/househo 1d 5.5 l 15 2.8 Number of people aged 15-49 2.3 l 5 1.4 Dependency ratio 1 .9 0 5 1.4 Frequencies of categorical variables: Improvements to dwelling: Gender of household heads: None 1 1 Male 38 One 14 Female 27 Two 18 Three 18 Frequencies of cropping systems: Frequencies of soil types: Sole Maize 22 Katondo 59 Gliricidia/maize 1 0 Makande 59 Gliricidia/maize/pigeonpea 37 M’chenga 58 Maize/pigeonpea 1 07 57 Table 2.3 Chi-Square test of independence for resource availability and gender and for the use ofthegliricidia/maize intercrop and available land and labor. y_ariables alpha d.f. table value test statistic“ Use of intercrop x available labor 0.05 1 3.84 0.148 0.10 1 2.71 0.148 Prosperity x gender 0.05 3 7.81 2.26 0.10 3 6.25 2.26 Use of intercrop x landholdings 0.05 l 3.84 4.86 0.10 1 2.71 4.86 Cropping system x soil type 0.05 6 12.6 5.7 0.10 6 10.6 5.7 ** The chi-square test statistic is generated by {I ((obs. value -— exp. value)"2)/exp. value Table 2.4 Two-way table of flequencies for Chi-square test of cropping system placement on farmer-defined soil types. No. Sole Gliricidia/ Gliricidia/maize/ Maize/ Fields maize maize pigeonpea pigeonpea Observed values: Katondo 59 *7(12)** 5 (8) 14 . (24) 33 (56) Makande 59 10(17) 2 (3) 14 (24) 33(56) M’chenga 58 5(9) 3 (5) 9 (16) 41 (71) *frequency of occurrence "Percent occurrence across all cropping system and soil categories 58 FIGURES ‘3" , Malosa EPA 1 Thondwe EPA Mpokwa EPA Chin-Id. . / / 1"“ , ’ m w W / x \ / \ WW I I \ ma. / \ M J". ' \ \ Dzaot Dzaone-EPA \ Nam ’ \ / \ \ ’/ F______.___—| uall / Figure 2.1 Map of Zomba district, with the study fields and villages. 59 N U’I N O a 0| Frequency d O 0| O 1 Comparative prosperity lGIiricidialmaize DOther Maize Figure 2.2 Comparative prosperity of households and use of the gliricidia/maize, as estimated by the number of improvements made to the family dwelling among tin roof, glassed windows, and fired bricks. 42 35 28 Frequency : Q \l l Female Male ElNon-gliricidia ' IGIiricidia/maize Figure 2.3 Gender of household head/s and use of the gliricidia/maize intercrop. 60 Frequency 0 1 2 3 Improvements to Dwelling 13 Male I Female I Couple Figure 2.4 Relationship of available resources and gender. People per household 16 14 12 10 cur-axon X -——X o .83: X 0 0X 0 o 1—DlX-fi : )DK 6 o -- *X-Clt-X—X-o-X X XX -—X—)§-o—X 9 X X X 0 0.5 1 1.5 2 2.5 3 Landholding (ha) 0 HH members (non-gliricidia) X HH members (gliricidia) Figure 2.5 Landholding size and number of people per household. 61 \l G 60 - —_—i __ 50 d—— —— >. ‘5 q, 40 -— —- 8- e 30 *— IL. 20 —— 10 0 a . . Katondo Makande M’chenga DMaizePigeonpea IMaize/Gliricidia/Pigeonpea IMaize/Gliricidia ISoIe Maize Figure 2.6 Frequency of cropping systems on soil types. 62 REFERENCES Agehara, S., and DD. Wamcke. 2005. Soil Moisture and Temperature Effects on Nitrogen Release flom Organic Nitrogen Sources 10.2136/sssaj2004.0361. Soil Sci Soc Am J 69:1844-1855. Ajayi, O.C., S. Franzel, E. Kuntashula, and F. Kwesiga. 2003. Adoption of improved fallow technology for soil fertility management in Zambia: Empirical studies and emerging issues. Agroforestry Systems 59:317-326. Akinnifesi, F., W. Makumba, G. Sileshi, O. Ajayi, and D. Mweta. 2007. Synergistic effect of inorganic N and P fertilizers and organic inputs flom Gliricidia sepium on productivity of intercropped maize in Southern Malawi. Plant and Soil 294:203-217. Anderson, D. no date. On-Farm Research Guidebook. University of Illinois. Barrios, E., and M.T. Trejo. 2003. Implications of local soil knowledge for integrated soil management in Latin America. Geoderma 111:217-231. ’ Barrios, E., R.J. Buresh, and J .I. Sprent. 1996. Organic matter in soil particle size and density flactions flom maize and legume cropping systems. Soil Biology and Biochemistry 28:185-193. Benson, T., J. Kaphuka, S. Kanyanda, and R. Chinula. 2002. Malawi, An Atlas of Social Statistics National Statistical Office, Government of Malawi and the lntemational Food Policy Research Institute. Bray, R.H., and LT. Kurtz. 1945. Determination of total, organic, and available forms of phosphorus in soil. . Soil Sci. 59:39—45. Cambardella, C.A., and E.T. Elliott. 1993. Carbon and Nitrogen Distribution in Aggregates flom Cultivated and Native Grassland Soils. Soil Sci Soc Am J 57:1071-1076. Chikowo, R., P. Mapfumo, a. P. Nyamugafata, and K.E. Giller. 2004. Maize productivity and mineral N dynamics following different soil fertility management practices on a depleted sandy soil in Zimbabwe 102:119-131. Chinangua, L.L.R. 2006. Adoption of soil fertility improvement technologies among smallholder farmers in southern Malawi, Norwegian University of Life Sciences, Oslo, Norway. 63 Chirwa, T.S., P.L. Mafongoya, and R. Chintu. 2003. Mixed planted-fallows using coppicing and non-coppicing tree species for degraded Acrisols in eastern Zambia. Agroforestry Systems 59:243-251. Cochran, W.G. 1954. Some methods for strengthening the common x2 test Biometrics 10:417-451. Dewees, PA. 1995. Trees on Farms in Malawi: Private Investment, Public Policy, and Farmer Choice. World Development 23:1085-1102. Ellis, F ., M. Kutengule, and A. Nyasulu. 2003. Livelihoods and Rural Poverty Reduction in Malawi. World Development 31 : 1495-1 51 0. Ettema, CH. 1994. Indigenous Soil Classification Systems Introduction to Ethnopedology. University of Georgia, Athens, GA. FAO. 1999a. Lecture Notes on the Major Soils of the World. FAO. 1999b. Lecture Notes on the Major Soils of the World [Online] http://www.fao.org/DOCREP/003/Y1899E/yl899e08a.htm. F EWSNET. 2007. Malawi Food Security Update Food Security Update. Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis, p. 383-411, In A. Klute, ed. Methods of Soil Analysis, Pt. 1. Physical and Mineralogical Methods Agronomy Monographs 9 - 2nd edition. American Society of Agronomy, Madison, WI. Gentry, L., M. David, K. Smith, and D. Kovacic. 1998. Nitrogen Cycling and tile drainage nitrate loss in a corn/soybean watershed. Agriculture, Ecosystems & Environment 68:85-97. Giller, K.E., W.D. Sakala, and PL. Mafongoya. 1997. Building Soil Nitrogen Capital in ‘Aflica, In R. J. Buresh, ed. Replenishing Soil Fertility in Aflica. Soil Science Society of America, Madison, Wisconsin. Haney, R.L., A.J. Franzluebbers, E.B. Porter, F.M. Hons, and DA. Zuberer. 2004. Soil _ Carbon and Nitrogen Mineralization: Influence of Drying Temperature. Soil Sci Soc Am J 68:489-492. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania, USA. Harawa, R., J. Lehmann, F. Akinnifesi, E. Femandes, and G. Kanyama-Phiri. 2006. Nitrogen dynamics in maize-based agroforestry systems as affected by landscape position in southern Malawi. Nutrient Cycling in Agroecosystems 75:271-284. Hildebrand, RE, and J.T. Russell. 1996. Adaptability Analysis Iowa State University Press, Ames, Iowa, USA. 64 Ikerra, S.T., J.A. Maghembe, P.C. Smithson, and R.J. Buresh. 1999. Soil nitrogen dynamics and relationships with maize yields in a gliricidia-maize intercrop in Malawi. Plant and Soil 211:155-164. Kamanga, B. 2002. Understanding the farmer's agricultural environment in Malawi International Maize and Wheat Improvement Center (CIMMYT). Kravchenko, A.N., G. P. Robertson, X. Hao, and D.G. Bullock. 2006. Management Practice Effects on Surface Total Carbon: Differences in Spatial Variability Patterns 98:1559-1568. Kwesiga, F., PK Akinnifesi, P.L. Mafongoya, M.H. McDermott, and A. Agumya. 2003. Agroforestry research and development in southern Aflica during the 1990s: Review and challenges ahead. Agroforestry Systems 59: 1 73-1 86. Lehmann, J., M. da Silva Cravo, and W. Zech. 2001. Organic matter stabilization in a Xanthic Ferralsol of the central Amazon as affected by single trees: chemical charicterization of density, aggregate, and particle size flactions. Geoderma 99:147-168. Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. Wolfinger, and O. Schabenberger. 2006. SAS for Mixed Models, Second Edition SAS Institute Inc., Cary, NC, USA. Lovett, J.C. 1999. Tanzanian forest tree plot diversity and elevation Journal of Tropical Ecology 15:689-694. Mafongoya, P., A. Bationo, J. Kihara, and B. Waswa. 2006a. Appropriate technologies to replenish soil fertility in southern Aflica. Nutrient Cycling in Agroecosystems 76:137-151. Mafongoya, P.L., B. J., J. Kihara, and BS. Waswa. 2006b. Appropriate technologies to replenish soil fertility in southern Aflica. Nutrient Cycling in Agroecosystems 76:137-151. Makumba, W. 2003. Nitrogen Use Efficiency and Carbon Sequestration in Legume Tree- based Agroforestry Systems. A Case Study in Malawi, Wageningen University and Research Center, Wageningen, The Netherlands. Makumba, .W., B. Janssen, O. Oenema, and F. Akinnifesi. 2006a. Influence of time of application on the performance of gliricidia prunings as a source of N for maize. Experimental Agriculture 42:51-63. Makumba, W., B. Janssen, O. Oenema, F.K Akinnifesi, D. Mweta, and F. Kwesiga. 2006b. The long-term effects of a gliricidia-maize intercropping system in Southern Malawi, on gliricidia and maize yields, and soil properties. Agriculture, Ecosystems & Environment 1 16:85-92. 65 Malawi Government Ministry of Agriculture, 1., and Livestock Development. 1996. Guide to Agricultural Production, Lilongwe, Malawi. Manderscheid, L. 1965. Significance Levels-005, 0.01, or ? Journal of Farm Economics 47:1381-1385. Maroko, J.B., R.J. Buresh, and PC. Smithson. 1998. Soil Nitrogen availability as Affected by Fallow-maize Systems on Two Soils in Kenya. Biology and Fertility of Soils 26:229-234. ' Marriott, E., and M. Wander. 2006. Total and Labile Soil Organic Matter in Organic and Conventional Farming Systems. Soil Science Society of America Journal 70:950- 959. McIntosh, J.L. 1969. Bray and Morgan soil test extractants modified for testing acid soils flom different parent materials. Agron. J. 61:259-265. Mehlich, A. 1984. Mehlich no.3 extractant: A modification of Mehlich no.2 extractant. Commun. Soil Sci. Plant Anal. 15:1409—1416. Moran, KK, J. Six, W.R. Horwath, and C. vanKessel. 2005. Role of Mineral-Nitrogen in Residue Decomposition and Stable Soil Organic Matter Formation. Soil Science Society of America Journal 69:1739-1736. (ed.) 1987. Third Workshop on Aflican Forage Plant Genetic Resources Nairobi, Kenya. lntemational Livestock Research Institute. Mtambanengwe, F ., and P. Mapfumo. 2005. Organic Matter Management as an Underlying Cause for Soil Fertility Gradients on Smallholder Farms in Zimbabwe. Nutrient Cycling in Agroecosystems 73:227-243. Mutsaers, I-1.J.W., ox Weber, P. Walker, and NM. Fischer. 1997. A Field Guide for On-Farm Experimentation IITA, CTA, Isnar. Mwato, I.L., A.B.C. Mkandawire, and SK. Mughogho. 1999. Combined inputs of crop residues and fertilizer for smallholder maize production in southern Malawi Aflican Crop Science Journal 72365-3 73. Mweta, D.E., F.K. Akinnifesi, J.D.K. Saka, W. Makumba, and N. Chokotho. 2007. Use of pruning and mineral fertilizer affects soil phosphorus availability and flactionation in a gliricidia/maize intercropping system. Aflican Journal of Agricultural Research 2:521-527. Nkamleu, GB, and V.M. Manyong. 2005. Factors Affecting the Adoption of Agroforestry Practices by Farmers in Cameroon. Small-scale Forest Economics, Management and Policy 4: 1 35-148. 66 Oorts, K, B. Vanlauwe, O. Cofie, N. Sanginga, and R. Merckx. 2000. Charge characteristics of soil organic matter flactions in a Ferric Lixisol under some multipurpose trees. Agroforestry Systems 48:169-188. Pattanayak, S., D. Evan Mercer, E. Sills, and J.-C. Yang. 2003. Taking stock of agroforestry adoption studies. Agroforestry Systems 57:173-186. Perry, J.N., P. Rothery, S]. Clark, M. Heard, and C. Hawes. 2003. Design, analysis and statistical power of the farm-scale evaluations of genetically modified herbicide- tolerant crops. Journal of Applied Ecology 40:17-31. Phiri, A., G. Kanyama-Phiri, and S. Snapp. 1999. Maize and sesbania production in relay cropping at three landscape positions in Malawi. Agroforestry Systems 47:153- 162. Plante, A.F., R.T. Conant, C.E. Stewart, K. Paustian, and J. Six. 2006. Impact of Soil Texture on the Distribution of Soil Organic Matter in Physical and Chemical Fractions. Soil Sci Soc Am J 70:287-296. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania. Reynolds, L. 2000. Malawi: Country Pasture/Forage Resource Profiles Grassland and Pasture Crops. FAO. Richter, D.d., Jr., M. Hofinockel, M.A. Callaham, Jr., D.S. Powlson, and P. Smith. 2007. Long-Term Soil Experiments: Keys to Managing Earth's Rapidly Changing Ecosystems 10.2136/sssaj2006.0181. Soil Sci Soc Am] 71 :266-279. Saxton, KB, and W.J. Rawls. 2006. Soil Water Characteristic Estimates by Texture and Organic Matter for Hydrologic SOlutions 70: 1 569-1578. Schroth, G. 1998. A review of belowground interactions in agroforestry, focussing on mechanisms and management options. Agroforestry Systems 43:5-34. Sherrod, L.A., G.A. Peterson, D.G. Westfall, and LR. Ahuja. 2005. Soil Organic Carbon Pools after 12 Years in No-Till Dryland Agroecosystems. Soil Science Society of America Journal 69:1 600-1 608. Sileshi, G., F. Akinnifesi, O. Ajayi, and F. Place. 2008. Meta-analysis of maize yield response to woody and herbaceous legumes in sub-Saharan Aflica. Plant and Soil 307:1-19. Six, J., P.A. Schultz, J.D. Jastrow, and R. Merckx. 1999. Recycling of sodium polytungstate used in soil organic matter studies. Soil Biology and Biochemistry 31:1193-1196. 67 Snapp, S. 1998. Soil Nutrient status of smallholder farms in Malawi. Communications in Soil Science and Plant Analysis 29:2571-2588. Snapp, S. 2008. Agricultural Systems: Agroecology and Rural Innovation for Development Academic Press, Elsevier. Snapp, S., G. Kanyama-Phiri, B. Kamanga, R. Gilbert, and K. Wellard. 2002a. Farmer and Researcher Partnerships in Malawi: Developing Soil Fertility Technologies for the Near-terrn and Far-term. Experimental Agriculture 38:411-431. Snapp, S.S., D.D. Rohrbach, F. Simtowe, and HA. Freeman. 2002b. Sustainable soil management options for Malawi: Can smallholder farmers grow more legumes? Agriculture, Ecosystems & Environment 91:159-174. Snedecor, G.W., and W.G. Cochran. 1967. Statistical Methods Iowa State University Press, Ames, Iowa, USA. Szott, L., C. Pahn, and R. Buresh. 1999. Ecosystem fertility and fallow firnction in the humid and subhumid tropics. Agroforestry Systems 47:163-196. Thangata, RH, and J .R.R. Alavalapati. 2003. Agroforestry adoption in southern Malawi: the case of mixed intercropping of Gliricidia sepium and maize. Agricultural Systems 78:57-71. Tittonell, P., K.D. Shepherd, B. Vanlauwe, and KB. Giller. 2008. Unravelling the effects of soil and crop management on maize productivity in smallholder agricultural systems of western Kenya--An application of classification and regression tree analysis. Agriculture, Ecosystems & Environment 123:137-150. Tripp, R. 2008. Agricultural Change and Low-Input Technology, In S. Snapp and B. Pound, eds. Agricultural Systems: Agroecology and Rural Innovation for Development. Elsevier. Vanlauwe, B., C. Gachengo, K Shepherd, E. Barrios, G. Cadisch, and CA. Palm. 2005. Laboratory Validation of a Resource Quality-Based Conceptual Framework for Organic Matter Management 10.2136/sssaj2004.0089. Soil Sci Soc Am J 69:1135-1145. ’ Vanlauwe, B., S. Aman, K. Aihou, B. Tossah, V. Adebiyi, N. Sanginga, O. Lyasse, J. Diels, and R. Merckx. 1998. Alley cropping in the moist savanna of West-Aflica: III. Soil organic matter flactionation and soil productivity. Agroforestry Systems 42:245-264. Vanlauwe, B., K Aihou, S. Aman, B.K Tossah, J. Diels, O. Lyasse, S. Hauser, N. Sanginga, and R. Merckx. 2000. Nitrogen and phosphorus uptake by maize as 68 affected by particulate organic matter quality, soil characteristics, and land-use history for soils flom the West Aflican moist savanna zone 30:440-449. Waddington, S., W.D. Sakala, and M. Mulugetta. 2004. Progress in Lifting Soil Fertility in Southern Aflica. Fourth lntemational Crop Science Congress. Wander, M. 2004. Soil Organic Matter Fractions and Their Relevance to Soil Function, p. 67-102, In F. Magdoff and R. R. Weil, eds. Soil Organic Matter and Sustainable Agriculture. CRC Press. Wander, M.M., S.J. Traina, B.R. Stinner, and SE. Peters. 1994. Organic and Conventional Management Effects on Biologically Active Soil Organic Matter Pools 58:1130-1139. WestcoScientific. 2007. Nitrate-Nitrite in Water, Waste Water, Soil Extracts and other aqueous samples SmartChem Method 1421-1006, Brookfield, Connecticut. Whiteside, M., and S. Carr. 1997. Services and policies needed to support sustainable smallholder agriculture in Malawi [Online]. Available by Agricultural Services Reform in Southern Aflica http://www.eldis.org/ fulltext/ malawi. pd f(verified May). 69 CHAPTER THREE EVALUATING THE IMPACT OF GLIRICIDIA SEPIUM ON SOIL ORGANIC MATTER IN SMALLHOLDER CROPPIN G SYSTEMS IN SOUTHERN MALAWI "629‘ T I 1" ll. ABSTRACT v2 Intercropping maize with Gliricidia sepium (Jacq.) Walp has been promoted in southern Malawi to increase maize yields and improve soil quality. This study investigated whether the gliricidia/maize intercropping increases soil organic matter and soil nutrient values compared to prevailing maize cropping practices in smallholders fields on various soil types at a range of elevations in southern Malawi. Other management factors included in the analysis were sand content of the soil, age and population density of gliricidia, use of nitrogen fertilizer, rainfall, and presence and population of pigeonpea/maize intercrop. Thirty-seven households using the gliricidia/maize intercrop were interviewed as well as twenty-eight that did not adopt the maize/gliricidia intercrop. One questionnaire investigated specific cropping practices used in intercrop fields and comparison fields. Another questionnaire recorded demographic information about the household. Soil samples were taken to a 20 cm depth in the center of the 2005/6 planting rows flom the fields categorized in the questionnaires. Analyses were performed on the soil samples for soil organic matter 70 (SOM), sand content, available P, exchangeable K and CEC. Soil organic matter values associated with the intercrop were not significantly different from other cropping systems. Sand content was negatively related to SOM and was the only factor tested that had a significant impact on SOM. When the gliricidia/maize intercrop fields were analyzed separately, soil type, elevation, and their interaction affected trends in soil organic matter. Trendlirres with positive slopes indicate a positive relationship between soil organic matter and elevation in the two finer soil types, Makande and Katondo, while the relationship was negative in the coarser M’chenga soils. For each 100m increase above sea level (asl) flom the mean elevation, Katondo and Makande soils increased in soil organic matter 10 and 4 g kg" or 1.1% and 0.46% respectively. Elevations among the fields in the study ranged flom 760m to 1064m. Fertilizer N was associated with a slight positive trend in soil organic matter. The addition ofpigeonpea, Cajanus cajan (L.), to the maize cropping system at elevations below 877m asl was associated with an increase of 1.8 g kg " or 11% in soil organic matter values compared to the sole maize cropping system The emergence of increasing trends in SOM correlated with increasing elevation and fineness of soil texture in fields with the gliricidia/maize intercrop may indicate that a SOM benefit will develop over time in fields at relatively higher elevations and finer soil textures. The trend toward increase in SOM with addition of pigeonpea may indicate a SOM benefit to this intercrop at relatively lower elevations. 71 INTRODUCTION Because maize is the staple crop in Malawi and in most of southern Aflica (Chirwa et al., 2003), enhancing production in this crop is a major factor in food security in the region. In Malawi, more than 50% of land farmed by smallholders is annually planted in maize with yields averaging only 1375 kg ha"(Benson et al., 2002). The price of fertilizer is the major reason smallholders in southern Malawi give for not using N fertilizers to improve their maize yields (Chinangua, 2006). Annual legumes such as pigeonpea Cajanus cajan (L.) Millsp. (van der Maesen), and agroforestry species such as T ephrosia volgelli produce leaf litter that has been shown to contribute as much as 50 kg ha" of nitrogen (N) to soil nutrient flows in on- farm studies (Snapp, 1998). Several leguminous trees and shrubs such as Sesbania sesban (L.) Merr., T ephrosia vogelii Hook, f., Gliricidia sepium (Jacq.) Walp. and Leucaena leucocephala (Lam) have been adapted and developed in southern Aflica to increase soil nitrogen (N) and soil organic matter (SOM) in maize-based cropping systems (Kwesiga et al., 2003). Agroforestry practices have a long history in Malawian agriculture, beginning with the indigenous Faidherbia albida. Dewees (1995) listed Faidherbia and mango (Mangifera indica) as the two trees in occupying the most space in farmer’s fields near Lilongwe in the mid-1990’s. In the same study, the plant population of the introduced species Gliricidia sepium (Jacq.) Walp., was higher than the plant population of Faidherbia, but occupied much less space. Intercropping of gliricidia with maize has been shown to increase maize yield in trials at the Makoka agricultural research station in Zomba district in southern Malawi (Akinnifesi et al., 2007). 72 The gliricidia/maize intercrop is promoted as being especially suited to southern Malawi, because the high population densities do not permit rotation of land out of food crops (Thangata and Alavalapati, 2003). Population densities in Zomba district range flom 100-200 persons km2 in the north to 200—400 persons km2 in the south. Increasing population densities and lack of sufficient land to maintain fallow periods have often been identified as important drivers of soil degradation in sub-Saharan Aflica. Benson et al. (2002) found that maize productivity in southern Malawi was lower than in the central and northern regions because high population pressures were required continual cropping, which depleted soil fertility. Tittonell et al. (2008) noted widespread negative nutrient balances on smallholder farms and large gaps between potential and actual maize yields as markers of unproductive soils in western Kenya. Mtambanengwe et al. (2005) have found that soil organic matter is often the single largest source of nutrients for growing crops. As such, characterization of SOM in smallholder fields is crucial in understanding soil nutrient status and potential productivity. Giller et al. (1997) noted the importance of temperature and moisture availability in controlling the rate of decomposition of SOM, and the importance of soil texture in controlling the extent to which decomposition eventually proceeds. In this study, elevation is used as a proxy for temperature. Lovett (Lovett, 1999), while studying tree plot diversity in Tanzania, found elevation and temperature to be strongly correlated. Thus, decreasing temperatures with increasing elevations slow the decomposition of SOM, as does decreasing available moisture, measured as annual rainfall. While the capital city, Lilongwe, is on the Central Aflican Plateau, the southern region is dominated by the final section of the East Aflican rift valley system (Msiska et 73 al., 1987). Elevations are much more variable, even away from uplifts such as Zomba and Mulange mountains, and range flom less than 100m above sea level (asl) to 1200m asl (Benson et al., 2002). Kamanga (2002) described seven soil types typically cited by smallholders in central and southern Malawi in describing land, and ranging flom high to low sand and clay content. The two most common soils were Katondo and M’chenga soils, which varied in texture flom sandy clay loam sandy loam Variations in these biophysical drivers of SOM cycling will be demonstrably important in modifying the m.” tar-h: effects of the gliricidia/maize intercrop on SOM values in smallholder’s fields. rug-*9- It is important that the promising technologies for enhancing maize production such as agroforestry intercrops and improved fallows are also tested on smallholder farms after they are identified at research stations. Franzel et al. (2001) note that replicated on- farm biophysical trials of agroforestry technologies are vital in evaluating the performance of a given technology across the typical range of biophysical conditions encountered by smallholders, but are expensive to maintain. Mutsaers has recommended an alternative method that treats a trial as a randomized complete block with a single replicate in each farmer’s field (Mutsaers et al., 1997). This permits each farmer to evaluate soil fertility management options under her/his own conditions and criteria, while allowing the trial to represent performance under a realistic range of conditions and ' practices. Hildebrand et al. (1996) has noted that with farmer-managed single replicate trials, Anova analysis tests only the main effect, rather than the main effect plus any main effect by environment interaction. They advocate using two replicates per field so as to capture the environmental effect, while acknowledging the difficulties inherent in implementation of multiple replicates in farmer-managed trials. However, the use of 74 mixed analysis allows both categorical variables, such as treatment effects, and continuous covariates, such as elevation to be tested within the same model (Littell et al., 2006). For each on-farm trial, the benefits of farmer management, including farmer preference for single replicate tests must be balanced against the researcher’s desire to test treatment by environment interactions. Because on-farm trials typically contain more variability in conditions than researcher-managed fields, it is also necessary to set an appropriate measure of statistical significance, or apha value for the trial. In researcher-managed plots on research stations, this is usually Pr > F 20050. Snedecor et al. (1967) argue that this value should be seen as a convention to be discussed, rather than an automatic rule. Manderscheid (1965) suggests that the researcher should consider the costs associated with Type One and Type Two errors, the prior probabilities of the hypothesis and alternative, and the size of the Type Two error associated with each significance level. Perry et al. (2003) designate an alpha of 0. l 000 for a field-scale studyin Great Britain, while Anderson (no date) argues that many real differences in treatment effects have been ignored in on-farm research and accepts alpha values up to 0.3000. In the current study, the bulk of the cost of a rejecting the null hypothesis is the extra labor associated with implementing the gliricidia/maize intercrop versus typical practices. The cost of accepting the null hypothesis when it is false is in lost yields when the intercrop is not implemented. Given the flagility of food security in southern Malawi, an increase in the alpha value beyond 0.050 is warranted. The on-farm research approach used in this study will provide an opportunity to study the effects of the gliricidia/maize technology within the smallholders biophysical and management constraints. It will also allow differences in SOM dynamics with 75 changes in elevation and soil type to become apparent. However, the effect on SOM of variability in soil types, available water, and management among the fields may mask SOM eflccts of the intercrop, especially in those fields which contain young gliricidia plantings. Because farmers often manage higher potential fields more intensively (Scoones and Toulmin, 1998) than low potential fields and many tend to use both organic and inorganic treatments in the same field (Omamo et al., 2002) it is often challenging to separate the effects of different interventions. The current study investigates (1) whether the gliricidia/ maize intercrop increases SOM and soil fertility values compared to prevailing maize cropping practices in smallholders fields on various soil types at a range of elevations, (2) whether increasing age or population density ofgliricidia is related to increased SOM values, and (3) whether population density of pigeonpea is related to increased SOM values. Other management factors included in the analysis were addition of fertilizer nitrogen (N), rainfall, and the presence of pigeonpea. 76 MATERIALS AND METHODS A groecology The rainfall pattern in Zomba district where this study was conducted, is unimodal with most rain falling between November and March (Figure 3.1). Three-year annual rainfall averages ranged flom 836mm to 1648mm for the areas studied. Soil types included in the study ranged flom sandy M’chenga soils, to sandy clay loam and sandy clay Katondo soils, and Makande soils with vertic properties. Maize cultivation among smallholders in Southern Malawi is typically done manually with hoes. Planting rows are built in October, before the onset of heavy rains and are usually perpendicular to any slope. After the onset of rains, usually in November or December, maize is planted at populations near 44,400 plants ha". The traditional planting pattern spaced rows at 90cm and planting stations within the rows at 90cm. Three seeds are typically planted per planting station which resulted in a maize population of 37,000 per hectare. This pattern has been partially displaced by a pattern with rows spaced at 75cm and single seeds planted at 30cm intervals which results in a maize population of 44,400 per hectare. Maize is typically weeded twice, once in January and again in February. The maize grain is harvested after drying down in the field, usually in May. Maize biomass is usually recycled within the fields. In southern Malawi, maize is commonly intercropped with legumes, the most common of which is pigeonpea. Pigeonpea is planted soon after maize emergence, and is harvested in September and October. 77 Gliricidia intercrop practices Recommendations for gliricidia/maize intercrops in southern Malawi call for gliricidia to be planted in alternating maize rows at 90 cm spacing within the row (Makumba, 2003), which results in a gliricidia population of 6,150 trees ha" with the 900m row spacing and 7,400 trees ha'lwith the 750m row spacing (Figure 3.2). The trees are usually pruned twice per year, during land preparation and again during the first weeding and leaves and tender twigs incorporated into the maize rows. A pruning during August is sometimes done to produce tender foliage for incorporation during land preparation. Survey methods Zomba district was chosen as the study location because of the high concentration of fields within the district and within individual villages, permitting possible village- level effects to be included in the analysis (Figure 3.1). Three other districts were considered, but had lOwer incidence of use of the gliricidia/maize intercrop. Thirty- seven of the households interviewed were chosen purposively because they were the households in the district currently using the gliricidia/maize intercrop. Twenty-eight fields were chosen flom among those near and similar in soil type to the gliricidia/maize fields, as a comparison among those households in each village which did not adopt the maize/gliricidia intercrop (Table 3.1). The non-adopter households chosen represented nineteen villages scattered across 35% of the area of the district. The proportions of fields in the three cropping system treatments were planned rather than random. Each gliricidia field was paired with a maize field that was managed 78 by the same household, usually adjacent, and as similar as possible in soil type. Another field similar in soil type flom a non-adopter household was also included as a comparison (Table 3.1). Comparisons of cropping systems in this chapter refer to three categories: (1) maize/gliricidia intercrop fields, (2) maize/pigeonpea or sole maize fields managed by the same household, and (3) two fields per village of maize/pigeonpea or sole maize managed by households not using gliricidia. The two questionnaires used in the survey were developed by the author at Michigan State University, and pre-tested and refined with the assistance of the study enumerator in Malawi. One questionnaire investigated specific cropping practices used in the fields that were sampled, and another recorded simple demographic statistics in the household. One of the demographic statistics concerned improvements in the quality of housing, with a score of zero representing a house without a metal roof, glassed windows, or fired bricks. ‘A score of three indicated addition of all of these elements. This score gave a rough estimate of the comparative prosperity of the household. Soil types were characterized in the questionnaire by local descriptors (Kamanga, 2002). Soil Sampling and analysis Soil samples were taken to a 20 cm depth in the center of the 2005/6 planting rows flom the fields categorized above using a straight-walled soil probe with 20m diameter. Eight samples were taken for each l/10 hectare of field area. Areas near houses were avoided when sampling to minimize variability flom additions of household ash. The samples flom each field were combined, air-dried, passed through a 2 mm sieve, and stored. Analyses for soil organic matter, available P, exchangeable K and CEC 79 were performed by A & L Great Lakes Laboratories. A sub-sample of the soil was dried overnight at 40°C and crushed with an Agvise flail-type grinder. Organic matter was determined by loss on ignition at 360° C and the data correlated with and reported as a Walkley-Black titration. Available phosphorus and exchangeable potassium (K+) were extracted according to Mehlich III (Mehlich, 1984), and analyzed by inductively-coupled plasma spectrometry (ICP). The P data was correlated to and reported as Bray P-l (Bray and Kurtz., 1945). The data for exchangeable K+ correlated to and reported as a 1 N ammonium acetate extraction (McIntosh, 1969). CEC was calculated flom the results for exchangeable cations. Sand content was determined after Gee and Bauder( 1986). Ten grams of soil was dispersed by reciprocal shaking for 8 hours with 5 g L" sodium hexametaphosphate. The dispersed samples were washed through a 53 micron sieve with deionized water. The material remaining on the screen was dried overnight at 105 °C and weighed. Field measurements and calculations Elevation was measured at the approximate center of each field with a Garmin Etrex Legend GPS. This GPS unit has an accuracy of :1: 15m, which was adequate to record field locations on a map. This instrument proved inadequate when measuring the area of fields of less than one-tenth hectare and this data is not reported. Because of the height of gliricidia and pigeonpea, random population measurement of gliricidia and pigeonpea with quadrats was not practical. A 4m by 4m section of each field was chosen 80 to be representative of the average population of gliricidia and pigeonpea in the field, and the populations of the target plants counted within the area. None of smallholders surveyed applied nitrogen (N) fertilizer prior to planting. The amount of N applied as a side-dress was calculated for each field using the concentration of N in urea or calcium ammonium nitrate applied, and the volume and spacing of the applications given by the smallholders. Monthly rainfall data for 2003- 2006 was provided by the four agricultural extension planning offices containing fields in the study. Rainfall for each field was estimated by summing the seasonal rainfall after the planting date given by the smallholder for 2005/6 cropping season. Statistical analysis methods Because the sampling was purposive rather than random it is possible that the sample membership and the properties of interest are not entirely independent. For example, those with higher income often have more education, and thus more access to information on innovations such as agroforestry practices. Because they have more information the probability that they will adopt agroforestry practices may be higher than the probability for a low income household. To the extent that the probabilities diverge, the two groups are not directly comparable. The information presented in the results section assumes that any divergent probabilities have less influence than the variables listed, for example that the necessity of intensifying a smaller than average size landholding influences the decision to adopt the gliricidia intercrop more than the difference in information access. 81 Ancovas were conducted using the PROC MIXED procedure in SAS (Littell et al., 2006) to determine if continuous variables such as soil sand content and categorical variables such as cropping system and soil type had significant influence on soil organic matter and soil nutrients. The Ancovas were structured with cropping system as a fixed effect. Village and household were designated as random effects. Other independent variables were designated as covariates. Soil type designations commonly used by smallholders in central and southern Malawi were included in the questionnaire (Kamanga, 2002), as well as sand content by weight. Sand content and soil type were always analyzed in separate models because smallholder’s soil type descriptors depended heavily on sand content. Ancova analyses in PROC MIXED combine categorical variables such as cropping system with continuous variables such as sand content in a fixed effects model. In the PROC MIXED analysis, standard errors are calculated considering both the fixed effects and random effect components of the model (Littell et al., 2006). Where covariates had no significant influence, the models reduced to Anovas using SAS Type 3 analysis, which uses expected mean squares to estimate variance components (Littell et al., 2006). The Ancovas were built parsimoniously to avoid over-fitting. No more than three explanatory variables were included in any model. The probability of a type one error, or alpha value, designated as acceptable is often greater in on-farm and farrn-scale studies than in researcher-managed plots. An‘ alpha value of0.1500 is used in this study, which is intermediate between those given by Perry et al. (2003) and Anderson (no date). 82 RESULTS Environmental variables The households and fields in this study were located between 15° 12' and 15° 37' S latitude and 35° 12' E and 35° 23' E longitude, in Zomba district, in southern Malawi. Sand content of soil ranged flom 45% to 86% across all fields, which corresponds to soil textural classes ranging flom sandy clay to loamy sand, assuming low silt content. Average sand content varied little across the four extension planning areas (EPA) (Table 3.2). Elevation of the fields in the study varied flom 760 to 1064 meters above sea level, in the upper half of the 100m to 1200m range given for Southern Malawi. Average elevations decreased 200m flom Thondwe EPA to Malosa EPA. Rainfall in the 2005- 2006 cropping season varied flom 496mm to 1370mm across all households, and flom 565m in Thondwe EPA to 1350mm in Malosa EPA, probably due to the influence of Zomba mountain, which rises to 2087 meters. Across all fields available to all households, soil types were divided evenly (Table 3.2, Chapter 3). Among the fields sampled for the cropping system comparison however, 36 were Katondo, 37 were Makande, and 28 were M’chenga. Katondo soils predominated in the Thondwe and Malosa EPA’s, M’chenga soils in Dzaone EPA, and Makande soils in Mpokwa EPA (Table 3.2). 83 Management variables Fertilizer N was added to eighty-seven percent of the maize fields sampled during the 2005/6 cropping season. Less than 7 kg ha" N was added to seventy percent of the fertilized fields in the study. Fertilizer use was highest in Thondwe EPA and lowest in Mpokwa EPA. Populations of gliricidia plants per hectare in the gliricidia/maize intercrop fields varied flom 420 to 7500 plants per hectare, averaging of 2300 plants per hectare across all fields. Only gliricidia populations in Thondwe EPA surpassed this" average, while gliricidia populations in the other three EPA’s were lower. Gliricidia plantings ranged in age from 4 to 12 years in age, averaging 8.1 years. Gliricidia plantings in Thondwe and Dzaone EPAs fell below this average, while those in Mpokwa and Malosa surpassed it. Pigeon pea populations ranged flom 0 to 6250 plants per hectare, with an average of 1360 plants per hectare. Pigeonpea populations were highest in Mpokwa EPA and lowest in Thondwe EPA. Response variables Soil organic matter values averaged 20.4 g kg" and ranged flom 10 to 49 g kg" across all fields. SOM values across EPA’s were similar, except for Malosa EPA. Malosa EPA had an average SOM of 17.7 g kg" which was 4.9 g kg" lower than Thondwe with of 22.6 g kg" . Malosa EPA included fewer M’chenga soils than the other EPA’s, and was at a lower elevation than the other EPA’s. The difference in average SOM values between Malosa and Thondwe EPA’s was 4.9 g kg" . Differences in SOM within the farms ranged flom 0 to 6 g kg" across all fields. Eight pairs of fields or 22% had differences in SOM values that exceeded the difference between EPA mean SOM values. The difference between SOM means for Thondwe and Mpokwa was 1.3 g 84 kg" Twenty-seven pairs of fields, or 73% had greater difference in SOM values than the difference in the mean SOM in Thondwe and Mpokwa. Bray P values ranged flom 4 to 135 mg kg", with a mean of33.6 mg kg" and a standard deviation of 25.2 mg kg". Bray P values in Thondwe and Dzaone EPA’s were lower than in Mpokwa and Malosa EPA’s. Exchangeable K+ values ranged flom 46 to 424 mg kg" with a mean of 169mg kg" and a standard deviation of 80 mg kg" . Exchangeable K+ values for Mpokwa EPA were markedly higher than those for the other EPA’s. CEC’s ranged flom 3.2 to 17.6 cmol kg", with an average and standard deviation of 7.1 and 2.6 cmol kg". Average CEC’s for Dzaone and Mpokwa EPA’s were higher than the overall average, while CEC’s for Thondwe and Malosa were lower. Multiple soilfertility enhancing practices The two most common soil-fertility-enhancing practices, aside flom the gliricidia/maize intercrop were incorporation of crop residues and the addition of N in fertilizer. The gliricidia/maize intercrop was usually used in addition to incorporation of crop residues and addition of fertilizer N. Crop residues were incorporated in 36 of the 37 or 97% of the maize/gliricidia fields, and in 36 of the 37 paired maize fields. Crop residues were incorporated in all of the 28 non-adopter maize fields. Fertilizer N was applied in 76% of the maize/gliricidia fields, and 89% of the paired and non-adopter fields, indicating that N flom gliricidia may have been considered a substitute for fertilizer N in some fields. In six of the thirty-seven, or 16% of the households using the maize/gliricidia intercrop, an average of 5.3 kg ha" of fertilizer N was applied to the maize field without the gliricidia intercrop, while no N was applied to the field with the intercrop. On 27 of the 37, or 73 % of households using the maize/gliricidia intercrop, an 85 average of 5.8 kg ha" of fertilizer N was applied at equal rates to the gliricidia intercrop and comparison fields. Livestock manure, compost, and legume rotation were used in 13%, 9%, and 1% of fields, respectively, and in similar proportions across the three cropping systems. Pigeonpea intercrops were not listed by smallholders as legume rotations, even though 56 of the 102 fields in the study, or 55%, contained pigeonpea. Thirty-eight percent of the gliricidia/maize intercrop fields contained pigeonpea, while 78% of the paired maize fields and 43% of the non—adopter maize fields contained pigeonpea. Soil Organic Matter Cropping system, sand content, soil type, elevation, fertilizer N, rainfall, population and age of gliricidia and presence and population of pigeonpea were tested in a SAS mixed model for effect on soil organic matter, Bray P, exchangeable K+, and CEC. Though SAS mixed analysis showed trends several in relation to Bray P, exchangeable K+ and CEC, the relationships were biologically counterintuitive, and the results are not reported. Variability of soil organic matter among villages Soil organic matter data is presented grouped by elevation and soil type in Table 3.3, because SOMttrends were associated with these explanatory variables. The significance of covariance parameters for the random variables of village, household nested within village, and residual variance are presented in Table 3.3. SOM variability among villages is similar but about 2/3 of the scale of that in individual fields (Table 3.2). When soils were characterized by sand content, the significance of variations in SOM 86 attributable to village effect were highest in the model that included elevation (Table 3.4). However, the residual SOM variability was greater than the village effect in each case. Variability attributable to village effect was less when soils were characterized by soil type than when characterized by sand content (Table 3.4), and residual SOM variation was greater. This indicates that variability in SOM associated with the village effect is much less significant (Pr>F = 0.0140) than the residual variability (Pr>F = 0.0003) between fields after variability flom the listed explanatory variables is taken into account. For this reason, the village effect is not explored further. The multiple soil fertility enhancing practices used on fields in this study are discussed in the following section. Subsequent sections discuss SOM effects of the explanatory variables when the full set of fields studied is included, followed by a discussion of the effects of the explanatory variables in the fields containing gliricidia, and finally in the fields containing no gliricidia. Soil organic matter and explanatory variables Soil organic matter values, associated with the three cropping systems were not significantly different (Pr>F = 0.8245 to 0.2591) across all fields (Table 3.3) in the study whether soils were characterized by sand content or by soil type. Increasing sand content was strongly and negatively related to SOM (Table 3.3, Figure 3.3). For each 10% increase in sand content, SOM decreased approximately 10 g kg", or 14% at the mean sand content. The non-gliricidia cropping systems were concentrated on M’chenga soils with higher sand content while the gliricidia cropping system was more common on soils with lower sand content (Figure 3.3). This is consistent with the trend seen in Chapter 3 that gliricidia placement was less common on the sandier M’chenga soil. When the sand 87 content of the three soil types was compared (Figure 3.4), soils labeled by the smallholders as Mchenga were concentrated at the higher sand contents while the Katondo soils were distributed at lower sand contents. A wide range of sand contents were found in Makande soils. When soil was characterized by sand content (Table 3.4), the soil organic matter trend increased with increasing annual rainfall through 830mm annually, and declined with increasing rainfall above 830mm. When soils were characterized by soil type (Table 3.4), soil organic matter showed an increasing trend with increasing elevation. At the mean elevation of 882m, an increase of 100m elevation resulted in a 2.9 g kg", or 14% increase in soil organic matter. Soil organic matter in fields with the gliricidia/maize intercrop When the gliricidia/maize intercrop fields were analyzed separately (Table 3.5), and soil was characterized by sand content, neither the age (Pr>F = 0.5440) nor the population (Pr>F = 0.9476) of gliricidia plantings had a significant influence on SOM. The number of soil fertility improvement methods used in a given field was added as a categorical variable, but proved to have little direct impact on SOM. Fertilizer N was associated with a slight positive trend in soil organic matter. At the mean rate of application, a 1 kg ha" increase in fertilizer N was associated with a 1.3% increase in SOM. Soil type, elevation, and their interaction had a significant effect on soil organic matter (Table 3.5). Trendlines with positive slopes (Figure 3.5) indicate a positive relationship between soil organic matter and elevation in the two finer soil types, Makande and Katondo, while the relationship was negative in the coarser M’chenga soils. 88 For each 100m increase flom the mean elevation, Katondo and Makande soils increased in soil organic matter 10 and 4 g kg" or 1.1% and 0.46% respectively, while M ’chenga soils decreased in SOM 2.3 g kg" or 0.26%. Elevations among the fields in the study ranged flom 760m to 1064m. Elevations in the Zomba district range flom 300m to 2100, with most of the area in the surrounding the study fields ranging from 600m to 1200m (Benson et al., 2002). Planting patterns of gliricidia showed evidence of having been established at the recommended population and spacing, and subsequently to have lost plants flom this planting pattern without replacement by most smallholders. Mean population per hectare was 2300, or one-third of the recommended plant population (Table 3.2). Plant populations were markedly low in plantings flom four to six years in age (Figure 3.6). Soil organic matter infields without the gliricidia/maize intercrop Increasing elevation was associated with an increasing trend (Pr>F =-0.1836 to 0.0703) of SOM in fields without the gliricidia intercrop (Table 3.5, Figure 3.7) when soils were characterized by soil types. Soil organic matter averages in sole maize and maize/pigeonpea cropping systems were very similar in aggregate, 20.40 g kg " and 20.37 g kg " respectively. However, the addition of pigeonpea to the maize cropping system at elevations below 877 meters was associated with an increase of 1.8 g kg " or 1 1% in soil organic matter levels compared to the sole maize cropping system (Figure 3.7). 89 DISCUSSION Ferralsols and Lixisols, which are common in southern Malawi, have comparatively low CECs and absolute levels of plant nutrients due to the predominance of high sand content (Figure 3.3) and low-activity clays (FAO, 1999a). This makes recurrent inputs of plant nutrients a pro-condition for continuous cultivation. Given the limited natural fertility of these soils, we would expect to be able to demonstrate the benefits of the gliricidia/maize intercrop, which provides a recurrent nutrient source, over sole maize. Though the expected SOM effect of the gliricidia intercrop was not evident, the expected biogeochemical drivers of SOM content were more evident in the gliricidia/maize intercrop fields than in the comparison fields without the intercrop. However, multiple soil-fertility-enhancing practices were used in the fields studied, so that SOM effects flom individual practices might interact with one another, making it difficult to separate individual effects. Tittonell et a1. (2008) found that farmers reinforced soil fertility variability in their fields by investing more resources on already fertile fields. As in many on-farm studies the differences in SOM attributable to the treatment studied were smaller than differences in SOM between fields, due to differing biophysical characteristics, such as elevation and soil texture, and to differing management that stems flom these characteristics. Mafongoya et al. (2006b) found that in Southern Aflica within farm and within field variability of soil nutrients is sometimes greater than mean differences across districts. In our study, within farm variability of SOM was greater than the largest difference in SOM between EPA’s in 22% of cases. Within farm variability of SOM was greater than the second largest difference in SOM among EPA’s in 73% of cases. Tittonell et al. (2008) found that management and 90 environmental effect may also be negatively reinforcing. For example, if a field with low soil fertility is planted late, it is also subject to greater soil erosion, lessening the soil fertility further. Study of such interactions requires multiple soil sampling at designated intervals and is beyond the scope of this study. Samplingframe issues The smallholders participating in this study are assumed to be representative of the l,“. -‘ I 1 larger population of smallholders in southern Malawi, However their villages were within ten kilometers of paved roads and within the same district as the Makoka research station, which should give them greater accessibility to gliricidia seedlings and knowledge of the cropping system. Soilfertility enhancing practices Makumba (2003) listed incorporation of crop residues in his description of typical cropping practices for maize-based cropping systems. These findings are consistent with our finding that 97% of sampled fields were managed with incorporation of crop residues in both maize/gliricidia intercrops and non-gliricidia fields. Because residue incorporation is used in all of the treatments, this important practice is not being reduced by the use of the gliricidia/maizer intercrop. It has not usually been possible for smallholders to apply the recommended rates of N fertilizer to maize. Price ratios of nitrogen to maize in Malawi may be twice those in Mozambique, which borders Malawi on the east (Mafongoya et al., 2006a). One of the advantages of the gliricidia/maize intercrop is addition of N to the soil in green leaf manure. (add Chirwa, 2006 reference and make sure it is in the diss. Endnote file) (Also 91 add Makumba on amount of N in gliricidia) . Because a lower proportion of maize/gliricidia fields received fertilizer N additions than the comparison fields (76% versus 89%), it is evident that some N fertilizer additions are being replaced with the gliricidia intercrop. Nitrogen fertilizer additions did not have significant explanatory value for soil organic matter values when used in a model (Table 3.4). In 16% of households, however, the intercrop replaced fertilizer N additions, giving growers the option of covering more of their maize cropping area with N additions flom either the intercrop or flom fertilizer N. Seventy-six percent of households added fertilizer N to the maize/gliricidia intercrop and the paired maize fields at the same rate, which had the potential to increase yields more than either treatment alone. Although a synergy between fertilizer N additions and the gliricidia intercrop in maize grain production has been demonstrated at nearby Makoka research station (Akinnifesi et al., 2007), this synergy did not carry through to SOM values, and was not detectable in this on-farm study. Thus, the practice of intercropping maize with gliricidia seems to be additive to the most common two soil- practices of incorporation of crop residues and addition of fertilizer N, because most of the smallholders used both these with all cropping systems. Pigeonpea is valued by smallholders as a food source during the dry season, and for soil enrichment (Snapp, 1998), though it was not classified as a soil improvement practice by either researcher or smallholders in this study. Chirwa (2003), found that the addition of the gliricidia to the maize/pigeonpea intercrop did not decrease the yield of pigeonpea grain, and predicted that gliricidia would not replace pigeonpea in local cropping systems. The fact that pigeonpea was used in 78% of paired maize fields and 92 only 38% of the gliricidia/maize fields might contradict Chirwa’s expectations, or it might reflect the smallholder’s tendency to place gliricidia more often on finer-textured Katondo and Makande soils, leaving the maize/pigeonpea intercrops to predominate on the coarser M’chenga soils. Soil organic matter and explanatory variables The results reported in chapter two demonstrate the maize/gliricidia intercrop to significantly increase soil organic matter compared to sole maize cropping systems. The presence of the maize/gliricidia in the smallholder’s fields in this study did not demonstrate a discemable effect on SOM (Table 3.3, Figure 3.3). However, the results reported in chapter two were reached after 15 years of intensive management at a research station, while the age of the gliricidia plantings in the fields averaged 8 years of age and less than half the recommended gliricidia population density. Soil texture greatly influences decomposition of soil organic matter. Giller et a1. (1997) note that while temperature and moisture availability control the rate of decomposition of SOM, texture controls the extent to which decomposition eventually proceeds. The tendency of clay colloids to slow decomposition of SOM may result both from adherence between colloid and organic molecules (Giller et al., 1997), and increased aggregation with increasing clay content (Vanlauwe et al., 1998). When the full set of fields studied were included in analysis, increasing sand content, and corresponding decrease of clay content of the soil was very strongly and negatively related to SOM content (Table 3.4, Figure 3.3). Descriptors of soil type common in central and southern Malawi were chosen to classify soil types to facilitate descriptions of current placement of the gliricidia/maize 93 intercrop and to simplify communication of any possible recommendations arising flom this research Kamanga (2002), interviewing smallholders in central and southern Malawi, characterized Katondo soils as similar to alfisols (USDA), M’chenga soils as very sandy and Makande soils as having vertic properties. When the sand content of the of the three soils was compared (Figure 3.4), the M’chenga and Katondo soils clustered in the high and low sand contents, respectively, as expected. The Lixisols (F AO, 1999b) listed as the most common soil type in Malawi (Kamanga, 2002) correspond roughly to -:— sure-:1 the Alfisol soil order in the USDA classification system. The soils classified as M’chenga by smallholders may correspond to an Alfisol with a thick, sandy A horizon. The Katondo soils may correspond to a Alfisol, in which the A horizon is partially or wholly eroded, and the B horizon is part of the layer annually tilled for production. The fact that the soils described as Makande extend across the range of sand contents was not expected. Soils with vertic properties, such as Makande, having high-activity clays, should have a greater protective effect on SOM than the low activity clays found in the Katondo soils or alfisols. In moisture-limited cropping systems (Figure 3.1), such as those in this study, increased rainfall leads to greater yields of plant biomass, which is returned to the soil as crop residues. Thus, the increasing SOM values with increasing rainfall may be liked to increased biomass additions. However, rainfall values above 830mm in this study were correlated with decreasing SOM. At higher rainfall values, both rate of biomass addition and rate of SOM breakdown are important influences on SOM values. Agehara et al.(2005) found that soil moisture levels between 50% and 70% maximized microbiological activity during irrcubations, maximizing mineralization of nutrients, and 94 speeding microbial breakdown of SOM. Therefore increased rainfall above 830mm may have increased the rate of SOM decomposition and resulted in decreased SOM values. Statistical analyses similar to those for SOM were carried out for Bray P and exchangeable K values and for CEC. In the few instances that discemable trends were identified, the trends were in contradiction to typical biological drivers and are not reported. Soil organic matter infields with the gliricidia/maize intercrop The fields with the maize/gliricidia intercrop were analyzed separately so as to be able to include age and population of the gliricidia planting as potential factors in SOM values. Gliricidia plantings showed evidence of having originally been established in the recommended planting patterns, then reduced in population. The markedly low populations of gliricidia planted flom the years 2000 to 2002 may have resulted flom insufficient rainfall for establishment. Akinnifesi et al. (2007) have reported below- average rainfall in this district in 2001/2 and 2002/3. This could have reduced the survival rate of trees in gliricidia plantings that were not well established, corresponding to plantings flom four to six years of age in our study. Gliricidia population and age had no significant effect on SOM. However, this result is consistent with the findings of Akinnifesi et al. (2007), in which SOM values were slightly reduced after nine years of monocropped maize, while they were stable after nine years of the maize/gliricidia intercrop. Sand content had a strong negative effect on soil organic matter. (Table 3.3) The most pronounced trend in SOM values of fields with the maize/gliricidia intercrop was 95 found when soils were characterized by soil type (Table 3.5). Soil type and the interaction of elevation and soil type show strong trends, and the effect of elevation on SOM approaches statistical significance. Giller et al. (1997) listed temperature and moisture availability as factors that controlled the rate of decomposition ’of SOM, and reviewed comparative research in England and Nigeria that determined that rates of SOM decomposition were doubled with an 8° to 9° C increase in temperature independent of differences in rainfall and ‘- 77:1 temperature patterns. Lovett (1999) found a strong correlation between elevation and temperature in Tanzania, and measured a 05° to 06° C decrease in mean monthly $.1an . . maximum and minimum temperatures for each 100m increase in elevation. Given a similar lapse rate in Zomba district, the 300m elevation range in this study would generate a difference in temperature only one- fourth that described by Giller. If the relationship of decomposition rates and temperature is linear, this could result in differences of 25% in rates of decomposition between soils at higher and lower elevations, based solely on temperature differences. Increasing clay content would be expected to increase SOM. Soils with vertic properties, such as Makande soils (Kamanga, 2002), would be expected to have a greater protective effect on SOM than the low-activity clays found in Katondo soils. The effect of soil type on the influence of elevation on SOM values is clear in Figure 3.5. The strongest effect of increasing elevation on SOM values is among thesoils characterized as Makande. The eflect of elevation on SOM is positive with Katondo soils containing low-activity clays, but the magnitude of the effect is lower, while the effect of elevation is reversed with sandy M’chenga soils. It may be that in Malawi’s ustic environment that 96 evapotranspiration is decreased with increasing elevations, leading to increased maize yields, and increased mining of soil nutrients and SOM. Soil organic matter in fields without the gliricidia/maize intercrop The effect of elevation on SOM in fields without the gliricidia intercrop is similar to the effect of elevation in fields with the intercrop (Table 3.4, Table 3.6). The interaction with soil type is not evident, however, in fields without the gliricidia intercrop. The pigeonpea plant is of smaller stature than the gliricidia plant, and average populations of pigeonpea were only 60% of populations ofgliricidia. Thus the biomass contributions and resulting SOM values would be expected to be lower. The SOM values, however, are similar (Figure 3.5, Figure 3.7). Aggregate SOM values for the sole maize and maize/pigeonpea cropping systems are also similar. But at elevations below 877m, addition of pigeonpea to the cropping system resulted in an increase of 11% in SOM over the sole maize system. This may be a result of the higher C:N values typical of pigeonpea leaf biomass. Makumba (2003) found that pigonpea leaf plus root biomass had a C:N ratio of 19 , compared to 16 for gliricidia leaf biomass. 97 CONCLUSIONS Among the explanatory variables, sand content had the most pronounced effect on SOM values followed by elevation and soil type. Use of the gliricidia intercrop did not result in discemable effect on SOM when all fields were included. The effect on SOM of elevation and soil type and their interaction were more discemable when only the fields including the gliricidia/maize intercrop were included. This suggests that the gliricidia is supplying a recurrent source of plant nutrients and SOM and that this supply of SOM allowed gradients of increasing SOM to develop with increasing elevation and fineness of soil texture. These results suggest that maize yield increases, when present, should form the first incentive for the use of the gliricidia intercrop, for these yield increases may be generally operative and more immediately valuable. The second incentive, recapitalization of SOM, should be described when combinations of increasing site elevation and increasing fineness of soil textures will maintain increased SOM values. 98 TABLES Table 3.1 Sampling flame for the on-farm study. EPA Cropping System No. vilgiges No. households No. fields Thondwe gliricidia 2 4 4 paired maize 4 non-gliricidia 3 3 Total 7 1 I Dzaone gliricidia 7 14 14 paired maize l4 non-gliricidia l 1 11 Total 25 39 Mpokwa gliricidia 4 8 8 paired maize 8 non-gliricidia 7 7 Total I 5 23 Malosa gliricidia 6 11 l 1 paired maize l 1 non-gliricidia 7 7 Total I 8 29 Grand total 19 65 102 99 Table 3.2 Mean and standard deviation of environmental, management, and response variables, divided by extension planning area. Variable Thondwe Dzaone , Mpokwa Malosa Environmental variables Sand content (%) 66 (9.7)* 69.4 (11.0) 71.8 (7.3) 69.6 (9.5) Elevation (m) 1053 (8.1) 925 (50.2) 822 (37.8) 808 (25.4) Rainfall (mm) 565 (55.0) 1099 (4.0) 904 (131.8) 1315 (90.5) Katondo soil (%) 63.6 31.6 8.7 51.7 Makande soil (%) 18.2 26.3 65.2 34.5 M’chenga soil (%) 18.2 42.1 26.1 13.8 Management variable Fertilizer (kg N ha") 11.0 (6.8) 5.6 (6.5) 4.1 (2.8) 6.2 (5.0) Gliricidia pop. (plants ha") 1408 (2067) 926 (1751) 554 (787) 1080 (1939) Gliricidia age (year) 5.0 (2.9) 7.6 (3.1) 11.0 (3.6) 8.7 (2.5) Pigeonpea pop. (plants ha") 741 (1173) 1497 (8057) 1711 (4972) 1088 (5130) Response variables Soil Organic Matter (g kg") 22.6 (6.7) 22.2 (10.4) 21.3 (7.4) 17.7 (5.8) Bray P (mg kg") 28.0 (24.9) 29.2 (18.9) 43.5 (37.6) 35.9 (25.4) Exchangeable K+ (mg kg") 114 (48.7) 181 (88.7) 207 (88.6) 153 (65.9) CEC (cmol kg") 5.6 (1.4) 8.3 (3.4) 8.1 (2.2) 5.8 (1.3) pH 5.9 (0.23) 6.1 (0.28) 6.4 (0.43) 6.1 (0.35) *Standard deviation is in parentheses following the mean of continuous variables 100 -3 rTn... '— “‘3'!“— Table 3.3 Mean and standard deviation of soil organic matter above_and below the mean elevation for the sub-group of fields, divided by soil type and cropping system Soil Organic Matter (g kg") (maize/ gliricidia fields) Elevation 5 877m Elevation > 877m Katondo 19.9 (6.7) 25.7 (9.4) Makande 17.4 (2.9) 29.3 (12.8) M’chenga 23.3 (3.7) 17.3 (6.8) (fields without gliricidia) Elevation 3 882m Elevation > 882m Sole maize 15.8 (4.8) 22.4 (10.8) Maize/pigeonpea 19.2 (6.04 22.8 L107) Table 3.4 The effect of cropping system, sand content, soil type, elevation, fertilizer N, rainfall, village, household nested within village and residual variance on soil organic matter and CEC in all fields Statistical Significance for Combinations of Variables Soil Organic Matter: Soil characterized by sand content Pr > F: Cropping system 0.8245 0.5984 0.7954 Sand content <.0001 <.0001 <.0001 Elevation 0.2453 Fertilizer N 0.6709 Rainfall 0. 1454 Covariance parameters: Pr > Z: Village ' 0.0140 0.0228 0.0720 Household(village) 0.0535 0.0739 0.0290 Residual 0.0003 0.0067 0.0002 Soil characterized by soil type Pr > F: Cropping system 0.2591 0.7915 0.3189 Soil Type 0.2716 0.6936 0.2019 Elevation 0.1626 Fertilizer N 0.5143 Rainfall 0.71 89 Significance of covariance parameters: Pr > Z: Village 0.0539 0.0368 0.0865 Household(village) <.0001 0.0007 <.0001 Residual <.0001 0.0009 <.0001 101 mama} Table 3.5 The effect of sand content, soil type, number of soil fertility methods, elevation, rainfall, fertilizer N, gliricidia population, and age of gliricidia on soil organic matter in fields with the gliricidia intercrop. Pr > F for Combinations of Variables Soil Organic Matter: Soil characterized by sand content Sand content <.0001 0.0012 <.0001 <.0001 <.0001 Soil fertility methods 0.6658 0.8826 0.2705 0.6553 0.6643 ElevatiOn 0.9124 Rainfall 0.6179 Fertilizer N 0.1472 m Gliricidia population 0.9476 . - Age of gliricidia 0.5440 3 - Soil characterized by soil types is . Soil Type 0.6164 0.6821 0.5941 0.6045 0.1241 £ Elevation 0.3094 0.2345 0.1860 0.1555 0.0746 Rainfall 0.8226 Fertilizer N 0.7996 Gliricidia population 0.4143 Age of gliricidia 0.5440 Soil Type x Elevation 0.1118 102 Table 3.6 The effect of sand content, soil type, elevation, rainfall, fertilizer N, pigeonpea population and age on SOM in fields without gliricidia. Pr > F for Combinations of Variables Soils characterized by sand content Sand Content <.0001 <.0001 <.0001 <.0001 <.0001 Cropping system 0.5569 0.4371 0.7499 0.6012 0.3944 Elevation 0.361 1 Rainfall 0. 168 l Fertilizer N 0.8246 Pigeonpea population 0.8713 Presence of pigeonpea 0.2600 Soils characterized by soil type Soil Type 0.4601 0.2492 0.9722 0.4581 0.3940 Elevation 0.1836 0.1586 0.0703 0.1743 0.1400 Cropping system 0.7680 Rainfall 0.5227 Fertilizer N 0.5430 Pigeonpea population 0.6520 Presence of pigeonpea 0.1235 103 ‘5: 4‘31?" _ L‘“'_‘."-‘_-""‘" 3 FIGURES 600 500 400 Rainfall (mm) W 8 I “a! “2.; -LL 1 I I Extension Planning Area Figure 3.1 Three-year average annual rainfall from the 2003 to 2006 cropping seasons in the four extension planning areas. ***O***O***O***O***O* } 75cm *k‘skalrilrakaika-kaseascik it 3% *0* it *0* it *0* * 9:021: it *l<>***l I 90cm ****************** * maize plant <3 gliricidia plant Figure 3.2 Planting pattern for the maize gliricidia intercrop given a 750m spacing between rows and single-seeded planting pattern within rows. 104 60 its" - x ‘ E 40 A-. E X X o ’5 X S 30 ”—‘T—T' i ‘ ‘ A! X K o ‘.X A go 20 '—: O o ’5 10 - '+ m P 0 I I t I I i" ' 40 50 60 70 80 90 5‘ Sand Content (%) 1"" OGIiricidia/maize XPaired maize ANon-adopter maize Figure 3.3 Soil organic matter as a firnction of sand content, by cropping system, including all fields. 40 30 20 10 Soil Organic Matter (g kg") 40 50 60 70 80 90 Sand Content (%) OKatondo XMakande AM'chenga Figure 3.4 Soil organic matter as a function of sand content, by soil type, including all fields. 105 (a) O .L O Soil Organic Matter (9 kg") re a 700 800 900 1000 1 100 Elevation (m) 0 Katondo I Makande A M'chenga — - Linear (Katondo) Figure 3.5 The effect of soil type and elevation 011 soil organic matter in fields containing the gliricidia/maize intercrop. 8000 '3. o E 7000 ' ' ‘c‘ e g 6000 . 5 5000 +— 5 e e e e ‘8 e 3000 g s o e _ zooo-—v——-y = 8 e e 8 e ‘9 1000 O i . . 0 I U I fi 3 5 7 9 11 13 Gliricidia Age Figure 3.6 The effect of age of gliricidia planting on gliricidia population. 106 50 40 30 20 10 Soil Organic Matter (g kg") . ’8 X x x x I» e. )9 700 800 900 1000 1100 Elevation (m) OSole Maize XPigeonpea Figure 3.7 Soil organic matter as a fiinction of elevation in fields without the gliricidia/maize intercrop, by cropping practice. 107 REFERENCES Agehara, S., and DD. Wamcke. 2005. Soil Moisture and Temperature Effects on Nitrogen Release flom Organic Nitrogen Sources Soil Sci Soc Am J 69:1844- 1855. Ajayi, O.C., S. Franzel, E. Kuntashula, and F. Kwesiga. 2003. Adoption of improved fallow technology for soil fertility management in Zambia: Empirical studies and emerging issues. Agroforestry Systems 59:317-326. Akinnifesi, F ., W. Makumba, G. Sileshi, O. Ajayi, and .D. Mweta. 2007. Synergistic effect of inorganic N and P fertilizers and organic inputs flom Gliricidia sepium on productivity of intercropped maize in Southern Malawi. Plant and Soil 294:203-217. Anderson, D. no date. On-Farm Research Guidebook. University of Illinois. Barrios, E., and M.T. Trejo. 2003. Implications of local soil knowledge for integrated soil management in Latin America. Geoderma 111:217—231. Barrios, E., R.J. Buresh, and J .I. Sprent. 1996. Organic matter in soil particle size and density flactions flom maize and legume cropping systems. Soil Biology and Biochemistry 28:185-193. Benson, T., J. Kaphuka, S. Kanyanda, and R. Chinula. 2002. Malawi, An Atlas of Social Statistics National Statistical Office, Government of Malawi and the lntemational Food Policy Research Institute. Bray, R.H., and LT Kurtz. 1945. Determination of total, organic, and available forms of phosphorus in soil. . Soil Sci. 59:39—45. Cambardella, C.A., and E.T. Elliott. 1993. Carbon and Nitrogen Distribution in Aggregates flom Cultivated and Native Grassland Soils. Soil Sci Soc Am J 57:1071-1076. Chikowo, R., P. Mapfirmo, a. P. Nyamugafata, and KB. Giller. 2004. Maize productivity and mineral N dynamics following different soil fertility management practices on a depleted sandy soil in Zimbabwe 102:119-131. Chinangua, L.L.R. 2006. Adoption of soil fertility improvement technologies among smallholder farmers in southern Malawi, Norwegian University of Life Sciences, Oslo, Norway. 108 Chirwa, T.S., P.L. Mafongoya, and R. Chintu. 2003. Mixed planted-fallows using coppicing and non-coppicing tree species for degraded Acrisols in eastern Zambia. Agroforestry Systems 59:243-251. Cochran, W.G. 1954. Some methods for strengthening the common )(2 test Biometrics 10:417-451. Dewees, PA. 1995. Trees on Farms in Malawi: Private Investment, Public Policy, and Farmer Choice. World Development 23:1085-1102. Ellis, F., M. Kutengule, and A. Nyasulu. 2003. Livelihoods and Rural Poverty Reduction in Malawi. World Development 31:1495-1510. Ettema, CH. 1994. Indigenous Soil Classification Systems Introduction to Ethnopedology. University of Georgia, Athens, GA. FAO. 1999a. Lecture Notes on the Major Soils of the World [Online] http://www.fao.org/DOCREP/003/Y1899E/yl899e08a.lrtm. FAO. 1999b. Lecture Notes on the Major Soils of the World. FEWSNET. 2007. Malawi Food Security Update Food Security Update. Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis, p. 383-411, In A. Klute, ed. Methods of Soil Analysis, Pt. 1. Physical and Mineralogical Methods Agronomy Monographs 9 - 2nd edition. American Society of Agronomy, Madison, WI. Gentry, L., M. David, K. Smith, and D. Kovacic. 1998. Nitrogen Cycling and tile drainage nitrate loss in a corn/soybean watershed. Agriculture, Ecosystems & Environment 68:85-97. Giller, K.E., W.D. Sakala, and PL. Mafongoya. 1997. Building Soil Nitrogen Capital in Aflica, In R. J. Buresh, ed. Replenishing Soil Fertility in Aflica. Soil Science Society of America, Madison, Wisconsin. Haney, R.L., A.J. Franzluebbers, E.B. Porter, F.M. Hons, and DA. Zuberer. 2004. Soil Carbon and Nitrogen Mineralization: Influence of Drying Temperature. Soil Sci Soc Am J 68:489-492. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania, USA. Harawa, R., J. Lehmann, F. Akinnifesi, E. Femandes, and G. Kanyama-Phiri. 2006. Nitrogen dynamics in maize-based agroforestry systems as affected by landscape position in southern Malawi. Nutrient Cycling in Agroecosystems 752271-284. Hildebrand, PE, and J.T. Russell. 1996. Adaptability Analysis Iowa State University Press, Ames, Iowa, USA. 109 Ikerra, S.T., J.A. Maghembe, P.C. Smithson, and R.J. Buresh. 1999. Soil nitrogen dynamics and relationships with maize yields in a gliricidia-maize intercrop in Malawi. Plant and Soil 211:155-164. Kamanga, B. 2002. Understanding the farmer's agricultural environment in Malawi lntemational Maize and Wheat Improvement Center (CIMMYT). Kravchenko, A.N., G. P. Robertson, X. Hao, and D.G. Bullock. 2006. Management Practice Effects on Surface Total Carbon: Differences in Spatial Variability Patterns 98:1559-1568. Kwesiga, F ., F.K. Akinnifesi, P.L. Mafongoya, M.H. McDermott, and A. Agumya. 2003. Agroforestry research and development in southern Aflica during the 19908: Review and challenges ahead. Agroforestry Systems 59:173-186. Lehmann, J ., M. da Silva Cravo, and W. Zech. 2001. Organic matter stabilization in a Xanthic Ferralsol of the central Amazon as affected by single trees: chemical charicterization of density, aggregate, and particle size flactions. Geoderma 99:147-168. Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. Wolfinger, and O. Schabenberger. 2006. SAS for Mixed Models, Second Edition SAS Institute Inc., Cary, NC, USA. Lovett, J .C. 1999. Tanzanian forest tree plot diversity and elevation Journal of Tropical Ecology 15:689-694. Mafongoya, P., A. Bationo, J. Kihara, and B. Waswa. 2006a. Appropriate technologies to replenish soil fertility in southern Aflica. Nutrient Cycling in Agroecosystems 76:137-151. ‘ Mafongoya, P.L., B. J., J. Kihara, and BS. Waswa. 2006b. Appropriate technologies to replenish soil fertility in southern Afiica. Nutrient Cycling in Agroecosystems 76: 1 37-151. ' Makumba, W. 2003. Nitrogen Use Efficiency and Carbon Sequestration in Legume Tree- based Agroforestry Systems. A Case Study in Malawi, Wageningen University and Research Center, Wageningen, The Netherlands. Makumba, W., B. Janssen, O. Oenema, and F. Akinnifesi. 2006a. Influence of time of application on the performance of gliricidia prunings as a source of N for maize. Experimental Agriculture 42:51 -63. Makumba, W., B. Janssen, O. Oenema, F.K. Akinnifesi, D. Mweta, and F. Kwesiga. 2006b. The long-term effects of a gliricidia-maize intercropping system in Southern Malawi, on gliricidia and maize yields, and soil properties. Agriculture, Ecosystems & Environment 110 l 16285-92. Malawi Government Ministry of Agriculture, 1., and Livestock Development. 1996. Guide to Agricultural Production, Lilongwe, Malawi. Manderscheid, L. 1965. Significance Levels--0.05, 0.01, or ? Journal of Farm Economics 47:1381-1385. Maroko, J.B., R.J. Buresh, and PC. Smithson. 1998. Soil Nitrogen availability as Affected by Fallow-maize Systems on Two Soils in Kenya. Biology and Fertility of Soils 26:229-234. Marriott, E., and M. Wander. 2006. Total and Labile Soil Organic Matter in Organic and Conventional Farming Systems. Soil Science Society of America Journal 70:950- 959. McIntosh, J.L. 1969. Bray and Morgan soil test extractants modified for testing acid soils flom different parent materials. Agron. J. 61 :259-265. Mehlich, A. 1984. Mehlich no.3 extractant: A modification of Mehlich no.2 extractant. Commun. Soil Sci. Plant Anal. 15:1409—1416. Moran, K.K, J. Six, W.R. Horwath, and C. vanKessel. 2005. Role of Mineral-Nitrogen in Residue Decomposition and Stable Soil Organic Matter Formation. Soil Science Society of America Journal 69:1739-1736. (ed.) 1987. Third Workshop on Aflican Forage Plant Genetic Resources Nairobi, Kenya. lntemational Livestock Research Institute. Mtambanengwe, F ., and P. Mapfirmo. 2005. Organic Matter Management as an Underlying Cause for Soil Fertility Gradients on Smallholder Farms in Zimbabwe. Nutrient Cycling in Agroecosystems 73:227-243. Mutsaers, H.J.W., G.K Weber, P. Walker, and NM. Fischer. 1997. A Field Guide for On-Farm Experimentation IITA, CTA, Isnar. Mwato, I.L., A.B.C. Mkandawire, and SK Mughogho. 1999. Combined inputs of crop residues and fertilizer for smallholder maize production in southern Malawi Aflican Crop Science Journal 7:365-373. Mweta, D.E., F.K Akinnifesi, J.D.K Saka, W. Makumba, and N. Chokotho. 2007. Use of pruning and mineral fertilizer affects soil phosphorus availability and flactionation in a gliricidia/maize intercropping system Aflican Journal of Agricultural Research 2:521-527. 111 Nkamleu, GB, and V.M. Manyong. 2005. Factors Affecting the Adoption of Agroforestry Practices by Farmers in Cameroon. Small-scale Forest Economics, Management and Policy 42135-148. Oorts, K, B. Vanlauwe, O. Cofie, N. Sanginga, and R. Merckx. 2000. Charge characteristics of soil organic matter flactions in a Ferric Lixisol under some multipurpose trees. Agroforestry Systems 48:169-188. Pattanayak, S., D. Evan Mercer, E. Sills, and J .-C. Yang. 2003. Taking stock of agroforestry adoption studies. Agroforestry Systems 57:173-186. Perry, J.N., P. Rothery, SJ. Clark, M. Heard, and C. Hawes. 2003. Design, analysis and statistical power of the farrn-scale evaluations of genetically modified herbicide- tolerant crops. Journal of Applied Ecology 40:17-31. Peters, P. 1996 Failed Magic or Social Context? Market Liberalization and the Rural Poor in Malawi. Center for lntemational Development at Harvard University. Phiri, A., G. Kanyama-Phiri, and S. Snapp. 1999. Maize and sesbania production in relay cropping at three landscape positions in Malawi. Agroforestry Systems 47:153- 162. Plante, A.F., R.T. Conant, C.E. Stewart, K. Paustian, and J. Six. 2006. Impact of Soil Texture on the Distribution of Soil Organic Matter in Physical and Chemical Fractions. Soil Sci Soc Am J 702287-296. (ed.) 2006. 18th World Congress of Soil Science, Philadelphia, Pennsylvania. Reynolds, L. 2000. Malawi: Country Pasture/Forage Resource Profiles Grassland and Pasture Crops. FAO. Richter, D.d., Jr., M. Hofinockel, M.A. Callaham, Jr., D.S. Powlson, and P. Smith. 2007. Long-Term Soil Experiments: Keys to Managing Earth's Rapidly Changing Ecosystems Soil Sci Soc Am J 71:266-279. Saxton, K.E., and W.J. Rawls. 2006. Soil Water Characteristic Estimates by Texture and Organic Matter for Hydrologic Solutions 70: 1569-1578. Schroth, G. 1998. A review of belowground interactions in agroforestry, focussing on mechanisms and management options. Agroforestry Systems 43:5-34. Sherrod, L.A., G.A. Peterson, D.G. Westfall, and LR. Ahuja. 2005. Soil Organic Carbon Pools after 12 Years in No-Till Dryland Agroecosystems. Soil Science Society of America Journal 69: 1 600-1 608. 112 Sileshi, G., F. Akinnifesi, O. Ajayi, and F. Place. 2008. Meta-analysis of maize yield response to woody and herbaceous legumes in sub-Saharan Aflica. Plant and Soil 307:1-19. Six, 1., PA. Schultz, J.D. Jastrow, and R. Merckx. 1999. Recycling of sodium polytungstate used in soil organic matter studies. Soil Biology and Biochemistry 31:1193-1196. Snapp, S. 1998. Soil Nutrient status of smallholder farms in Malawi. Communications in Soil Science and Plant Analysis 29:2571-2588. Snapp, S. 2008. Agricultural Systems: Agroecology and Rural Innovation for Development Academic Press, Elsevier. Snapp, S., G. Kanyama-Phiri, B. Kamanga, R. Gilbert, and K. Wellard. 2002a. Farmer and Researcher Partnerships in Malawi: Developing Soil Fertility Technologies for the Near-term and Far-term. Experimental Agriculture 38:411-431. Snapp, S.S., D.D. Rohrbach, F. Simtowe, and HA. Freeman. 2002b. Sustainable soil management options for Malawi: Can smallholder farmers grow more legumes? Agriculture, Ecosystems & Environment 91 :159-174. Snedecor, G.W., and W.G. Cochran. 1967. Statistical Methods Iowa State University Press, Ames, Iowa, USA. Szott, L., C. Palm, and R. Buresh. 1999. Ecosystem fertility and fallow function in the humid and subhumid tropics. Agroforestry Systems 47:163-196. Thangata, P.H., and J.R.R. Alavalapati. 2003. Agroforestry adoption in southern Malawi: the case of mixed intercropping of Gliricidia sepium and maize. Agricultural Systems 78:57-71. Tittonell, P., KD. Shepherd, B. Vanlauwe, and KE. Giller. 2008. Unravelling the effects of soil and crop management on maize productivity in smallholder agricultural systems of western Kenya--An application of classification and regression tree analysis. Agriculture, Ecosystems & Environment 123:137-150. Tripp, R. 2008. Agricultural Change and Low-Input Technology, In S. Snapp and B. Pound, eds. Agricultural Systems: Agroecology and Rural Innovation for Development. Elsevier. Vanlauwe, B., C. Gachengo, K. Shepherd, E. Barrios, G. Cadisch, and CA. Palm. 2005. Laboratory Validation of a Resource Quality-Based Conceptual Framework for Organic Matter Management Soil Sci Soc Am J 69:1 135-1145. Vanlauwe, B., S. Aman, K. Aihou, B. Tossah, V. Adebiyi, N. Sanginga, O. Lyasse, J. Diels, and R. Merckx. 1998. Alley cropping in the moist savanna of West-Aflica: 113 III. Soil organic matter flactionation and soil productivity. Agroforestry Systems 42:245-264. Vanlauwe, B., K Aihou, S. Aman, B.K Tossah, J. Diels, O. Lyasse, S. Hauser, N. Sanginga, and R. Merckx. 2000. Nitrogen and phosphorus uptake by maize as affected by particulate organic matter quality, soil characteristics, and land-use history for soils flom the West Aflican moist savanna zone 30:440-449. Waddington, S., W.D. Sakala, and M. Mulugetta. 2004. Progress in Lifting Soil Fertility in Southern Africa. Fourth lntemational Crop Science Congress. Wander, M. 2004. Soil Organic Matter Fractions and Their Relevance to Soil Function, p. 67-102, In F. Magdoff and R. R. Weil, eds. Soil Organic Matter and Sustainable Agriculture. CRC Press. Wander, M.M., S.J. Traina, B.R. Stinner, and SE. Peters. 1994. Organic and Conventional Management Effects on Biologically Active Soil Organic Matter Pools 58:1130-1139. WestcoScientific. 2007. Nitrate-Nitrite in Water, Waste Water, Soil Extracts and other aqueous samples SmartChem Method 1421-1006, Brookfield, Connecticut. Whiteside, M., and S. Carr. 1997. Services and policies needed to support sustainable smallholder agriculture in Malawi [Online]. Available by Agricultural Services Reform in Southern Africa http://www.eldis.org/fulltcxt/malawi.pdf (verified May). 114 APPENDIX 115 Evaluating the Impact of Gliricidia Sepium on Soil Quality with Smallholder Farmers in the Chinyanja Triangle Household Survey Household code number (number consecutively, 1, 2, 3,. . .through the whole study) May we ask you some questions about your household and farming practices ? 1:1 yes C] no (tick) (We are not allowed to interview people less than 18 years of age.) Farmer’s name: Village: EPA: District: Date: Interviewer’s name: General questions: How many years have farmers in this village planted gliricidia? (# years) (May be less than perfectly precise, and may also be related to the year an important public event occurred.) Have many farmers used gliricidia over these years? Ci yes 13 no Of those farmers that began with gliricidia, how many have continued to use it? [:1 almost all [1 half Cl quarter 1:1 less than a quarter (tick one box) Household questions: House status (tick all that apply, do not ask): 1:1 metal roof [:1 thatch roof 1:1 fired bricks 1:1 unfired bricks [:1 Windows/glass panes 13 Windows/wood shutters 116 Table 4.1 Household Members: (to be asked of the farmer or a knowledgeable elder). Name of Sex Age Participates in fieldwork (on any plot) househOId Pl ' h # f h d h ' h k ° h ears ease wntet e o ours per ay eac person mlg twor Int 6 member )( y field, or write: 1=M l = too young or too old 2=F 2 = works off-farm Name Cod Age Code e Household livestock: chickens goats pigs cattle (other) Land available for use by the household (May be asked while at the field being sampled) (write in number of animals) 117 Table 4.2 Field Size: (to be asked of the farmer or a knowledgeable elder). Number Size of Domi Typical crop, Typical soil Rating of field Field nant intercro , or rotation fertilit p 3’ soil on that field treatments t e on that field , ‘ .yp A = local maize variety Please rate ”1° (Please May be"“ In B = improved maize 1 = crop fields (1 = number: comparison to field variety residues best, 2 = next 1,2,3...in the size ofthe 1 = sole maize 2 ___ livestock best...) each filed that we 2 : maize/gliricidia manure according to - lntercrop 2 - - h od t' household) are sampllng. _ , , 3 fertilizer t e pr uc lve 1=katon 3 _. maize/plgeonpea 4 = gliricidia capacity of do, /gllr1c1dla lntercrop 5 = other th 'I , _ 4 = maize intercrop with e 80' 5 1n a 2—maka . agroforestry 0d ( pigeon pee or other annual 6 ___ compost m era e nde, legume 7: annual production 3=mche 5 = vegetables legume year. nga. . (7) i {023000 1‘ l' rotation or 4=mtsrl — ot er crops, p ease lst intercrop o, 8 = other, please list other names, please list Number percentages Code Code Code Rating 118 Evaluating the Impact of Gliricidia Sepium on Soil Quality with Smallholder Farmers in the Chinyanja Triangle Farm Field Survey Household code number (flom hh questionnaire) Farmer’s name Field or garden code number (number consecutively, 1,2,3,. . .for each household) Matching information flom ICRAF survey? 1:1 yes 13 no * questions that should be omitted for ‘control’ fields ** questions that should be omitted if ICRAF survey was done on this field. GPS data: (flom center of field) Longitude: Latitude: Altitude: “Measure the field sampled with the GPS, and draw a picture of the field on the back of this questionnaire. Size of field: ha Field type: 1: gliricidia intercrop + inorganic N [I gliricidia intercrop — inorganic N Clcontrol/gliricidia adopter + inorganic N 1:1 control/gliricidia adopter — inorganic N Dcontrol/gliricidia non-adopter + inorganic N 1:1control/gliricidia non-adopter — inorganic N (tick) May we take a sample of the soil and tell you more about your soil type? 1:1Yes, 1:1No (tick)May we take some of this soil to the office to analyze filrther? DYes, DNO (tick) Take soils samples the field (in ‘X’ pattern), sampling flom the maize rows. Take 3 samples of 10 cores per sample. Soil sampling is done in : 1:1 rows flom last year’s crop, 1:1 rows flom next year’s crop (tick) Rows are: [3 tops scraped to cover maize biomass, El tops not scraped (tick) 119 Plate pH: Texture: Table 4.3 Soil Characteristics (of field sampled): Name of soil Dominan Tillage Infiltration Runoff Available Fertili t soil of rainfall water ty ”my 1: “one holding t= Write the Mark an 2 = 1 = very fast 2 = small ca ac it highly Chichewa names of ‘x’ in the somewhat 2 = somewhat amount 1 13h. y fame the main soils in square of easy fast 3 = '2 _‘gh 2 = this field. the 3 = 3 = somewhat moderate ‘ moderat dominant 501““th slow amount moderately ely soil in this difficult 4 = very slow 4 = much hrgh fertile field 4 = 3 ‘ 3 = difficult moderately poorly low fertile 4 = very low 4 = exhaust ed Name mark code code code code code Which type of soil is the best (farmer’s point of view)? "This field is: exhausted (I) relatively deep and fertile (2) shallow, rocky and/or (point of view of the interviewer, do not ask, please circle) 120 Table 4.4. Soil Fertility methods (of field sampled): Years 2005/6 2004/5 2003/4 Methods 1 = crop residues 2 = livestock manure 3 = fertilizer (fill table below) 4 = gliricidia 5 = other agroforestry 6 = compost 7: annual legume rotation or intercrop 8 = other, please list Enter each response on a separate line. Advantages of the methods 1 = high maize yield 2 = fast maize growth 3 = permanent, cannot lose 4 = improves household nutrition 5 = low labor requirement 6 = suppresses weeds 7 = other, please list 8 = easy to till 9 = reduces pests May enter more than one response. Disadvantages of methods 1 = slow maize growth 2 = shortage of livestock manure 3 = extra labor 4 = high price 5 = low cash flow at planting 6 = soil depletion 7 = dries the soil 8 = intercrop competes with maize 9 = other, please list 10=builds up pests May enter more than one response, in order of importance. *Year that gliricidia was planted (please circle): 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 *Table 4.5. Gliricidia biomass cycle: (Please mark ‘i’ for years in which the tree was immature, ‘m’ for years in which the tree was mature, and ‘d’ for years in which the tree was declining due to age/disease.) 2005 20 06 1996 1997 1998 1999 2000 2001 2002 2003 2004 MOO-d 121 *Table 4.6 Gliricidia pruning: Ye Annual Pruning type Leaf Processing and Timing ar Pruning Incorporation 1 = branches cut, How many 1 = pnmed leaves separated l = Whole leaves applied over whole days passed once 2 = branches lefi on soil surface between 2 = pruned tree, leaves taken 2 =.Wholc leaves, in or under planting prunlng 8119 twice 3 = other, please ridge Incorporatlon 3 = pruned list 3 = Whole leaves, spot applied ? thrice 4 = Powdered leaves applied over . 4 = not pruned May include whole soil surface May Include 5 = other different answers 5 = Powdered leaves, in or under dlfferent please list for different planting ridge answers for prunings, separated 6 = Powdered leaves, spot applied different by commas 7 = Powdered, mixed with fertilizer, prunings, ’ and spot applied, separated by commas 200 5/6 200 4/5 200 3/4 122 Table 4.7. Was fertilizer listed in table 3? (tick) 1:1No DYes, please fill table Year Type Timing Application Amount l=compo l = basal . l = bottle-cap applied to each maize station und D, 2 = top-dress l = spot‘applled 2 = spoonful applied to each maize statlon 2=Urea 2 = applled ‘0 3 = handful applied to each maize statlon 3=C AN ridge 4 = fertilizer and manure mixed 4:23.21. May include 3 = broadcast on 5 = applied in a line along the ridge 0.+4S, both basal WhOle surface (please wrlte how many kg for this field) 5=Othcr and top- 6 l= broadcasthover wholisufrfac: f M . te owmany gortls1c ) (please dress Ma - , (p ease wrr - y Include 7 = other lease 5 ‘i 1181) different answers p p“ fy for basal and top- May include different answers for basal and dress, separated top-dress, separated by commas by commas. 2005/6 2004/5 2003/4 123 Table 4.8. Please fill the following information about the cropping system( in field sampled): Year Crop or Variety Spacing Yield Yield Incorpo- Notes intercrop name difference ration 1=25cm x In addition to 1m.; . gliricidia, if Please list 2: 90cm May What, w°°'° Wha,‘ High or resent“ variety x 90cm. includ the yleld 0f portlon 0f low [1): SOlc.maize name 3=1mx1m c this field have crop rainfall 2-maize + an d/ or the difi‘ere been wrth sole resrdues Unable _ , _ 9 pigeon pea Chichewa 4=other "t maize. were t to weed 3=maize + name. (please answe mcorpora Crop other Ie me SPCCIfI’) rs for l=qualter ed? diseases 4: tobafcuo intercr 2=half Other 5=other , May 0P8, 3:53“ l“Wane? croppin (please list) include maize 2=ha|f g issues different “Sted 3=three (please answers firSt- quarter describ for 4:8" e) intercrops, = maize 50 kg listed first. = 70 kg 90 kg 2005 /6 2004 /5 2003 /4 **Table 4.9. Please fill this table about weeding practices: Year Date of maize planting Number of Time of weeding weedings “d r May be approximate in 2005 15! 2 :2 and 2004. 2005/6 2004/5 2003/4 124 Do you have any suggestions on how the agroforestry technologies can be improved? Do you any question to ask us? Thank you very much for taking the time to give us this sample and information. Farmer report sheet: Name of farmer: Plate pH: Texture: Size of the field 125 IVERSITY LlBRARiES w I n 20 8963 "'iliiliiiiliili . . ................ ------- ------ .......... ............. ..... ..... I - I . 4 ........................... ................. ..... ....... ...........................