Gas emissions of applied 1 5~ were measured beneath a soil cover daily following saturation of Vertisol and Alfisol soils repacked in pots to the original field bulk density and held at three temperatures (5, 15 or 30°C) with or without addition of wheat straw. Collective gas emissions over 57, 43 and 15 days at 5, 15 and 30°C respectively were compared with the 1 5~ loss determined by mass balance.Loss measured by gas emissions (15N2 and 15Nz0) ranged from 36% to 152% of the denitrification loss as determined by 1 5~ mass balance. In the absence of added straw, measurement by gas emissions was consistently less than loss by 15N balance. Where straw was added, 15N loss by gas emissions was overestimated, probably because of a smaller headspace (0.3 L) than considered desirable (1-1.5 L) for emission measurements.Potential denitrification rates, in the presence of added straw, were similar for the Vertisol and Alfisol. Decreasing temperature slowed potential rates of denitrification from -2.5 kg ha-' day-' at 30°C to 0 . 8 kg ha-' day-' at 15'C and 0.4-0.5 kg ha-' day-' at 5OC. Decreasing temperature prolonged the period of waterlogging following a saturating event. Thus, collective loss of 15N was considerable even at the lower rates of denitrification at 5OC (52-76% over 57 days) or 15OC (87-92% over 43 days).Straw addition (10.5 t ha-') to the Vertisol, which contained no visible plant residues from previous crops, more than doubled the losses of applied 15N. In the absence of straw, rates of denitrification and immobilization were similar in magnitude, 0.97, 0.26 and 0.16 kg ha-' day-' for 30, 15 and 5OC respectively. Very rapid loss of applied 1 5~ in the presence of added straw led to decreases in immobilization of applied 15N, highlighting the potential effects of the much higher maximum rates for denitrification than for immobilization.The N20 emissions generally represented the smaller fraction (<25%) of denitrification emissions, becoming smaller as temperature was increased. As a proportion of emissions due to denitrification, N2O emissions were very low (<0.5% Vertisol, <3% Alfisol) in the presence of added straw.

3~ 4bL ROOT EXUDATION OF SORGHUM AND UTILIZATION OF EXUDATES BY NITROGEN-FIXING BACTERIA nummnq-Qual>l*t#\e dltTercncvr ~n ~h c v>luh!i. crud.ilr.5 at rccdl8np1 vfdllcrcl>l ,vsphiir~ tFliipliurr! ',t.,rlr,rl gclnltipo %err. dn~luc,~li,~tcd b! \rrt,tl#oc, in eroulh to1 LII,IIIII~VPIIII ~,~<~.~IC-ion~iininp mc~llil I\ol,~tc\ brncd ~r thcir ihl!~r> lo iittllic ~h c c~ud.tlc (rum i giscn penolipt \'.tr#duw1 n ,hc amount. ,>I roluhlr c.trho#i rrudcd h, .~xcocrll\ givuil wdlinp. 111 thr d8lIcrcnl penvl\per w.n .d\a cih\rncJ Houc\er, !he \.\nnl>c,tl ~n pl.~~~l-.~rr~ct~tdd .ICCI!ICIIC r~d l l~l l~l l ~<11\1!) 01 nni, (11.1111. i:,,>p8rillunl liprderatn 4481 d8d nil! mnrrl thc irnrllon is e u~d i l l~o n iof \i,li~hlc org;ln!i. i.iuh<~n liclr the irrsrt8i-n 11. rhllil) ol the ~hi,l~le Ir, uulb~c >hi. sxudrlc\ In icmz-\ol#d medlr I*~RODI CTI OI plas~hou\c, and 11, ficld erper>nienl$ (ICRISAT. ' A a u p d on dn) V Mcnnt or I8 RPIIYIIII~, 1SFI % 10 npflurh n n can drp 4, 6. and 0 Mean, ul rk i day am mrm ' Plant m&ha on day V Munr of b 10 rrpbcalxonb (S t 1 L n u r mrnlal~on rocRElrntr acttnt) (CIPI 11 '3 ml r c r-U 85 u~t h n l P < 0 0s ~l n~(~r p t 1) \I hoot r l r-093 upllhlPnx P~O O I. r,,n,, tap, 2) "i mnl r t r-0 81 "Q,h.n, P co.05 acltnt) Cxpt 21 n hwi n r =O%b w l k n ! P coo$. acbnl? (*lip1 ti n o w n r C r-04JNS. DhOI) (Up, ?I Yi OMnk r t-0 6SNs. acunl) (up, I1 n .rd rl l a 1 r-09s spnifwnl P <001

Measurement of gas emissions from denitrification of applied N has been restricted because of the lack of a convenient method. Recently a method using an electric are to measure 15N contents of dinitrogen (N2) and nitrous oxide (N2;O) in air has been developed. Gas emissions from denitrification of applied 15N were determined using this method for gas analysis of the 15N2 and 15N2O captured beneath an air-tight soil cover. Loss of 15N was calculated from gas emission measurements by two methods, accumulation of daily emissions and from the peak 15N emission value by assuming linear increase and decrease over the period of emissions. Losses estimated at low emissions with incomplete soil saturation were similar (1.9 - 5.6% 15N applied) for the two methods. Losses estimated at higher emissions with complete soil saturation were higher when calculated using peak emission values (14.8 - 28.5%) rather than accumulated daily emissions (9.5 - 18.7%). Losses estimated by emissions were compared with 15N loss estimated by mass balance at the completion of two successive soil saturations. As daily cover duration was shortened, gas emission estimates of loss more closely approximated total gaseous 15N as estimated from unaccounted for 15N in the15N balance. With shortest cover duration (15 min day-1) there was close agreement (94% estimated from peak emissions) with 15N loss estimated by 15N balance. A strategy for quantitatively estimating 15N loss by emission measurements is suggested.

Field experiments were conducted during autumn and winter (April-July) at four locations on Vertisol or Alfisol soils on the Darling Downs of Queensland in 1988 and 1989 to determine 15N losses when soil was saturated after applications of 15N labelled nitrate-N prior to sowing winter cereal crops. Losses of applied 15N were quantified by either gas emission or mass balance measurements on microplots (0.043 m2) confined laterally to a depth of 110 or 260 mm. At each field location, two experiments were established, one on a soil containing little visible crop residue where winter cereal had been harvested the previous November and another site containing residues of a recently harvested sorghum crop. Because shallow (110 mm) confinement was found to be unsatisfactory for both gas emission and mass balance measurement of 15N losses, comparison of the two methods was not applicable at one of the four field locations. Loss estimates for the six field sites by accumulating daily gas emissions averaged 80.7 � 33.4% (range 43-132%) of that estimated by mass balance. Loss estimates from peak emission measurements were generally closer to that estimated by mass balance 100.8� 39.9% (range 56-169%). Loss of applied 15N (40 kg N ha-1) when soils were saturated in April was several-fold more (19-29 kg N ha-1)) than that lost when soils were saturated in July (3.9-6.4 kg N ha-1)). Loss of 15N following saturation during July 1988 was similar in magnitude to the quantity of 15N apparently immobilized into soil organic forms (5.8-6.0 kg N ha-1)). Sorghum residues returned in March, or wheat straw added in December prior to a long period of dry weather, promoted loss of 15N applied prior to soil saturation in April or July. Alternatively, where residues of a previous winter cereal had decomposed considerably, loss of applied 15N was much lower than where sorghum residues had been added prior to saturations in April (15.3 cf. 28.6 kg N ha-1)) or July (3.9 cf. 6.4 kg N ha-1)).

No abstract