View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

provided by ICRISAT Open Access Repository

Plant Soil (2011) 343:67–81

DOI 10.1007/s11104-011-0753-7

REGULAR ARTICLE

Effects of organic and mineral fertilizer nitrogen

on greenhouse gas emissions and plant-captured carbon
under maize cropping in Zimbabwe
Farai Mapanda & Menas Wuta &
Justice Nyamangara & Robert M. Rees

Received: 23 September 2010 / Accepted: 21 February 2011 / Published online: 12 March 2011
# Springer Science+Business Media B.V. 2011

Abstract Optimizing a three-way pact comprising N2O-N hm−2, 711 to 1574 kg CO2-C hm−2 and−2.6 to
crop yields, fertility inputs and greenhouse gases may 5.8 kg CH4-C hm−2 from six treatments during season
minimize the contribution of croplands to global II with the highest fluxes. The posed hypothesis that
warming. Fluxes of N2O, CO2 and CH4 from soil composted manure-N may be better placed as a
were measured under maize (Zea mays L.) grown mitigation option against soil emissions of GHG than
using 0, 60 and 120 kg N hm-2 as NH4NO3-N and mineral fertilizer-N was largely supported by N2O
composted manure-N in three seasons on clay fluxes during the wet period of the year, but with high
(Chromic luvisol) and sandy loam (Haplic lixisol) level of uncertainty. Nitrogen addition might have
soils in Zimbabwe. The fluxes were measured using stimulated both emissions and consumption of CH4
the static chamber methodology involving gas chro- but the sink or source strength depended highly on
matography for ample air analysis. Over an average of soil water content. We concluded that the application
122 days we estimated emissions of 0.1 to 0.5 kg of mineral-N and manure input may play an important
role with reference to global warming provided the
Responsible Editor: Per Ambus.
season can support substantial crop productivity that
may reduce the amount of N2O loss per unit yield.
F. Mapanda (*) Confidence in fluxes response to agricultural man-
Department of Research and Specialist Services, Chemistry
agement is still low due to sporadic measurements
and Soil Research Institute,
P.O. Box CY550, Causeway, Harare, Zimbabwe and limited observations from the southern African
e-mail: [email protected] region.

M. Wuta
Keywords Cattle manure . Greenhouse gas . Maize .
Department of Soil Science and Agricultural Engineering,
University of Zimbabwe, Mineral fertilizer . Plant captured carbon
P.O. Box MP167, Mt. Pleasant, Harare, Zimbabwe

J. Nyamangara
Introduction
International Crops Research Institute for the Semi-Arid
Tropics, Matopos Research Station,
P.O. Box 776, Bulawayo, Zimbabwe Maize production and mineral fertilizer use in
southern Africa are likely to become more closely
R. M. Rees
linked than they have been in the immediate past
Scottish Agricultural College,
West Mains Road, (Heisey and Mwangi 1996) mainly due to expansion
Edinburgh EH9 3JG, UK of maize production in line with current national
68 Plant Soil (2011) 343:67–81

policies to reduce food imports. This linkage is likely ern Africa largely because the organic matter content
to cause a shift in the local and regional nitrogen (N) of soils from arable fields in the region is on average
and carbon (C) balance as it may affect both primary less than 2% (i.e., <1% soil organic C) (Nyamangara
productivity and soil emissions of greenhouse gases 2002). For Zimbabwe at least 70% of total arable land
(GHGs). Equally important are organic soil amend- is covered by the highly weathered sandy soils
ments such as cattle manure that are extensively used (Anderson et al. 1993), with low physical protection
in most smallholder farming systems in Africa. of soil organic C. On the other hand, solid manure
Economic reasons, compatibility with capacity, soil may be a potential mitigation strategy that can reduce
conditioning and pollution reduction are often used to GHG emission from croplands (e.g. Desjardins et al.
justify the use of organic fertilizers (e.g. Odhiambo 2005). This option can be improved by manure pre-
and Magandini 2008; Svotwa et al. 2009). However, treatment by composting (Petersen 1999), although
there have been limited observations from Africa to the composting itself may be an important source of
support that the use of these organic soil amendments GHG (Hao et al. 2001). Many of these worries have
as sources of plant-N may also reduce soil emission however not been confirmed in southern Africa.
of GHGs such as nitrous oxide (N2O) and methane The latest (2006) IPCC guidelines for national
(CH4) compared with the use of mineral-N fertilizers. GHG inventories gives a default N2O emission factor
The lack of work addressing GHG emissions on the for N additions from both mineral fertilizers and
basis of agricultural productivity rather than cultivated organic amendments as 0.01 kg N2O-N kg−1 applied
area alone has been lamented by Flynn and Smith N (de Klein et al. 2006). This factor is a revision
(2010), and such missing link may relate to Burneya et downward from the 1997 IPCC factor (1.25% of
al. (2010)’s suggestion that investment in yield im- applied N). Similarly for CH4, a baseline emission
provement compares favourably with other commonly factor of 1.3 kg CH4 hm−2 day−1 is available in the
proposed climate change mitigation strategies. The use same guidelines only as a starting point, with scaling
of manure-N to increase food security and as a climate factors of 0, 0.25, 0.28 and 1, for upland, drought
change mitigation strategy has been widely debated in prone, regular rain-fed and continuously flooded
Africa and elsewhere. It is generally accepted that water regimes, respectively. Zimbabwe and many
under the current regional farming systems’ setup the other African nations under the United Nations
use of cattle manure alone may not sustain the food Framework Convention on Climate Change have
requirements of southern Africa. The reason is partly adopted this emission factor approach for reporting
that an overwhelming majority of farmer households national GHG inventories. However, considering the
own none or small herds of cattle. Cattle manure use limited number of observations made from southern
has often resulted in relatively low maize yields in the Africa during this period, it is highly likely that there
short-term, as a result of N-immobilization in soil (e.g. were few data inputs from this region.
Ma et al. 1999; Nyamangara et al. 2003a). Such The objectives of our study were therefore to
response could imply low plant-capturing of atmo- determine the fluxes of N2O, CO2 and CH4 from soils
spheric CO2 which may be important in the short term. under maize, and to relate these fluxes to both N-
In the long term however, Lasco et al. (2006) source (composted cattle manure and mineral fertilizer)
highlighted that for annual crops increase in biomass and crop productivity. We hypothesised that composted
in a single year is assumed equal to biomass losses manure-N is better placed as a mitigation option
from harvest and mortality in that same year, thus there against soil emissions of GHG than mineral fertilizer-
may be no net accumulation of biomass C. N that has generally been implicated in the depletion of
The use of inorganic N-fertilizers has been impli- soil organic C and ultimate decrease in agronomic
cated by Khan et al. (2007) and Mulvaney et al. efficiency of applied-N. An equivalent N-rate from
(2009) in loss of soil organic C through increased composted cattle manure would therefore yield less
respiration and GHG emissions. In addition, this loss GHG emissions from croplands under maize
would also decrease soil productivity and agronomic because of the possible longer N residence time in
efficiency of applied-N, and would therefore shift the soil that normally gives manure some of its residual
N and C balance in favour of GHG emissions. Such effects on the preceding crops (e.g. Nyamangara et
experience would be far more detrimental for south- al. 2003a).
Plant Soil (2011) 343:67–81 69

Materials and methods plots were ploughed to about 0.2 m soil depth and
disced before sowing each season. Sowing positions
Study sites were marked at 0.9 m×0.45 m spacing using hoes,
and a locally common maize variety (SC513, with
The study was conducted during Cropping Seasons I 57 days to silk and 126 days to maturity (Seed-Co
(2006/2007), II (2007/2008) and III (2008/2009) at 1998)) was sowed (to two plants per position).
the University of Zimbabwe Farm (UZ-Farm) located Mineral-N fertilizer (NH4NO3, 34.5% N) was applied
15 km north of Harare (31° 00′ 48″ E; 17° 42′ 24″ S), at 0, 60 and 120 kg N hm−2, 50% at sowing and the
and the Domboshawa Training Centre (DT-Centre) remaining 50% at six weeks after sowing (next to
located 30 km north-east of Harare (31° 00′ 30″E; 17° crop) each season. Manure (60 and 120 kg N hm−2)
42′ 03″S). The UZ-Farm and DT-Centre experience was strip applied at sowing alongside each sowing
cold–dry winters and hot–wet summers (subtropical). line and buried. In addition, annual basal dressings of
The amount and distribution of rainfall received at the P (30 kg hm−2, as single super phosphate) and K
two sites during the study period are shown in Fig. 1a,b. (30 kg hm−2, as muriate of potash) were applied in
January was the wettest of all rainy months, especially sowing holes on all experimental plots before sowing
in season II (2007/2008) where close to 40% and the seed. Basal dressing concurred with sowing to
30% of total rainfall were received in January at the minimise mechanical operations, and the seed was
UZ-Farm and DT-Centre, respectively. The warmest placed at sufficient distance from fertilizers within
period of each year was October–December each hole to reduce salt stress during crop emergence.
(Fig. 1c,d). The red clay soil at the UZ-Farm is This method of application is the most common for
classified (FAO) as Chromic luvisol derived from maize in both commercial and smallholder farming
dolerite, while the brown sandy-loam soil at DT- systems where graduated fertilizer cups or some
Centre is Haplic lixisol derived from granite. Other fertilizer metering devices on traction implements
site characteristics and soil properties are given in are used for spot application in southern Africa. The
Table 1. rates applied closely related to the local recommen-
dations of 300–400 kg hm−2 of N: P: K (8.0: 7.0: 6.3)
Experimental treatments and management at sowing and 250–300 kg hm−2 NH4NO3 from four
or six weeks after sowing for a maize target yield of
Two identical experiments were conducted on crop- 6000–7000 kg hm−2 (ARI 2002). Weed control was
lands that had been fallowed for at least 10 years, one done manually using hand hoes.
at the UZ-Farm and the other at the DT-Centre. The
treatments were: Sampling
1. No fertilizer-N and no manure-N applied (control)
Gaseous emissions from soil were trapped using
2. NH4NO3−N, applied at 60 kg hm−2
open-bottom and transparent polythene chambers
3. NH4NO3−N, applied at 120 kg N hm−2
with an area of 0.40 m x 0.28 m, (0.2 m high) and a
4. Cattle manure-N, applied at 60 kg N hm−2 plus
net volume of 0.019 m3, using a method similar to
NH4NO3−N at 60 kg N hm−2 (combination)
that reported by Rees et al. (2006) and Mapanda et al.
5. Cattle manure-N, applied at 60 kg N hm−2
(2010). Each chamber was placed above the sampling
6. Cattle manure-N, applied at 120 kg N hm−2
area (located randomly) within a plot, and positioned
Experimental plots (5 m×6 m) were laid out with in a way that it protruded into the intra-row spacing
six treatments arranged in a Randomized Complete with its width covering more than half of this spacing.
Block Design with four replicates. Manure from This was partly to get as close as possible to where
pasture-fed beef cattle was collected and aerobically the fertilizer and manure were applied (without
heap composted under a shed for three months disturbing crop roots), and also to use crop canopy
(September–November). Before weighing and land to reduce excessive heating of chambers from direct
application the composted manure was analysed for sunlight. To avoid gas loss between the soil and
total-N and C content (Table 2), and water content for chamber edges a small chisel was used to fasten this
moisture correction in each season. Experimental seal with surrounding soil, taking precaution not to
70 Plant Soil (2011) 343:67–81

(a)UZ-Farm
120 P; F1 F2 P; F1 P; F1 F2
W H
105 Event symbol L H H
W L F2 L W
90
75 sum= 872 mm
60
sum = 596 mm sum = 997 mm
45
30
Rainfall (mm day -1)

15
0
120
(b) DT-Centre P; F1
P; F1 F2 H P; F1 F2
H W
105 L H
W L F2 L W
90 L = Ploughing
75 P = Planting
sum = 802 mm
F1=1st fertilization
60
F2=2nd N-fertilization sum = 619 mm sum = 1154 mm
45
W=Weeding
30 H =Harvesting
15 Gas sampling
0
(c) UZ-Farm maximum
35
minimum
30
25
20
15
10
Air temperature (°C)

5
0
35
(d) DT-Centre
30
25
20
15
10
5
0
-5
.

Month: J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M
Year: 2006 2007 2008 2009

Fig. 1 Daily rainfall received and some key events timelines at the UZ-Farm (a) and DT-Centre (b), and the daily maximum and
minimum air temperatures at the respective sites (c, d) in 36 months of the study starting June (J) 2006
Plant Soil (2011) 343:67–81 71

Table 1 General character-

istics of the study sites, Characteristic UZ-Farm DT-Centre
including climate data cov-
ering the period 2005 – Altitude, m 1505 1540
2010 and soil properties Mean annual rainfall, mm yr-1 748±176 824±226
(mean±standard deviation) Mean maximum temperature, °C 26.1±0.7 27.2±0.7
in the 0 – 0.1 m depth at the
UZ-Farm and DT-Centre in Mean minimum temperature, °C 12.4±0.4 10.4±1.9
Zimbabwe Soil type, FAO Chromic luvisol Haplic lixisol
Soil parent material Dolerite Granite
Slope,% 2-3 2-3
Soil pH 6.5±0.1 5.4±0.2
Bulk density, g cm-3 1.27±0.03 1.78±0.03
Soil organic C,% 1.63±0.23 0.76±0.10
Cation exchange capacity, cmol(+) kg-1 9.5±1.5 5.2±1.4
Clay content,% 52±1.4 15±1.4

disturb the surface inside the chamber perimeter. The (Rochette and Eriksen-Hamel 2008), causing non-
emitted gas in each chamber was collected into pre- linearity and systematic underestimation in the fluxes
evacuated 20 cm3 glass vials using a 50 ml-graduated (Kroon et al. 2008). Gas sampling from the plots of
syringe, once immediately after securing the chamber different treatments was done almost at the same time,
on the sampling area and once after one-hour of with a 2 to 3 minutes lag in sampling from one chamber
trapping the gas. In addition, a separate linearity to the next. This time limit however could not allow for
test was latter carried out once by similarly chamber replication within each plot, and partially
collecting the gas from three chambers at 10 addressed the effect of potential diurnal variations in
minutes intervals for 60 minutes and plotting the the fluxes, especially for CO2 which would then be
fluxes against time. negligible (Tang et al. 2005; Yang et al. 2010).
Gas sampling was carried out at intervals shown in Soil samples were collected, initially and following
Fig. 1a,b, in order to compare the relative treatment each gas sampling, using a bucket auger at 0–0.15 m
effects on GHG fluxes during the rainy season only depth, and at two places within an area where a gas
since high moisture content during this period would be chamber was placed in each plot. Soil temperature
expected to trigger more considerable GHG fluxes than was measured in-situ at three randomly selected
during the dry period (Rees et al. 2006; Dick et al. positions within each plot using digital soil thermom-
2008). It was assumed that GHG fluxes during the dry eters with 0.1 m long stainless steel probes, while air
seasons would be very low and comparable across temperature was measured at a height of approximately
treatments due to a possible low microbial activity in 1 m above ground. Both soil temperature and air
dry soils and insignificant sources of CH4 for uptake by temperature were measured at the first and second gas
soil microorganisms. In addition, sampling during the collection time from each plot. Relative to temperatures
dry period using the employed chamber insertion inside the chambers (sporadically measured soon after
method would increase chances of gas leakage or the second gas collection) air temperatures were lower
contamination by lateral diffusion beneath the base by between 0.8–11.7°C (mean: 7.9°C). The samples

Table 2 Total C and N, and

the C/N ratio of the N- N-source Total C Total N C/N
sources used in the experi- g kg-1 g kg-1
mental plots at the two
study sites during the three Cattle manure (Season I: 2006/2007) 326 16 20
cropping seasons Cattle manure (Season II: 2007/2008) 312 12 26
Cattle manure (Season III: 2008/2009) 280 13 22
NH4NO3 (Seasons I, II, and III) – 345 –
72 Plant Soil (2011) 343:67–81

were analysed for water content (SWC), mineral-N indicator (bromocresol-methyl red) solution. The
(NH4+-N and NO3−-N), pH and organic C. distillate (50 ml) was titrated with 0.005 M H2SO4
The maize crop was harvested at 22–24 weeks in a micro-burette. Nitrate-N was determined in the
after sowing in each season, by cutting all above same sample by adding Devarda’s alloy to reduce
ground crop material within a net plot area of 3 m x NO3--N to NH4+-N and distilling again into fresh
3 m. Harvested materials were subdivided into grain, boric acid, followed by titration with 0.005 M H2SO4.
shelled-cob and maize residue, and weighed for The SWC was determined gravimetrically as soil
aboveground biomass. Sub-samples from each subdi- weight loss on oven drying at 105°C till constant
vision were collected for moisture correction and weight. Soil pH was measured in a distilled water
estimation of total plant N and C. suspension with a pH meter (model: Corning 215)
after an hour of mechanical shaking, while total
Analysis of samples organic C was extracted by wet oxidation using
concentrated H2SO4 and K2Cr2O7, and determined
Nitrous oxide, CO2 and CH4 were quantified by Gas by titrating with Fe(NH4)2(SO4)2·6H2O solution.
Chromatography (GC model: Hewlett Packard 5890, Plant moisture and C analyses were carried out
Series II, Avondale, PA, USA) at Scottish Agricultural using the methods also described by Okalebo et al.
College (SAC) in the UK. Nitrous oxide was (1993). Dry matter was determined by weight
determined using an Electron Capture Detector main- difference upon oven drying at 70°C till constant
tained at 380°C; CO2 was analysed using a Thermal weight. Total N content was determined using the
Conductivity Detector, and CH4 using a Flame semi-micro Kjeldahl method (Bremmer and Mulva-
Ionisation Detector. The least detected level values ney 1982), while plant C was estimated from organic
for the specific gases were 0.24 ppm for N2O, matter (loss in weight after slow ignition in a muffle
350 ppm CO2 and 0.84 ppm CH4. The fluxes of furnace at 650°C over 12 hours). A conversion factor
N2O, CH4 and CO2 were calculated as the differences of 1.724 was used to convert organic matter to
in concentration between sampling time zero and organic C based on the assumption that organic
sampling after one hour. However, a four point matter contains 58% organic C (i.e., g organic
sampling in time would be more suitable than this matter/l.724=g organic C) (Okalebo et al. 1993;
two-point approach in future studies in order to Nelson and Sommers 1996). Three replicates were
reduce uncertainties in the estimated fluxes. Atmo- made by sub-sampling each sample in the laboratory.
spheric pressure and temperature during the time of Before ashing the samples were pre-heated in an oven
sampling were taken into account in the calculations. at 70°C to correct for any absorbed moisture.
In order to check the integrity of sample containers
being used in this study, batches of standard samples Data analysis
were sent from the UK to Zimbabwe and returned
(within 10 weeks) with field samples for analysis. The The linearity test conducted using samples collected
concentrations of 16 vials containing CO2, CH4 and at 10 minutes intervals for 60 minutes gave some
N2O were consistently between 92–100% of the coefficients of regression (r2 values) of between 0.88
original value, with a mean of 98%. and 0.97 (average: 0.92), thus on average about 8% of
Soil water content (SWC), NH4+-N and NO3−-N error originating from deviations of linearity. A total
analyses were carried out immediately from fresh measurement uncertainty of at least 20% was esti-
samples, while pH and organic C analyses were done mated from the study (chamber height: 5%; time:
after air drying of samples and sieving through a temperature: 3%; linearity: 8%; transport: 2%; GC
2 mm and a 0.5 mm sieve, respectively. Soil analyses calibration and measurements: 2%) based on theoretical
were done using the methods described by Okalebo et confidence estimate (e.g. Rochette and Eriksen-Hamel
al. (1993). Mineral-N was extracted by mechanically 2008; Venterea 2010). Homogeneity of variance and
shaking 5 g of a soil sample with 50 ml of 2 M KCl normality tests were carried out on gas, soil and plant
for one hour, and filtering. Ammonium-N was data using the Levene’s and Kolmogorov-Smirnov’s
determined after steam distillation of the extract in Tests, respectively, at 5% level. Two-way analysis of
MgO that is, trapping the NH4+−N in boric acid plus variance was carried out to establish any significant
Plant Soil (2011) 343:67–81 73

UZ-Farm (clay) DT-Centre (sandy loam)

50 50
45 (a) No applied-N (control) (b)
-2 45
40 NH4NO3-N, 60 kg hm
µg N 2O-N m-2 hour -1

40
35 NH4NO3-N, 120 kg hm-2
35
30 Combined, 60 + 60 kg N hm-2 30
25 Manure-N, 60 kg hm-2 25
20 Manure-N, 120 kg hm-2 20
15 15
10 10
5 5
0 0
-5 -5

90 (c) 90 (d)
mg CO 2-C m-2 hour -1

80 80
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
150 150
120 (e) 120 (f)
µg CH 4-C m-2 hour -1

90 90
60 60
30 30
0 0
-30 -30
-60 -60
-90 -90
multiply multiply
-120 -120
µg CH4-C x 10 µg CH4-C x 10
-150 -150 (season 2 only)
(season 2 only)
-180 -180
.

.
Jan FebDec Jan Feb Mar Apr Nov Dec Jan FebMar Jan Feb Dec Jan Feb Mar Apr Nov Dec Jan FebMar
SeasonI Season II Season III SeasonI Season II Season III

Fig. 2 Soil emissions (and consumption) of N2O (a, b), CO2 (c, d) and CH4 (e, f) from the experimental plots of six treatments at the
UZ-Farm and DT-Centre in 3 cropping seasons

treatment and season effects (P<0.05). However, performed using the Pearson correlation coefficients,
most of the data did not meet all assumptions of the while regression analysis was conducted to measure
Fisher-founded Analysis of Variance even after the relative importance of each soil factor on GHG
transformation, and in such cases the Kruskal- fluxes. Fluxes of CO2 and CH4 were not considered
Wallis one-way analysis of variance by ranks for the ratios of yield to fluxes useful to judge
(Genstat 2003) was used to establish any significant agriculture management impact with reference to
treatment responses (P<0.05). The Mood’s Median global warming because soil respiration is not
Test was also used in significance tests on CO2 and included in the national GHG inventory but primarily
CH4 fluxes data because it was more robust against an indication of soil biological activity, while CH4
outliers than the Kruskal-Wallis Test. contained very high standard errors. Genstat 7.2
A pair-wise separation of significantly different (Discovery Edition, Lawes Agricultural Trust UK)
treatment means was done using the Mann-Witney and SPSS 8.0 (SPSS Inc., USA) statistical packages
test. Bivariate correlation analysis (two-tailed) was were used in the statistical analysis of data.
74 Plant Soil (2011) 343:67–81

Results UZ-Farm and DT-Centre, respectively, in Season II.

These translated to 0.5% and 0.3% of applied N,
Soil N2O fluxes respectively, during this wet period representing about
a third of the year, but with a high level of
The pattern of N2O fluxes in the few measurements uncertainty. Sole manure-N contributed relatively
taken was largely similar for all treatments with a few fewer N2O emissions (0.1% of applied N from
exceptions of those treatments that received NH4NO3- 120 kg N hm−2 at the UZ-Farm and <0% for the
N that had up to 32 μg N2O-N m−2 hour−1 at the UZ- remaining sole manure treatments). The results also
Farm, and up to 30 μg N2O-N m−2 hour−1 at the DT- showed with low confidence that the treatment with a
Centre (Fig. 2a,b). Spatial variability in N2O fluxes was combination of N-sources lost 0.04% and 0.01% of
considerably high but much of the highest N2O fluxes applied N as N2O-N at the UZ-Farm and DT-Centre,
particularly on plots that received NH4NO3-N were respectively, during the wet period of Season II while
noted between December and January (rainfall peak the 60 kg manure-N hm−2 treatment contributed even
period, Fig. 1) of Season II at the UZ-Farm. A less N2O emissions than the control plots that
considerable decrease in N2O emissions (P <0.05) was received only 30 kg P hm−2 and 30 kg K hm−2
noted in Season III, in which no emissions exceeded annually at both sites (Table 3). In Season III
10 μg N2O-N m−2 hour−1 in all measurements,. however, the same control plots and 120 kg manure-
The total relative emissions of N2O from the six N hm−2 gave relatively higher N2O emission inven-
treatments, coarsely estimated over a period of 122 tories than the rest of the treatments.
and 116 days during Seasons II and III, respectively
are shown in Table 3. No linear relationship could be Soil emissions of CO2
established between N-rate and N2O emissions but
the effect of N-source was apparent with NH4NO3 Soil respiration (mean ranges: 1.8–64.6 mg m−2 hour−1
contributing emission increments of up to 0.31 and CO2-C at UZ-Farm, and 0.1–90.3 mg m−2 hour−1 CO2-
0.19 kg N2O-N per season from 60 kg N hm−2 at the C at the DT-Centre), like N2O showed some consid-

Table 3 Total soil emissions or consumption of N2O, CO2 and measurements in the second (122 days) and third (116 days)
CH4 from six treatments at two sites, estimated as area under cropping seasons, respectively
the curve of the mean (and median) fluxes from the 5 and 6

Treatment University of Zimbabwe Farm Domboshawa Training Centre

N2O-N CO2-C CH4-C N2O-N CO2-C CH4-C

g hm-2 kg hm-2 kg hm-2 g hm-2 kg hm-2 kg hm-2

Season II: 2007-2008

0 kg N hm-2 (control) 103 (108) 711 (520) 0.15 (1.28) 324 (308) 774 (786) 5.84 (8.12)
-2
60 kg hm (NH4NO3-N) 408 (383) 750 (908) -0.41 (−1.58) 515 (383) 1155(1142) -1.70 (−0.15)
120 kg hm-2 (NH4NO3-N) 289 (266) 1241(1347) -0.65 (−3.50) 408 (361) 1574(1408) -5.53 (−2.50)
60+60 kg hm-2 (combined) 157 (121) 742 (771) -2.51 (−2.16) 335 (306) 967 (936) -2.61 (−2.48)
60 kg hm-2 (manure-N) 59 (45) 896 (903) -2.58 (−4.35) 257 (243) 928 (887) 0.16 (−1.17)
120 kg hm-2 (manure-N) 262 (135) 1219(1176) 1.08 (0.66) 267 (246) 1191(1015) -1.32 (−2.72)
Season III: 2008-2009
0 kg N hm-2 (control) 63 (93) 384 (375) 0.07 (−0.16) 50 (39) 285 (297) -0.18 (0.05)
60 kg hm-2 (NH4NO3-N) 22 (16) 419 (412) -0.02 (−0.14) 32 (10) 162 (133) -0.87 (−0.55)
120 kg hm-2 (NH4NO3-N) 18 (17) 452 (475) -0.72 (−0.34) 31 (26) 234 (167) -0.33 (−0.53)
60+60 kg hm-2 (combined) 37 (37) 639 (641) 0.02 (−0.13) 32 (31) 359 (345) -0.09 (−0.21)
-2
60 kg hm (manure-N) 34 (38) 621 (611) -0.40 (−0.04) 45 (34) 340 (339) -0.57 (−0.40)
120 kg hm-2 (manure-N) 67 (26) 601 (647) -0.81 (−1.17) 58 (62) 410 (409) -0.44 (−0.11)
Plant Soil (2011) 343:67–81 75

erable spatial variability and largely similar patterns −0.3 to−773 μg CH4-C m−2 hour−1) at the DT-Centre
across the treatments (Fig. 2c,d). The median CO2 (Fig. 2e,f). Methane fluxes showed high spatial
emissions however responded to N-rate (P<0.05) from variability and varied considerably across seasons
both N-sources. Soil respiration showed a positive linear (P <0.05), with Season II having the highest fluxes.
relationships (P<0.05) with NH4NO3-N rate (r2, 0.41 A considerable decrease in CH4 fluxes was noted from
and 0.74 at the UZ-Farm and DT-Centre, respectively) all treatments in February at both sites (emissions: 1.9–
and with manure-N rate (r2, 0.46 and 0.66 at the 8.7 μg m−2 hour−1; consumptions:−0.1 to−8.3 μg m−2
UZ-Farm and DT-Centre, respectively) in Season II, and hour−1 CH4-C), following a major weeding phase and
this was agreeable from at least 73% of this season’s a mid-season rainfall break (Fig. 1a,b). Methane fluxes,
data. In Season III the highest CO2 emissions were from unlike other GHG, responded bilaterally to N-rate, but
plots that received manure-N, while plots with NH4NO3 with a high level of uncertainty. When CH4 emissions
had lower soil respiration for the greater part of the and consumption were analysed separately, the magni-
beginning of Season III (Fig. 2c,d) and overall at the tude of each increased with N-rate (P <0.05). At both
DT-Centre (Table 3). sites there were positive linear relationships between
CH4 consumption and NH4NO3-N rate in Seasons II
Soil emissions and consumption of CH4 and III (r2, 0.15–0.53), and also between CH4
consumption and manure-N rate (r2, 0.12–0.47). The
The average CH4 emissions (and consumption) ranged relationships were more distinct at the UZ-Farm than at
from 1.7–1130 μg CH4-C m−2 hour−1 (consumption: DT-Centre. Similarly, CH4 emissions (medians),
−0.1 to−565 μg CH4-C m−2 hour−1) at the UZ-Farm, increased with NH4NO3-N rate at the UZ-Farm (r2,
and 0.5–1197 μg CH4-C m−2 hour−1 (consumption: 0.70) in Season II, but the relationship was poor in

UZ-Farm (clay) DT-Centre (sandy loam)

30 30
26 (a) 26 (b) No applied-N (control)
NH4NO3-N, 60 kg hm-2
22 22
SWC (%)

NH4NO3-N, 120 kg hm-2

18 18 Combined, 60 + 60 kg N hm-2
Manure-N, 60 kg hm-2
14 14
Manure-N, 120 kg hm-2
10 10
6 6
2 2

32 32
Soil temperature (°C)

30 30
28 28
26 26
24
(c) (d)
24
22 22
20 20
18 18
.

Jan Feb Dec Jan Feb Mar Apr Nov Dec Jan Feb Mar Jan Feb Dec Jan Feb Mar Apr Nov Dec Jan Feb Mar
Season I Season II Season III Season I Season II Season III

Fig. 3 Soil water content (SWC) (a, b) and soil temperature (c, d) from the experimental plots of six treatments at the UZ-Farm and
DT-Centre in 3 cropping seasons
76 Plant Soil (2011) 343:67–81

Table 4 Means (and ranges) of NH4+-N, NO3--N, pH and organic carbon (OC) of soils from experimental plots of different treatments
measured during the time of gas sampling in Seasons II and III at the UZ-Farm and DT-Centre

Treatment University of Zimbabwe Farm (UZ-Farm) Domboshawa Training Centre (DT-Centre)

NH4+-N NO3--N pH OC/% NH4+-N NO3--N pH OC

mg kg-1 mg kg-1 % mg kg-1 mg kg-1 %

Season II (2007/2008)
No N applied 8.1 (2.3-17) 2.3 (1.0-4.1) 5.6 (5.2-6.3) 1.5 (1.2-1.9) 8.5 (1.1-16) 3.2 (0.2-13) 5.0 (4.8-5.2) 0.6 (0.3-0.8)
(control)
NH4NO3-N, 8.6 (4.2-16) 3.6 (1.0-9.5) 5.7 (5.5-5.9) 1.6 (1.1-1.9) 11 (0.8-19) 3.5 (0.5-14) 4.8 (4.4-5.5) 0.6 (0.3-0.7)
60 kg hm-2
NH4NO3-N, 8.4 (2.2-13) 2.8 (1.1-5.1) 5.7 (5.3-6.0) 1.4 (1.0-1.9) 7.5 (0.9-15) 4.3 (0.5-16) 4.8 (4.4-5.4) 0.6 (0.5-0.7)
120 kg hm-2
Combined, 6.8 (2.5-15) 3.5 (1.8-7.6) 5.7 (5.1-6.1) 1.4 (1.0-1.8) 11 (0.9-18) 2.8 (0.3-11) 4.8 (4.5-5.1) 0.5 (0.3-0.7)
120 kg hm-2
Manure-N, 8.1 (2.4-21) 2.6 (1.5-4.3) 5.7 (5.3-6.1) 1.5 (1.2-1.9) 6.2 (0.1-17) 2.3 (0.1-7.6) 5.0 (4.8-5.6) 0.6 (0.4-0.8)
60 kg hm-2
Manure-N, 12 (3.9-24) 4.5 (1.1-9.7) 5.7 (5.2-6.4) 1.5 (1.1-2.3) 6.6 (0.7-16) 4.0 (0.6-12) 4.9 (4.6-5.3) 0.6 (0.4-0.7)
120 kg hm-2
Season III (2008/2009)
No N applied 4.4 (0.1-13) 4.1 (1.8-8.4) 5.5 (5.3-5.8) 1.4 (1.1-1.7) 9.5 (2.5-25) 1.5 (0.1-3.4) 4.8 (4.5-5.1) 0.5 (0.3-0.5)
(control)
NH4NO3-N, 7.2 (3.1-12) 7.1 (0.1-14) 5.5 (5.2-5.8) 1.6 (1.4-1.8) 6.8 (3.3-10) 3.1 (0.1-6.6) 4.8 (4.4-5.3) 0.5 (0.4-0.7)
60 kg hm-2
NH4NO3-N, 10 (1.1-24) 5.7 (1.4-11) 5.6 (5.3-5.9) 1.5 (1.3-1.7) 8.5 (5.8-13) 2.5 (0.1-7.6) 4.6 (4.4-5.1) 0.6 (0.4-0.7)
120 kg hm-2
Combined, 4.8 (1.0-10) 4.7 (1.8-7.7) 5.6 (5.5-5.8) 1.5 (1.2-1.9) 6.7 (1.8-12) 3.7 (0.1-8.8) 4.9 (4.8-5.2) 0.6 (0.4-0.9)
120 kg hm-2
Manure-N, 6.6 (0.1-27) 5.6 (2.1-9.3) 5.7 (5.6-5.8) 1.5 (1.3-1.8) 5.0 (3.4-6.7) 2.7 (0.1-6.5) 4.9 (4.6-5.2) 0.6 (0.3-1.0)
60 kg hm-2
Manure-N, 6.4 (0.1-14) 8.2 (3.9-11) 5.6 (5.3-5.9) 1.7 (1.5-2.0) 7.6 (2.5-12) 3.3 (0.1-8.3) 4.9 (4.5-5.2) 0.6 (0.4-0.8)
120 kg hm-2

Season III at the same site and in both seasons at the than to applied treatments (P>0.05) (Fig. 3a,b). The
DT-Centre (r2, <0.2). SWC ranged from 6.9–24.2% at the UZ-Farm and 3.3–
The relative net CH4 fluxes from the six treat- 13.3% at the DT-Centre, while soil temperature in the
ments, coarsely estimated over a period of 122 and 0–0.1 m depth ranged from 19.3–32.8°C at the UZ-
116 days during Seasons II and III, respectively, are Farm and 23.0–30.7°C at the DT-Centre. The average
also given in Table 3. Without disregarding the and ranges of soil mineral N (NH4+-N and NO3--N),
limited confidence as a result of the high uncertainty pH and organic C from the six treatments during
the application of NH4NO3-N, manure-N and combi- Seasons II and III are given in Table 4. Emissions of
nation of these could have largely increased the N2O, CO2 and CH4 were significantly correlated with
capacity of the soils at both sites to act as CH4 sinks SWC, soil temperature and pH (P<0.05), but not with
in both Seasons II and III, with the exception of the other measured soil factors. The regression analysis
120 kg manure-N hm−2 in Season II at UZ-Farm. revealed a few distinct relationships between GHG
fluxes and SWC or soil pH at the DT-Centre (Fig. 4).
Soil factors-GHG fluxes relationships Consumption of CH4 decreased with increasing SWC
above 9% which is the general water holding capacity
The variations in soil water content (SWC) and soil of this soil (Nyamangara et al. 2003b; but see Vogel et
temperature during the seasons were more closely al. 1994), while CH4 emissions increased considerably
related to seasonal rainfall and temperature patterns towards and above this capacity.
Plant Soil (2011) 343:67–81 77

Crop productivity-GHG fluxes relationships further increase in yield per unit loss of N2O-N. This
was generally consistent at both sites and in both
The measured crop biomass (grain and aboveground Seasons II and III.
residues), total-N uptake and captured C in maize
from six treatments over three seasons at the two sites
are shown Fig. 5. In all seasons the productivity of Discussion
maize significantly increased (P<0.01) with increas-
ing NH4NO3-N rate at both sites, but was less Despite the sporadic measurements presented in this
responsive to manure-N rate above the 60 kg N paper, the study was among the few reported in the
hm−2. Season II had the least yields, particularly at the southern Africa region (e.g. Chikowo et al. 2004;
DT-Centre where much of the rains were concentrated Rees et al. 2006) and more so on fertilizer input
in January (Fig. 1b), while the highest yields were management since much of the available information
found in Season III at both sites. This was one major has largely been descriptive or only model-based.
contrast with GHG emissions in which the highest Much of the attention was therefore focused on
fluxes were found in Season II. Total N removed as relative rather than absolute GHG responses to
grain ranged from 18–71 kg hm−2 at the UZ-Farm and treatment and budgets. The posed hypothesis that
6–66 kg hm-2 at the DT-Centre, and constituted about composted manure-N may be better placed as a
66–87% and 43–83% of total N uptake at the UZ- mitigation option against soil emissions of GHG than
Farm and DT-Centre, respectively. The amount of C mineral fertilizer-N was largely supported with high
partitioned to grain (825–3190 kg hm-2 at UZ-Farm uncertainty during Season II, particularly for N2O
and 272–2651 kg hm-2 at DT-Centre), stover (1340– fluxes during the wet period of the year (Table 3).
3762 kg hm-2 at UZ-Farm and 359–2703 kg hm-2 at However, for a rational compromise between N2O
DT-Centre) and shelled-cob (186–562 kg hm-2 at UZ- emissions and crop productivity NH4NO3-N was
Farm and 63–392 kg hm-2 at DT-Centre), also relatively more promising (Table 5) on condition that
increased with NH4NO3-N rate. the extremes of rainfall intensity and poor rainfall
Linking maize productivity (Fig. 5) with N2O distribution are not encountered, as was noted during
emissions (Table 3) to give some estimated ratios of Season III.
grain yield, total grain N and total grain C to N2O The soil emissions of N2O reported in this study
fluxes revealed with a high level of uncertainty that were generally low and consistent with other GHG
the application of NH4NO3 largely resulted in flux studies in Africa involving continuous cereal or
relatively lower N2O emissions per unit of produce cereal/legume rotations on regular croplands and in
compared with the control treatment (Table 5). some agroforestry systems (e.g. Chikowo et al. 2004;
Similarly, manure-N application had the same effect Baggs et al. 2006; Dick et al. 2008; Halvorson et al.
at 60 kg N hm-2, but doubling this rate did not show a 2008). It is possible that the sampling design used

800 90 90
emissions
600 (a) (b) (c)
Sole NH4NO3-N Sole manure-N
80 80
y = 10.2x 2 - 137x + 454
400
CO 2 -C (mg m-2 hour-1 )

r 2 = 0.52 70 70
CO 2 -C (mg m-2 hour-1)

y = 13.0x 2 - 120.3x + 286.6

CH 4 -C (µg m-2 hour-1 )

200 r 2 = 0.46 60 y = -9.4x 2 + 115.7x - 312.0

60
0 r 2 = 0.34
50 50
-200
40 40
-400 consumption
30 30
-600 2
y = -14.5x + 291x - 1486
-800 2
r = 0.45 20 20
-1000 10 10
-1200 0 0
0 3 6 9 12 15 18 21 2 3 4 5 6 7 2 3 4 5 6 7
SWC (%) soil pH Soil pH

Fig. 4 Relationships between: CH4 emissions/consumption and SWC (a), CO2 emissions and pH of soils that received NH4NO3-N
(b), and CO2 emissions and pH of soils that received manure-N (c) at the DT-Centre
78 Plant Soil (2011) 343:67–81

was not particularly good at fully identifying hotspots broadcasting under continuous cereals applied urea
of N2O production associated with the spot applica- (50 kg hm-2, 46% N) gave up to 1.5 kg N2O-N hm-2
tion of fertilizer used, and long path-length methods year-1, while added manure gave up to 0.9 kg N2O-N
(not available in this study) may be more suitable in hm-2 year-1 (Dick et al. 2008). Apparently, this was
such a design. However, in terms of area coverage one of the very limited studies in Africa that looked
(assuming each hotspot cover a 0.05–0.08 m radius into both wet and dry seasons and estimated annual
beyond the reach of a chamber, maize stem position N2O emissions of 1.54 kg N hm-2 from mineral
included), the missed area represented only 2–5% of a fertilizer application and 0.97 kg N hm-2 from animal
hm2 (calculated from the total sowing stations in a manure. However, similar urea broadcasting at 60 kg
hm2). In a study in Mali where N was applied by hm-2 by Kimetu et al. (2006) only yielded up to

------------------------ UZ-FARM (clay soil) ------------------------ ------------------- DT-Centre (sandy loam soil) -------------------
16000
(a) (b)
14000

12000
Dry weight (kg hm -2)

= 2 x SEM (for totals)

10000

8000

6000

4000

2000

(c) (d)
100
N uptake (kg N hm -2)

80

60

40

20

(e) (f) T1: No applied-N (control)

8000
T2: NH4NO3-N, 60 kg N hm-2
plant captured C (kg C hm -2)

7000 T3: NH4NO3-N, 120 kg N hm-2

6000 T4: Combined, 60 + 60 kg N hm-2
5000 T5: Manure-N, 60 kg N hm-2
-2
4000 T6: Manure-N, 120 kg N hm
Shelled-cob
3000
Stover
2000 Grain
1000
0
T1
T2
T3
T4
T5
T6

T1
T2
T3
T4
T5
T6

T1
T2
T3
T4
T5
T6

T1
T2
T3
T4
T5
T6

T1
T2
T3
T4
T5
T6

T1
T2
T3
T4
T5
T6

Season I Season II Season III Season I Season II Season III

Fig. 5 Partitioned aboveground plant biomass (a, b), N-uptake (c, d) and C captured (e, f) by maize grown on experimental plots
under six treatments at the UZ-Farm and DT-Centre
Plant Soil (2011) 343:67–81 79

2.9 μg N2O-N and an annual total of 0.1 kg N2O-N CH4 consumption. In our study there was no signifi-
hm-2 that is less than some of the N2O inventories cant correlation between NH4+-N and CH4 fluxes. A
from our study. There could therefore be a problem of review by Bodelier and Laanbroek (2004) however,
heterogeneity in N inputs for area based measure- showed that N can stimulate both emissions and
ments and in some parts of Europe some grazed consumption of CH4 in soil, and indicated that
pastures can have cattle urine patches that add mineral-N seems to be a prerequisite for the occurrence
1000 kg N hm-2, and yet many N2O measurements of CH4 consumption and might initiate and stimulate
studies are carried out using randomly located the enzymatic machinery in a yet unknown way.
chambers (Rees RM 2010, personal communication).
Methane fluxes under the six treatments were the
most inconsistent as N input stimulated both emissions Conclusion
and consumption of CH4 with Seasons II and III at both
sites. The overall net fluxes (Table 3) would therefore The level of uncertainty with reference to the
reflect the relative net influence that also strongly response of GHGs to mineral-N and manure-N was
depended on SWC variability within each season very high due to sporadic measurements and the
(Fig. 4a). The increase in CH4 emissions by soil with limited number of observations from the region in
addition of water and N were also noted by Mosier et general. Without disregarding this limited confidence
al. (2003), who attributed this to increased soil the posed hypothesis that composted manure-N may
respiration, conditions under which a microbial popu- be better placed as a mitigation option against soil
lation developed that tended to increase the utilization emissions of GHG than mineral fertilizer-N was
of atmospheric CH4. However, Vanitchung et al. largely supported by N2O fluxes during the wet
(2005), who reported CH4 emissions of up to period of the year. It may be possible to reduce the
61.5 mg CH4-C m-2 day-1 from N-fertilized upland overall contribution of NH4NO3-N to N2O emissions
soils under maize, argued that NH4+-N inhibits CH4 per unit amount of productivity by increasing the
consumption and high SWC can limit O2 and CH4 fertilizer rate to 120 kg N hm-2, the local general
diffusion between atmosphere and soil, thus, reducing recommendation for N fertilizer for maize, provided

Table 5 Estimated ratios of grain yield, total grain-N and total grain-C to soil emissions of N2O over an average period of 122 and
116 days in Seasons II and III, respectively at two sites

Treatment UZ-Farm DT-Centre

Yield Total-N Total-C Yield Total-N Total-C

kg g-1 N2O-N

Season II (2007/2008)
0 kg N hm-2 (control) 14 0.17 8.0 1 0.02 0.8
-2
NH4NO3-N, 60 kg hm 7 0.08 3.7 2 0.02 0.9
NH4NO3-N, 120 kg hm-2 15 0.20 8.6 4 0.06 2.1
Combined, 120 kg hm-2 29 0.35 17 5 0.05 2.6
Manure-N, 60 kg hm-2 43 0.50 25 4 0.05 2.4
Manure-N, 120 kg hm-2 12 0.16 6.8 4 0.05 2.2
Season III (2008/2009)
0 kg N hm-2 (control) 42 0.47 25 31 0.40 18
NH4NO3-N, 60 kg hm-2 188 2.36 108 100 1.32 58
NH4NO3-N, 120 kg hm-2 290 3.93 164 148 2.13 86
Combined, 120 kg hm-2 149 1.93 86 118 1.61 68
Manure-N, 60 kg hm-2 116 1.35 67 70 0.88 40
Manure-N, 120 kg hm-2 71 0.86 41 55 0.68 32
80 Plant Soil (2011) 343:67–81

the season can support substantial crop productivity. org/ifacontent/download/35735/516582/version/1/file/

2010_ifa_greenhouse_gas.pdf. Accessed 23 August 2010
In future studies it will be necessary to increase
Genstat (2003) GenStat for Windows. Release 4.23DE Discov-
temporal resolution, number of gas samples per flux ery Edition. VSN International Ltd Hemel Hempstead, UK
and spatial representativeness of GHG measurements. Halvorson AD, Grosso SJD, Reule CA (2008) Nitrogen, tillage,
and crop rotation effects on nitrous oxide emissions from
Acknowledgements This study was co-sponsored through irrigated cropping systems. J Environ Qual 37:1337–1344
grants from the European Union (NitroEurope Project No. Hao X, Chang C, Larney FJ, Travis GR (2001) Greenhouse gas
017841) and the International Foundation for Science (Grant emissions during cattle feedlot manure composting. J
No. C/4300-1) for which the authors are extremely grateful. Environ Qual 30:376–386
Heisey PW, Mwangi W (1996) Fertilizer use and maize
production in Sub-Saharan Africa. CIMMYT Economics
Working Paper 96–01, Mexico, DF
References Khan SA, Mulvaney RL, Ellsworth TR, Boast CW (2007) The
myth of nitrogen fertilization for soil carbon sequestration.
Anderson IP, Brinn PJ, Moyo M, Nyamwanza B (1993) J Environ Qual 36:1821–1832
Physical resource inventory of the communal areas of Kimetu JM, Mugendi DN, Bationo A, Palm CA, Mutuo PK,
Zimbabwe – An Overview. NRI Bulletin No. 60 National Kihara J, Nandwa S, Giller K (2006) Partial balance of
Resources Institute, Chatham nitrogen in a maize cropping system in humic nitisol of
ARI (2002) Simplified production factsheets on selected crops Central Kenya. Nutr Cycl Agroecosyst 76:261–270
grown in Zimbabwe. Agronomy Research Institute, Kroon PS, Hensen A, van den Bulk WCM, Jongejan PAC,
Department of Agricultural Research and Extension Vermeulen AT (2008) The importance of reducing the
Publication, Harare systematic error due to non-linearity in N2O flux measure-
Baggs EM, Chebii J, Ndufa JK (2006) A short-term investiga- ments by static chambers. Nutr Cycl Agroecosyst 82:175–
tion of trace gas emissions following tillage and no-tillage 186
of agroforestry residues in western Kenya. Soil Tillage Res Lasco RD, Ogle S, Raison J, Verchot L, Wassmann R, Yagi K
90:69–76 et al (2006) Cropland. In: Eggleston H, Buendia L, Miwa
Bodelier PLE, Laanbroek HJ (2004) Nitrogen as a regulatory K, Ngara T, Tanabe K (eds) 2006 IPCC Guidelines for
factor of methane oxidation in soils and sediments. FEMS National Greenhouse Gas Inventories. Institute for Global
Microbiol Ecol 47:265–277 Environmental Strategies, Japan
Bremmer JM, Mulvaney CS (1982) Nitrogen – Total. In: Page Ma BL, Dwyer LM, Gregorich EG (1999) Soil nitrogen
AL (ed) Methods of soil analysis, Agron. Monogr. 9, 2nd amendment effects on nitrogen uptake and grain yield of
edn. ASA and SSSA, Madison, pp 595–624 maize. Agron J 91:650–656
Burneya JA, Davis SJ, Lobella DB (2010) Greenhouse gas Mapanda F, Mupini J, Wuta M, Nyamangara J, Rees RM
mitigation by agricultural intensification. In: Robertson (2010) A cross-ecosystem assessment of the effects of land
GP, Kellogg WK (eds) Proceddings of the National cover and land use on soil emission of selected greenhouse
Academy of Science of the United States of America, gases and related soil properties in Zimbabwe. Eur J Soil
MI, pp 12052–12057 Sci. doi:10.1111/j.1365-2389.2010.01266.x
Chikowo R, Mapfumo P, Nyamugafata P, Giller KE (2004) Mosier AR, Pendall E, Morgan JA (2003) Effect of water addition
Mineral N dynamics, leaching and nitrous oxide and nitrogen fertilization on the fluxes of CH4, CO2, NOx,
losses under maize following two-year improved and N2O following five years of elevated CO2 in the
fallows on a sandy loam soil in Zimbabwe. Plant Colorado Shortgrass Steppe. Atmos Chem Phys 3:1–6
Soil 259:315–330 Mulvaney RL, Khan SA, Ellsworth TR (2009) Synthetic
de Klein C, Novoa RSA, Ogle S, Smith KA, Rochette P, Wirth nitrogen fertilizers deplete soil nitrogen: A global dilemma
TC et al (2006) N2O emissions from managed soils, and for sustainable cereal production. J Environ Qual
CO2 emissions from lime and urea application. In: 38:2295–2314
Eggleston H, Buendia L, Miwa K, Ngara T, Tanabe K Nelson DW, Sommers LE (1996) Total carbon, organic carbon,
(eds) 2006 IPCC Guidelines for National Greenhouse Gas and organic matter. In: Page AL et al. (eds) Methods of
Inventories. Institute for Global Environmental Strategies, Soil Analysis, 2nd edn. Agron. Monogr. 9. ASA and
Japan SSSA, Madison, WI, pp 961–1010
Desjardins LR, Smith W, Grant B, Campbell B, Riznek R Nyamangara J (2002) Soil organic matter status in Southern
(2005) Management strategies to sequester carbon in Africa in relation to sustainable soil tillage and crop
agricultural soils and to mitigate greenhouse gas emis- production. In: Chuma E (ed) African Conservation
sions. Clim Change 70:283–297 Tillage Network Year Book. IES-GTZ, Harare, pp 44–
Dick J, Kaya B, Soutoura M, Skiba U, Smith R, Niang A, Tabo 50
R (2008) The contribution of agricultural practices to Nyamangara J, Piha MI, Giller KE (2003a) Effects of combined
nitrous oxide emissions in semi-arid Mali. Soil Use cattle manure and mineral nitrogen on maize N uptake and
Manage 24:292–301 grain yield. Afr Crop Sci J 11:289–300
Flynn HC, Smith P (2010) Greenhouse gas budgets of crop Nyamangara J, Bergstrom LF, Piha MI, Giller KE (2003b)
production – current and likely future trends. International Fertilizer use efficiency and nitrate leaching in a tropical
Fertilizer Industry Association, Paris. http://www.fertilizer. sandy soil. J Environ Qual 32:599–606
Plant Soil (2011) 343:67–81 81

Odhiambo JJO, Magandini VN (2008) An assessment of the Tang J, Baldocchi DD, Xu L (2005) Tree photosynthesis
use of mineral and organic fertilizers by smallholder modulates soil respiration on a diurnal time scale. Glob
farmers in Vhembe district, Limpopo province, South Change Biol 11:1298–1304
Africa. Afr J Agric Res 3:357–362 Vanitchung S, Harvey NW, Chidthaisong A (2005) Effects of
Okalebo JR, Gathua KW, Woomer PL (1993) Laboratory methods land use on methane oxidation in tropical upland soils.
of soil and plant analysis: a working manual. Tropical Soil Thaiscience, Hua Hin, Thailand. http://www.thaiscience.
Biology and Fertility Programme. Soil Society of East Africa info/Article%20for%20ThaiScience/Article/4/Ts-4%
Technical Publication No. 1. Marvel EPZ, Nairobi 20effects%20of%20land%20use%20on%20methane%
Petersen SO (1999) Nitrous oxide emissions from manure and 20oxidation%20in%20tropical%20upland%20soils.pdf.
inorganic fertilizers applied to spring barley. J Environ Accessed 22 August 2010
Qual 28:1610–1618 Venterea RT (2010) Simplified method for quantifying theoret-
Rees RM, Wuta M, Furley PA, Li CS (2006) Nitrous oxide ical underestimation of chamber-based trace gas fluxes. J
fluxes from savanna (miombo) woodlands in Zimbabwe. J Environ Qual 39:126–135
Biogeogr 33:424–437 Vogel H, Nyagumbo I, Olsen K (1994) Effect of tied ridging
Rochette P, Eriksen-Hamel NS (2008) Chamber measurements and mulch ripping on water conservation in maize
of soil nitrous oxide flux: are absolute values reliable? Soil production on sandveld soils. Tropenlandwirt 95:33–
Sci Soc Am J 72:331–342 44
Seed-Co (1998) Productive farming seed manual 1998/99. Yang L, Yan J, Cai Z (2010) Effects of N-applications
Seed-Co Company, Harare and photosynthesis of maize (Zea mays L.) on soil
Svotwa E, Baipai R, Jiyane J (2009) Organic farming in the small respiration and its diurnal variation. Front Agric China
holder farming sector of Zimbabwe. J Org Syst 4:8–14 4:42–49