Temperature, wetting cycles and soil texture e•ects …directory.umm.ac.id/Data Elmu/jurnal/S/Soil Biology And...Temperature, wetting cycles and soil texture e•ects on carbon and

Temperature, wetting cycles and soil texture e�ects on carbonand nitrogen dynamics in stabilized earthworm casts

M. McInerney*, T. Bolger

Department of Zoology, University College Dublin, Bel®eld, Dublin 4, Ireland

Accepted 9 September 1999

Abstract

Two soils, one a silty loam soil (Kilkea) and the other a clayey soil (Leitrim), and stabilized earthworm cast derived from

both soils, were subjected to two wet-drying regimes (4 d and 16 d wetting intervals) at two temperatures (108C and 208C). Eachtreatment was sampled regularly for CO2 emissions and dissolved organic carbon, ammonium and nitrate were measured inleachate. Production of CO2 from the clayey soil was 20±40% less than from the silty loam at both temperatures and underboth cycles, while production was 40% less at the lower temperature. Amounts of dissolved organic carbon (DOC) in leachate

from the silty loam soil were greater than from the clayey soil, while, in general, greater amounts of DOC leached fromuningested soil than from earthworm cast. Treatments with the silty loam soil lost approximately 30±40 times more total N thancorresponding clayey soils under the 4 d cycle, and 12±16 times more under the 16 d cycle, thus illustrating the extent of

nutrient protection in clay soils, but also the reduction of this e�ect with super-saturation and prolonged drying (16 d cycle).Nutrient losses suggest greater protection of nutrients in cast at ®rst, followed by a reduction in protection (weakening ofaggregate structure) in later cycles. These results have important implications for nutrient losses in the ®eld, since signi®cant and

complex interactions were evident between the experimental factors. To improve our understanding of soil organic matter(SOM) dynamics, interactions between climatic, soil and cast-mediated changes (i.e. increased organic±inorganic interaction andburial) need to be incorporated into future models of soil decomposition. # 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Earthworms; Wet-drying cycles; Microaggregation; Organic matter; Stabilization; Carbon; nitrogen; Leachate; Dissolved organic car-

bon; Respiration

1. Introduction

Many factors, which operate at vastly di�erentspatial and temporal scales (Wardle and Lavelle,1997), a�ect the sizes of soil organic carbon pools andglobal CO2 ¯ux (Bolin et al., 1979; Swift et al., 1979;Stevenson, 1986). The role of soil texture (silt/clay con-tent) in the formation of organic±inorganic complexesand soil aggregates, and the protection of organic com-pounds from biological degradation has long been ap-preciated (Mortland, 1970). By slowing and modifyinghumic transformations (Anderson et al., 1981) soils

with higher silt/clay content retain higher proportionsof residual organic C and N (Jenkinson, 1971; Soren-sen, 1972; Ladd et al., 1975).

Increasing silt/clay content also increases waterholding capacity, so that soil texture interacts with cli-mate in controlling these ecosystem processes (Burkeet al., 1989). This interactive complexity is furtherincreased by earthworm cast production and its poten-tial for variability with age and state.

Microaggregates containing a core of organic matter(Shipitalo and Protz, 1989) are formed during transitthrough the earthworm's gut and once these organic±inorganic matrices are egested as casts, bonds betweenorganic and mineral components are strengthenedwhen these materials are brought into close associ-ation, due to water loss. Thus earthworm cast struc-

Soil Biology & Biochemistry 32 (2000) 335±349

0038-0717/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.

PII: S0038-0717(99 )00158 -3

www.elsevier.com/locate/soilbio

* Corresponding author. Fax: +353-1-706-1152.

E-mail address: [email protected] (M. McInerney).

ture can enhance aggregation (Blanchart et al., 1990),structural stability (De Vleeschauwer and Lal, 1981)and tensile strength (McKenzie and Dexter, 1987;Zhang and Schrader, 1993) of soils. However, thisstability is greatly in¯uenced by water content. Notsurprisingly, studies on fresh casts (Shipitalo andProtz, 1988; Marinissen and Dexter, 1990; Hindell etal., 1994a,b) have pointed out that their physical prop-erties, particularly relating to structural stability, canvary with age and wetting/drying cycles. Thus thecompact structure of casts may result in temporary im-mobilisation, but also under certain moisture/tempera-ture conditions, the pulsed release of nutrients (Lavelleand Martin, 1992; Wardle and Lavelle, 1997). Marinis-sen and Dexter (1990) found that one drying/rewettingcycle increased the stabilisation of casts. Hindell et al.(1994b) concluded that the wetter the soil and thelonger it stays wet, the longer casts remain in a disper-sible state, and that this could increase nutrient losses.The microstructural state of casts can also be in¯u-enced by their position (buried or surface) in the soil,since buried casts can remain in a more moist con-dition than surface casts. In relation to this, Oades(1993) noted that earthworms in¯uence the rate, extentand spatial development of the soil drying phase.

Thus, over the medium- to long-term, cast structuremay enhance (O'Brien and Stout, 1978) or block (Mar-tin, 1991) the overall mineralization of soil organicmatter. With large volumes of soil being ingested byearthworms (Lavelle, 1978; Lee, 1983) and given thatthe casts di�er physically and chemically (and micro-biologically) from the soil from which they are derivedthey represent an important source of soil heterogen-eity (Anderson, 1988).

Previous studies have focused on short-term physicaland chemical changes in fresh casts. Few have concen-trated on aged, stabilised casts, though the potentialfor cast persistence in soil has been well documented(Lavelle, 1978; Lee, 1983). We examined the inter-actions between wetting/drying cycles, temperature,silt/clay content on aged and already stabilised earth-worm casts (i.e. medium to long-term decomposition)and the subsequent C and N dynamics of these soilaggregates. In particular it examined the extent towhich respired carbon (CO2±C) and dissolved organiccarbon (DOC) were a�ected by these treatments.

2. Materials and methods

2.1. Soils

The soils we used were collected from the top 7±15cm of the mineral horizons at two sites in Ireland. TheKilkea soil (Typic Hapludalf) was taken from a Nor-way spruce stand in Kilkea, Co. Kildare. It had a pH

of 4.7, water holding capacity of 36.0%, a C contentof 3.825%, total N content of 0.315% and an organicmatter content of 8.62%. The particle sizes were20.81% clay, 38.42% silt and 40.77% sand. The Lei-trim soil was a heavy gley soil (Gley, Al®c Hapla-quept) from a Sitka spruce forest site at Lisgavneen,Co. Leitrim. The soil had a pH of 4.8, water holdingcapacity of 29.8%, C content of 2.41%, total N con-tent of 0.182% and an organic matter content of5.83%. The soil contained 48.17% clay, 42.75% siltand 9.08% sand. Prior to their use in these exper-iments the soils were air dried, homogenised (macroag-gregates broken up, large roots, macro-organicparticles and stones removed) and sieved (2 mm).

2.2. Earthworm cast and control soil production

Earthworms (Lumbricus terrestris Linneaus) exhib-ited normal feeding and burrowing behaviour whilebeing maintained at a mean temperature of 152 38Cand 80±100%RH in containers with oak leaf litter(Quercus petraea Mattuschka, Liebl.) and soil. Theresulting casts were collected and stored at 48C untilthe experiments began. Earthworm casts producedusing the Kilkea soil had an organic matter content1.17% greater than the initial soil. To ensure that non-cast treatments had a similar organic matter content,shredded oak litter was incorporated by gentle handmixing of soil, water and oak (shredded to <2 mm) toform macroaggregates. The resulting mixture for bothcast and control soil contained 9.79% organic matter.A similar procedure was adopted for Leitrim soils(6.99% organic matter, 1.48% derived from incorpor-ated litter). Oak litter, cast and parent soil organicmatter contents were estimated as `loss-on-ignition'.The percentage of clay, silt and sand in earthwormcast was found to be similar to that of uningested soiland aggregate structure of both casts and uningestedsoil (controls) was similar (dia 0.4±1.3 cm approxi-mately).

2.3. Incubation units and procedure

Carbon, nitrogen and water ¯uxes were examinedusing laboratory based incubation units. These trans-parent polystyrene Nalgene ®ltration units (Cat. No.FD240-10), consisted of an upper chamber, used tohold the soil and sand layers, and a lower chamberwhere leachate was collected. Eighty grams of cast orcontrol soil (derived from Kilkea and Leitrim soils)were added to these units above a 20 g layer of acidwashed sand (i.e. bottom layer).

The unit was sealed to ensure that it was completelyair tight during the respiration measurements. A headspace of approximately 75 ml was present above thesoil in the upper chamber. Gas may have di�used

M. McInerney, T. Bolger / Soil Biology & Biochemistry 32 (2000) 335±349336

through the soil and sand layers into the lowerchamber but such losses were expected to be minimal.

A 40 mm ®lter separated the unit chambers thus pre-venting sand from passing through. It was envisagedthat the sand in turn would prevent larger particulateorganic carbon (POC) from leaching, but allow dis-solved organic carbon (DOC) to percolate into thelower chamber of the experiment unit.

The cast and control soils were stored for 6 weeks at148C with a water content of 60±80% water holdingcapacity. This was to ensure that the initial short termpulse of mineralization associated with fresh earth-worm casts had ended, and that steady state, medium-to long-term mineralization rates were in e�ect. Thecast and control were then dried at 208C for 2 d andstored at 48C. This enabled a similar conditioning in-cubation to be set for all treatments.

To ensure that drying and cold storage did notadversely a�ect microbial development during the ex-periment, a comparison of CO2 production (at 208Cand 60±80% water holding capacity) for nonstoredand dried/cold stored/reinoculated treatments was car-ried out. No signi®cant di�erences were evidentbetween these treatments after 10±12 d. Thus, all treat-ments were maintained at 208C and 60±80% waterholding capacity for a period of 14±16 d before the ex-periment began.

In total, 16 treatments, with 4 replicates, were usedin this experiment as outlined in Table 1. Cast andcontrol soil treatments derived from both the Kilkeaand Leitrim soils were placed in incubation units,which were then subject to two wet/drying regimes (4d and 16 d wetting intervals) and two temperatureregimes (108C and 208C).

For the 4 d cycle, using a Pasteur pipette to simulate

rainfall (droplet only), 30 ml of distilled water wereadded at the beginning of each cycle, while 120 mlwere added at the start of each 16 d cycle. Thus anequal amount of water was added under either cycleover a 16 d period. The experiment lasted for 80 d forall treatments, that is, ®ve 16 d cycles and twenty 4 dcycles. Each treatment was sampled regularly for CO2

emissions while DOC, ammonium and nitrate weremeasured in leachate.

2.4. Analyses

Moisture content (%) was estimated by drying at1058C (Allen, 1989; Anderson and Ingram, 1993).`Loss-on-ignition' was used to estimate percentage or-ganic matter (Allen, 1989). Samples of 2±4 g weredried at 1058C, ground for 10 min at 900 oscillationsminÿ1 using a RETSCH grinder (type MM2), beforetotal C, H and N of cast and soils were measured by¯ash combustion followed by GC using a CARLO-ERBA 1106 automatic analyser.

CO2 production was measured by syringe removalof a 5 ml air sample from the experiment units' headspace (75 ml) after a 15±20 min incubation. Thesamples were then injected into a modi®ed Ciras-SC(PP Systems) infra red gas analyser. Samples weretaken 3 times during the ®rst 24 h (2, 4 and 8 h afterwetting), and on the second (24 h) and third (48 h)days, after wetting. For the 16 d wetting interval, CO2

was further sampled at 2±4 d, 4±8 d after the wettingphase and on d 14 of the cycle.

Leachate was collected approximately 2 h after theend of the water application i.e. when no more waterdroplets were coming from upper chamber of theunits. Immediately after the sample was taken, DOC(in leachate ®ltered through a 0.45 mm Millipore(HAWP 01300)) was estimated by measurement ofabsorbence at 400 nm and comparison with a standardcurve obtained using a TOC analyser. Samples werethen frozen for a maximum of 2 weeks before concen-trations of nitrate-N, chloride and sulphate weremeasured using ion chromatography (Dionex QicAS4A separator column) and concentrations of ammo-nia-N were measured using a Tecator Flow InjectionAnalyser FIAstar 5010 using the method described inapplication notes ASN 50-01/84. Samples were ®lteredthrough a 0.2 mm membrane ®lter to remove particu-late matter and organic material. Three replicates wereanalysed for ammonia-N and nitrate-N, and four repli-cates were sampled for DOC. The results were ana-lysed using a four ®xed factor (temperature, wettingcycle, cast and soil type) model of analysis of variance.All analysis was carried out using SAS (StatisticalAnalysis Software, 1992).

All the results are expressed as amounts gÿ1 soil car-

Table 1

Treatments used in this experiment. The 4 d and 16 d wetting cycles

had 30 ml and 120 ml water applications respectively

Soil type Temperature (8C) Wetting cycle (d) Cast/no cast

Kilkea 10 4 cast

Kilkea 10 16 cast

Kilkea 20 4 cast

Kilkea 20 16 cast

Leitrim 10 4 cast

Leitrim 10 16 cast

Leitrim 20 4 cast

Leitrim 20 16 cast

Kilkea 10 4 no cast

Kilkea 10 16 no cast

Kilkea 20 4 no cast

Kilkea 20 16 no cast

Leitrim 10 4 no cast




M. McInerney, T. Bolger / Soil Biology & Biochemistry 32 (2000) 335±349 337

bon in order to compensate for the fact that theamounts of organic matter varied between the soils.

3. Results

3.1. CO2 production

Following the addition of water to the experimental

units there was always an immediate and sharpincrease in the amount of CO2 evolved. In general,maximum CO2 output occurred 2±4 h after the ad-dition (Fig. 1). Following this peak there was a rapiddecline for a period of 6±10 h followed by a slowerdecline over the remaining 3, or 15, d of the cycle.

Irrespective of treatment, the production of CO2

was always greater at 208C than at 108C, with the pro-duction at 108C being approximately 40% less. The

Fig. 1. Loss of CO2 (ml CO2 gÿ1 soil C hÿ1) from cast (Ð) and control (- - -) for Kilkea soil treatments during the 4 d wetting cycle. Bars rep-

resent the mean2S.E.

Table 2

Table of F values for the response of carbon and nitrogen parameters to temperature (T), wetting cycle (W), soil (S) and cast (C) treatments.�P<0.05, ��P<0.01, ��P<0.001

Treatment d.f. F values d.f. F values

CO2 water DOC NO3N NH4N total N

S 1, 48 175.97�� 446.97�� 562.95�� 1, 32 247.35�� 52.74�� 209.08��

W 1, 48 26.55�� 7965.15�� 391.18�� 1, 32 13.63�� 17.79�� 24.71��

T 1, 48 1076.09�� 821.82�� 25.2�� 1, 32 8.53�� 1.73 1.03

C 1, 48 1.81 0.96 15.63�� 1, 32 0.79 2.08 2.14

S�W 1, 48 72.89�� 46.51�� 231.04�� 1, 32 10.17�� 10.41�� 16.26��

S�T 1, 48 31.07�� 48.08�� 28.19�� 1, 32 9.14�� 1.59 1.24

S�C 1, 48 0.15 7.33�� 12.1�� 1, 32 0.64 2.46 2.22

W� T 1, 48 1.2 534.96�� 61.23�� 1, 32 0.02 0.36 .23

W�C 1, 48 2.24 1.33 2.33 1, 32 6.17� 1.65 5.62�

T�C 1, 48 1.3 2.35 0.27 1, 32 < 0.01 0.02 0.01

S�W� T 1, 48 19.27� 4.64� 0.6 1, 32 0.09 0.46 0.38

S�W�C 1, 48 2.46 3.21 2.88 1, 32 6.29� 1.63 5.66�

W� T�C 1, 48 0.46 3.28 1.24 1, 32 4.11 0.05 1.28

S�T�C 1, 48 0.39 0.1 2.74 1, 32 < 0.01 0.02 0.01

S�W� T�C 1, 48 0.28 0.84 3.2 1, 32 4.27� 0.08 1.27


production from the clayey Leitrim soil was 5±35%

less than that from the Kilkea soil (Fig. 2). Although

the interaction between wetting regime and tempera-

ture was not signi®cant, there was a signi®cant inter-

action between soil, wetting regime and temperature

�FSWT3 � 19:27�� (Table 2). The response to tempera-

ture was most evident for the Kilkea soil under the 4 d

cycle where the production of CO2 was 1.85 times

greater at 208C. The increase for the Leitrim soil was

only 1.6. However, this di�erence in response was not

evident under the 16 d cycle where both soils showed

an approximately 1.7-fold increase in production.

Therefore it is apparent that soil type has a signi®cant

bearing on the relationship between C mineralization

and temperature/moisture regime. It is also apparentthan cast had no signi®cant e�ects, but Q10 values for

cast treatments were lower than for the controls,

suggesting that the cast aggregate structure limited de-

Fig. 2. Sum of CO2 (ml gÿ1 CO2 soil C) measured over the course of the experiment for (a) Kilkea soil and (b) Leitrim Soil. Bars represent the

mean2S.E. Ec10 is earthworm cast at 108C, Ec20 is earthworm cast at 208C, Co10 is control at 108C and Co20 is control at 208C.


composition rates (at least at high temperatures)(Table 3). In the clayey Leitrim soil such di�erenceswere less and only present under the 4 d cycle, high-lighting the overriding in¯uence of clay content on mi-crobial respiration.

3.2. Leachate volume

Several of the factors interacted signi®cantly in

Table 3

Q10 values from experimental treatments

Soil cycle Kilkea Leitrim

4 d 16 d 4 d 16 d

Cast 1.80 1.69 1.57 1.72

Control 1.89 1.74 1.60 1.72

Average 1.85 1.72 1.58 1.72

Fig. 3. Cumulative volume (ml) of water leached from (a) Kilkea and (b) Leitrim treatments. Bars represent the mean2S.E. Ec10 is earthworm

cast at 108C, Ec20 is earthworm cast at 208C, Co10 is control at 108C and Co20 is control at 208C.


determining the volumes of leachate produced. Greatervolumes (1±3 times) of water were leached from allsystems under the 16 d cycle (Fig. 3). However, therewere obviously smaller volumes of leachate at 208Cthan at 108C with the 4 d cycle, but this was not soobvious with the 16 d cycle �FWT � 534:96��). The soiltype interacted with temperature and wetting regime inthat there was a reduced volume of leachate for bothsoils at higher temperatures under the 4 d cycle but thevolumes of leachate from the clayey Leitrim were notreduced at 208C under the 16 d cycle. In other words

the Kilkea soil retained more water than the Leitrimsoil at the higher temperature when there was a signi®-cant period without water input, i.e. 16 d cycle.Although the interaction between cast and soil typewas signi®cant, its contribution to the variance wasminimal.

3.3. Dissolved organic carbon

None of the third or fourth order interactions weresigni®cant for DOC. However, many of the factors

Fig. 4. Weighted average amounts of DOC (mg DOC gÿ1 soil C) lost from (a) Kilkea and (b) Leitrim treatments. Bars represent the mean2S.E.

Ec10 is earthworm cast at 108C, Ec20 is earthworm cast at 208C, Co10 is control at 108C and Co20 is control at 208C.


interacted in pairs. The e�ect of temperature and wet-

ting regime di�ered between the soils. The average

amounts leached from the Kilkea soil responded to the

temperature and wetting factors, with more DOC

being lost under the 16 d cycle and at 108C�FWT � 61:23��). The Leitrim soil did not respond in a

consistent manner to these factors (Fig. 4). Under the

4 d cycle both soils lost approximately twice as much

DOC at 108C than at 208C. However, under the 16 d

Table 4

Ratio of CO2±C-to-DOC from experimental treatments

Soil cycle Kilkea Leitrim

4 d 16 d 4 d 16 d

Cast 108C 15.81 8.74 20.26 32.97

Control 108C 15.25 7.97 14.64 21.15

Cast 208C 58.90 14.63 52.80 39.62

Control 208C 67.23 15.17 42.27 20.88

Fig. 5. (a) Weighted average amounts of NO3±N (ng NO3±N gÿ1 soil C) lost from (a) Kilkea and (b) Leitrim treatments. Bars represent the



cycle the Kilkea soil did not show a marked response

to temperature while the response of the Leitrim soil

varied between cast and noncast treatments.

The e�ect of the casts was to reduce the losses of

DOC from the Leitrim soil. However, for the loamy

Kilkea soil, as indicated above, the e�ect was depen-

dent on the temperature. At 108C less DOC was lea-

ched from the casts, but at 208C there was more

leachate.

3.4. Losses of carbon in CO2 and DOC

In general, the ratios of CO2±C/DOC losses fromLeitrim soils were higher than from Kilkea soils underthe 16 d cycle (Table 4). This was due mainly to con-siderably lower losses of DOC with this cycle. Thereverse was evident under the 4 d cycle, though thedi�erences were not as considerable.

The e�ect of the 16 d wetting/drying cycle relative

Fig. 6. (a) Weighted average amounts of NH4±N (ng NH4±N gÿ1 soil C) lost from (a) Kilkea and (b) Leitrim treatments. Bars represent the



to the 4 d cycle was to reduce the CO2±C-to-DOCratio in the Kilkea soil. For the Leitrim soil the e�ectwas more complex. At 208C the 16 d wetting/dryingcycle reduced the ratio, while the reverse was true at108C.

Temperature had a considerable in¯uence on carbonlosses. With the 4 d cycle, the amount of C lost asDOC was less at 208C than at 108C. The reverse wastrue for the amount of C lost as CO2. This resulted ina 4- and a 3-fold di�erence in the ratio of CO2±C-to-DOC between temperatures for the Kilkea and Leitrim

soils, respectively. A similar trend was evident with the16 d cycle (both soils), though the di�erence was con-siderably lower, ranging between 5 and 40%, mainlydue to higher losses of DOC with this cycle.

In general, cast treatments had lower DOC losses,and thus a higher CO2±C-to-DOC ratio than controlsoils. The one exception was in the case of the Kilkeasoil under the 4 d cycle at 208C. In this case the oppo-site was true due to a combination of lower CO2 andhigher DOC losses. Di�erences in CO2±C-to-DOCratios between temperatures were reduced in cast treat-

Fig. 7. Mean total nitrogen loss (ng total N gÿ1 soil C) in leachate from (a) Kilkea and (b) Leitrim treatments. Ec10 is earthworm cast at 108C,Ec20 is earthworm cast at 208C, Co10 is control at 108C and Co20 is control at 208C.


ments (Kilkea soil) for both cycles. However, this wasparticularly obvious with the 4 d cycle. The same e�ectwas evident with the 4 d cycle for Leitrim soils, butthe reverse was true with the 16 d cycle. The percen-tage of the C loss in the form of DOC ranged between1.7 and 15.5% for Kilkea soils and 2.3 and 6.9% forLeitrim soils.

Thus, wet/drying intensities, temperature, soil tex-ture and cast structure all in¯uenced the ratio of car-bon lost in CO2 and DOC in this experiment.

3.5. Nitrate-N and ammonium-N

The amount of nitrate-N leached was dependent oncomplex interactions between all four experimental fac-tors FSWTC � 4:27�: Under the 4 d cycle, amounts ofnitrate-N in leachate from the Kilkea soil werebetween 40 and 60 times greater than those from theLeitrim soil (Fig. 5). The same trends were evident forthe 16 d cycle but the sizes of the di�erences were less.In addition, more was leached under the 16 d cyclethan under the 4 d cycle. The relationship betweenwetting cycle and cast varied between the soils�FSWC � 6:29�). With the Leitrim soil the response tocasts was una�ected by wetting cycle, i.e. the amountswere generally higher coming from casts. This was alsotrue for the Kilkea soil except when incubated at 208Cunder the 4 d cycle, i.e. less nitrate was leached fromcasts produced from the loamy soil at 208C under the4 d cycle. The e�ect of temperature was very depen-dent on the other factors and while there may be ageneral pattern of increase at the higher temperaturethe opposite was true for the clayey Leitrim soil underthe 16 d cycle where the amounts leached were reducedat 208C.

The amount of ammonium-N leached was depen-dent on the soil type, the wetting regime and the inter-action between them. In general, the amounts ofammonium-N in leachate from the Kilkea soil underboth wetting regimes were more than 10 times greaterthan those from the Leitrim soil (Fig. 6). In addition,the amounts leached under the 16 d cycle were signi®-cantly larger than those under the 4 d cycle. While nosigni®cant e�ect of the casts was seen in the am-monium-N dynamics, several patterns do appear toemerge which are masked by the large within treat-ment variation. For example, more ammoniumappears to be leached from the Kilkea soil casts, par-ticularly under the 16 d cycle, while the opposite wastrue for the Leitrim soil.

Overall, treatments with the Kilkea soil lost approxi-mately 25±35 times more total N than correspondingLeitrim soils under the 4 d cycle, and 10±20 timesmore under the 16 d cycle (Fig. 7). This illustrates theextent of nutrient protection in Leitrim soils, but alsothe reduction of this e�ect with super-saturation fol-

lowing prolonged drying (16 d cycle, 120 ml water ap-plication). Comparing wet/drying cycles,approximately one-third less total N was leachedduring the 4 d cycle than the 16 d cycle in Leitrimsoils, while this di�erence was just 60% in the Kilkeasoil.

4. Discussion

It is apparent from our study that temperature, fre-quency of wetting and drying, soil texture and castaggregates in¯uence the dynamics of C and N in soil.In addition, and perhaps more importantly, there aresigni®cant interactions between these factors.

4.1. Carbon loss

The production of CO2 was lower in the clayey soil,greater at 208C for both soils, while the e�ect of wet-ting cycle was only apparent for the clayey soil. Forboth soils (and for both cycles) there was a rapidincrease in CO2 production during the ®rst 2±8 h afterwetting followed by a decrease over the next 12±24 hof the cycle. Van Gestel et al. (1993) stated that afterremoistening dried soils, available (labile) componentsare assimilated and transformed into new biomass andpartly into CO2. Franzluebbers et al. (1994) found asimilar pattern to CO2 losses, when plant matter,mixed with a loamy soil, was subjected to 5 d drying±wetting cycles. However, in their study the slower ratesof change (both increases and decreases) may havebeen due to less water addition (relative to water hold-ing capacity), while in our experiment, water additionwas greater than water holding capacity for bothcycles. This would have resulted in less disruption oforganic±mineral bonds and, as a result, less nutrientswould have been made available for microbial activity.De Bruin et al. (1989) found a more rapid decrease inCO2 production from their treatments by comparisonwith our results which may have been due to limitedamounts of labile organic material in their treatments,since plant litter was not incorporated into their treat-ments. For the initial 2 months of our experiment CO2

output decreased slowly in all treatments as was thecase in studies by De Bruin et al. (1989) and Fran-zluebbers et al. (1994) though in the latter, the rate ofdecrease in CO2 production on each successive cyclewas considerably greater than in our experiment.

Van Gestel et al. (1993) stated then that the ¯ush ofC mineralization caused by drying/rewetting dependson the size of the soil organic pool, the quality of theorganic matter, the extent to which it is protected byclay adsorption and on the properties of the soil biota.Certainly for the Leitrim soil, greater carbon stabilis-ation due to high silt/clay content (Mortland, 1986),


may have lead to a reduction in CO2 production inthis soil (Jenkinson, 1971; Sorensen, 1972; Martin andHaider, 1986) and may have had an overriding in¯u-ence on respiration rates. This idea is supported by thefact that more than 15% as much CO2 was producedper g soil C from the Kilkea soil than from the Leitrimsoil at both temperatures and in both cycles. The wet-ting cycles on the other hand dampened variationsbetween soils (see Q10 values, Table 3) while casts hadno signi®cant e�ect on respiration in our experiment.

Losses of DOC were not so in¯uenced by tempera-ture but the wetting cycle had a considerable in¯uencein the loamy Kilkea soil. In general, slightly greateramounts of dissolved organic C were leached fromuningested soil than from earthworm casts. Thesedi�erences were particularly evident for the clayey Lei-trim soil, where they occurred at both temperaturesand both wetting/drying cycles. With the less clayeyKilkea the di�erences were not so obvious. Duringintense supersaturation of the soil, bonding may beweakened considerable by the replacement (duringswelling) of cation bridges with weaker water bridges,a reversal of the process outlined by Shipitalo andProtz (1989) for cast formation. As a result, the di�er-ences between the strength of the cast and uningestedsoil aggregates (McKenzie and Dexter, 1987; Marinis-sen and Dexter, 1990) may be reduced or eliminated,resulting in losses of previously-protected organic mat-ter and nutrients (Hindell et al., 1994b). Alternatively,moisture content below a critical minimum (after pro-longed drying) may weaken the structural integrity(through denaturation of organic bonds, Hindell et al.,1997), to the point where di�erences between DOClosses in leachate from cast and control are eliminated(when rewetting occurs). It is interesting to note then,that during the initial phases (®rst three cycles) of the16 d cycle (Kilkea soil), cast treatments leached lessDOC than the control at 108C, but on the fourth cyclethe trend was reversed. At 208C there were never anydi�erences. This would suggest that the single waterapplication (120 ml) had a more pronounced e�ect at208C, and that both supersaturation and drying inten-sity played a part in the disruption and dispersal of or-ganic matter in cast aggregates.

The cumulative volumes of leachate were greaterfrom the Leitrim soil for both cycles, indicating thatthe Kilkea soil retained more water than the Leitrimsoil for these wetting/drying regimes. This in turnsuggests that water bridges (during supersaturation)and potential swelling capacity may have had a greaterin¯uence in the Kilkea rather than in the Leitrim soils(the latter did not show a breakdown in aggregation).

Visually, it was clear that, in Leitrim soils, the castmacroaggregate structure maintained its integrity. Acombination of organo±mineral bonding (Shipitaloand Protz, 1989) and inorganic cementing (Tisdall and

Oades, 1982) may have increased cast aggregatestrength to such an extent that the increased water ap-plication (120 ml) or increased intensity of drying (16d cycle) failed to reduce the protection a�orded to or-ganic matter in the cast. Hindell et al. (1997) foundthat the dispersion index of air-dried casts was not sig-ni®cantly increased by 5 wet/drying cycles in a hardsetting, silty loam soil. Van Gestel et al. (1991) statedthat DOC variations between soil leachates were dueto di�erences in CEC and microporosity (both ofwhich are dependent or in¯uenced by compaction andclay content) providing a level of protection duringdesiccation. Thus, overall concentrations and amountsof DOC in leachate from the Kilkea soil were 1±3times greater than from the Leitrim soil.

The percentage of the C lost as DOC (1±15%) cor-responds with other studies of C loss in forest systems.Bunnell et al. (1977) estimated, using modelled datafrom several countries, that losses due to microbial res-piration are usually in the range of 70±90%, the restbeing due to leaching of DOC and disintegration,while A.H. Magill (unpubl. Masters thesis, Universityof New Hampshire, 1996) monitored amounts of CO2

and DOC in leachate from seven species of hardwoodand softwoods and found that between 6 and 39% oftotal C loss was leached as DOC. In studies by Cronan(1985), 54±68% of the forest ¯oor C loss was due torespiration, while DOC leaching accounted for theremaining 32±46% of C loss.

4.2. Nitrogen loss

Nitrogen (both nitrate-N and ammonium-N) lossesin leachate from Kilkea soils under the 16 d cycle werehigher from the casts than the control (at both tem-peratures, the di�erence being greater at the highertemperature). We can assume that initial pools oflabile or mobile N were greater in the cast (Barley andJennings, 1959; Haimi and Huhta, 1990; Scheu, 1993;Parkin and Berry, 1994). In addition, supersaturation(120 ml) may have reduced any protection or stabilis-ation e�ect inherently present in cast aggregates.Nitrate-N losses (from both cast and control) increasedin the ®rst month and then decreased, while othernutrients (e.g. ammonia) showed a steady decline. Theincrease can be attributed to nitri®cation of existingammonium pools and/or greater mineralization of or-ganic N. Jensen (1997) indicated that organic N couldbe assimilated directly by the microbial biomass with-out entering the soil inorganic N pool.

Nutrient losses under the 4 d cycle were not as con-sistent as under 16 d cycle. No di�erences weredetected at 108C but at 208C more nitrate was leachedfrom the control, perhaps indicating greater nitri®ca-tion in this treatment. Alternatively, unlike the 16 dcycle (where breakdown in cast structure was impli-


cated in nutrient release, see DOC section above), thecast structure during the 4 d cycle may have main-tained greater protection/stabilisation of organic ma-terial and thus N mineralization would have beenhigher in the control.

For the Leitrim soils (both cycles), higher losses ofchloride, sulphate (results not shown) and ammoniumwere evident in leachate from the control for the ®rst2±3 sampling dates. However, during the remainder ofthe experiment, this trend was either reversed or castand control lost nutrients at the same rate. Thissuggests greater protection of nutrients in cast at ®rst,followed by a reduction in protection (weakening ofaggregate structure) in later cycles. The same is true ofnitrate losses in the 16 d cycle, but not the 4 d cycle,where greater nitri®cation may have occurred in thecast than the control, a similar trend to that of seen inKilkea soils with the 16 d cycle.

The in¯uence of temperature on N transformationsand speci®cally nitri®cation in forestry ecosystems hasbeen reviewed extensively (Focht and Verstraete, 1977;Killham, 1990; Quemada and Cabrera, 1997; Carnoland Ineson, 1999). The latter study found that thetemperature optimum for nitrate concentrations in lea-chate from undisturbed soil cores to be 118C. How-ever, it was pointed out that disturbed soils inevitablyhave increased nutrients and aeration. Increased nutri-ent availability can lead to higher nitri®cation optima(ranging between 20 and 408C). Thus, nitri®cation(and associated leaching of nitrate) in our experimentis likely to have had an optimum closer to 208C, whichis consistent with leachate results from Kilkea soils inboth cycles and in Leitrim soils during the 4 d cycle.However, for Leitrim soils with the 16 d cycle, higherclay content, in conjunction with greater `cementing'stabilisation (Hindell et al., 1997) at 208C, may havebeen the dominant in¯uence on nitrifying bacteria,resulting in lower nitrate losses than at 108C. This cor-responds with the results of CouÃ teaux and Sallih(1994) who found that the greatest sequesterisation of15N was in soils with the highest cation exchange ca-pacity, the highest clay content and the highest organicmatter content.

In Leitrim soils the e�ect of cast structure on theleaching of total N was masked by the more in¯uentiale�ect of high clay content. However, in the Kilkeasoils, the cast structure lost more total N than controlsoils during the 16 d cycle (after aggregate destabilisa-tion), while the reverse occurred at 208C during the 4d cycle (due to less aggregate destabilisation).

5. Conclusion

The high levels of contingency in the outcomes ofecological interactions have recently been highlighted

(Lawton, 1999; Bolger et al., in press). The large num-bers of signi®cant interactions, between all the factorsexamined in this experiment, provide further convin-cing evidence of such contingency.

The loamy Kilkea soil lost greater amounts of Cand N than the clayey Leitrim soil per g soil C. Intensewetting cycles increased C losses (as DOC) from theloamy Kilkea soil, and C (as CO2 form) and total Nlosses from the Leitrim soil.

The e�ects of temperature and cast formation onnutrient losses were less clearcut and dependent on soiltype and wetting cycle. Temperature increased CO2

losses and reduced DOC losses (during less intensewetting cycles), but had limited in¯uence on total Nlosses. Time series results suggest implications fornutrient losses from casts in the ®eld. With increasingintensity of wet/drying cycles, casts, it is predicted, willlose more nutrients (C, N) than uningested soils. Butat higher temperatures (0208C) with less intense wet/drying cycles, casts will retain more nutrients thanuningested soils, particularly in soils which have a highsilt/clay content. Thus, while both surface and buriedcast material will reduce nutrient losses in leachaterelative to uningested soil, losses are expected to behigher in surface cast, since surface material is exposedto wet/drying cycles of greater intensity than buriedcasts. To improve our understanding of SOMdynamics, cast mediated changes (i.e. increased or-ganic±inorganic interaction and burial) need to be in-corporated into future models of soil decomposition,perhaps by enhancement of the soil texture functionalong with litter incorporation.

Acknowledgements

The work was supported by the EnvironmentalResearch and Development Programme of the Com-mission of the European Union (Vamos Programme).We would like to thank Sylvia Dolan and Pat Nevillefor their technical help.

References

Allen, S.E., 1989. Chemical Analysis of Ecological Materials, 2nd ed.

Blackwell Scienti®c, Oxford.

Anderson, D.W., Saggar, S., Bettany, J.R., Stewart, J.W.B., 1981.

Particle size fractions and their use in studies of soil organic mat-

ter. 1. The nature and distribution of forms of carbon, nitrogen

and sulfur. Soil Science Society of America Journal 45, 767±772.

Anderson, J.M., 1988. Spatiotemporal e�ects of invertebrates on soil

processes. Biology and Fertility of Soils 6, 216±227.

Anderson, J.M., Ingram, J.S.I., 1993. Tropical Soil Biology and

Fertility, 2nd ed. CAB International, UK.

Barley, K.P., Jennings, A.C., 1959. Earthworms and soil fertility.

Australian Journal of Agricultural Research 10, 364±370.

Blanchart, E., Lavelle, P., Spain, A.V., 1990. E�ects of biomass and


size of Millsonia anomala (Oligochaeta, Acanthodrilidae) on par-

ticle aggregation in a tropical soil in the presence of Panicum

maximum (Gramineae). Biology and Fertility of Soils 10, 113±

120.

Bolger, T., Heneghan, L.J., Neville, P. Invertebrates and nutrient

cycling in coniferous forest ecosystems: spatial heterogeneity and

conditionality. In: Coleman, D.C., Hendrix, P.F. (Eds.),

Invertebrates as Webmasters. CABI, London, in press.

Bolin, B., Degens, E.T., Duvigneaud, P., Kempe, S., 1979. The glo-

bal biogeochemical carbon cycle. In: Bolin, B., Degens, E.T.,

Kempe, S., Ketner, P. (Eds.), The Global Carbon Cycle, pp. 1±

44.

Bunnell, F.L., Tait, D.E.N., Flanagan, P.W., Van Cleve, K., 1977.

Microbial respiration and substrate weight loss. I. A general

model of the in¯uences of abiotic variables. Soil Biology &

Biochemistry 9, 33±40.

Burke, I.C., Yonker, C.M., Parton, W.J., Cole, C.V., Flach, K.,

Schimel, D.S., 1989. Texture, climate, and cultivation e�ects on

soil organic matter content in US grassland soils. Soil Science

Society of America, Journal 53, 800±805.

Carnol, M., Ineson, P., 1999. Environmental factors controlling

NO3ÿ leaching, N2O emissions and numbers of NH4

+ oxidisers in

a coniferous forest soil. Soil Biology & Biochemistry 31, 979±990.

CouÃ teaux, M.-M., Sallih, Z., 1994. Fate of inorganic 15N in the pro-

®le of di�erent coniferous forest soils. Biology and Fertility of

Soils 17, 101±107.

Cronan, C.S., 1985. Comparative e�ects of precipitation acidity on

three forest soils, carbon cycling responses. Plant and Soil 88,

101±112.

De Bruin, B., Penning De Vries, F.W.T., Van Broekhoven, L.W.,

Vertregt, N., Van De Geijn, S.C., 1989. Net nitrogen mineraliz-

ation, nitri®cation and CO2 production in alternating moisture

conditions in an unfertilized low-humus sandy soil from the

Sahel. Plant and Soil 113, 69±78.

De Vleeschauwer, D., Lal, R., 1981. Properties of worm casts under

secondary tropical forest regrowth. Soil Science 132, 175±181.

Focht, D.D., Verstraete, W., 1977. Biochemical ecology of nitri®ca-

tion and denitri®cation. Advancees in Microbial Ecology 1, 135±

214.

Franzluebbers, K., Weaver, R.W., Juo, A.S.R., Franzluebbers, A.J.,

1994. Carbon and nitrogen mineralization from cowpea plants

part decomposing in moist and in repeatedly dried and wetted

soil. Soil Biology & Biochemistry 26, 1379±1387.

Haimi, J., Huhta, V., 1990. E�ects of earthworms on decomposition

process in raw humus forest soil. A microcosm study. Biology

and Fertility of Soils 10, 178±183.

Hindell, R.P., McKenzie, B.M., Tisdall, J.M., Silvapulle, M.J.,

1994a. Relationships between casts of geophagous earthworms

(Lumbricidae, Oligochaeta ) and matric potential. 1. Cast pro-

duction, water content, and bulk density. Biology and Fertility of

Soils 18, 119±126.

Hindell, R.P., McKenzie, B.M., Tisdall, J.M., Silvapulle, M.J.,

1994b. Relationships between casts of geophagous earthworms

(Lumbricidae, Oligochaeta ) and matric potential. 2. Clay dis-

persion from casts. Biology and Fertility of Soils 18, 127±131.

Hindell, R.P., McKenzie, B.M., Tisdall, J.M., 1997. In¯uence of dry-

ing and ageing on the stabilisation of earthworm (Lumbricidae )

casts. Biology and Fertility of Soils 25, 27±35.

Jenkinson, D.S., 1971. Studies on the decomposition of 14C labelled

organic matter in soil. Soil Science 3, 64±70.

Jensen, E.S., 1997. Nitrogen immobilization and mineralization

during initial decomposition of 15N-labelled pea and barley resi-

dues. Biology and Fertility of Soils 24, 39±44.

Killham, K., 1990. Nitri®cation in coniferous forest soils. Plant and

Soil 128, 31±44.

Ladd, J.N., Oades, J.M., Amato, M., 1975. Microbial biomass

formed from 14C, 15N-labeled plant material decomposing in soils

in the ®eld. Soil Biology & Biochemistry 13, 119±126.

Lawton, J., 1999. Are there general laws in ecology? Oikos 84, 177±

192.

Lavelle, P., 1978. Les vers de terre de la Savane de Lamto (CoÃ te

d'Ivoire), peuplements, populations et fonctions dans l'ecosys-

teme. Revue d'Ecologie et de Biologie du Sol 22, 463±472.

Lavelle, P., Martin, A., 1992. Small-scale and large-scale e�ects of

endogeic earthworms on soil organic matter dynamics in soils of

the humid tropics. Soil Biology & Biochemistry 24, 1491±1498.

Lee, K.E., 1983. The in¯uence of earthworms and termites on soil

nitrogen cycling. In: Lebrun, P., AndreÂ , H.M., de Medts, A.,

GreÂ goire-Wibo, C., Wauthy, G. (Eds.), New Trends in Soil

Biology. Ottignies-Louvain-la-Neuve, Dieu-Brichart, pp. 35±48.

Marinissen, J.C.Y., Dexter, A.R., 1990. Mechanisms of stabilisation

of earthworms casts and arti®cial casts. Biology and Fertility of

Soils 9, 163±167.

Martin, J.P., Haider, K., 1986. In¯uence of mineral colloids on turn-

over rates of soil organic matter. In: Huang, P.M., Schnitzer, M.

(Eds.), Interactions of Soil Minerals with Natural Organics and

Microbes. Soil Science Society of America, Madison, pp. 283±

304.

Martin, A., 1991. Short- and long-term e�ects of the endogeic earth-

worms Millsonia anomala (Omodeo) (Megascolecidae,

Oligochaeta) of tropical savannahs, on soil organic matter.

Biology and Fertility of Soils 11, 234±238.

McKenzie, B.M., Dexter, A.R., 1987. Physical properties of casts of

the earthworms Aporrectodea rosea. Biology and Fertility of Soils

5, 152±157.

Mortland, M.M., 1970. Clay-organic complexes and interactions.

Advances in Agronomy 22, 75±117.

Mortland, M.M., 1986. Mechanisms of adsorption of nonhumic or-

ganic species by clays. In: Huang, P.M., Schnitzer, M. (Eds.),

Interactions of soil minerals with natural organics and microbes.

Soil Science Society of America, Madison, pp. 59±76.

O'Brien, B.J., Stout, J.D., 1978. Movement and turnover of soil or-

ganic matter as indicated by carbon isotope measurements. Soil

Biology & Biochemistry 10, 309±317.

Oades, J.M., 1993. The role of biology in the formation, stabilization

and degradation of soil structure. Geoderma 56, 377±400.

Parkin, T.B., Berry, E.C., 1994. Nitrogen transformations associated

with earthworm casts. Soil Biology & Biochemistry 26, 1233±

1238.

Quemada, M., Cabrera, M.L., 1997. Temperature and moisture

e�ects on C and N mineralization from surface applied clover

residue. Plant and Soil 189, 127±137.

Scheu, S., 1993. Cellulose and lignin decomposition in soils from

di�erent ecosystems on limestone as a�ected by earthworm pro-

cessing. Pedobiologia 37, 167±177.

Shipitalo, M.J., Protz, R., 1988. Soil and water management and

conservation. Factors in¯uencing the dispersability of clay in

worm casts. Soil Science Society of America, Journal 52, 764±

769.

Shipitalo, M.J., Protz, R., 1989. Chemistry and micromorphology of

aggregation in earthworm casts. Geoderma 45, 357±374.

Sorensen, L.H., 1972. Stabilisation of newly formed amino acid

metabolites in soil by clay minerals. Soil Science 114, 5±11.

Stevenson, F.J., 1986. Cycles of Soil. Wiley, New York.

Statistical Analysis Software, 1992. SAS User's Guide: Statistics.

SAS Institute Inc, USA.

Swift, M.J., Heal, O.W., Anderson, J.M., 1979. Decomposition in

Terrestrial Ecosystems. In: Studies in Ecology, Vol 5. Blackwell

Scienti®c Publications, Oxford.

Tisdall, J.M., Oades, J.M., 1982. Organic matter and water-stable

aggregates in soils. Journal of Soil Science 33, 141±163.

Van Gestel, M., Ladd, J.N., Amato, M., 1991. Carbon and Nitrogen

mineralization from two soils of contrasting texture and microag-


gregate stability, in¯uence of sequential fumigation, drying and

storage. Soil Biology & Biochemistry 23, 313±322.

Van Gestel, M., Merckx, R., Vlassak, K., 1993. Soil drying and

rewetting and the turnover of 14C-labelled plant residues. First

order decay rates of biomass and nonbiomass 14C. Soil Biology &

Biochemistry 25, 125±134.

Wardle, D.A., Lavelle, P., 1997. Linkages between soil biota, plant

litter quality and decomposition. In: Cadisch, G., Giller, J.E.

(Eds.), Driven by Nature, Plant Litter Quality and

Decomposition. CAB International, Wallingford, pp. 107±124.

Zhang, H., Schrader, S., 1993. Earthworms e�ects on selection physi-

cal and chemical properties of soil aggregates. Biology and

Fertility of Soils 15, 229±234.


Documents

Temperature, wetting cycles and soil texture e•ects …directory.umm.ac.id/Data Elmu/jurnal/S/Soil Biology And...Temperature, wetting cycles and soil texture e•ects on carbon and