cnriinaguensis il...

I

i

t j

r

! S

o f f

a

e

i )

e

1.

rc'

.t

.1

n

3 1

I

Pedobiologia 43,835-841 (1999) Pedo Urban & Fischer Verlag biologia http:l/www.urbanfischer.de/joLlmals/pedo

Effect of exclusion of the anecic earthworm Martiodrilzls cnriinaguensis Jiménez and Moreno on soil properties and plant growth in grasslands of the eastern plains of Colombia

Thibaud Decaëns', Juan José Jiménez? and Patrick avellel il I Laboratoire d'Ecologie des Sols Tropicaux, I.R.D. / Unidad Suelos y Plantas, C.I.A.T., 32 Av. Varag- nat, 93 143 Bondy cedex, France 2 Departamento de Biología Animal I, Facultad de Biología, Universidad Complutense, 28040 Madrid, Spain

Accepted: 2s. April 1999

Summary. The effects of a deep-burrowing earthworm species, Mnrtìoclrillrs cnrilnnguetz- sis, on soil fertility and plant growth were assessed in two grasslands in the Eastern Plains of Colombia. This species aestivates deep in the soil during the dry season. Fibre glass netting was used to prevent colonisation of the top soil at the onset of the humid season. Effects of earthworm absence on soil properties and plant biomass were analysed with principal component analysis.

Experimental manipulation succeeded in preventing M. carimnguensis activity in the top soil. The larger part of the variance in our data (29.35 %) was explained by the effect of pasture type and age on soil structure and organic matter. A lesser part (10.8 l %) was due to the effect of legumes on soil physical properties. The effect of earthworm absence on soil properties and plant growth explained 16.50 % of the total variance. M. cnririznguerzsis biomass was associated with low soil compaction, high C contents, low AI saturation, high herbaceous biomass and low weed biomass. These results support the general knowledge of how earthworms can affect soil fertility and plant growth.

Key words: Anecic earthworms, soil fertility, soil ecology, tropical grasslands, earthworm manipulations

Introduction

Most studies of the impact of earthworms on soil have been conducted in short-term pot experiments (e.g. Pashanasi et al. 1992; Spain et al. 1992; Derouard et al. 1997). Despite the significant advances this approach has made in the understanding of the role of earthworms in soil processes, there is still a lack of understanding of how these organisms may influence ecosystem processes in in-sim field conditions (Bohlen et al. 1995).

Our objective was to assess an innovative method (Baker 1998) to selectively exclude anecic earthworms (sensu Bouché 1977) from small experimental units. The efficiency of the method is discussed in relation to its ability to eliminate the large anecic earthworm Marti-

Correspondence to: T. Decaëns, Present address: Laboratoire d'Ecologie, UFR Sciences, Université de Rouen, 7682 1 Mont Saint Aignan Cedex, France, E-nid: thibaud.decaens @univ-rouen.fr

odriliis cariningiieizsis Jiménez and Moreno from the top-soil of two grass-legume based pastures in the Eastern Plains of Colombia. At Carimagua, populations of this species are en- hanced in intensive pastures and dramatically depleted in annual crops (Jiménez et al. 1998a). The effects of the disappearance of this species on soil properties and plant growth was assessed and compared with the effects of, other environmental factors.

Materials and Methods

Stiirly site

Tlie study was carried out at the CIAT-CORPOICA research station of Carimagua, Eastern Plains, Meta, Colombia (4" 37' N, 7 I " 19'W). The climate is humid tropical with an annual mean temperature and rainfall of 26°C and 2300 nini respectively (CIAT data). Native vegetation is a patchwork of open savannas and gallery forests. Soils are well aggregated Oxisols with high acidity and very low chemical fer ti li ty.

Experinzentrrl plots at id desigil

This experiment was carried out i n two pastures cacli grazed by cottle with 2.0 nnimal unit. ha-[: (A) A 3 year-old pasture of ßrtrchitrria /iii/~iitlicol~~, Arcrcliis pintoi, Sty1o.strrithc.s ctipit(itti and Ccwtraceiria CICLI- t$Xiiirri; (B) A one year-old posture of friri icii /~r /ti~r.vhiiim and Art/chis piiltoi.

The background ecology of the earthworm species of Cariningun are available in Jiménez et al. ( I 998a, b). All species aestivate during the dry season. The anecic M. cririiiiti,~irc~~isi.s burrows deep into the soil profile (60-100 cm) and falls into diapause until the next wet scason (Jim6nez et al. 1998b). Our idea was to prevent individuals from returning to the top soil at the onset ol' the rains. As otlicr species aestivate in superficial soil layers, we expcctcd this experiment to have spccitïc effects on 121. carima- g~1erl.si.s.

Twclve small experimental units were established in each piisture at tlie end of the dry season (early March 1996) and submitted to grazing pressure. Tlie protocol is synthcsiscd i n Fig. I: (I) ii 40 X 40 X 30 cni soil monolith was dug out with a spadc; (2) the monolith sidcs (except the top sidc)l were wrap- ped up in a fibre glass netting of 0.5 min mesh; (3) the monolith was placed i n a hole, and surface earthworm movements were avoided by surrounding thc units with a buried metallic plate; (4) an area was delimited where individuals of M. rcrrirrin,pirc.,isi.~ (located by the presence of fresh casts, Jiménez et al. 1998b) were systematically controlled by injecting 0.01 g of active carbofuran in their burrow. Twelve plots were also established without fibre glass netting as controls. Plot location was randomly chosen in the field, avoiding the field edges.

Soil and vegetation were sampled before, 6 months and 18 months after the experiment was established (respectively October 1995, August 1996 and June 1997). At each date, four units were randomly chosen in each pasture. Saniples aimed at characterising vegetation (shoot and root biomass), soil properties (bulk density, aggregate mean weigh density, penetration resistance. sorptivity, infiltration, chemical properties) and earthworm (density and biomass), ant and termite populations (number of chambers). Methods were standardised methods recommended by the TSBF (Tropical Soil Biology and Fertility) program (Anderson & Ingram 1989). All saniples'were repeated 4 fold.

Statistics

The effect of the experimental manipulation on both earthworm biomass and the abundance of ant and termite chambers was tested using the Fisher PLSD test. Asymmetry of data frequency distribution was reduced by Box-Cox transformation (Sokal & Rohlf 1995) with the software Vernorm (Legendre &k Vaudor 1991). The Bonferonni procedure for nested tests was used and tlie adjusted 0.05,O.Ol and 0.001 significance levels were respectively: 0.008,0.002 and 0.0002.

A Principal Component Analysis (PCA) was performed with ADE software (Thioulouse et al. 1997) to extract the principal source of variance in the environmental data sets. We used a matrix of 40 lines (samples) X 23 columns (soil and vegetation variables). The biomass of each earthworm species and the abundance of ant and termite chambers were projected on the PCA axes as additional columns. Simple linear regressions were used to test the correlations between faunal parameters and PCA axes.

Ch SUI

Fig diil

Re:

On1 stuc cies 0.0 arth

i ass( ace( higl Ord two

onr

mas satu tent sam was gro!

P mes

I $

sed en- {a). as-

ta, ld :n al

i

1

I

b

,

...... ................. Fibre glass netting

Sampling surface,

Chemical surface

v ...Y ....... i ' 2 p

plate

Fig. 1. Dcscription of'the sitnipling procedure. Numbers correspond to the chronological order of the clill'erent m~unil~iilation steps, and refer to the fi111 descriptions in the text

Results

Only one individual oiM. carirriagcrensis was found in the experimental units at the end of the study. Due lo the large size of hf. cnrirr~iyuensis, slight decreases in the density of this species resulted in sharp decreases in overall earthworm biomass (from - 89 % to - 100 96, p < 0.01) (Fig. 2a, 2b). Differences observed for other earthworm species (Fig. 2a, 2b) and social arthropods (not shown) were not significant.

Axis 1 of the PCA accounted for 29.33 % of the total inertia (Table 1). The correlation ratio associated with the variables showed negative relationships between (i) high graminea, herbaceous and root biomass, high bulk density in the 5-10 cm layer, high penetration resistance, high pH and available nutrient content, and (ii) high aluminium saturation (Fig. 3a, Table 1). Ordination of the sample scores showed a separation between (i) samples taken in the first two dates and those taken later (Fig. 3b), and (ii) samples taken in the pasture A and in the pasture B (Fig: 3c). This axis was interpreted as the effect of pasture characteristics and age on soil physical and chemical properties.

Axis 2 accounted for 16.52 % of the total inertia (Table 1). It opposed (i) high weed biomass, high bulk density, high penetration resistance, high total N content and high aluminium saturation to (ii) high herbaceous (legume and graminea) and root biomass, high total C content and high mean diameter of soil aggregates (Fig. 3a, Table 1). It clearly segregated samples taken inside the exclusion units with those taken in the control units (Fig. 3d), and was defined as the effect of the exclusion of M. curimuguensis on soil fertility and plant growth.

Axis 3 accounted for 10.81 % of the total inertia. It was driven by the abundance of legumes in the plant cover and the soil hydraulic properties (sorptivity and infiltration) (Table 1).

Pedobiologia 43 (1999) 6 837

80

c! E 4 60 e x .- v

c .- a Ocnercdrilidae sp $ 40

20

P

€

d O

Om 6m- 6mt 18m- 18m+ ' Om 6m- 6m+ 18m- 18mt

Pasture A Pasture B

Fig. 2. Density (a) and biomass (b) of each earthworm species in the experimental units with (+) and without (-) netting, at each sampling date and in the two pastures (N = 8 for Om and 4 for the other sampling dates; ind. = individuals; fm = fresh matter; O ni = before; 6 m = 6 months; 18 ni = 18 months)

Om 6m- 6m+ 18m- 18m+ Om 6m- 6m+ 18m- 18mt

Pasture A Pasture B

Table 1. Relative inertia for cacti ol' the thrcc I'i;.st axes (in brackets) nntl correlation r a h nssocioted with the variubles inclutlcd in thc lorinciixil corrcsnondencc nnnlvsis ( ITA) kiriablcs Cotlc Asis 1 A.ri.s 2 /\.tis 3

(29.33 %I) (I 6.52 Yr) ( I0.S I %)

Vqettrtion clicrrcIctoisti~s Graininca biomnss Legume biomass Weecl bioinass % of legumes Herbaceous biomass Root bioniass Soil cheriiical properties (0-5 CIII)

Total C content Total N content Total P content AI saturation Ca content Mg content K content Soil physical properties Bulk density (0-5 cm) Bulk density (5-10 cm) Mean weigh diameter of aggr. (0-10 cm) Pen,etration resistance (0-15 cm) Penetration resistance (15-30 cm) Penetration resistance (30-60 cm) Penetration resistance (0-60 cm) Sorptivity Infiltration constant

PH

GU LI3 WB %L H ß R ß

PH C% N% TP Al Ca

K

BDO-5 BD5-IO

Ms

MWD PRO-I 5 PRI5-30 PR30-60 PRO-60 S A

-0.75 I -0.173 -0.086 0.080

-0.748 -0.707

-0.786 -0. I68 -0.188 -0.738

0.625 - -0.8 17 -0.803 -0.336

-0.281 -0.612 -0.054 -0.623 -0.65 1 -0.464 -0.703 -0.164 -0.174

-0.447 -0.039 -0.402 -0.709

0.492. -0.044 -0.101 -0.796 -0.464 -0.191 -0.500 -0.245

0.159 O. 1 14 -0.359 0.018

0.532 -0.089 O. I90 0.030 0.493 -0.360

-0.332 0.217 -0.347 0.262 -0.152 0.25 1

0.5 18 -0.131 0.5 15 -0.09 I

-0.377 0.167 0.538 0.205 0.499 -0.181 0.400 -0.198 0.579 -0.021

-0.171 -0.645 0.030 -0.579

838 'Pedobiologia 43 (1999) 6

This axis was interpreted as the effect of legumes on soil physical properties. Samples were not ordinated along this axis according to a logical scheme as was the case for the first two axes (not shown).

When projected on the plan defined by the two first axes of the PCA, the biomass of M. cn- riiiingiieizsis was negatively correlated with axis 2 (Fig. 3a). This negative relationship was confirmed by the regression coefficients (r = 0.63; p < 0.001). The biomass of other earth worm species and the abundance of social insect chambers were located near the axes inter- sections, showing that these variables were not correlated with any of them (Fig. 3a). No significant correlations were found between faunal biomass and the third PCA axis.

! flillt


Discussion

Field experiments are pertinent for evaluating earthworm effects on soil parameters and plant growth. This may be very useful in assessing the role of earthworms on primary production in both natural and managed ecosystems (Bohlen et al. 1995). In our study, preventing verti- cal movements of earthworm at an appropriate time and depth was an efficient method to eliminate a targeted anecic species.

The main factors influencing soil properties were pasture type and age. This may be due to plant specific effects, soil protection by the plant cover, litter quality and production, and the cumulative effects of animal trampling that may be different according to the history and the nature of pastures (Sánchez et al. 1991; Gijsnian 1996; Gijsanan & Thomas 1996). The effect of legumes on soil physical properties was of third importance. This may be explained by the effect of root channels that increase soil niacroporosity and permeability (Mytton et al. 1993).

Disappearance of M. crrrinznguensis resulted i n significant changes in soil properties and plant biomass, although this impact was of secondary importance when compared with pasture effect. The decrease in earthworm biomass was associated with soil degladation (in- creased soil compaction antl aluniinium saturation, decreased carbon content) and herbaceous biomass. Decrease of plant biomass, i n turn, favoured opportunistic weed species. These results are consistent with the general knowledge or how cxthworms can influence soil properties and plant growth (see reviews by Lee 19S5; La1 1988; Edwards & Bohlen 1996; La- velle 1997). I t demonstrates that the loss of otic species i n an earthworm coninlunity, when it is associated with an important decrease i n biomnss, may result i n significant loshes in ecosystem function. The rapidity and intensity of the effects obscrvctl i n our study, ho\vever, may havc been accentuated by intensive animal trampling. In nnturnl untlistiirbed conditions, earthworm impocts on soil pt~occsscs tire known to be pcrsistcnt on a largcr time scale (i.(>. se- veral nionths to years) (Blanchnrt 1992).

Earthworm populations are highly sensitive to land usc practiccs (sec, e.g. Lavelle &: Pas- hnnsi 1989; Recldy et al. 1995). At Carimagita, native species biomass is greatly incrcased in man-made pasturc or, i n contrast, clrnmaticnlly dccrcascd i n annual crops (DccaEns et al. 1994; Jiménez ct til. 1998a). Such impacts may hnvc tlctcrmin:ini conscqucnccs on soil fiinc- tion antl primary production (Lavcllc 1996; Lavcllc 1997). Attcnlion shoultl now be paid to managing ccirthworm populations i n tropical agroccosystcms i n orclcr to prol'it from their impacts on soil fertility and enhance the sustainability of agricultural procluction.

Acltnowledgenients

The authors thank R. J. Thomas, D. K. Friesen, E. Amézquita, C. G. Meléndez and N. Asakawa (CIAT) for technical and financial support, J. P. Rossi (I.R.D.) and two anonytnoLrs referees for useful revision of an early version of this paper.

References

Anderson, J. M., Ingrani, J. S. (eds.) ( I 989) Tropical Soil Biology and Fertility, a hindbook of methods - 2nd Edition. CAB International, Oxford, United Kingdom, 221 p.

Baker, G. H. (1998) The ecology, management, and benefits of earthworms in agricultural soils. with particular reference to Southern Australia. In: Edwards, C. A. (ed.) Earthworm Ecology. St. Lucie Press, Boca Raton, USA, pp. 229-257.

Blanchart, E. (1992) Restoration by earthworms (Megascolecidae) of the macroaggregate structure of a destructured savanna soil under field conditions. Soil Biology and Biochemistry 24, 1587-1594.

Bohlen, P. J., Parmelee, R. W., Blair, J. M., Edwards, C. A., Stinner, B. R. (1995) Efficacy of methods for manipulating earthworm populations in large-scale field experiments in agroecosystems. Soil Biology and Biochemistry 27, 993-999.

Bouché, M. B. (1977) Statégies lombriciennes. In: Lohm, U., Persson, T. (eds.) Soil Organisms as Component of Ecosystems. Ecological Bulletin (Stockholm), Stockholm, pp. 122-132.

840 Pedobiologia 43 (1999) 6

Decaëns, T., Lavelle, P., Jiménez, J. J., Escobar, G., Rippstein, G. (1994) Impact of land management on soil macrofauna in the Oriental Llanos of Colombia. European Journal of Soil Biology 30, 157-168.

Derouard, L., Tondoh, J., Vilcosqui, L., Lavelle, P. (1 997) Effects of earthworm introduction on soil processes and plant growth. Soil Biology and Biochemistry 29,541-545.

Edwards, C. A., Bohlen, P. J. (1996) Biology and Ecology of Earthworms - 3rd Edition. Chapman & Hall, London, United Kingdom, 426 p.

Gijsinan, A. J. (1996) Soil aggregate stability and soil organic matter fractions under agropastoral systems established in native savanna. Australian Journal of Soil Research 34,89 1-907.

Gijsnian, A. J., Thomas, R. J. (1996) Evaluation of some physical properties of an Oxisol after conver- sion of native savanna into legume-based or pure grass pasture. Tropical Grasslands 30,237-248.

Jiménez, J. J., Moreno, A. G., Deciiëns, T., Lavelle, P., Fisher, M., Thomas, R. J. (199Sa) Earthworm communities in nritive savannas and man-made pastures of the Eastern Plains of Colombia. Biology ancl Fertility of Soils 28, 101-1 IO.

Jiménez, J. J., Moreno, A. G., Lavelle, P., Decaëns, T. (1998b) Population dynamics and adaptive strategies of Mrrrriodrihs cori~~~nglrerlsis (Oligochaeta, Glossoscolecidae), a native species from the well-drained savannas of Colombia. Applied Soil Ecology 9, 153-1 GO.

Lal, R. (1 988) Effects of nincrofxina on soil properties in tropical systems. Agriculture, Ecosystems and Environnmts 24, 101-1 16.

Lnvcllc, P. ( 1996) Diversity of soil fnunn and ecosystem function. Biology International 33, 3-1 6. Lavelle, P. (1997) Faunal activities and soil processes: adaptive strategies that determine ecosystem

function. Advances i n Ecological Research 27,93-132. Lavclle, P., Pashnnsi, B. ( 19S9) Soil macrofauna and land management in Peruvian Amazonia (Yurinia-

guns, Loreto). Pedobiologia 33,283-409. Lec, K. E. (1985) Earthworm. Their Ecology ancl Relationships with Soils and Land Use. Academic

Press, Orlando, U.S.A., 41 1 p. Legenclrc, P., Vaudor, A. ( I99 I ) The R package: niultiditnentionil analysis, spatial an:ilysis. Université

de Montréal, Département des Sciences Biologiques, Montréal, Canada. Mytlon, L. R., Cresswell, A., Colbourn, P. (1993) Improvenient in soil structure associated with white

clover. Crass and Forage Science 4S, 84-90. Pnshonnsi, B., Melendez, G., Szott, L., L:ivelle, P. (1 992) Effect of inoculation with the endogeic earth-

worm Pofiroscoh core~/irur~r.s (Glossoscolecidae) on N availability, soil microbial biomass and the growth of three tropical fruit tree scedlings i n pot experiments. Soil Biology and Biochemistry 24,

Recldy, M. V., Kumar, V. P. K., Reddy, V. R., Balashouri, P., Yule, D. E, Cogle, A. L., Jangawad, L. S. (1 995) Earthworm biomass response to soil management in a semi-arid tropical AI fisol agroeco- systcm. Biology and Fertility of Soils 19, 317-321.

Sdnchez, P. A., Castilla, C. E.. Alegre, J. C. (1991) Grazing pressure effects on the pasture degradation process. In: TropSoil Technical Report, 1958-1 989, CIAT, Cali, Colombia, pp. 182-186.

Sokal, R. R., Rohlf, E J. (1995) Biometry: the principles and practice of statistics in biological research - 3rd Edition. W. H. Freeman and Company, New York, USA.

Spain, A. V., Lavelle, P., Mariotti, A. (1992) Stimulation of plant growth by tropical earthworms. Soil Biology and Biochemistry 24, 1629-1 633.

Thioulouse, J., Chessel, D., Doledec, S., Olivier, J. M. (1997) ADE 4: a multivariate analysis and gra- phical display software. Statistics and Computing 7,75-83.

1655-1 659.


cnriinaguensis il...

Documents

Transcript of cnriinaguensis il...