Serhat ALBAYRAK, Hüsamettin BALK˙IS & Melih Ertan Ç˙INAR

Shallow-water soft bottom macrozoobenthic communities from Edremit Bay (NE Aegean Sea)

Halil BOLU & ˙Inanç ÖZGEN

Life History and Biology of Diloba caeruleocephala (Figure of Eight) (Lepidoptera; Noctuidae)

DEKONINCK W, DE KONINCK H, BAUGNÉE J-Y & J-P MAELFAIT

Ant biodiversity conservation in Belgian calcareous grasslands : active management is vital

SGARDELIS

Is duration of organic management reflected on nematode communities of cultivated soils?

SNOEKS, Essetchi Paul KOUAMÉLAN & Gouli Gooré BI _{AN INTERNATIONAL JOURNAL PUBLISHED BY}

Volume 137 (2) – July 2007

ISSN: 0777-6276

Belgian Journal of Zoology

All manuscripts should be sent electronically to BJZ@ua.ac.be.

SUBSCRIPTION INFORMATION TO THE BJZ

Membership of the Royal Belgian Zoological Society includes a free subscription to the Belgian Journal of Zoology. Reduced annual dues exist for regular members, PhD students (up to four years after graduation) and undergraduate students. Membership includes the free subscription to the Belgian Journal of Zoology, as well as a subscription to the zoologist’s mailing list. Rates for annual dues can be found on the Society website (see below). Payments are accepted only in these currencies by cheque via a Belgian Bank or via an international money order directed to “Royal Belgian Society of Zoology, B-1000 Brussels” on account number 000-0049113-31 (Fortis Bank). For international transfers: Fortis account number: 001-1870002-15, IBAN: BE83 0011 8700 0215, BIC: GEBABEBB, Fortis Bank NV, Warandeberg 3, B-1000 Brussels (Belgium). Payments through VISA or MASTERCARD are no longer accepted.

Back issues/volumes and a price list are available on request to the treasurer (erik.verheyen@naturalsciences.be). For large orders a substantial reduction can be offered.

Membership and subscription Requests for exchange, and other library matters

Dr. Erik Verheyen (erik.verheyen@naturalsciences.be) Dr. Laurent Meese (laurent.meese@natuurwetenschappen.be) Treasurer Library Administrator Royal Belgian Institute of Natural Sciences Royal Belgian Institute of Natural Sciences Vautierstraat 29 Vautierstraat 29 B-1000 Brussels B-1000 Brussels Belgium Belgium

The Belgian Journal of Zoology is published with the financial support of the “Direction générale de l’Enseignement supérieur de la Recherche scientifique” and of the “Universitaire Stichting van België/Fondation Universitaire de Belgique”.

General information about the Society and the Belgian Journal of Zoology,
including detailed instructions for authors, can be found on
http://www.naturalsciences.be/institute/associations/rbzs_website/rbzs/

The Belgian Journal of Zoology publishes, in English or French, original papers, reviews and notes in all fields of zoology. Papers in French must include English translations of the title and of the abstract.

Manuscripts must contain significant new findings and must not have been published elsewhere nor be simultaneously under consideration for any other publication outlet. Single-species descriptions will not be accepted unless they have broader relevance such as implications for phylogeny or biogeography.

Authors benefit from a waiver of page charges for up to eight printed pages for members and six pages for non-members. Additional pages are charged to authors.

PREPARATION OF YOUR MANUSCRIPT

The page size of the Belgian Journal of Zoology is 17 x 24.6 cm, in two columns of 8,15 cm. Please keep this in mind, particularly with regard to the size of figures and tables.

Manuscripts must be submitted in electronic version only (double spaced, with numbered pages and margins of at least 2.5 cm), as well as the original figures and tables. Figures and tables should be kept separate from the text; the legends should be at the end of the manuscript. Manuscripts have to be submitted to the editorial office at BJZ@ua.ac.be.

Text.	Papers should be written in clear, concise language and consist of an abstract (summarizing the essential results and con
	clusions), introduction, material and methods, results and discussion. Sections and sub-sections should not be numbered,
	but authors may suggest a printing format to give structure to the article. The title must be sufficiently informative.
	Please provide a short running title and (additional) key words (separated by commas). The full first name (first name
	first) and the address of each author should be given, as well as the e-mail address of the first or the corresponding
	author. Keep following matter in mind: units should have the international format, and no space separates the value and
	the unit; do not use a space before interpunction, but do use a space after it.
Figures.	Figures are referred to in the texts by “Fig. n” (capital F and full stop) or “Figs n-m” (no full stop). Drawings should be
	clearly readable after reduction to page width, column width or in between. Photographs should be sharp and with good
	contrast. Label size should be adequate and the same in all illustrations after reduction. Magnification or reduction
	should be given by a scale bar. Photocopies of photographs or figures are not acceptable. Indicate the preferred location
	of each figure in the manuscript. An electronic version of all illustrations must be provided (see below). Inclusion of
	colour prints is at additional charge. Legends to the figures come on a separate page.
Tables.	The sizes of tables should take account of page width and length, or column width. Tables are referred to in the text by
	“Table n” (with capital T). The legend to a table is given on top of the table. Indicate the preferred location of each table
	in the manuscript.
Electronic version.	An electronic version of the text should be provided, exclusively by e-mail as an attached file, preferably in MSWord
	or RTF. Files of figures and tables should remain separated from the text file. Figure files should have the JPG, PICT
	or TIFF format and those scanned should have a resolution of at least 300 dpi. Tables and graphs may be sent as an
	Excel file. Do not incorporate figures in a Word document because this may result in loss of quality. All files should
	have the name of the first author and should begin with the MS number once it has been allocated after submission (see
	below).
Species names.	Species names and the names of genera should be in italics. Names of species (and names of higher taxa) should be fol
	lowed, on the first occasion of use, by the name of the author who described the species or introduced the taxon name
	(in lower case) and the year. This rule does not apply for names of plants. If a table of all the species names used in the
	text is provided, the author names can be mentioned there. Do not use the genus name to indicate a species or individ
	uals of a species.
References.	Author names in text citations should be in small capitals and in the following format: NAVAS (1996), LAUDER & LIEM
	(1980) for two authors or WILSON et al. (1987) for more than two authors. Consecutive references must be separated by
	a semicolon (;). All authors quoted in the text should be found in the reference list.
	Examples of literature citations:
	* Paper in a journal:NAVAS CA (1996). Implications of microhabitat selection and pattern of activity on the thermal
	ecology of high elevation neotropical anurans. Oecologia, 108: 617-626. (Abbreviation of the
	journal according to the “World list of Scientific Periodicals”.)
	* Paper in a book: MALLEFET J, VANHOUTTE P & BAGUET F (1992). Study of Amphipholis squamata luminescence.
	In: ALERA-LIACI & CANICATTI TI (eds), Echinoderm Research, L. Balkema, Rotterdam: 125-130.
	* BELLAIRS: BELLAIRS R (1991). Egg incubation: its effects on an embryonic development in birds and rep
	tiles. Cambridge University Press, Cambridge.

Unpublished reports, master theses, etc… are not considered publications and should be referred to as “unpublished data” or “personal communication” in the text. A report can be given in a footnote. Ph.D. theses are accepted in the reference list. Author names that only appear in the text after a taxon, do not appear in the reference list.

Short notes. Short notes do not exceed two printed pages (incl. figures and tables –approx. 10 000 characters minus the figures and references). They are the appropriate form for new findings of temporary importance, comments on papers published in the journal, important new records of fauna, curiosa, etc. They should be written as a continuous text without the various divisions of regular papers (and without an abstract). References for notes must be indicated in the text by numbers. Short notes will be reviewed as normal papers.

PROCESSING

Manuscripts should be sent to the editor. On receipt of the manuscript, the corresponding author is notified by e-mail, and will receive the number under which the manuscript has been registered. Please mention this MS# first in the subject of e-mails and in the names of all files subsequently sent to the editor. The MS is then sent to at least two referees, and a reply may be expected six to eight weeks after receipt. If the MS is accepted, the author is requested to prepare the final version, taking into account the remarks of the referees. Only one galley proof is sent as a PDF file, to the corresponding author. The proof must be carefully corrected and sent back without delay.

Publication delay is between three and nine months, but is highly dependent of the willingness of the referees and the collaboration of the authors.

Corresponding author : e-mail: serhatal@istanbul.edu.tr

INTRODUCTION

Benthic organisms including a variety of feeding modes can provide a link between substratum and water column predators. They are important food resources especially for demersal fishes. Since most benthic organisms are sessile or sedentary, the analysis of benthic communities gives a tool to determine the dimension of environmental stresses and if those stresses are natural or due to pollutants (POCKLINGTON & WELLS, 1992; PANCUCCIPAPADOPOULOU et al., 1999).

Macrobenthic communities inhabiting the Greek coasts of the Aegean Sea were intensively studied (ZARKANELLAS & KATTOULAS, 1982; KISELEVA, 1983; ZENETOS et al., 1991; SIMBOURA et al., 1998; PANCUCCI-PAPADOPOULOU et al., 1999). Researches at the Turkish coasts were mainly focused in İzmir Bay (ÖNEN, 1983; ERGEN et al., 1994; ERGEN & ÇİNAR, 1994; ÇİNAR et al., 1998).

Edremit Bay is one of the most important fishery regions of the Aegean Sea. It is located between the longitudes E26º34’34’’-E26º56’44’’ and latitudes N39º18’41’’-N39º34’45’’. This region supports dense populations of many demersal fishes and is also suitable for trawling (KOCATAŞ & BILECIK, 1992). However, increasing number of summer houses, light food industry and fisherman ports are major factors threatening the ecosystem of the bay.

The aim of this study is to assess macrozoobenthic species inhabiting different depths of Edremit Bay and their distributional patterns according to major environmental variables. The present study also analyzed the structure and ecological features of macrobenthic communities prevailing in the area where no detailed study has been previously performed on this subject.

MATERIALS AND METHODS

This study was conducted in Edremit Bay in October 2002 (Fig. 1). Macrobenthic samples were obtained by means of a 0.1m² Van Veen grab from four transects, namely A, B, C and D, perpendicular to the coast line. The bottom was sampled at five stations corresponding to 1, 5, 10, 20 and 30m depths at each transects. These four transects and five depths at the shores of different human population densities were chosen as representatives of the ecological situation of the bay. Transect A is located at the shore of most dense population throughout the year, transects B and C are at the shores of summer houses and transect D is at the shore of non-settled area. Three replicates were collected at each depth.

All benthic samples were sieved through a 0.5mm mesh sized sieve and then the retained material was fixed with 4% formaldehyde-sea water solution. In the laboratory, macrobenthic organisms were sorted into main taxonomic groups, and then identified to species level and counted.

Small portions of surface sediment were removed from grab samples for analysis of silt-clay percentage (FOLK, 1974). The bottoms of stations are sand or silty-clay, except for B2 and D2, where the bottom was covered with the phanerogames Cymodocea nodosa (Ucria) Ascherson and Posidonia oceanica (Linnaeus) Delile, respectively.

For the determination of temperature, salinity and dissolved oxygen of sea water, water samples were taken by means of a 3 liter water sampler just above the sea bottom. The temperature was measured by thermometer on the water sampler, salinity by Mohr-Knudsen method (IVANOFF, 1972) and dissolved oxygen by Winkler method (WINKLER, 1888).

A total of 139 macrozoobenthic taxa were identified during the course of this study. The number of species at stations ranged from 9 to 43 and the number of specimens from 43 to 579 (per 1m^-2). The most abundant taxonomic groups in terms of number of species and number of specimens were Polychaeta (44.6% of species, 42.7% of specimens) and Mollusca (26.6% of species, 37.3% of specimens), respectively.

Half of the total number of taxa (67 species) appeared only once in samples. Among these, Polychaeta was represented by a total of 30 species and was followed by Crustacea (17 species), Mollusca (13 species) and Echinodermata (7 species). Only three species can be classified as constant in the area; Nephtys hombergii occurred at 13 stations, Ampelisca diadema at 11 stations and Amphiura chiajei at 10 stations. Twenty species were common and others rare.

Species represented by high number of individuals in the area were Nephtys hombergii (maximum density: 87ind.m^-2), Corbula gibba (maximum density: 83ind.m²), Rissoa splendida (maximum density: 63ind.m^-2), Notomastus latericeus (maximum density: 60ind.m^-2), Nephthys incisa (maximum density: 57ind.m^-2), Processa sp. (maximum density: 50ind.m^-2), Mellinna palmata (maximum density: 40ind.m^-2), Nucula nitidosa (maximum density: 40ind.m^-2) and Ampelisca diadema (maximum density: 37ind.m^-2).

When values of biotic parameters according to depths and transects are examined, it can be seen that the lowest mean value of specimen density per 1m² was found at 1m depth (65±10.4 SE) and transect D (169±32.1 SE), the highest at 20m depth (299±113.1 SE) and transect C (278±106.5 SE). As for the mean number of species the lowest scores were also calculated at 1m depth (11±0.9 SE) and transect A (14±1.5 SE) and the highest at 5m (27±9.8 SE) and transect B (21±6.3 SE). The lowest mean Shannon-Weaver diversity index value was determined at 1m depth (3.1±0.1 SE) and transect A (3.3±0.1 SE), the highest at 5m depth (3.9±0.5 SE) and transects B and C (3.7±0.2 and 0.3 SE). The lowest mean evenness value was calculated at 10m (0.85±0.04 SE) and transect A (0.89±0.02 SE), and the highest at the deepest depth, 30m (0.93±0.01 SE) and transect C (0.93±0.01 SE). The phanerogames Cymodocea nodosa and Posidonia oceanica found at B2 and D2 (5 m) greatly increased the number of species and diversity index values estimated at 5m. The high values of standard error calculated at this depth were mainly due to the effect of these phanerogames.

The nMDS and cluster analysis (Fig. 2) based on the fourth-root-transformed species abundances in each station revealed six distinct species assemblages in the area.

TABLE 2

DISCUSSION

The Aegean Sea is considered as oligotrophic (MIHALATOU & MOUSTAKA-GOUNI, 2002), and such areas have low diversity index but high evenness index values (PANCUCCİ-PAPADOPOULOU et al., 1999). Samples taken in this study presented generally medium community diversity values whereas evenness values were high. Similar results were reported from South Evvoikos Gulf (NW Aegean Sea) by SIMBOURA et al. (1998), but higher diversity index values from Rhodes Island (SE Aegean Sea) by PANCUCCI-PAPADOPOULOU et al. (1999). Moreover, ERGEN & ÇINAR (1994) also determined higher diversity index values at the south-eastern Aegean Sea in comparison to the northern part of the Aegean Sea. PANCUCCI-PAPADOPOULOU et al. (1999) attribute this situation to the prevailing current regime sediment sorting and subtropical character of the southern Aegean Sea.

A positive correlation was determined between the number of individuals and depth (p<0.05). ZENETOS et al. (1991) and PANCUCCI-PAPADOPOULOU et al. (1999) reported a negative correlation between depth, and species number and abundance. However, ZENETOS et al. (1991) collected materials from 75-200m depths and indicated that the highest number of species occurred at the shallowest depth (75m) and the lowest at the deepest one (200m), whereas PANCUCCI-PAPADOPOULOU et al. (1999) obtained materials from 30-400m depths and they observed higher values of biotic variables mainly at depths between 30-70m. In addition, MACKIE et al. (1997) and ÇİNAR (2003b) concluded that the nature of substratum (sediment composition) and depth were the major environmental variables influencing the distribution of polychaetes from the southern Irish Sea and northern Cyprus, respectively. ÇİNAR (2003b) studied syllid compositions of habitats from 0 to 600m along the northern Cyprus and stated that the number of syllid species and individuals, and diversity and evenness index values were negatively correlated with depths. In our study, the mean number of species attained its maximum level at 5m because of the presence of the phanerogames and then gradually decreased to 30m depth. However, the mean diversity index values are fairly constant among depths.

Among the species encountered in this study, the polychaetes Lumbrineris latreilli and Notomastus latericeus, the amphipod Ampelisca diadema, and the bivalves Nucula nitidosa and Myrtea spinifera are known to be typical species of fine or mixed sediments; the polychaete Melinna palmata, and the bivalves Acanthocardia paucicostata and Corbula gibba are characteristic species of muddy bottoms (ZARKANELLAS & KATTOULAS, 1982; POPPE & GOTO, 1993; SIMBOURA et al., 1998). All of these species were generally represented by high number of individuals at the stations.

The macrobenthic community structure identified during this study indicates that a relatively undisturbed condition prevails in the area. A total of four stations had a diversity value between 4.1 and 5, twelve stations between 3.1 and 4, and four stations between 2.7 and 3. Soft bottom benthic habitats can be classified based on community diversity index as bad (0<H’≤1.5), poor (1.5<H’≤3), moderate (3<H’≤4), good (4<H’≤5) and high (H’>5) (SIMBOURA & ZENETOS, 2002). According to the above classification, 20% of the stations are in good condition, 60% in moderate and only 20% in poor condition. However, some polychaetes, which are considered as indicator species of semi-polluted or transitional zones of the Mediterranean Sea when they form high-density populations, such as Lumbrineris gracilis, L. latreilli and Monticellina heterochaeta (PEARSON & ROSENBERG, 1978; SIMBOURA et al., 1998; ERGEN et al. 2002), exist in the area. These species were frequently found at stations but generally no high abundance levels of these species were encountered. The bivalve Corbula gibba can attain high abundance at the periphery of the severely polluted environment (PEARSON & ROSENBERG, 1978; ZARKANELLAS & KATTOULAS, 1982). This species was represented by high number of individuals in the study and particularly dominated the stations A2, A3, B3 and B4, together with the other tolerant species Lumbrineris gracilis. Diversity index values also were lower than four at these stations. Other stations of transect A were partly dominated by species Lumbrineris gracilis, Notomastus latericeus and Abra alba, which can survive in organically polluted bottoms. Moreover, the lowest mean Shannon-Weaver diversity index value was determined within this transect. The above mentioned stations are located at the innermost region of the bay and seem to be affected by increasing human settlements and the food industry in the area.

ACKNOWLEDGEMENTS

This work was supported by the Research Fund of the Istanbul University, project number: 1614 / 30042001.

REFERENCES

CLARKE KR & WARWICK RM (2001). Change in marine commu

nities: an approach to statistical analysis and interpretation,

2^nd edn. PRIMER-E: Plymouth. ÇİNAR ME (2003a). Ecological features of Syllidae (Polychaeta)

from shallow-water benthic environments of the Aegean

Sea, eastern Mediterranean. Journal of Marine Biological

Associations of the United Kingdom, 83: 737-745.

149.

6: 107-118.

INTRODUCTION

Diloba caeruleocephala (L.) (Lepidoptera: Noctuidae: Dilobinae) is a polyphagus pest, with almond, apple, pear, plum, peach, apricot, and cherry among its most important hosts. First instars attack newly opened buds of the host, while older instars feed along the main leaf veins. Damage to fruit was documented by MAÇAN (1986). BODENHEIMER (1958) discussed the importance of this species to fruit trees in Turkey.

Diloba caeruleocephala is reported from all Europe, Lebanon, Israel, North Africa, the Russian states of federation and Asia (CAYROL, 1972; DOLLMAN, 1958; MÜLLER, 1953; POPOV, 1962). MAÇAN (1986) reported it from the southeastern and eastern Anatolia region of Turkey; and KANSU (1995) and NIZAMLIOĞLU (1961) documented its distribution to Ankara and Istanbul.

Because of its considerable economic importance and given that details of the life history are poorly documented, we studied the biology of this species in the laboratory. The results are presented in this paper.

MATERIALS AND METHODS

Diloba caeruleocephala adults (n=50) were captured in the vicinity of Diyarbakır in the autumn of 2002. Eggs were obtained by placing the females in a screen cage (40x30x30cm) with 15cm lengths of Amygdalus communis L. shoots. Newly laid eggs were removed daily, counted, and kept in petri dishes (10cm diameter) on moist filter paper. Larvae were fed with freshly cut host-plant leaves and the larvae were transferred to new leaves every second day. The development of larvae was observed daily, and shed larval head capsules were collected, measured and preserved in 70% ethyl alcohol. Pupae were harvested daily and transferred to a new cage (40x30cm) containing a potted host plant. The larvae and pupae were weighed individually, and pupal length was also measured. The colony was maintained under controlled laboratory conditions at 26^oC with 16: 8 (L: D) photoperiod. All data were subjected to analysis of variance (ANOVA), and means were separated using Fisher’s Least Significant Difference (LSD) (P<0.05).

RESULTS

Fig. 1. – A composite photo of egg stages of Diloba caeruleocephala (Scale=4X)

TABLE 1 the underside of the leaf, typically spinning some silk web _{Measurements of head capsule length,} on the leaf. They chew a small amount of leaf tissue on _{weight, and length of each larval instar} the underside of the leaf, creating a small pit, which is _{of Diloba caeruleocephala (n=15) and} gradually enlarged as the larvae continue to feed.

Means within a column followed by a different letter are different (P<0.05) (Fisher’s Least Significant Difference tests).

The first instar is light brown to dark brown, with long setae over the body (Fig. 2). The head capsule in the first instar has many long setae brownish in colour with two large brown patches. The legs and prolegs are light brown and tarsal segments were black to brown. The anal prolegs are dark brown. Antennae are black, with dark black basal area. The larvae eat their eggshells and aggregate on

The second instar is black with yellow subdorsal bands. Each segment has a row of short, branching small spines. The head is cream with two black dorsal stripes extending posteriorly (Fig. 3).

Pupae

Pupae are initially soft and light tan in colour with speckled black and brown markings (Fig. 4).

The 10th segment of the pupa bears the cremaster by which the pupa attaches to a substrate. The pupae are in cocoons made of soil or leaf debris (Fig. 5). They meas-

Fig. 6. – The female and male of Diloba caeruleocephala (Scale=1X)

The wingspan is 38±0.4mm in females and 34±0.5mm in males (n=25). Mating pairs often rest together for 4-6 hours. Females started laying eggs about 3 days after mating and each laid from 130-174 eggs (n=25). Adults survived about 2 weeks in the laboratory.

The duration from egg to adult emergence is 27-35 days at 26^oC, 60±5% RH, 16: 8 (L: D) photoperiod in the laboratory.

ACKNOWLEDGMENT

REFERENCES

INTRODUCTION

The calcareous grasslands of northwestern Europe comprise the dry grasslands on limestone and chalk as well as those on calcareous loess. These grasslands originated after felling of primordial forest, which had already started in prehistoric times. Since these habitats are semi-natural, they require some form of management by grazing or mowing to prevent afforestation. They were once widespread in the hilly calcareous regions of Europe, but because of changes in agricultural practice their extent has decreased dramatically. Since World War II urbanisation, abandonment of grazing practice and fertilisation resulted in a dramatic decrease of these grasslands in Belgium. Everywhere in Europe this habitat type is severely pressured by threats that impinge upon several ecological communities. A shift in land use is one of the most important among these. Moreover the remaining areas are often strongly fragmented. This poses threats of extinction and inbreeding for the remaining characteristic arthropod populations (WALLISDEVRIES et al., 2002; VANDEWOESTIJNE et al., 2005).

Calcareous grasslands, especially the ones on dry slopes and plateaus, are known as being rich in plant diversity and they often contain many rare species (MITCHLEY & XOFIS, 2005). Regions which have this rich floristic diversity are often biodiversity hot spots for a wide variety of invertebrates. In Belgium the calcareous grasslands of the Viroin valley are such habitats (ADAM et al., 1991; HONNAY et al., 2004; BUTAYE et al., 2005a,b). They are catalogued as a Belgian hot spot for several groups, i.e. Heteroptera (DELESCAILLE et al., 1991), Saltatoria (DECLEER et al., 2000), butterflies (GOFFART & DE BAST, 2000; VANDEWOESTIJNE et al., 2005), bees and wasps (CLAESSENS, 1992; LONGO et al., 1992) and spiders (BARA, 1991).

In Belgium the ant fauna of these habitats is rather poorly investigated. Elsewhere in Europe calcareous grasslands and other similar thermophilic grasslands are known to have a rich ant fauna often including very rare species (MABELIS, 1983; GALLÉ, 1986; SEIFERT, 1996; MEYER-HOZAK, 2000; BRASCHLER & BAUR, 2003; 2005; OTTONETTI et al., 2006). In the northern part of the country (Flanders), calcareous grasslands are scarce and the one that was recently well studied at the Sint-Pietersberg, was indeed a hot spot of ant biodiversity (more than 25 species) in Flanders (DEKONINCK et al., 2003; 2005). In the southern part of Belgium (Wallonia), calcareous grasslands are more widespread but their ant fauna is poorly known. GASPAR (1965; 1971) investigated the habitat type in the Famenne region, whilst the Viroin valley has been studied by LEMAL (1978) and later by DELESCAILLE et al. (1991). But others have never been sampled for their ant fauna. Recently, two localities in the Viroin valley, respectively at Treignes and Vaucelles, were studied by BAUGNÉE (2004). He recorded Camponotus piceus (Leach, 1825) for the first time in Belgium and suggested that this interesting region still has a lot of entomological “secrets” to reveal. Hence, when a new pitfall sampling campaign on xeromorphic and rocky soils (Xerobromion=XB and Mesobromion=MB) in the Viroin valley was started, rare, extinct or new species for the Belgian ant fauna were expected. As different nature management types, microhabitats and vegetations were sampled, this study also enables us to evaluate the influence of the type or the lack of management in these grasslands on the occurring ant communities.

MATERIALS AND METHODS

The Viroin valley is located in southwestern Belgium, mainly in the municipality of Viroinval (province of Namur). The Viroin, a river of medium size (about 25km long and 5-10m width) flows between a succession of hills and joins with the Meuse near Givet in the French Ardennes. The geological formation dates from the Devonian period and occurs as limestone hills (‘Tiennes’ in French) in the landscape. In the Viroin valley two major types of calcareous grasslands are present: mesomorphic calcareous grasslands with five different plant communities, and xeromorphic grassland with three different plant communities (see HONNAY et al., 2004; BUTAYE et al., 2005a,b). The mesomorphic calcareous grasslands in the Viroin valley belong to the Mesobromion and show an intermediate position between the Central European and Atlantic calcareous grassland communities (BUTAYE et al., 2005a). The xeromorphic calcareous grassland communities on the other hand show strong affinities with the calcareous vegetation of Central Europe and southern Europe. The Viroin valley is a favourite region for entomologists and other naturalists because some of the most notable Belgian calcareous grasslands occur there (DUVIGNEAUD et al., 1990).

TABLE 1

Codes, locality, description, management, exposition, distance to other habitat and area of the sampled localities (with temperature: +=cold, ++=warm, +++=hot sites)

TABLE 2

Vegetation characteristics at each site (given in percentage ground cover or height in cm) (TdL=Tienne du Lion; Cha=Chalaine; FdC=Fondry des Chiens; RTr=Roche Trouée; TaP=Tienne aux Pauquis; TdR=Tienne des Rivelottes and Aba=Les Abannets).

Management RESULTS

In most calcareous grasslands in the Viroin valley, management measures for biodiversity conservation are ^{Faunistics and general results} of recent date. Only in three sites a continuous management is carried out (Table 1). In Roche Trouée and Cha-During the study 30 different species were found laine both situated in Nismes, short intensive summer (37.5% of the Belgian Formicidae (DEKONINCK et al., grazing is used to control tall grass encroachment and 2006)). Temnothorax albipennis was recorded for the first consists of 20-25 sheep that are put in a 25 x 25m mobile time in Belgium. The ant fauna composition of the differ-raster “as long as needed”. In Tienne de Rivelottes the ent sites is presented in Table 3. Lasius niger (Linnaeus, management regime during the last years was mowing in 1758) the most common Belgian ant, was not found dur-September. ing the project. Two species were found in each of the

TABLE 3

Species composition at the different sites (x=only found in isolated specimen or very small numbers, xx=abundant, xxx=very abundant) with Belgian status (CO=common ant species, RA=rare and VR=very rare ant species) and between () their habitat preference (Ca=calcareous grasslands and rocky habitats, Xe=xeromorphic grasslands and other hot habitats, Wo=woodlands, Bo=bogs, He=heathlands), (TdL=Tienne du Lion; Cha=Chalaine; FdC=Fondry des Chiens; RTr=Roche Trouée; TaP=Tienne aux Pauquis; TdR=Tienne des Rivelottes and Aba=Les Abannets).

Axis 1

Fig. 1. – Detrended Correspondence Analysis (DCA) diagram (Axis 1 and 2) with 14 most dominant species (percentages explained on Axis 1=58% and on Axis 2=22%). Species are presented with * and their codes (see Table 3) and the sites presented with ‹ in a different colour according to their management (TdL=Tienne du Lion; Cha=Chalaine; FdC=Fondry des Chiens; RTr=Roche Trouée; TaP=Tienne aux Pauquis; TdR=Tienne des Rivelottes and Aba=Les Abannets and WO=woodland, MB=Mesobromion and XB=Xerobromion).

The abundances of two species are significantly positively correlated with the first axis: Lasius platythorax (r=0.882, p<0.01) and Myrmica ruginodis (r=0.748 p<0.05). This axis is also positively correlated with the cover of grasses (r=0.901, p<0.01), cover of Brachypodium pinnatum (r=0.848 p<0.01) and with average vegetation height (r=0.817, p<0.05). Axis 1 is significantly negatively correlated with the cover of herbs (r=-0.883, p<0.01) and the abundance of Lasius alienus (r=-0.813, p<0.05). No species or environmental variable is significantly correlated with the second axis. We can conclude that the main differences in species composition are explained by the gradient going from xeromorphic towards mesomorphic environmental characteristics as was also found by BUTAYE et al., 2005a for plants.

Indicator species with a significant (p<0.05) IndVal higher than 50, were searched for in a posteriori groups of pitfalls. Indicator species for a group of pitfalls with the same management regime (3 regimes: No management, Short intensive summer grazing and mowing in September) and for the groups obtained by DCA (Group 1, 2 and 3) are presented in Table 4. For Group 1 from the DCA, five common less thermophilic ant species typical for forest, have a significant IndVal higher than 50: Formica pratensis, Myrmica ruginodis, Lasius platythorax, Lasius fuliginosis and Leptothorax acervorum. For the Xerobromiums and Mesobromiums we can distinguish a group (Group 2) with two common (Myrmica sabuleti and Tetramorium impurum) and one rare xeromorphic species (Myrmica schencki) as indicator species and a second group indicator species (for Group 3) with mostly rare and very rare xeromorphic ant species as Aphaenogaster subterannea, Lasius jensi and Myrmecina graminicola and the common ant Lasius mixtus.

For groups of pitfalls with the same management regime we also found indicator species. Lasius platythorax, an indicator for the less thermophilic ant species group, seems also an indicator for no management activity. Seven rare and xeromorphic species also appear to react positively to short intensive summer grazing management. Lasius jensi, Lasius umbratus, Aphaenogaster subterranea, Stenamme debile and Temnothorax nylanderi have a significant IndVal higher than 50 for the management regime mowing in September. However as this management regime was only present in one site, these results have to be interpreted with caution.

DISCUSSION

During this study 30 ant species were recorded. If some other neighbouring and not calcareous grassland localities (for example anthropogenic localities, dense forests, …) would have been sampled too, this list would have even been longer, e.g. Lasius brunneus (Latreille, 1798) and Lasius niger are very common and widespread species in their typical habitats in this region. Temnothorax nigriceps Mayr, 1855; Solenopsis fugax (Latreille, 1798); Camponotus ligniperda (Latreille, 1802) and Lasius emarginatus (Olivier, 1792) are also known for the region, but their colonies are only locally found (LEMAL, 1978; DELESCAILLE et al. 1991). Some other ant species were only recently found in the Viroin region (BAUGNÉE, 2002; 2004): Plagiolepis vindobonensis Lomnicki, 1925; Camponotus piceus and Temnothorax parvulus (Schenck, 1852). Until now, a total of 40 ant species is known from the region. This is almost half of the Belgian ant fauna (DEKONINCK et al., 2006). Hence, the region can be considered as a hot spot for ant biodiversity in Belgium.

Lasius niger, one of the most common Belgian ants, was not found during the sampling campaign. In the xeroand mesomorphic grassland sites studied here, this eurytopic ant species is replaced by another, less common Lasius s. str. species: Lasius alienus. Like L. alienus, also Tapinoma erraticum is found on each of the sites in high abundance. Both species can be considered as common for the region. They are typical for calcareous and other rocky habitats in Wallonia and are almost not present in Flanders (BOER et al., 2003; DEKONINCK et al., 2003; 2005).

Not only in this particular region, but also elsewhere in Belgium and Europe, it is likely that in calcareous grasslands with an increasingly dense and high grass layer due to a lack of or an inadequate management, endangered xerophilic ant species will be replaced be mesophilic, rather common species. As management to restore the former open xeromorphic grasslands sites by mowing, grazing and selective clearing started only recently in the Viroin valley, two types of ant monitoring can be very useful. Regularly mapping the nests of Formica species and typical xerophilic ant species that can be recognised in the field as Aphaenogaster subterranean and Temnothorax interruptus and also monitoring by pitfall trapping (3-6 pitfalls during 6 months May-October) for all xeromorphic and rare species that can not be recognized in the field, can be good instruments to evaluate management effects (BRASCHLER & BAUR, 2003; 2005; ENGLISCH et al., 2005). This may also give us more insights in the habitat preferences of the many rare ant species occurring there. We therefore recommend starting an ant monitoring of these grasslands of European importance.

ACKNOWLEDGEMENTS

REFERENCES

microspatial distribution of insects: a case study of carabids

in an alluvial forest. Ecography, 24: 470-482.

138: 133-134.

40: 301-314.

58.

28: 651-658.

CLAESSENS C (1992). Faunistique comparée et écologie des Hyménoptères Apoïdes de Treignes (Belgique, province de Namur). Mémoire de Licence, Université de Mons-Hainaut.

DAUBER J & WOLTERS V (2005). Colonization of temperate grasslands by ants. Basic and Applied Ecology, 6: 83-91.

DECLEER K, DEVRIESE H, HOFMANS K, LOCK K, BARENBRUG B & MAES D (2000). Voorlopige atlas en Rode Lijst van de sprinkhanen en krekels van België (Insecta Orthoptera). Werkgroep Saltabel, I.N en K.B.I.N., rapport Instituut voor Natuurbehoud 2000/10, Brussel, 75 pp.

DEKONINCK W & VANKERKHOVEN F (2001). Eight new species for the Belgian ant fauna and other remarkable recent records (Hymenoptera Formicidae). Bulletin de la Société royale belge d'Entomologie, 137: 36-43.

DEKONINCK W, VANKERKHOVEN F & MAELFAIT JP (2003). Verspreidingsatlas en voorlopige Rode Lijst van de mieren van Vlaanderen. Rapport van het Instituut voor Natuurbehoud 2003/07, Brussel 191 pp.

DEKONINCK W, BOER P & MAELFAIT JP (2004). Lasius platythorax Seifert, 1991 as a host of several Chthonolasius species, with remarks on the colony foundation of the parasites (Hymenoptera: Formicidae). Myrmecologische Nachrichten,

6: 5-8.

DEKONINCK W, MAELFAIT JP, VANKERKHOVEN F & GROOTAERT P (2005). Remarks on the distribution and use of a provisional red list of the ants of Flanders (Formicidae, Hymenoptera). In: PROCTER D & HARDING PT (eds), JNCC Report No. 367 Proceedings of IN Cardiff 2003, Red Lists for Invertebrates: their application at different spatial scales – practical issues, pragmatic approaches, 74-85.

DEKONINCK W, MAELFAIT JP, VANKERCKHOVEN F, BAUGNÉE JY & GROOTAERT P (2006). An update of the checklist of the Belgian ant fauna with comments on new species for the country (Hymenoptera, Formicidae). Belgium Journal of Entomology, 8: 27-41.

DELESCAILLE LM, HOFMANS K & WOUÉ L (1991). Les réserves naturelles du Viroin. Parcs Nationaux XLVI, 4-68.

DESENDER K & MAELFAIT JP (1986). Pitfall trapping within enclosures: a method for estimating the relationship between the abundance of coexisting carabid species (Coleoptera, Carabidae). Holarctic Ecology, 9: 245-250.

DUFRÊNE M & LEGENDRE P (1997). Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs, 61: 345-366.

DUTOIT T & ALARD D (1997). Mineral contents of chalk grasslands in relation with sheep grazing involved in conservation management systems. Revue d’Ecologie, 52: 9-20.

DUVIGNEAUD J, SAINTENOY-SIMON J, WOUÉ L, CLESSE B, DEWITTE T & HOFMANS K (1990). L'herborisation de la Société royale de Botanique de Belgique à Roly et dans le parc naturel Viroin-Hermeton, le dimanche 26 juin 1988: les problèmes de gestion des pelouses thermophiles. Belgian Journal of Botany, 123: 45-62.

ENGLISCH T, STEINER FM & SCHLICK-STEINER BC (2005). Fine-scale grassland assemblage analysis in Central Europe: ants tell another story than plants (Hymenoptera: Formicidae; Spermatophyta). Myrmecologische Nachrichten, 7: 61-67.

FISCHER SF, POSCHLOD P & BEINLICH B (1996). Experimental studies on the dispersal of plants and animals on sheep in calcareous grasslands. Journal of Applied Ecology, 33: 1206-1222.

GALLÉ L (1986). Habitat and Niche Analysis of Grassland Ants (Hymenoptera: Formicidae). Entomologia Generalis, 11 (3/ 4): 197-211.

GASPAR C (1965). Etude myrmécologique d’une région naturelle de Belgique: la Famenne. Survey des fourmis de la région (Hymenoptera, Formicidae). Bulletin de l’Institut agronomique et de la Station de Recherche de Gembloux, 32 (4): 427-434.

GASPAR C (1971). Les fourmis de la Famenne. Bulletin de l’Institut royal des Sciences naturelles de Belgique, 47(20): 1-116.

GOFFART P & DE BAST B (2000). Atlas préliminaire des papillons de jour de Wallonie. Publication du Groupe de Travail Lépidoptères, 80 pp.

GREENSLADE PJM (1964). Pitfall trapping as a method for studying Carabidae (Coleoptera). Journal of Animal Ecology, 33: 301-310.

HALSALL NB & WRATTEN D (1988). The eficiency of pitfall traps for polyphagous predatory Carabidae. Ecological Entomology, 13: 293-299.

HONNAY O, BUTAYE J, ADRIAENS D & DE HERT K (2004). De kalkgraslanden van de Viroinvallei: plantengemeenschappen en de invloed van beheer en habitatversnippering. Natuur.focus, 3(4): 135-141.

LEHOUCK V, BONTE D, DEKONINCK W & MAELFAIT JP (2004). The distribution of ant nests (Hymenoptera, Formicidae) in coastal grey dunes of Flanders (Belgium) and their relation to myrmecochorous plants. Belgian Journal of Zoology, 134(1): 95-102.

LEMAL C (1978). Inventaire des Formicidae de la région de Treignes (Insectes Hyménoptères) et préliminaires à leur étude écologique. Mémoire de Licence, Université Libre de Bruxelles, 95 pp + annexes.

LONGO F, WAHIS R & BARBIER Y (1992). Les Hyménoptères vespiformes des pelouses sèches de Treignes. Les Cahiers des Réserves Naturelles, 1: 67-71.

MABELIS AA (1983). De invertebratenfauna van de Zuidlimburgse kalkgraslanden. Mieren (Hymenoptera: Formicidae)

II. Natuurhistorisch maandblad, 72: 33-37.

MAELFAIT JP (1996). Spiders as bioindicators. Bioindicator Systems for Soil Pollution. In: VAN STRAALEN NM & KRIVOLUTSKY DM (eds), pp. 165-178, Kluwer Academic Publishers, Dordrecht.

MAELFAIT JP & BAERT L (1975). Contribution to the knowledge of the arachno- and entomofauna of different woodhabitats. Part I. Sampled habitats, theoretical study of the pitfall method, survey of the captured taxa. Biologisch Jaarboek Dodonaea, 43: 179-196.

MC CUNE B & MEFFORD MJ (1999). PC-ORD. Multivariate Analysis Ecological Data, Version 4, MjM Software Design, Gleneden Beach, Oregon, USA, 237 pp.

MC CUNE B & GRACE JB (2002). Analysis of Ecological Communities. MjM Software Design, Glenden Beach, Oregeon, USA, 300 pp.

MEYER-HOZAK C (2000). Population biology of Maculinea rebeli (Lepidoptera: Lycaenidae) on the chalk grasslands of Eastern Westphalia (Germany) and implications for conservations. Journal of Insect Conservation, 4: 63-72.

MITCHLEY J & XOFIS P (2005). Landscape structure and management regime as indicators of calcareous grassland habitat condition and species diversity. Journal for Nature Conservation, 13: 171-183.

OTTONETTI L, TUCCI L & SANTINI G (2006). Recolonization Patterns of Ants in a Rehabilitated Lignite Mine in Central Italy: Potential for the Use of Mediterranean Ants as Indicators of Restoration Processess. Restoration Ecology, 14: 60

66.

RETANA J & CERDÁ X (2000). Patterns of diversity and composition of Mediterranean ground ant communities tracking spatial and temporal variability in the themal environment. Oecologia, 123: 436-444.

SCHLICK-STEINER BC, STEINER FM & STEINER HM (2005). Effect of Extensification of Coppice Management on Central European Ant Assemblages (Hymenoptera: Formicidae). Entomologia Generalis, 28(1): 23-37.

Institute for Biological Research “Sinisa Stankovic” Bulevar despota Stefana 142, 11060 Belgrade, Serbia.

Corresponding author : E-mail: miravel@ibbi.ibiss.bg.ac.yu; Tel: (+381-11) – 615-043; Fax: (+381-11) – 2761-433

ABSTRACT. In the present study, an additional combination of end-points was used to evaluate the effects of industrial development on the natural population of the black-striped mouse (Apodemus agrarius), previously estimated using cytogenetic and morphometric assays. Developmental stability was assessed by determining the level of fluctuating asymmetry (FA) and the total amount of phenotypic variability (PV). Body weight (BW) and body length (BL) were also used as indices of body size, while the quotient between observed and expected values of body weight presented an index of body condition (BCI). The experimental design employed in this study is unique because only a few studies have been conducted with A. agrarius and because no investigator has previously used this combination of end-points in an environmental quality monitoring study. FA was increased in the polluted area for foramen parietalis, foramen dentale, foramen palatinum and foramen angularis. Males from the polluted site had significantly higher FA for foramen parietalis compared with male mice from the reference area. Juvenile animals from the polluted area had significantly higher FAs for foramen dentale and foramen palatinum compared to adults. However, when all the foramina were considered together, a three-way analysis of variance revealed that there was no significant interaction between the factors (sex x age category x site) on the FA (average value). A comparison of PVs indicated higher values in the polluted area for all analyzed characters. For foramen parietalis and foramen dentale, PVs were significantly greater in the polluted site than in the control area. Within the same sex, PVs were significantly greater in mice from the polluted site compared with mice from the control area. The results also indicated that mice trapped at the contaminated site had a reduced body size and poorer body condition. Adult mice exhibited better body condition than juveniles, as revealed by the significantly higher BCI values. Finally, to investigate the relative importance of the three potential biomarkers, a multivariate analysis of variance (MANOVA procedure) was performed. The MANOVA’s results revealed the significant effects of age category, site and interaction for sex x site on FA (average value), size and body condition. In conclusion, the results indicated that despite A. agrarius’s high tolerance to contaminants it may be an important species for environmental quality evaluation studies.

KEY WORDS : body size, body condition, developmental stability, fluctuating asymmetry, foramina.

INTRODUCTION

Typically, the processes of evaluating effects of polluted environments on natural populations of small mammals have been based on a single type of end-point (e.g. cytogenetic, morphometric or morpho-physiological techniques).

Mice from the genus Apodemus have been shown to be relevant pollution bioindicators (BERRY, 1975; ABRAMSSON-ZETTERBERG et al., 1997; STOPKA & MCDONALD, 1999; ADRIAN et al., 2002; DAMEK-POPRAWA, 2003; IEARDI et al., 2003; METCHEVA et al., 2001; TOPASHKAANCHEVA et al., 2003). The consequences of such exposure on the biology of animals have been assessed using cytogenetic assays, such as: micronuclei frequencies (ABRAMSSON-ZETTERBERG et al., 1997; IEARDI et al., 2003), chromosomal aberrations frequencies (METCHEVA et al., 2001; TOPASHKA-ANCHEVA et al., 2003; VELICKOVIC, 2004) and sperm abnormalities (IEARDI et al., 2003) or morpho-physiological assays, such as: morphological (ADRIAN et al., 2002; DAMEK-POPRAWA, 2003), morphometric (BERRY, 1975; NUNES et al., 2001b; STOPKA & MCDONALD, 1999; VELICKOVIC, 2004) and hematological parameters (NUNES et al., 2001a; TOPASHKA-ANCHEVA et al., 2003).

In the present study an additional combination of endpoints was used to evaluate the effects of industrial development on the natural population of the black-striped mouse (Apodemus agrarius, Pallas 1771) previously estimated using cytogenetic and morphometric assays (VELICKOVIC, 2004). Here, besides measuring developmental stability (DS), I also examined the consequences of disturbed environment on body size and body condition.

Developmental stability (DS), (MATHER, 1953; THODAY, 1955; or “developmental homeostasis” LERNER, 1954) refers to the ability of an individual to produce a consistent phenotype in a given environment (GRAHAM et al., 1993). Reduced DS can result from wide variety of environmental and/or genetic perturbations (VALENTINE & SOULÉ, 1973; VALENTINE et al., 1973; SIEGEL & DOYLE, 1975a; 1975b; 1975c; SIEGEL et al., 1992; YABLOKOV, 1986; CLARKE, 1992; 1993).

Fluctuating asymmetry (FA), directional asymmetry (DA) and antisymmetry (AS) are the three recognized types of asymmetry in morphological traits. Each is characterized by a different combination of the mean and variance of the distribution of (R & L) differences (VAN VALEN, 1962; PALMER & STROBECK, 1986; PALMER, 1994), where R and L represent the measurement of the right and left sides of bilaterally symmetrical traits respectively.

MATERIALS AND METHODS

This study was performed at two areas in northern Serbia. The contaminated site (Pancevo) is the site of a large petrochemical and fuel storage complex. It is located approximately 15km northeast of the capital (Beograd) and includes an ammonia plant (founded in 1962), a factory for chemical fertilizers (founded in 1975), and a crude oil refinery (founded in 1968). Here it is necessary to indicate certain chemical compounds, their metabolites and unwanted by-products, because they represent some of the most dangerous chemical pollutants producing long-term negative effects on the environment, human health and living organisms. About 700 tons of liquid ammonia (NH₃) and 500 tons of urea (CH₄N₂O) are processed daily, and about 8x10³ tons of hydrochloric acid (HCl, 33%), 109,600 tons of vinyl chloride monomer (VCM, C₂H₃Cl), 200x10³ tons of ethylene dichloride (C₂H₄Cl₁₂), 4850x10³ tons of crude oil products are produced annually. At the petrochemical plant, chlorinated solvents such as trichloromethane, tetrachloromethane, trichloroethane, dichloroethene, trichloroethene, and others closely associated with the unwanted by-products of PVC (polyvinyl chloride) production, were found in both soil and groundwater samples (GOPAL & DELLER, 2002). Additionally, VCM is a human carcinogen known to cause liver and blood tumors. VCM released into the atmosphere is protected from polymerization and remains dangerous for a long period for the environment and to humans. The combustion of VCM generates cyclic dioxins that are internationally recognized cancer inducers. Soot is extremely carcinogenic; its toxicity is attributable to high concentrations of polycyclic aromatic hydrocarbons; they are metabolically transformed by aquatic and terrestrial organisms are into carcinogenic and mutagenic metabolites.

Statistical treatment
for developmental stability indices

Developmental stability was assessed using quantitative characters presented as a number of paired foramina in the skull. Foramina, the small openings for nerves and blood vessels, were counted macroscopically under a binuclear microscope (16x magnification) on the right and left sides of the cleaned skull without a knowledge of where the animal had been trapped; the skulls were cleaned in a dermestid beetle (Dermestes maculatus) colony.

The abbreviations for foramina are (Fig. 1): foramen parietalis (F₁), foramen dentale (F₂), foramen palatinum

(F₃), foramen alveolaris maxillaris (F₄), foramen mentale (F₅), foramen angularis (F₆).

This set of meristic (non-metric) characters was selected because it has been shown that meristic characters are amenable to analysis using quantitative genetic methods (FALCONER, 1989) and because developmental stability (measured as FA) has been assessed using the numbers of paired foramina in the skull (BERRY, 1969; ZAKHAROV & YABLOKOV, 1990; PANKAKOSKI et al., 1992; PERTOLDI et al., 1997).

Two indices were calculated separately for each character to describe developmental stability: (1) FA₄: index 4 of PALMER & STROBECK (1986), PALMER (1994) and (2) the total phenotypic variability of foramen numbers (PV), ZAKHAROV (1984).

(1)

FA4 (foramina) ^{= Var (R}i ^{- L}i⁾

(2)

PV (foramina) ^{= Var (r}i^+li^),

where (R_i - L_i) = signed asymmetry, R_i and r_i = value of the character on the right side of the skull, L_i and l_i = value of the character on the left side of the skull, Var = variance.

Before proceeding with the asymmetry analysis, statistical tests were carried out to detect the features confounding the analyses of FA (PALMER & STROBECK, 1986; PALMER, 1994) such as: directional asymmetry (DA) and antisymmetry (AS).

Firstly, the deviation from the normality of the (R_i -L_i) distributions was assessed using the Kolmogorov-Smirnov test of normality. These distributions were tested for significant skewness (g₁) and kurtosis (g₂) according to SOKAL & ROHLF (1981). Secondly, a one-sample t-test for a departure of the mean of (R_i - L_i) from an expected mean of zero was performed. Thirdly, for each character within samples a simple linear regression of absolute asymmetry ⏐(R_i - L_i)⏐on (R_i + L_i)/2 was assessed.

Because the FA and PV values are variances, statistical comparisons between samples can be detected by comparing the heterogeneity of variances. Although this was implemented by performing the F-test, a sequential Bonferonni correction (RICE, 1989) was applied to avoid “false” significant results. The Friedman test was used to compare the medians of variances.

For each mouse, the measurements of body size were recorded as body length (BL) calculated by subtracting the tail length from the total length (1mm accuracy) and body weight (BW) measured to the nearest 0.1mg (embryo weights were deducted from the body weight of pregnant females). The expected weight for a given length was obtained from a linear equation of the logarithms of BW and BL (in males and females, using an analysis of covariance for slopes). Subsequently, an index of body condition (BCI) was calculated for each mouse as the quotient between observed and expected values of BW, (NUNES et al., 2001a).

All these variables were checked for normal distribution using the Kolmogorov-Smirnov test of normality.

The analyses of the main effects of site, sex, age category and all possible interactions of these variables on BW, BL and BCI were performed using three-way ANOVA (analysis of variance) procedures.

RESULTS

For the three (R_i-L_i) distributions considered in this study, normality was rejected: (in Cer: foramen mentale d=0.374, P<0.01; in Pancevo: foramen mentale d=0.371, P<0.01 and foramen angularis d=0.165, P<0.05).

The results of the regression analyses between ⏐R_i-L_i⏐and (R_i+L_i)/2 were significant: at Cer (for foramen alveolaris maxillaris, r=0.400, P=0.000; for foramen mentale, r=0.747, P=0.000 and for foramen angularis, r=0.260, P=0.016) and at the Pancevo site (for foramen parietalis r=0.830, P=0.000; for foramen dentale r=0.677, P=0.000; for foramen palatinum r=0.273, P=0.027; for foramen alveolaris maxillaris r=0.399, P=0.001, and for foramen mentale r=0.628, P=0.000).

For cases of significant linear regression of absolute asymmetry ⏐(R_i - L_i)⏐on (R_i + L_i)/2 value, the index chosen to measure FA (FA₄) was recalculated as the value of (R_i-L_i) divided by (R_i+L_i)/2, index 6 (FA₆ of PALMER & STROBECK, 1986; PALMER, 1994)

where (R_i - L_i) = signed asymmetry, R_i = value of the character on the right side of the skull, L_i = value of the character on the left side of the skull, Var = variance.

To ensure the corrections’ successful implementation, I again correlated the corrected (R_i-L_i) values [(R_i-L_i)/ (R_i+L_i)] with ((R_i+L_i)/2).

The results of the t-test showed that there were no significant departure of the mean of (R_i-L_i) from the expected mean of zero in both samples (no DA asymmetry is present, all P>0.05). Three out of twelve (R_i-L_i) distributions were significantly skewed (two in Cer: for foramen parietalis, g₁=-35.671, P<0.01, and for foramen angularis g₁=2.145, P<0.05, and one in Pancevo: for foramen parietalis g₁=12.445, P<0.01). A significant platikurtic or bimodal distribution (-g₂ value) was not found; thus no antisymmetry is present.

Higher values of FA were found in the polluted area for all characters except for foramen alveolaris maxillaris. The F-test results also showed that in five out of six characters the variances describing FA differ significantly between the two areas (P_r=0.05). Specifically, for four out of these five characters (foramen parietalis, foramen den-tale, foramen palatinum, and foramen angularis), FAs were higher (developmentally less stable) in the contaminated area than in the reference area. For foramen mentale however, FA was significantly higher in the control area, Table 1.

TABLE 1

Comparison of fluctuating asymmetry (Asymmetry) and total phenotypic variability (Variability) in foramen numbers between the polluted (Pancevo=P_AN.) and control (Cer=C) areas.

N =sample size, NS =statistically not significant, Pr =revised probability values (results of the sequental Bonferonni correction), * = significant at P<0.05, ** = significant at P<0.01

Concerning age categories, a statistically significant difference in FA levels was observed between juvenile and adult mice from the polluted area. Juvenile animals had significantly higher FAs for two characters (F_foramen 8

Reference

Contaminated

Table 2. The average FA value in foramen numbers was calculated by dividing the sum of FAs by the total number of characters analyzed. ⁰

_{TABLE 2} Developmental stability indices

^{Results of the three-way analysis of vari-}Fig. 2. – Comparison of developmental stability indices ance relating the FA (average value) with (fluctuating asymmetry = FA and total phenotypic variability age category, sex and site. = PV) in A. agrarius in the polluted and control areas. The

When sexes were separated and tested for PV, the ^{Table 3.} results indicated a difference between the sites within When CVs were computed by age category, site and sexes. Within both sexes, the PVs were significantly sex, the highest value was detected for the index of body higher in mice from the polluted area compared with mice condition in adult males from the reference site from the unpolluted site (for males: F_{foramen parietalis} (CV=19.719), while the smallest value exists for body =7.354, P=0.008, and for females: F_{foramen dentale} =4.524, length in juvenile males from the polluted area P=0.038). (CV=3.351).

TABLE 3

Descriptive statistics for body measurements in Cer (reference) and Pancevo (contaminated) sites

N = sample size, Mean = mean value, S.D. = standard deviation, Min. = minimum, Max. = maximum, CV = coefficient of variation, BW = body weight, BL= body length, BCI = index of body condition

Linear equations of the logarithms of BW and BL were obtained for each sex separately: for males (y=2.083x– 6.555, r²=0.730, P=0.000) and for females (y=1.597x– 4.369, r²=0.709, P=0.000). The models obtained were highly significant for both sexes (P=0.000). However their expected power is low: 30.1% and 26.8% of the variance explained for males and females, respectively (Fig. 3a and Fig. 3b).

The effects of sex, age category, site and interactions of these variables on BW, BL and BCI were assessed by using the ANOVAs (Table 4). No effects of possible interactions of sex, age category and site on BW, BL and BCI values were observed. Significant differences were however detected in BW between age categories: (19.665_(g)vs. 15.574 _(g), for adults and juvenile individuals, respectively; F= 45.686, P=0.000).

In BL, significant differences were observed between sexes: (94.623 _(mm) vs. 92.677 _(mm), for males and females respectively; F=4.465, P=0.040), between age categories

(96.286 _(mm) vs. 88.717 _(mm), for adults and juvenile individuals, respectively; F=48.268, P=0.000) and between sites (95.520 _(mm) vs. 92.348 _(mm), in the control area and the polluted area, respectively; F=16.770, P=0.000).

BCI in adults was significantly higher than in juveniles

(6.663 vs. 5.589; F=38.783, P=0.000) and in mice from the reference site than in the polluted area (6.403 vs. 6.234; F=5.757, P=0.020).

* = statistically significant at P<0.05

Finally, to investigate the relative importance of the three potential biomarkers, a multivariate analysis of variance (MANOVA procedure) was performed. The MANOVA’s results revealed the significant main effects of age category and site on FA (average value), size and body condition, Table 5. The MANOVA’s results also showed significant interaction for sex x site. Significant test results for the MANOVA procedure were based on F statistics derived from Wilks’ lambda.

^*=P<0.05; ***=P<0.001

DISCUSSION

Several researchers have examined the variability in the numbers of paired foramina as an expression of developmental stability (measured as FA) and have found differences between natural populations (BERRY, 1969; ZAKHAROV & YABLOKOV, 1990; PANKAKOSKI et al., 1992). Specifically, these authors found that a disruption in developmental stability was observed in animals affected by different types of environmental pollutants (but see SCHANDORFF, 1997).

The results presented here demonstrated that the level of FA in the polluted area was increased for all foramina except for foramen alveolaris maxillaries, and the developmental stability of the same characters decreased.

The fact that FA levels were significantly different between sites within the same sex indicates that it is possible that certain pollutants affect sexes differentially. The study also indicated that the progeny of animals with long-term exposure to environmental pollutants had significantly higher levels of FA for foramen dentale and foramen palatinum. Thus, the increased FA in mammals seems to be a consequence of stress experienced by mothers and fetuses during pregnancy or lactation (SIEGEL & DOYLE, 1975b) and that the developmental disturbances occurred during the prenatal life or at the early stages of development. On the other hand, the use of energy for stress tolerance that could otherwise be used for developmental control, growth, reproduction, and survival (HOFFMAN & PARSONS, 1989; PARSONS, 1990; ALEKSEEVA et al., 1992; OZERNYUK et al., 1992) impose a distinct developmental stress.

Interestingly, in a 3-way ANOVA correcting for age and/or sex, no effect of site on average FA is found. One of possible explanations would be that there is no replication, and thus by definition the individuals within the two sampling sites are no real replicates.

It is worth remarking that the skulls of rodents have been frequently employed in FA studies; the lower jaw of a mouse has long served as a model system for studying the development and evolution of complex morphological structures (KLINGENBERG, et al., 2003). The measurements that are classically recorded for the mandible (mandible mass, mandible size, maximum width and length of the lower molars, the height and length of the mandible) indicated that increased FA levels in the lower jaw (and associated tooth) dimensions in mice responded to various kinds of stress (SIEGEL & DOYLE, 1975c; SIEGEL et al., 1977; LEAMY, 1984; ATCHLEY et al., 1984; PANKAKOSKI et al., 1992; GILEVA & NOKHRIN, 2001; VELICKOVIC, 2004). However for all of these studies suggesting that stress can increase FA, there are others where no effects, or even the opposite effects, of stress on FA have been observed (LEAMY et al., 1999a; MARKOW, 1995; WOODS et al., 1999). Thus, the results obtained in my studies and in other investigations corroborate the potential use of FA on both metrical and meristic characters, as a good diagnostic procedure for assessing stress tolerance. Concurring with this view, LEAMY & KLINGENBERG (2005) suggest that FA may play some kind of role as a fitness indicator.

The data presented here also indicates that both sexes had significantly higher PVs in the polluted area. Interestingly, when all foramina were considered together, the PV value was significantly higher in the polluted area than at the control site. This finding is consistent with similar results in an earlier study (PANKAKOSKI et al., 1992), whose authors revealed that the total phenotypic variance is greater in the common shrew (Sorex araneus, Linnaeus 1758) associated with heavy metal pollution than in the reference areas.

The present study therefore indicates that developmental stability is one of the most general characteristics in individual developmental processes and is also a sensitive parameter that can be used for the biomonitoring of natural populations’ conditions (SOULÉ, 1967; JONES, 1987; ZAKHAROV, 1981; 1987). An analysis of developmental stability may therefore potentially reveal changes in a population’s condition, the changes not yet reflected by disturbances of individual viability (YABLOKOV, 1986). Moreover, by studying developmental stability it might also be possible to assess the synergetic effects of toxic compounds or interactions between pollutants and other stressful factors (ZAKHAROV et al., 1991), which are either complex or impossible to study by other means.

Certain data from the body size and body condition analyses found in this study are similar to findings reported by NUNES et al. (2001a) for the Algerian mouse (Mus spretus, Lataste 1883) inhabiting an area contaminated with heavy metals. However, the data from M. spretus has been used more as a reference than for direct comparisons. Mice (from both studies) trapped at the contaminated sites have reduced BW as well as BL values. In NUNES et al. (2001a), all analyzed Mus spretus, except non-reproductive males, weighed less at the metal-polluted site compared with mice inhabiting the unpolluted area. The reduction of bone lengths and tooth dimensions in laboratory rats raised under effects of various stresses, such as extreme temperatures or noise, has been demonstrated by DOYLE et al. (1977). On the other hand, lead poisoning in laboratory rats results in decreased body weight (GOYER et al., 1970). Thus, the small sizes of mice found in contaminated areas may be explained by the negative effects of chemical pollutants from the environment. However, other causes, such as nutrition quality/availability, cannot be ruled out.

M. spretus and A. agrarius mice from the reference areas exhibited better body condition than mice from the polluted sites, revealed by the significantly higher BCI value. Furthermore, the same trend was observed for adults from the investigated localities; compared with juveniles, the adults exhibited better body condition than the juveniles. This observation can be explained as a consequence of normal ontogenetic development; because juveniles reach adult length before they reach adult mass, they will normally go through a period in which they are long relative to their mass. Another possibility may result from the deposition of fat reserves during the breeding season (MILLAR, 1975).

The present study indicated that developmental stability indices, combined with body size measures as well as an index of body condition can be used to detect, analyze and evaluate the effects of environmentally-generated stress factors on natural populations of small mammals. Therefore, I propose the use of this combination of endpoints as an ecotoxicological biomarker in pollution monitoring.

In conclusion, the results obtained from previous researches and the present study demonstrated that despite A. agrarius’s high tolerance to contaminants (PETROV, 1992), it may be an important species for environmental quality monitoring.

For future work, it would be interesting to compare the results of investigations performed on Apodemus agrarius with new data obtained from other species of the genus Apodemus and/or other species of small mammals indigenous to Serbia.

However, it would be important to consider the following: 1) to increase number of investigated localities in order to avoid pseudoreplication. According to HURLBERT (1984) pseudoreplication is defined as the use of inferential statistics to test for treatment effects with data from experiments where either treatments are not replicated (though samples may be) or replicates are not statistically independent 2) that each polluted site should be scored for quantitative variables such as types and concentrations of environmental pollutants 3) a larger sample size is necessary because increases statistical power (COHEN, 1988; YEZERINAC et al., 1992).

ACKNOWLEDGEMENTS

REFERENCES

140: 109-129.

GILEVA EA & NOKHRIN DYU (2001). Fluctuating asymmetry in cranial measurements of east European voles (Microtus rossiameridionalis Ognev, 1924) from the zone of radioactive contamination. Russian Journal of Ecology, 32: 39-44.

GOPAL S & DELLER N (2002). Precision Bombing, Widespread Harm: Two Case Studies of the Bombings of Industrial Facilities at Pancevo and Kragujevac during Operation Allied Force Yugoslavia, 1999. - Takoma Park, Maryland: Institute for Energy and Environmental Research, 11: 1-10.

GOYER R, LEONARD D, MOORE J, RHYNE B & KRIGMAN M (1970). Lead dosage and the role of the intra-nuclear inclusion body. Arch. Envir. Health, 20: 705-711.

GRAHAM J, FREEMAN D & EMLEN J (1993). Developmental stability: a sensitive indicator of populations under stress. In: LANDIS WG, HUGHES JS & LEWIS MA (eds), Environmental Toxicology and Risk Assessment, STP 1179. Philadelphia, PA, USA, pp. 136-158.

GREEN A (2001). Mass/length residuals: measures of body condition or generators of spurious results? Ecology, 82: 14731483.

HALLGRIMSSON B, WILLMORE K & HALL BK (2002). Canalization, developmental stability and morphological integration in primate limbs. Yearb. Phys. Anthropol., 45: 131-158.

HAYES J & SNOKWILER J (2001). Morphometric indicators of body condition worthwhile or wishful thinking? In: SPEAKMAN J (ed), Body composition analysis of animals: a handbook of non-destructive methods, Cambridge University Press, Cambridge: 8-38.

HOFFMAN A & PARSONS P (1989). Evolutionary Genetics and Environmental Stress, Oxford University Press, Oxford.

HURLBERT S (1984). Pseudoreplication and the design of ecological field design. Ecological Monographs, 54: 187-211, Ecological Society of America.

IEARDI LA, ZIMA J, ALLEGRA F, KOTLANOVA E, CAMPANELLA L, GROSSI R & CRISTALDI M (2003). Evaluation of genotoxic damage in wild rodents from a polluted area in Czech Republic. Folia Zoologica, 52: 57-66.

JAKOB E, MARSHALL S & UETZ G (1996). Estimating fitness: a comparison of body condition indices. Oikos, 77: 61-67.

JONES J (1987). An asymmetrical view of fitness. Nature, 325: 298-299.

KLINGENBERG CP, MEBUS K & AUFFRAY JC (2003). Developmental integration in a complex morphological structure: how distinct are the modules in the mouse mandible? Evol. Develop., 5: 522-531.

KOTIAHO J (1999). Estimating fitness: comparison of body condition indices revisited. Oikos, 87: 399-400.

LARCHER W (2000). Temperature stress and survival ability of Mediterranean scerophyllous plants. Plant Biosystems, 134: 279-295.

LEAMY L (1984). Morphometric studies in inbred and hybrid house mice. V. Directional and fluctuating asymmetry. American Naturalist, 123: 573-579.

LEAMY LJ, DOSTER MJ & HUET-HUDSON YM (1999a). Effects of methoxychlor on directional and fluctuating asymmetry of mandible characters in mice. Ecotoxicology, 8: 63–71.

LEAMY L & KLINGENBERG CP (2005). The genetics and evolution of fluctuating asymmetry. Annu. Rev. Ecol. Evol. Syst.,

36: 1-21.

LEARY R & ALLENDORF F (1989). Fluctuating asymmetry as an indicator of stress: implications for conservation biology. Trends Ecol. Evol., 4: 214-217.

LERNER IM (1954). Genetic homeostasis. Wiley, New York. MARKOV TA (1995). Evolutionary ecology and developmental instability. Annu. Rev. Entomol., 40: 105-120. MATHER K (1953). Genetical control of stability in development. Heredity, 7: 297-336. METCHEVA R, TEODOROVA S & TOPASHKA-ANCHEVA M (2001). A comparative analysis of the heavy metal and toxic ele

ments loading indicated by small mammals in different Bulgarian regions. Acta Zoologica Bulgarica, 53: 61-80.

MILLAR J (1975). Tactics of energy partitioning in breeding Peromyscus. Canadian Journal of Zoology, 53: 967-976.

MILLAR J & HICKLING G (1990). Fasting endurance and the evolution of mammalian body size. Func. Ecol., 4: 5-12.

MØLLER AP & SWADDLE JP (1997). Asymmetry, Developmental Stability, and Evolution. New York: Oxford University Press. p. 291.

NUNES A, AUFFRAY JC & MATHIAS M (2001a). Morphological and hematological parameters in the Algerian mouse (Mus spertus) inhabiting an area contaminated with heavy metals. Environmental Pollution, 113: 87-93.

NUNES A, DA LUZ MATHIAS M & CRESPO A (2001b). Developmental instability in a riparian population of the Algerian mouse (Mus spretus) associated with a heavy metal polluted area in Central Portugal. Archives of Environmental Contamination and Toxicology, 45: 515-521.

OZERNYUK N, DYOMIN V, PROKOFYEV E & ANDROSOVA I (1992). Energy homeostasis and developmental stability. Acta Zool. Fenn., 191: 167-175.

PALMER A (1994). Fluctuating asymmetry analyses: a primer in developmental instability: Its origins and evolutionary implications, In: MARKOW TA (ed), Developmental instability: its origins and evolutionary implications, Dordrecht, Kluwer: 335-346.

PALMER A & STROBECK C (1986). Fluctuating asymmetry: measurement, analysis, patterns. Ann. Rev. Ecol. Syst., 17: 391-421.

PANKAKOSKI E, VÄISÄNEN R & NURMI K (1987). Variability of muskart skulls: measurement error, environmental modification and size allometry. Sist. Zool., 36: 35-51.

PANKAKOSKI E, KOIVISTO I & HYVATINEN H (1992). Reduced developmental stability as an indicator of heavy metal pollution in the common shrew Sorex araneus. Acta Zool. Fenn.,

191: 137-144. PARSONS P (1990). Fluctuating asymmetry: an epigenetic measure of stress. Biol. Rev., 65: 131-145. PARSONS P (1992). Fluctuating asymmetry: a biological monitor of environmental and genetic stress. Heredity, 68: 361-364.

PERTOLDI C, LOESCHCKE V, BO MADSEN A & RANDI E (1997). Developmental instability in the Eurasian Otter (Luttra luttra) in Denmark. Ann. Zool. Fennici, 34: 187-196.

PETROV B (1992). Mammals of Yugoslavia – insectivores and rodents. Nat. Hist. Mus. Belgrade: 37. RADWAN J (2003). Inbreeding not stress increases fluctuating asymmetry in the bulb mite. Evolutionary ecology research,

5: 287-295. RICE W (1989). Analyzing table of statistical tests. Evolution,

43: 223-225. SCHANDORFF S (1997). Developmental stability and the harbor seal epizootic in 1988. Ann. Zool. Fennici, 34: 167-175.

SCHULTE-HOSTEDDE A, MILLAR J & HICKLING G (2001). Evaluating body condition in small mammals. Canadian Journal of Zoology, 79: 1021-1029.

SCHULTE-HOSTEDDE A, ZINNER B, MILLAR J & HICKLING G (2005). Restitution of mass-size residuals: validating body condition indices. Ecology, 86: 155-163.

SIEGEL M & DOYLE W (1975a). The differential effects of prenatal and postnatal audiogenic stress on fluctuating dental asymmetry. Journal of Experimental Zoology, 191: 211-214.

SIEGEL M & DOYLE W (1975b). The effects of cold stress on fluctuating asymmetry in the dentition of the mouse. Journal of Experimental Zoology, 193: 385-389.

SIEGEL M & DOYLE W (1975c). Stress and fluctuating limb asymmetry in various species of rodents. Growth, 39: 363

369.

SIEGEL P, SIEGEL MI, KRIMMER EC, DOYLE WJ & BARRY H (1977). Fluctuating dental asymmetry as an indicator of the stressful prenatal effects of ⁹ tetrahydrocannabinol in the laboratory rat. Toxicology and Applied Pharmacology, 42: 339

344.

I (1991). Developmental stability and population dynamics in the common shrew, Sorex araneus. American Naturalist,

138: 797-810.

¹ Laboratoire d’Hydrobiologie, UFR-Biosciences, Université de Cocody-Abidjan, 22 BP 582 Abidjan 22, Côte d’Ivoire

² Laboratoire d’Ichtyologie, Musée Royal de l’Afrique Centrale, 13, B-3080 Tervuren, Belgique et Laboratoire d’Anatomie Comparée et Biodiversité, Katholieke Universiteit Leuven, Belgique.

RÉSUMÉ. La présente étude a consisté à évaluer en cage flottante (1m³) l’effet de l’introduction d’un système de refuge (bandelettes plastiques) sur la survie et la croissance des alevins de Heterobranchus longifilis. Trois séries d’expériences ont été menées. La première [poids moyen initial (P_mi): 0.9±0.1g; densité: 100 individus/m³] a porté sur l’évaluation de l’effet de la coloration (blanche, bleue, noire) du système de refuge sur la survie et la croissance. Après 90 jours d’élevage, aucune influence (p>0.05) de la coloration des bandelettes qui constituent le système de refuge n’a été observé sur le taux de survie (Ts) et le poids moyen final (Pmf). Par contre, comparativement aux cages sans système de refuge, ces paramètres ont été nettement améliorés (p<0.05) avec l’utilisation du dispositif de refuge (respectivement 70.9-74.0% et 16.2-17.1g contre 64% et 11.8g pour les cages sans refuge). La deuxième expérience (durée: 90 jours; Pmi: 0.9±0.1g) a permis de suivre, pour les deux types de cages testées, l’effet de quatre densités de mise en charge (100, 200, 500 et 1000 individus/m³) sur la survie et la croissance. D’une manière globale, le système de refuge a produit les meilleurs résultats. La densité de 100 individus/m³ a donné les meilleurs taux de croissance et de survie. La troisième expérience (durée: 60 jours; densité: 100 individus/m³) a porté sur la détermination de l’effet du système de refuge et du poids moyen initial (13.2±1.3g, 25.2±2.5g, 40.3±4.0g et 60.3±6.0g) sur la survie et la croissance dans les deux types de cages testés. Les résultats de cette expérience ont permis de mettre en évidence deux types d’effets du système de refuge: (1) un effet positif chez les poissons de poids compris entre 13.2 et 29.2g et (2) un effet négatif chez les poissons de poids compris entre 60.3 et 92.0g.

MOTS CLÉS : Heterobranchus longifilis, système de refuge, survie, croissance

MATÉRIEL ET MÉTHODES

et al. (1996; 1997). Ces différents taux ont été réajustés tous les 15 jours après une pêche de contrôle. Tous les jours, les alevins morts flottant à la surface de l’eau ont été comptés et examinés individuellement. Ont été considérés comme victimes d’une mortalité naturelle, les individus morts dont le corps ne présentait aucune trace de lésion. Quant aux poissons morts dont le corps était incomplet ou lésé (cannibalisme de type I) et ceux ayant disparu (cannibalisme de type II), ils ont été considérés comme victimes de cannibalisme (HECHT & APPELBAUM, 1988; BARAS et al., 1999).

TABLE 1

Composition de l’aliment commercial selon le fabricant FACI (Fabrication d’Aliments Composés Ivoiriens: 18 BP 686 Abidjan 18, Côte d’Ivoire).

Les caractéristiques physico-chimiques du milieu d’élevage ont été mesurées 3 fois par semaine à l’aide d’un oxymètre de modèle WTW OXY 330 (pour la température et le taux d’oxygène dissous), d’un pH-mètre de modèle WTW pH 330 (pour le pH) et d’un disque de Secchi de diamètre 30cm (pour la transparence). Les caractéristiques physico-chimiques mesurées tout au long des trois expériences sont consignées dans le Tableau 2. Les valeurs enregistrées sont comprises dans la fourchette admise en pisciculture (WESTERS, 1979; MCDONALD, 1983; DÉLINCÉ, 1992).

En fin d’expérience, le contenu de chaque cage a été entièrement compté et pesé (au g près) et au moins 30% de la population a été pesée individuellement (au 1/10g près). A partir des données récoltées, le taux de survie Ts (%), le taux de mortalité Tm (%), le taux de cannibalisme Tc (%), le poids moyen final Pmf (g), la croissance journalière Cj (g/j), le coefficient de variation du poids Cv (%) et le taux apparent de conversion alimentaire Taca ont été calculés selon les formules suivantes:

TABLE 2

Caractéristiques physico-chimiques enregistrées au cours de l’élevage de Heterobranchus longifilis dans le lac de barrage d’Ayamé de septembre 2003 à mars 2004 (Temp=Température; O₂=Taux d’oxygène dissous;

Trp=Transparence; min=valeur minimale; max=valeur maximale).

Ts (%)=nf/ni × 100, avec nf=nombre final de poissons et ni=nombre initial de poissons.

Tm (%)=nm/ni × 100, avec nm=nombre final de poissons retrouvés morts.

Tc (%)=nd/ni × 100, avec nd=nombre de poissons ayant disparus de la cage.

Pmf (g)=Pt/nf, avec Pt=poids total (g).

Cj (g/j)=(Pmf-Pmi)/t, avec Pmf=poids moyen final (g), Pmi=poids moyen initial (g), et t=durée de l’élevage en jours.

Cv (%)=100 × (écart type/poids moyen).

Taca=Pn/A, avec Pn: poids total sec de l’aliment distribué (g), A: biomasse produite (g)

Les résultats sont présentés sous forme de moyennes ± écarts types entre triplicats. Les comparaisons ont été réalisées en procédant à une analyse de variance (Anova, expérience 1) ou de covariance (Ancova, expérience 2). Lorsque ces tests révélaient une différence significative, des comparaisons ‘‘Post Hoc’’ (Least Significant Difference: LSD) ont été exécutées. Les moyennes des traitements de l’expérience 3 ont été comparées deux à deux à l’aide du test de Student. La relation entre la densité (expérience 2) et les paramètres de croissance et de survie a été étudiée à l’aide de l’analyse de la régression linéaire. Dans tous ces tests statistiques, les différences ont été considérées significatives au seuil de 5%. Les analyses ont été effectuées à l’aide du programme STATISTICA

6.0 (Statsoft, Inc.).

RÉSULTATS

Expérience 1

Étude de l’effet du système de refuge et de sa coloration sur la survie et la croissance des alevins de Heterobranchus longifilis:

Expérience 2

TABLE 4

Résultats de la comparaison (analyse de covariance: Ancova) de la variation du poids moyen final (Pmf), du coefficient de variation du poids (Cv final) et du taux de survie (Ts) en fonction de la densité de mise en charge de Heterobranchus longifilis (poids initial: 0.9g) élevé pendant 90 jours en cage flottante sans système de refuge et en cage flottante avec système de refuge. Les valeurs de p marquées de (*) indiquent des différences significatives.

Expérience 3

Étude comparative de la survie et de la croissance entre les cages aménagées et les cages non aménagées pour des alevins de poids moyens initiaux différents (13.2±1.3g, 25.2±2.5g, 40.3±4.0g et 60.3±6.0g):

Après 60 jours d’élevage, pour les alevins de poids compris entre 13.2 et 29.2g, les valeurs les plus intéressantes de Pmf, de Ts et de Cv ont été enregistrées dans les cages aménagées (Tableau 6). Chez les poissons de Pmi égal à 13g, les paramètres (Pmf, Cv et Ts) enregistrés présentent des différences significatives (Test de Student, p<0.05). Pour les poissons de poids variant entre 25.2 (Pmi) et 46.0g (Pmf), seul le Ts en cages aménagées (plus intéressante valeur: 94.0±1.0%) est statistiquement diffé

rent (Test de Student, p<0.05) de celui enregistré en cages non aménagées. Chez les individus de 40.3 (Pmi) à 65.0g (Pmf), la comparaison des paramètres (Pmf, Cv et Ts) enregistrés dans les cages aménagées à ceux obtenus dans

TABLE 6

Paramètres de survie et de croissance en fonction du poids moyen initial (Pmi) de mise en charge de Heterobranchus lon-

gifilis élevé pendant 60 jours en cage flottante munie de système de refuge (bandelette plastique blanche) dans le lac de barrage d’Ayamé (Pmf=poids moyen final, Cv=coefficient de variation et Ts=taux de survie). Pour chaque Pmi, dans une même colonne, les valeurs portant des lettres différentes sont significativement différentes (Test de Student, p<0.05); les écarts-types ont été calculés entre triplicats.

Fig. 2. – Influence de la densité de mise en charge (100, 200, 500 et 1000 individus/m³) sur les paramètres de survie et de croissance des alevins de Heterobranchus longifilis (Poids moyen initial 0.9g) élevés pendant 90 jours en cages flottantes munies de système de refuge et en cages flottantes sans système de refuge. Pour chaque paramètre, des lettres différentes au dessus des histogrammes indiquent une différence significative p<0.05 (Post Hoc: LSD). Les barres verticales représentent les écarts entre réplicats n=3.

DISCUSSION

Les travaux de LEGENDRE (1991) ont montré que de meilleurs taux de croissance de H. longifilis pouvaient être obtenus dans des structures sans système de refuge (étang, enclos, cage-enclos): chez des juvéniles de poids moyen 0.77g élevés en étang pendant 70 jours à la densité de 10 individus/m², une croissance journalière de 0.30g/j a été observée; chez des poissons de poids moyen supérieur à 10g, cet auteur a enregistré des gains de poids journaliers variant entre 2.74 et 3.90g/j en cage-enclos (3 à

8.5 individus/m²), 3.52g/j en étang (0.7 individus/m²), 4.27g/j en enclos (2 individus/m²) et 3.24g/j en bassin en béton (7.5 individus/m²). Les valeurs obtenues dans ces structures seraient surtout dues au fait qu’elles ont en commun une productivité naturelle importante composée de détritus organiques, débris végétaux et animaux, graines, fruits, coquilles de gastéropodes, larves de batraciens, insectes, vertébrés aquatiques et proies planctoniques (GILLES et al., 2001). Selon MICHA (1973) et OTÉMÉ et al. (1996), H. longifilis est un poisson omnivore à ten-dance carnassière. Par conséquent, dans ces structures d’élevage, en plus de l’aliment artificiel, ce poisson bénéficie des ressources naturelles. Les espèces élevées en cages flottantes par contre ne dépendent que de l’apport d’aliment exogène (COCHE, 1978).

En définitive, le système de refuge proposé dans la présente étude pour l’amélioration de la survie et de la croissance des alevins de H. longifilis élevés en cage flottante s’est révélé efficace. L’efficacité de cette structure reste toutefois limitée aux individus de taille inférieure ou égale à 29.2g.

REMERCIEMENTS

Les auteurs tiennent à remercier Dr MARC LEGENDRE, Mr SYLVAIN GILLES et trois arbitres anonymes qui ont contribué à l’amélioration de la qualité du manuscrit. Cette étude s’inscrit dans le cadre du projet "Optimisation du potentiel aquacole du lac de barrage d’Ayamé: Pisciculture en Cages Flottantes" conduit par l’Université de Cocody-Abidjan (Côte d’Ivoire) et le Musée Royal de l’Afrique Centrale (Belgique) et financé par la Coopération Technique Belge (CTB).

RÉFÉRENCES

APPELBAUM S & KAMLER E (2000). Survival, growth, metabo

lism and behaviour of Clarias gariepinus (Burchell 1822)

early stages under different light conditions. Aquacult. Eng.,

22: 269-287.

BARAS E, TISSIER F, PHILIPPART JC & MÉLARD C (1999). Sibling cannibalism among juvenile vundu under controlled conditions. II. Effect of body weight and environmental variables on periodicity and the intensity of type II cannibalism. J. Fish Biol., 54: 106-118.

BARAS E, MAXI MY & MELARD C (2000). Sibling cannibalism in dorada under experimental conditions. J. Fish Biol., 57: 1021-0136.

BARAS E & D’ALMEIDA AF (2001). Size heterogeneity prevails over kinship in shaping cannibalism among larvae of sharp-tooth catfish Clarias gariepinus. Aquat. Living Resour., 14: 251-256.

BROWN C & BRAITHWAITE V (2004). Size matters: a test of boldness in eight populations of the poeciliid Brachyraphis episcope. Anim. Behav., 68: 1325-1329.

CALOW P (1985). Adaptative aspects of energy allocation. In: TYLER P & CALOW P (eds), Fish Energetics New Perspectives, Croom Helm, London: 13-31.

CAVAILLES M, KONAN K & DOUDET T (1981). Essai d’élevage de poissons en cage flottante en eaux continentales. CTFT, Div. Rech. Piscic., Bouaké (Côte d’Ivoire): 1-28.

COCHE AG (1978). Revue des pratiques d’élevage de poissons en cage dans les eaux continentales. Aquaculture, 13: 157

189.

DÉLINCÉ G (1992). The ecology of the fish pond ecosystem with special reference to Africa. Kluwer Academic Publishers, Boston Dordrecht London.

FONTAINE F & LE BAIL PY (2004). Domestication et croissance chez les poissons. Prod. Anim., 17: 217-225.

GILLES S, DUGUE R & SLEMBROUCK J (2001). Manuel de production d’alevins du silure africain Heterobranchus longifilis. Maisonneuve & Larose, Paris.

HAYLOR GS (1991). Controlled hatchery production of Clarias gariepinus (Burchell 1822): growth and survival of fry at high stocking density. Aquac. & Fisher. Manag., 22: 405

422.

HECHT T & APPELBAUM A (1988). The culture of sharptooth catfish Clarias gariepinus in Southern Africa. S. Afr. Nat. Sc. Progr. Rep., 153: 1-133.

HEM S, LEGENDRE M, TREBAOL L, CISSÉ A, OTÉMÉ ZJ & MOREAU Y (1994). L’aquaculture lagunaire. In: DURAND JR, DUFOUR P, GUIRAL D & ZABI SGF (eds), Les milieux lagunaires, Tome II: Environnement et ressources aquatiques de Côte d’Ivoire, ORSTOM, Paris: 455-505.

HENGSAWAT K, WARD FJ & JARUTJARAMORN P (1997). The effect of stocking density on yield, growth and mortality of African catfish (Clarias gariepinus Burshell 1822) cultured in cages. Aquaculture, 152 (1-4): 67-76.

HOSSAIN MAR, BEVERIDGE MCM & HAYLOR GS (1998). The effects of density, light and shelter on the growth and survival of African catfish (Clarias gariepinus Burchell 1822) fingerlings. Aquaculture, 160: 251-258.

LEGENDRE M (1991). Potentialités aquacoles des Cichlidae (Sarotherodon melanotheron, Tilapia guineensis) et Clariidae (Heterobranchus longifilis) autochtones des lagunes de Côte d’Ivoire. Thèse de Doctorat, Université Montpellier II.

LEGENDRE M, SLEMBROUCK J, KERDCHUEN N & OTÉMÉ ZJ (1991). Evaluation d’une méthode extensive d’alevinage des Clariidae en cages implantées en étangs. ORSTOM, Montpellier.

MCDONALD DG (1983). The effect of H⁺ upon the gills of freshwater fish. Can. J. Zoolog., 61: 991-703. MÉLARD C (1986). Les bases biologiques de l’élevage intensif du tilapia du Nil. Cah. Ethol. Appl., 6: 1-224.

INTRODUCTION

In the past 20 years, considerable attention has been paid to nematodes, demonstrating that these ubiquitous members of the soil community reflect change in ecological structure and function of soils in ways more predictable and efficient than for other soil flora or fauna (FISCUS & NEHER, 2002). The ecological significance of nematodes lies in their high abundance and diversity, variety of trophic types and reproductive strategies (YEATES, 2003). They have been used to evaluate soil conditions under different crops (WIDMER et al., 2002), under various agricultural practices, such as tillage (FU et al., 2000), fallow farming (VILLENAVE et al., 2001), use of pesticides (LIANG et al., 2001), fungicides (VILLENAVE et al., 2004) or herbicides (YEATES et al., 1999), and to distinguish the effects of physical and chemical disturbance of cultivated soils (FISCUS & NEHER, 2002).

Furthermore, nematodes have been used for evaluating soil quality under organic and conventional agriculture (FRECKMAN & ETTEMA, 1993; FERRIS et al., 1996; NEHER, 1999; NEHER & OLSON, 1999; MULDER et al., 2003; GARCÍA-ÁLVAREZ et al., 2004 among others). However, only a few studies compare agroecosystems managed organically for different time periods, and are mostly confined to comparisons of crop yields or to soil parameters other than nematodes (FRIEDEL & GABEL, 2001; MARTINI et al., 2004; MONOKROUSOS et al., 2006). To our knowledge, the only studies relating nematode communities to duration of organic management are those of BERKELMANS et al. (2003) who monitor changes of nematode communities in sites managed organically from 4 to 12 years, and of VAN DIEPENINGEN et al. (2006) who compare conventional to certified organic systems of less and more than five years of organic management. Thus, there is a lack of data regarding possible trends of succession within communities initiated after changing management regime from conventional to organic.

The main goal of this study was, therefore, to explore responses of soil nematodes along a 5 year gradient from conventional to organic cultivation. According to the Regulation 2092/91/EEC, the period of conversion from a conventional to an organic production system is at least two years for annual crops and three for perennial ones. The agricultural systems included in our study were conventional, organic under conversion (1 and 2 years) and certified organic (4 and 5 years) cultivations of Asparagus officinalis. We selected them in order to minimize the effect of other factors that are known to affect nematode populations even more than agricultural practice per se, such as the cultivated species and duration of cultivation (YEATES et al., 1999) and the soil type (VAN DIEPENINGEN et al., 2006). Moreover, asparagus cultivations are perennial and not additionally stressed by annual plantings, and thus the history of agricultural practices is better reflected on soil conditions. Nematode communities were studied in terms of trophic and generic structure, life strategy and diversity.

Apart from changing management regime from conventional to organic, seasonal agricultural practices, input of fertilizers (conventional or organic), mechanical disturbance of soil, crop harvesting etc, are shared between management types, constituting periodic disturbances to the soil ecosystem. These disturbances are quite important for shaping nematode communities. The effect of the one may mask the effect of the other (FISCUS & NEHER, 2002), but most importantly they may mask the progressive effect of the transition from conventional to organic management. Thus, a second goal of our study was to explore short-term nematode responses to seasonal agricultural practices and compare them to the long-term responses to changing management regime. This was achieved by conducting samplings in periods following all steps within the annual crop growing cycle of asparagus.

MATERIALS AND METHODS

The study area is a cultivated plain of Northern Greece, at about 60km north west of Thessaloniki (7.5m a.s.l.), where asparagus is traditionally one of the main products. Due to higher price of organic agricultural products and financial support by the state, during the last decade a lot of asparagus cultivations were converted to organic. Soils are classified as luvisol (FAO) with a clay-loam texture (37% sand, 33% silt, 30% clay). The climate is transient between mediterranean and continental. Mean annual precipitation for a 10-year period was 485mm and mean annual temperature 14°C. January was the coldest month (4°C) with the highest precipitation (58mm), while the warmest (24°C) and driest (16mm) month was July.

The experimental plots of our study belonged to agroecosystems with different management history, i.e. conventional, organic under conversion and certified organic cultivations of the perennial Asparagus officinalis. Conventional sites covered about 9,000m² and were cultivated with asparagus for more than 6 years. Organic ones were conventionally cultivated for several years before conversion to organic farming, and covered in total about 30,000m² of the wider study area. At the start of the study the certified organic sites were managed organically from 4 to 5 years [O4, O5], while the transitional organic sites were managed organically from 1 to 2 years [O1, O2]. Asparagus cultivations, either organic or conventional, are not tilled, but during February, i.e. one month before the beginning of the harvest season, the soil between field rows is mounded up over the asparagus plants, for producing white spears. The soil mounds are smoothed away during June, i.e. after the end of harvest. Thus, all plots are subject to the same physical disturbance and differ only regarding the type of fertilizers and weed control. Fertilization takes place in summer, with manure and licensed organic fertilizers in one case and synthetic fertilizers in the other (Table 1). Regarding weed control, which also takes place in summer, it is done with hand-hoeing in case of organic cultivation, while in conventional cultivation it is done once every summer by means of Linuron for broadleaved grasses (0.2 l /1,000m²) and Fluazifop-p-butyl for grass weeds (0.2 l /1,000m²).

Our sampling scheme represented a full factorial design of 5 agricultural systems x 4 dates x 4 replicate plots. The latter were randomly dispersed covering 250m² each. At each plot we took a composite sample of 10 soil cores 2cm in diameter and 20cm in depth. Samples were

taken in the rows of the asparagus plantations, and more specifically close to the edge of them, so that regardless the seasonal soil mounding, all year samples corresponded to the same soil depth relative to the cultivated plants. Soil sampling was conducted four times throughout a year; in March (asparagus spears start growing and the harvest begins), in May (end of harvest), in October (the fields are left with no agricultural activities for 2 – 3 months) and in December (two weeks after cutting of the aboveground parts of asparagus plants which are then left in the field to decompose). The time span between samplings was long in relation to nematode life span, so that nematode population dynamics decouple nematode counts across time, creating sufficiently independent data.

Nematodes were extracted from 150mL of each composite soil sample. Before taking this subsample, the soil was gently mixed by hand and soil aggregates were broken up. For extraction, we used the modified Cobb’s sieving and decanting method proposed by S’JACOB & VAN BEZOOIJEN (1984), according to which a cotton-wool filter is used in the last step. After counting total abundance of nematodes, we fixed them with formaldehyde 4%. Later on from each sample we selected randomly at least 150 nematodes and identified them to the genus level in most cases, using the identification key of BONGERS (1994).

Soil bulk density and pH was measured once, at the beginning of samplings. The former was not found to differ among study sites, ranging from 1.06 to 1.24g/cm³. Values of pH decreased gradually and significantly from

7.99 in [C] to 7.94, 7.90, 7.83 and 7.51 in [O1], [O2], [O4] and [O5] respectively. On each sampling occasion, soil water content (% dry weight) was also estimated from each soil sample taken. It was found to differ seasonally, being higher in December (24.22%) and lower in May (20.36%).

Nematode taxa were assigned to trophic groups according to YEATES et al. (1993), classified along the colonisation-persistence gradient (c-p values) following BONGERS (1990) and BONGERS & BONGERS (1998), and arranged to functional guilds (portions of particular trophic groups exhibiting the same c-p value) according to BONGERS & BONGERS (1998) and FERRIS et al. (2001).

Data analysis

All analyses aim to estimate and compare the nematode responses to changing management regime and to seasonal agricultural practices, in terms of trophic structure, nematode indices, generic structure and composition and diversity.

For analyzing frequencies of the different nematode trophic groups at the different agricultural systems, we used the log-linear analysis of count data of categorical variables. This method fits log-linear models to the combinations of categorical variables, herein trophic group and site, to find out possible associations between them. More specifically, it estimates the randomly expected frequency of every combination of trophic group and site and compares it to the observed frequency.

For testing differences of nematode indices due to the long-term effect of changing management regime as well as the short-term effect of seasonal agricultural practices, we used two-way ANOVA (site x date). LSD post hoc comparisons where performed, when significant differences were revealed. Prior to analyses, data were examined for normality, homogeneity of variance (Levene’s test) and independence between variance and mean. Since values of nematode indices are bounded, an arc-sinus transformation was used. Kruskal-Wallis test was used in the case where the prerequisites of normality were not met even after transformation.

In order to compare the progressive effects of changing management regime to those of seasonal agricultural practices on the generic structure of nematode communities, all samples and nematode genera abundances were ordinated by means of correspondence analysis (CA), while sample ordination scores for Axis 1 and Axis 2 were analyzed by MANOVA (site x date). When significant differences were revealed, post hoc comparisons were performed by means of LSD-test. Abundances of nematode genera were log+1 transformed prior to CA, while prior to MANOVA data were examined for normality, homogeneity of variance (Levene’s test) and independence between variance and mean.

For assessing the diversity of nematode communities, we used the method of diversity ordering proposed by RENYI (1961). Renyi’s parametric index of order a shows varying sensitivity to the rare and abundant species of a community, as the scale parameter a changes (RICOTTA, 2000). For each community it provides a profile of the most widely used diversity indices. For a=0, the index equals log species number, for a=1, it equals Shannon’s index, for a=2, it equals Simpson’s index. For a tending to infinite, the index is most sensitive to the abundant species of a community. Thus, when diversity profiles differ in the range of low a values, this is due to the number of species. In the range of high a values, differences between communities are due to presence of abundant species. When diversity profiles intersect, the communities may be ordered differently by different diversity indices.

In order to study changes of generic composition, we used the IndVal (Indicator Value) method of DUFRENE & LEGENDRE (1997). This method assigns indicator (characteristic) species to a site or a group of sites, on the basis of species relative abundance and relative frequency of occurrence in the samples. For the classification of sample units required by IndVal, we used a complete linkage hierarchical classification tree based on 1-Pearson r distance measure. The statistical significance of the species indicator values was also evaluated by the IndVal program by means of a randomization procedure.

For log-linear analysis, ANOVA, MANOVA, Kruskal-Wallis and tree clustering we used the SPSS software package (version 11). For Correspondence analysis we used CANOCO (version 3.10) software package (TER BRAAK, 1988), while diversity ordering was performed by means of DivOrd (TOTHMERESZ, 1995).

RESULTS

TABLE 2

The contribution of individual trophic groups (%) and the total nematode abundance (±SE) in the studied agricultural systems. [C]: conventional, [O1], [O2], [O4], [O5]: managed organically for 1, 2, 4 and 5 years respectively.

Fig. 1. – Coefficients of the Log-Linear model for each combination of trophic group and site. When the confidence intervals of the coefficient include 0 the nematode population of the specific trophic group at the specific site does not differ from the overall average. Positive values of the coefficient indicate that the specific trophic group appears at the specific site more frequently than randomly expected, while the inverse happens for negative values. Codes as in Table 2.

Nematode indices ally from the older to the recent organic and further to conventional cultivation. Regarding sampling date, all Mean values of MI, PPI, EI, SI and CI at the different indices except SI displayed significant differences. MI sites and seasonal samplings are presented in Table 3, was higher in October, i.e. in a period long after any agritogether with the results of two-way ANOVA (site x cultural practice. PPI was higher in December, i.e. when date). Significant changes regarding sampling site were the above ground parts of asparagus plants were cut and observed only in the case of PPI, which increased gradu-left in the field to decompose. EI was higher in March, i.e.

Generic structure

The ordination of all samples along the two first axes of Correspondence Analysis is depicted in Fig. 2, while the results of MANOVA (site x date) on Axis 1 and Axis 2 ordination scores are given in Table 4. The effects of changing management regime on generic structure of the nematode community were significant at both axes, while those of seasonal agricultural practices were significant only along the second axis. Samples from [C] were clearly and significantly distinct from [O4] and [O5] samples, forming a gradient from conventional to recent and further to older organic cultivations along the first axis. Samples taken during the harvest season when the soil was mounded, i.e. March and May, were ordinated on the lower half of the plot, differing significantly from October and December samples, which were ordinated mostly on the upper half corresponding to samplings carried out during the post-harvest season when soil was flat. Because of these seasonal differences in generic structure, we provide separate results for the harvest and the post-harvest season in all subsequent analyses of the data.

TABLE 3

Mean values (±SE) of Maturity Index, Plant Parasitic Index, Channel Index, Enrichment Index, Structure Index and results of two-way ANOVA (site x date) or Kruskal-Wallis test (for SI). Superscripts a, b, c, d indicate differences revealed by LSD-test, (*: P<0.05, **: P<0.001, MS: Mean square). Codes as in Table 2.

TABLE 4

Results of MANOVA (site x date) for Axis 1 and Axis 2 sample ordination scores. Superscripts a, b, c, d indicate differences revealed by LSD-test, (*: P<0.05, **: P<0.001, ***: P<0.0001, MS: Mean square). Codes as in Table 2.

Fig. 3. – Diversity profiles of nematode communities in the studied agricultural systems during (a) the post-harvest season and (b) the harvest season. Codes as in Table 2.

The diversity profiles of the nematode communities for all studied systems are shown in Fig. 3. In the post-harvest season (Fig. 3a), diversity was higher in [O5], [O4] and gradually decreased to [O2], [O1] and further to [C], reflecting a gradient of a long-term management regime. In the harvest season (Fig. 3b), the diversity of recent organic plots increased, while [C] was even less diverse than in the previous season. In both seasons, differences of diversity were not due to the number of genera, since the profiles differed mainly in the range of high values of the scale parameter a. This means that differences of diversity were mainly due to the distribution of nematode numbers among genera. Indeed, the contribution of the most abundant genera of the organic sites did not exceed 22%. In [C] the community displayed a quite different structure due to the strong dominance of a single genus, namely Helicotylenchus, which accounted for 40% of the community in the post-harvest season. The strong dominance of this semiendoparasite in [C] was even higher (50%) in the harvest season, causing the observed decrease of diversity in this case.

A tree clustering of sites is presented in Fig. 4. Each site is represented twice; once for the post-harvest and once for the harvest season, coded as P and H respectively. Moreover, we provide the maximum indicator value of each genus at the site or cluster of sites to which it is assigned. Changing management regime from conventional to organic cultivation caused profound modifications in generic structure and composition, since organic plots were separated from conventional, regardless of season. The most recent organic plots constituted an exception since they were clustered either with the

Fig. 4. – Dendrogram presenting the indicator genera assigned by IndVal to each combination of site and season. H and P before site codes stand for “harvest season” and “post-harvest season” respectively. The Indicator Value of each genus for each cluster is given, while the trophic group and c-p value are indicated in parentheses. BA: Bacterivore, FU: Fungivore, PA: Phytoparasite, NP: Non parasitic plant feeder, PR: Predator, OM: Omnivore (** P<0.01).

conventional or with the organic plots depending on the season of sampling. At a second level, older organic cultivations [O4] and [O5] were grouped together, without nevertheless masking the effect of seasonal agricultural practices on nematode communities. On the other hand, this seasonal effect was not revealed in the case of conventional plots, since [HC] and [PC] were grouped together with the lowest linkage distance.

According to IndVal method, from the total of 61 genera recorded, 19 could be considered eyrytopic, since they were found at all sites in both seasons. Among these genera no phytoparasites were found. The highest IndVal values were displayed by Acrobeloides, Aphelenchus and Chiloplacus, which were well presented at all samples constituting together about 28% of the total nematode community. It seems that the major separation of the organic from the conventional cluster was mainly due to two genera; the bacterivore Heterocephalobus characterized the organic cluster, while the phytoparasite Helicotylenchus was assigned to the conventional cluster with the highest indicator value. The second most important genus characterizing the conventional cluster was the phytoparasitic Tylenchorhynchus, followed by the predator Eudorylaimus, which however was of very low abundance. Another genus that may be considered important bioindicator, is the phytoparasite Paratylenchus, which was recorded only in the older organic cultivations.

Data depicted in Fig. 5, concerning the abundances of the two main indicator genera, namely Helicotylenchus and Heterocephalobus, revealed a completely opposite trend of abundance change along the gradient from conventional to recent and further to older organic cultivations.

Fig. 5. – Abundance of Helicotylenchus and Heterocephalobus in the studied agricultural systems in (a) the post-harvest season and (b) the harvest season.

DISCUSSION

The investigation of nematode communities in soils with different history of management practices revealed several differences at various levels of community analysis. The most important finding from the analysis of trophic structure was the gradual reduction of plant feeders from conventional to the older organic cultivation, while the exactly opposite trend was revealed in the case of nematodes that feed on decomposers, i.e. bacteria and fungi. Our results support the hypothesis of VAN DIEPENINGEN et al. (2006) that the differences between the two management types are more gradual than black and white, and are in agreement with the findings of several authors

(e.g. NEHER & OLSON, 1999; GARCÍA-ÁLVAREZ et al., 2004), who reported an increase of decomposer feeders and especially bacterivores under organic cultivation.

In our study, the ratio of plant feeders to decomposer feeders was 1:0.7 under conventional cultivation, changed approximately to 1:2.6 in the recent and to 1:3.8 in the older organic plots, reflecting gradual changes between the “grazing food web” and the “detritus food web” of PETERSEN & LUXTON (1982) or the “root-herbivore” and “decomposer” guilds of BRUSSAARD (1998). This shift probably resulted from the different form of nutrient input in the cultivated plots. Organic fertilizers are expected to favour bacterivores and fungivores, since they act via the microbial soil component. Feeding activity of these groups stimulates decomposition and nitrogen mineralization (FERRIS et al., 2004). Indeed in a parallel experiment carried out in our experimental area, MONOKROUSOS et al. (2006) found that N-mineralization rates were lower in conventionally cultivated plots and increased gradually to older organic plots.

Nematode indices were found useful mostly for distinguishing seasonal agricultural practices. Maturity Index (MI) was higher during October, i.e. the season with least disturbance, since fields were left with no agricultural activities for several months. The growth period of asparagus stops in autumn and during winter nutrients are translocated from the fern to the rhizome from where they will be remobilised into new spear growth in spring (FAVILLE et. al., 1999). This nutrient translocation seems to favour phytoparasites leading to a higher Plant Parasitic index (PPI) in December.

Channel Index (CI) was higher in October and December, i.e. during the post-harvest season. This indicates a shift from a more bacterial to a more fungal mediated decomposition pathway (FERRIS et al., 2001). During this period, there were amendments of plant biomass, originating either from dead weed residues after summer control or from aboveground asparagus parts, which were cut in late November and left in the fields to decompose. Plant residues constitute amendments of high C/N ratio, favouring populations of fungivores (BOHLEN & EDWARDS, 1994). Two weeks before the beginning of spear growing the soil was mounded and thus the plant residues were incorporated into the soil. Their advanced decomposition renders this organic source more labile, favouring the enrichment opportunist bacterivores with c-p value 1 (FERRIS & MATUTE, 2003). The increase of the Ba-1 functional guild resulted in significantly higher values of the Enrichment Index (EI) during March.

The distinction between the post-harvest and the harvest season was also evident from the analyses of community structure at the generic level, but mostly in the older organic plots. This was mainly due to changes in abundance of genera which were classified as eurytopic. For example Panagrolaimus, Diploscapter and Bursilla, which all belong to the Ba-1 functional guild, increased in the harvest season, while the Fu-2 Aphelechoides and Paraphelenchus and the root hair feeder of c-p 2 Filenchus displayed higher abundance in the post-harvest season. Heterocephalobus (Ba-2) which characterized the organic plots displayed a significantly higher abundance during this season. On the other hand, seasonal patterns were not pronounced in conventional plots. The predominance of Helicotylenchus (c-p 3 phytoparasite) in these plots during both seasons diminished the importance of any other changes in the contribution of the remaining species. It seems that a strong pressure as the one imposed by long-term conventional management masks the results of less intense pressures, as the ones imposed by short-term management practices.

CONCLUSION

ACKNOWLEDGMENTS

REFERENCES

47.

235.

FERRIS H, BONGERS T & DE GOEDE RGM (2001). A framework for soil food web diagnostics: extension of the nematode faunal analysis concept. Appl. Soil Ecol., 18: 13-29.

FERRIS H & MATUTE MM (2003). Structural and functional succession in the nematode fauna of a soil food web. Appl. Soil Ecol., 23: 93-110.

FERRIS H, VENETTE RC & SCOW KM (2004). Soil management to enhance bacterivore and fungivore nematode populations and their nitrogen mineralisation function. Appl. Soil Ecol.,

25: 19-35. FISCUS DA & NEHER DA (2002). Distinguishing sensitivity of free-living soil nematode genera to physical and chemical disturbances. Ecol. Appl., 12: 565-575. FRECKMAN DW & ETTEMA CH (1993). Assessing nematode communities in agroecosystems of varying human intervention. Agr. Ecosys. Env., 45: 239–261. FRIEDEL GK & GABEL D (2001). Nitrogen pools and turnover in arable soil under different durations of organic farming. I:

pool sizes of total nitrogen, microbial biomass nitrogen, and potentially mineralizable nitrogen. J. Plant Nutr. Soil Sci.,

164: 415-419. FU A, COLEMAN DC, HENDRIX PF & CROSSLEY JRDA (2000). Responses of trophic groups of soil nematodes to residue application under conventional tillage and no-till regimes. Soil Biol. Biochem., 32: 1731-1741. GARCÍA-ÁLVAREZ A, ARIAS M, DÍEZ-ROJO MA & BELLO A (2004). Effect of agricultural management on soil nematode trophic structure in a Mediterranean cereal system. Appl. Soil Ecol., 27: 197-210. ILIEVA-MAKULEC K (2000) Nematode fauna of a cultivated peat meadow in relation to soil depth. Ann. Zool., 50: 247-254. LIANG WJ, LAVIAN I & STEINBERGER Y (1999). Dynamics of nematode community composition in a potato field. Pedobiologia, 43: 459-469. LIANG W, LAVIAN I & STEINBERGER Y (2001). Effect of agricultural management on nematode communities in a Mediterranean agroecosystem. J. Nematol., 33: 208-213. LIANG WJ, LAVIAN I, PEN-MOURATOV S & STEINBERGER Y (2005a). Diversity and dynamics of soil free-living nematode populations in a Mediterranean agroecosystem Pedosphere, 15 (2): 204-215. LIANG WJ, ZHANG XK, Li Q, JIANG Y, OU W & NEHER DA (2005b). Vertical distribution of bacterivorous nematodes under different land uses. J. Nematol, 37 (3): 254-258. MARTINI EA, BUYER JS, BRYANT DC, HARTZ TK & DENISON RF (2004). Yield increases during the organic transition: improving soil quality or increasing experience? Field Crop. Res., 86: 255-266. MCSORLEY R & GALLAHER RN (1995). Effect of Yard Waste Compost on Plant-Parasitic Nematode Densities in Vegetable Crops. J. Nematol, 27(4S): 545-549. MONOKROUSOS N, PAPATHEODOROU EM, DIAMANTOPOULOS JD & STAMOU GP (2006). Soil quality variables in organically and conventionally cultivated field sites. Soil Biol. Biochem, 38(6): 1282-1289. MULDER CH, DE ZWART D, VAN WIJNEN HJ, SCHOUTEN AJ & BREURE AM (2003). Observational and simulated evidence of ecological shifts within the soil nematode community of agroecosystems under conventional and organic farming. Funct. Ecol., 17: 516-525. NEHER DA (1999). Nematode community in organically and

conventionally managed agricultural soils. J. Nematol, 31: 142-154.

NEHER DA & OLSON RK (1999). Nematode communities in soil of four farm cropping management systems. Pedobiologia,

43: 430-438. PAPATHEODOROU EM, ARGYROPOULOU MD & STAMOU GP (2004). The effects of large- and small-scale differences in soil temperature and moisture on bacterial functional diver

sity and the community of bacterivorous nematodes. Appl. Soil Ecol., 25: 37-49.

PETERSEN H & LUXTON M (1982). A comparative analysis of soil fauna populations and their roles in decomposition processes. Oikos, 39: 287-388.

RENYI A (1961). On measures of entropy and information. In: NEYMAN J (ed). Proceedings of the 4^th Berkeley Symposium on Mathematical Statistics and Probability, vol 1. University of California Press: 547-561.

RICOTTA C (2000). From theoretical ecology to statistical physics and back: Self-similar landscape metrics as a synthesis of ecological diversity and geometrical complexity. Ecol. Model., 125: 245-253.

S’JACOB JJ & VAN BEZOOIJEN J (1984). A manual for practical work in nematology. Department of Nematology, Wageningen Agricultural University. Wageningen, The Netherlands.

TER BRAAK CJF (1988). CANOCO – a FORTRAN program for canonical community ordination. Technical Report: LWA88-02. Agricultural University Wageningen.

TISDALE SL & NELSON WL (1975). Soil Fertility and Fertilizers, 3^rd ed. MacMillan, New York.

TOTHMERESZ B (1995). Comparison of different methods for diversity ordering. J. Veg. Sci., 6: 283-290.

VAN DIEPENINGEN AD, DE VOS OJ, KORTHALS GW & VAN BRUGGEN AHC (2006). Effects of organic versus conventional management on chemical and biological parameters in agricultural soils. Appl. Soil Ecol., 31: 120-135.

VESTERGÅRD M (2004). Nematode assemblages in the rhizosphere of spring barley (Hordeum vulgare L.) depended on fertilisation and plant growth phase. Pedobiologia, 48: 257

265. VILLENAVE C, BONGERS T, EKSCHMITT K, DJIGAL D & CHOTTE JL (2001). Changes in nematode communities following cultivation of soils after fallow periods of different length. Appl. Soil Ecol., 17: 43-52. VILLENAVE C, EKSCHMITT K, NAZARET S & BONGERS T (2004). Interactions between nematodes and microbial communities in a tropical soil following manipulation of the soil food web. Soil Biol. Biochem., 36: 2033-2043. WASILEWSKA L (1997). Soil invertebrates as bioindicators, with

special reference to soil-inhabiting nematodes. Rus. J. Nematol., 5: 113-126.

WIDMER TL, MITKOWSKI NA & ABAWI GS (2002). Soil organic matter and management of plant-parasitic nematodes. J. Nematol., 34: 289-295.

YEATES GW (2003). Nematodes as soil indicators: Functional and biodiversity aspects. Biol. Fertil. Soil., 37: 199-210.

YEATES GW, BONGERS T, DE GOEDE RGM, FRECKMAN DW & GEORGIEVA SS (1993). Feeding habits in nematode families and genera - an outline for soil ecologists. J. Nematol, 25: 315-331.

YEATES GW, WARDLE DA & WATSON RN (1999). Responses of soil nematode populations, community structure, diversity and temporal variability to agricultural intensification over a seven-year period. Soil Biol. Biochem., 31: 1721-1733.

ZOLDA P (2006). Nematode communities of grazed and ungrazed semi-natural steppe grasslands in Eastern Austria. Pedobiologia, 50: 11-22.

Received: April 21, 2006 Accepted: March 20, 2007

INTRODUCTION

Flooding habitats, such as river banks and floodplains, temporary streams, salt marshes, creeks, inundated forests, and mangroves, are unique ecosystems with complex land-water interactions. They are inhabited by faunas of aquatic, as well as amphibious and terrestrial species, adapted to fluctuations of the water table. Terrestrial arthropods are by far the most species-rich group and play an important role in the ecosystem (DESENDER & MAELFAIT, 1999). Research on the terrestrial component of tidal marshes and other flooding habitats (DESENDER & MAELFAIT, 1999; ADIS & JUNK, 2002; BONN et al., 2002; KRUMPALOVA, 2002) is seriously hindered by the lack of an adequate and efficient technique, such as pitfall trapping, a good method of sampling because of its simplicity and ease of operation. It is an effective and cheap way of qualitatively surveying the ground surface-active arthropods over long periods of time, and allows for comparison of assemblages in different habitats (GREENSLADE, 1964). Even when it comes to diversity and gradient analysis of terrestrial arthropods in tidal marshes, sampling techniques are limited to aspirator and hand capturing with a very restricted effort of catch in time (DESENDER & MAELFAIT, 1999).

We developed a pitfall trap suitable for habitats that are regularly or unpredictably inundated. The trap is conceived in such a way that it is enclosed and protected in an air bell from the moment the rising water reaches the rim of the trap. Later on, when the water table falls back under the rim, the original situation is restored. So pitfall trapping continues as long as the trap is not submerged, regardless the time and the frequency of inundation. Other automated sampling methods are not satisfactory for two reasons: a floating pitfall trap captures only the animals on the water surface (GRAHAM et al., 2003), whereas the classic pitfall trapping method is restricted to the permanently dry portions of the habitat (MILFORD, 1999; TAJOVSKY, 1999; WISHART, 2000; IRMLER et al., 2002; DEIDUN et al., 2003; JOHNSON, 2003). This paper focuses on three consecutive items: (1) the description of a new type of trap, developed for temporarily submersed areas; (2) evaluation of its usefulness by literature search; and (3) an evaluation of its usefulness by its application in a salt march. The general design of the pitfall is described, adaptable to all types of habitats according to the specific needs of the researcher.

Archisotoma pulchella (Moniez, 1890) was lacking in the Belgian list of Collembola (JANSSENS, 2004), although it is a common species of intertidal mud flat communities in the surrounding countries (STERZYNSKA & EHRNSBERGER, 2000). Using our sampling technique, we found not only that species, but moreover we were able to reconstruct some characteristics of the population dynamics of the springtail, even when only one trap was used over a short period of time. So it is finally proved that this type of pitfall trapping yields data that generate new information, adding to knowledge of the biodiversity of flooding habitats and the ecology of their species.

MATERIALS AND METHODS

Fig. 1 presents the principle of a submersible pitfall trap before (Fig. 1A) and during (Fig. 1B) inundation, while Fig. 1C is a technical depiction of the instrument. Fig. 2 shows a photograph of the instrument with (Fig. 2A) and without (Fig. 2B) the cone, as well as the trap as used on the study area (Fig. 2C). When the water level rises (Fig. 1Be), the pitfall (1Ba) is left dry in the air bell (Fig. 1Bf) under the cone (Fig. 1Ad). The pitfall is exposed again when the water surface drops below the rim of the container (Fig. 1Aa, Ba). Its rim has to be elevated above the substrate (Fig. 1Ag, Bg) higher than the base of the cone (Fig. 1Ad), since the air of the bell compresses when the water table rises. In our design the cone is made from a reverse plastic funnel, closed at the top with silicone glue. The difference in level between the rim of the trap and the base of the air bell inside the cone is realized by the sample holder (Fig. 1Ac), a concrete ring that fits around the trap. The instrument requires tight fixation to the substrate, to withstand uplift of the air bell during inundation. This is realized by attaching the cone (Fig. 1Cd) with rubber bands (Fig. 1Ci) to a steel support structure anchored into the substrate by a soil auger (Fig. 1Ch). This steel structure consists of three brackets on radial arms (Fig. 1Cj) welded to the top of the auger (Fig. 1Ch). A 10% formalin solution is used as a fixative (Fig. 1Ab, Cb) for the captured animals.

RESULTS AND DISCUSSION

The rising water surrounding the pitfall trap causes an air bell in the trap as well as under the protecting cone. Moreover, the pressure on the air bell will increase in accordance with the rising water column above the instrument. When the height between the rim of the cone and the water surface is substantial, one needs a careful evaluation of the compression of the air bell and the corre

sponding increase of the water level approaching the rim

of the trap. Inaccurate evaluation causes flooding of the

trap and loss of the sample. Those who are less mathemat

ically inclined will arrive at acceptable results by trial and

error, when simulating the trap in the appropriate environ

pha, class Collembola; Syn.: Isotoma pulchella Moniez, 1890). It is a new species for the Belgian fauna.

A 45
40
35
30

25

20

15

10

5

Citations (%) Pitfalls (%)

40

B

fall, inundating the sample.

The results of the literature survey are visualised as a

ranking in Fig. 3. By splitting up the references into habi

tat types (Fig. 3A) one observes that pitfall trapping is

practised to a considerable extent in forests (35.1%) and agricultural ecosystems (31.0%); to a lesser degree in grasslands like pastures (9.3%), but only 7.2% in potentially or regularly submerged habitats. This confirms our suspicion that information based on pitfall trapping is biased towards forests and some preferred habitats, neglecting others including many types of inundated biotopes.