Version classiqueVersion mobile

Dynamique et gestion des pelouses calcaires de Haute-Normandie

 | 
Thierry Dutoit

Quatrième partie. Gestion conservatoire par le pâturage/Conservation management by grazing

Chapitre second. Mineral contents of chalk grasslands in relation with conservation management: a case study in Upper-Normandy (France)

Gestion conservatoire et teneurs minérales des pelouses calcicoles : un cas d'étude en Haute-Normandie (France)

Résumé

De nombreuses études ont été menées sur la gestion par le pâturage comme outil pour conserver la biodiversité des pelouses calcicoles en Europe mais son impact sur la composition minérale de la végétation est encore inconnu. Afin de mieux cerner les relations qui existent entre les teneurs minérales et la gestion, les espèces végétales dominantes de deux parcelles à gestion contrastée (abandonnée et pâturée) ont été échantillonnées mensuellement pour analyses minérales. Ces analyses ont également été réalisées sur des échantillons moyens des phytocénoses ainsi que des feuilles et écorces des arbustes présents dans les parcelles. La composition floristique des parcelles a été inventoriée par la méthode des quadrats permanents. Les principaux résultats sont: des différences spécifiques, temporelles et spatiales dans les teneurs minérales; une carence en oligo-éléments dans les espèces consommées; une réserve minérale dans les espèces refusées et les arbustes. Nous faisons ensuite le point sur les plans de gestion, qui doivent particulièrement tenir compte de cette ressource nonutilisée et la régénération du stock d'éléments minéraux pour la réalisation d'un système de pâturage soutenable.

Texte intégral

INTRODUCTION

1In western Europe, calcareous grasslands represent threatened species-rich ecosystem. They are considered as the result of Neolithic human forest clearance and were used as sheep-walks and cultivated land for many centuries (Smith 1980). Nowadays, since the modern agricultural revolution in the fifties, these grasslands have lost their former significance as a food resource for agriculture and were progressively abandoned. This appeared especially because of the agricultural impediments of these grasslands in the general productive context (Willems, 1978). As a consequence, natural serial changes occurred in the absence of grazing (Willems, 1990), leading to natural succession which proceeded to plagioclimax of scrub and/or species poor-coarse grasslands. The need of management to conserve these communities by sheep grazing has been discussed in general terms by Wells (1965) and is now applied experimentally in northern Europe (Arlot & Hesse, 1981; NCC, 1982; Rosén, 1982; Delescaille et al, 1991; Maubert & Dutoit, 1995).

2Although the changes in floristic richness and diversity which occur in grazed or mown grassland are well known (Wells, 1969a; Hillier et al., 1990), there are surprisingly few quantitative data available on mineral contents of chalk grassland turf and plant (Milton & Davies, 1947; Smith et al., 1971; Scoppola et al., 1984). Nevertheless, the effects of grazing on plant diversity strongly depend on both the intensity of grazing and the palatability of the dominant plants.

3The latter factor is not only determined by the plant species involved, but also by the breed of the grazers and periods of grazing (During & Willems, 1984). In consequence, we clearly need to expand our information on the effects of the management regimes on such variables as soils and plant composition and on the ability of chalk grasslands to provide sufficient nutrients to support grazing animals.

4In this study, we focused on the mineral composition of chalk grassland species and turf in two plots of the same grassland with different management: one is abandoned since 1950 and the other one is intensively grazed by sheep since 1984. Particular attention was also paid on mineral contents of leaves and barks of some scrubs which generally invade the abandoned grassland and initiate the succession to woodland.

STUDY SITE

5The study was carried out on semi-natural chalk grasslands located in the Saint-Adrien nature reserve 15 km south of Rouen (1°5'3"E, 49°19'22"N), in the Seine valley in northern France (Fig. 1). The vegetation in the reserve consists of a mixture of different age stands of chalk grassland succession on Cretaceous chalk. Two plots were chosen for further investigations on mineral contents of chalk grassland species.

6The first plot corresponds to a fenced area of 0,7 ha, which is continuously grazed since 1984 with a mean stocking rate of 4 texel sheep.ha-1.year-1. The vegetation of the grassland corresponds to a short heavily grazed grassland dominated by Festuca lemanii.

7The second plot corresponds to a tall grassland abandoned since 1950 and dominated by tussock grasses, e.g. Brachypodium pinnatum and Sesleria albicans with encroachment by some scrub species, e.g. Acer pseudoplatanus, Betula pubescens, Clematis vitalba, Cornus sanguinea, Corylus avellana, Crataegus monogyna, Fraxinus excelsior, Ligustrum vulgare, Prunus spinosa, Rosa canina, Viburnum lantana.

8More details of the vegetation of the site can be found in Liger (1952) and De Foucault & Frileux (1988).

MATERIALS AND METHODS

9Each month during the growing season (from may to September 1992), an handful of the dominant species and a quadrat of 1 m2 (4 plots of 50x50cm) were randomly harvested in the two plots. The vegetation and individual plants were clipped at 3 cm from ground level and dried at 70° C for 24 hours. In June 1992, five soil samples were also collected in the two plots. Leaves and barks of scrubs were respectively sampled in august 1992 and in January 1993 because these periods correspond to the maximum of sheep consumption (Hillegers, 1983-84). For the statute of plant species as far as sheep consumption is concerned (Tab. 1 & 2), reference is made to Hillegers (1993) and on personnal observations.

10A point quadrat method was used to investigate the vegetation in the two plots (Stampfli, 1991). For each plot (grazed, ungrazed), two permanent 1 m2 quadrats were marked for observations in homogeneous areas. For each of the four quadrats, point frequency records of 100 regularly spaced points (2 cm), consisting of 2 rows at regular distance of 33 cm were made with a needle which was long enough to contact even the tallest plants. Special care was given to prevent disarranging by the observer. These quadrats were used to assess the Specific Contribution (SC%) of dominant plant species, plant richness and diversity. The SC% is the ratio between the frequency of a specie to the total of frequencies of all the species recorded.

Fig. 1: Map showing the study area in the Saint-Adrien nature reserve. [1] grazed plot. [2] ungrazed plot

11The plants which occur outside the quadrats were also noted in order to record the whole plant community. This was undertaken in June 1992 with complementary visits in order to monitor the plants which occurred later in the year.

12Plant and soil analyses were carried out in the laboratory of Grassland Ecology (University of Louvain, Belgium). Chemical analyses of plants were performed on dry samples. Total mineral contents (CT) was obtained by calcination (450° C) and nitric acid digest. After filtration, insoluble ash (Cl) are quantified. Soluble ash (CS) were measured by atomic absorption spectrophotometry for K, Na, Mg, Ca, Fe, Cu, Zn, and Mn. Phosphorous was obtained by the same digestion and colorimetric determination using ammonium metavanate (Lambert, 1992).

13The pH of soil in a slurry with water was determined using a glass electrode. Nitrogen and carbon were analysed using respectively Kjeldahl and Anne methods. Analysis of extractable major(K2O, P2O5, Na2O, MgO, CaO) and trace elements (Fe, Cu, Zn, Mn) were realised by atomic absorption spectrophotometry after extraction with ammonium acetate and EDTA (pH = 4,65).

14Data about mineral concentration of soils and vegetation samples were analysed by the mean of variance analysis. Data about mineral contents of plant species were analysed by the mean of PCA with the help of STAT-ITCF software.

RESULTS

15The species identified and analysed in the above-ground vegetation of the two plots are listed in Table 1.

16In the grazed plot, intensive grazing has led to an important contribution of F. lemanni and Achillea millefolium while B. pinnatum and S. albicans are more rare than in the abandoned plot. Some annuals species (Briza media, Linum catharticum, Medicago lupulina, etc.) appeared in the sward because they can seed in the space of bare ground created by sheep trampling. In the abandoned grassland, the recovering of B. pinnatum prevents the regeneration of these annuals and only long-lived species can persist (Genista tinctoria, Teucrium chamaedrys). Consequently, in the grazed plot, there is an increase of plant evenness (0,74) in the opposite of the abandoned grassland where competitive species overtop the plant community.

17Results of soil analysis are given in Table 3. There is no difference between chemical soil composition of the two plots except pH and Fe. The chemical analysis show low levels of Fe and Cu in relation with the low content of the chalk for these elements. Likewise, turf analysis show that trace elements of the two plots are particularly low notably for Cu, which content is significantly under the threshold of animal deficiency (Tab. 4). At the scale of the turf, there is no significant difference between mineral contents of the grazed and the ungrazed plot.

Species

Code in PCA

Plot 1.
Grazed SC%

Plot 2.
Ungrazed SC%

Poaceae

Brachypodium pinnatum (g)

Bpi

15.68

28.52

Sesleria albicans (g)

Sal

1.78

15.77

Festuca lemanii (g)

Fle

20.41

2.01

Koeleria macrantha (g)

-

3.55

1.01

Briza media (g)

-

0.59

+

Bromus erectus (g)

-

0.30

Poa pratensis (g)

Ppr

0.59

Trisetum flavescens (g)

-

7.40

0.67

Avenula pubescens (g)

-

0.67

Fabaceae

Lotus comiculatus (g)

Lco

0.30

Medicago lupulina (g)

Mlu

6.80

Anthyllis vulneraria (g)

Avu

0.30

Ononis spinosa (u)

Osp

0.30

Genista tinctoria (u)

-

2.35

Others families

Bryophytes (u)

-

2.37

9.06

Achillea millefolium (g)

Ami

15.98

6.71

Hieracium pilose lla (u)

-

6.21

3.02

Carex flacca (g)

Cfl

4.44

5.03

Leontodon hispidus (u)

Lhi

2.66

0.67

Sanguisorba minor (g)

Smi

0.30

0.67

Galium pumilum (g)

-

0.30

2.01

Teucrium chamaedrys (u)

Tch

0.30

13.09

Seseli libanotis (u)

Sli

2.37

0.34

Campanula rotundifolia (u)

Cro

1.18

+

Leontodon hyoseroïdes (u)

-

0.59

Linum catharticum (u)

Lca

1.18

Leucanthemum vulgare (g)

-

1.18

Ranunculus bulbosus (u)

Rbu

2.37

Origanum vulgare (u)

Ovu

+

2.01

Knautia arvensis (u)

-

1.34

Vincetoxicum hirundinaria (u)

-

1.34

Cornus sanguinea (u)

-

1.01

Pulsatilla vulgaris (u)

Pvu

+

0.34

Centaurea scabiosa (u)

Csc

2.35

Daucus carota (g)

Dca

+

Veronica teucrium (u)

Vte

+

Thymus praecox (u)

Tpr

+

Asperula cynanchica (u)

Acy

+

Helianthemum nummularium (u)

-

+

Digitalis lutea (u)

.

+

Hypericum perforatum (u)

Hpe

+

Teucrium scorodonia (u)

Tsc

+

Eryngium campestre (u)

-

+

Galium mollugo (g)

-

+

Stachys recta (u)

-

+

Primula officinalis (u)

-

+

Scabiosa columbaria (u)

Sco

+

Polygala calcarea (u)

Pca

+

Bupleurum falcatum (u)

Bfa

+

Hieracium murorum (u)

Hmu

+

Solidago virgaurea (u)

Svi

+

Plant richness

32

35

Plant evenness (Shannon index)

0.74

0.67

Tab. 1: Floristic composition of the two plots, (g) grazed species, (u) ungrazed species, following (Hillegers (1984). (+) neighbouring plants occuring outside the quadrats. See text for the calculation of SC%

Species

Code in PCA

Leaves

Code in PCA

Barks

Acer pseudoplatanus

1

+/- grazed

A

Grazed in late winter

Betula pubescens

2

Heavily grazed

B

Ungrazed

Clematis vitalba

3

Heavily grazed

C

Ungrazed

Cornus sanguinea

4

Heavily grazed

D

Grazed in late winter

Corylus avellcma

5

+/- grazed

E

Grazed in late winter

Crataegus monogyna

6

Heavily grazed

F

Grazed in late winter

Fraxinus excelsior

7

Heavily grazed

G

Grazed in winter

Ligustrum vulgare

8

+/- grazed

H

Ungrazed

Prunus spinosa

9

+/- grazed

I

Grazed in late winter

Rosa canina

10

+/- grazed

J

Ungrazed

Viburnum lantana

11

+/- grazed

K

Grazed in late winter

Tab. 2: Consumption statute of leaves and barks of scrubs by sheep following Hillegers (1984)

Tab. 3: Comparison of soils analysis (n=5) of the two plots (significance: **p<0.01; * p<0.05; NS Not Significant)

Trace elements

Fe
ppm

Cu
ppm

Zn
ppm

Mn
ppm

Threshold of animal deficiency

4-5

7

45

45

1. Grazed plot

80,1**

5.8*

35.9 NS

39.6 NS

2. Ungrazed plot

69,8**

5.7*

37.7 NS

29.6 NS

Tab. 4: Comparison of the mean concentration (n=5) of trace element in turf of the two plots with treshold of animal deficiency following Jarrige (1988). (significance : ** n<0.01 ; * n<0.05 ; NS Not Significant)

18Specific plant mineral analyses show some differences in the distribution of each element between dominant species. Multivariate analyses on the mean concentration matrix show the mineral that influence variation in the dominant plants analysed in the two plots (Fig. 2). The first three axes account for almost 70% of total variation contribution with 37,9% for axis 1 and 19,3% for axis 2.

19The first axis is strongly influenced by trace elements Fe (corr. 0,83), Cu (0,87), Zn (0,88) and Mn (0,86). Species with a high content of trace elements correspond to the ungrazed species of plot 1 e.g. Ranunculus bulbosus, Thymus praecox, Veronica teucrium (Fig. 3). Axis 2 is more correlated with Ca (0,79), CT (0,74) and Mg (0,66). Species with high content of these elements correspond to Fabacea e.g. Ononis spinosa, Anthyllis vulneraria and ungrazed species present in the two plots e.g. Origanum vulgare, Leontodon hispidus, Seseli libanotis. Some species, which are eliminated by trampling in plot 1, are also rich on these elements and correlated with axis 2 (e.g. Centaurea scabiosa, Teucrium scorodonia). At the opposite of the two axes appear graminoïds (e.g. B. pinnatum, F. lemanii, S. albicans, Carex flacca) which are especially low for all the mineral except CI.

20Projection of scrubs data as supplementary individuals (i.e. mineral contents of leaves and barks) shows that barks have low amount for the majority of minerals except Cl (Ca in the case of C. monogyna). Only leaves of A. pseudoplatanus are particularly correlated with axis 1 (0,65) compared with the leaves and barks of the other scrub species.

21PC A carried out on mineral contents of leaves and barks for scrubs shows that the first axis (32,1%) is strongly correlated with CT (0,80), P (0,80), CI (0,69), Mg (0,74) and K (0,61). Leaves are positively correlated with this axis except those of B. pubescens and C. avellana (Fig. 4). At the opposite, barks are negatively correlated with this axis (except A. pseudoplatanus) as are mineral contents on Na (-0,51) and Zn (-0,21). As for Axis 2 (18,9%), it is strongly correlated with Ca (0,77). Grazed leaves and barks, particularly rich on Ca, CT and CI, are positively correlated with this axis (e.g. C. sanguinea,, C.monogyna, C. vitalba, F. excelsior). In the opposite, ungrazed barks (e.g. B. pubescens, C. vitalba, L. vulgare, R. canina) and +/-grazed leaves (e.g. V. lantana, L. vulgare, R. canina, P. spinosa) are negatively correlated with axis 2. The higher concentration of Na and Zn in barks compared with leaves can be seen in Figure 5.

Fig. 2: PCA carried out on the mean concentration (n=5) of plant mineral contents in the two plots (circle = grazed; square = ungrazed) of a chalk grassland. The code of herbaceous species is given in table 1. Keys for scrub species (I to 11 for leaves and A to K for barks) are given in tab. 2. See text for the definition of groups

Fig. 3: Comparison of the mean concentration (n-5) of plant mineral contents in a grazed (plot 1) and an ungrazed (plot 2) chalk grassland. The code of herbaceous species is given in tab. 1

Fig. 4: PCA carried out on mineral contents of leaves (1 to 11) and barks (A to K) of scrubs which invade chalk grasslands. Key for scrub species are given in tab. 2. See text for the definition of groups

Fig. 5: Mineral contents of leaves (I to 11) and barkes (A to K) of scrubs which colonised chalk grasslands. Keys for scrub species are given in tab. 2

DISCUSSION

22The weakly yield of most chalk grassland turf and unpalatibility of their dominant species is a well-known phenomenon (Williamson, 1976; Hillegers, 1984). As regard to our results and threshold of animal deficiency currently admitted by agronomists (Jarrige, 1988), mineral contents of chalk grassland turf and plants are also particularly low notably on trace elements. This could be linked with the scarcity and the unavailability of these elements in the chalk and shallow rendzina (Dutil, 1972).

23If the use of livestock can be a good implement to restore floristic diversity of abandoned chalk grassland (Bakker et al, 1983; Bacon, 1990; Alard et al, 1991), the grazing systems must take into account the low nutritive value of chalk grassland turf and the availability of the different elements in palatable and unpalatable herbaceous species and scrubs.

24Sheep grazing have a good incidence for plant diversity and subsequently mineral diversity. For example, B. pinnatum which is dominant in the abandoned plot, is also one of the more poor species. If grazing leads to a higher diversity in the plant community (Alard et al., 1994), it allows also the occurrence of new species which represent new mineral sources (Table 2). This can be of interest provided that they are eaten by animals. Nevertheless, several years of heavy grazing with improved sheep (Texel) have been necessary to broke the mat of dead material and reduce the coarse grasses (B. pinnatum, S. albicans). Then, there is a risk of mineral deficiency for these animals, which stay a long time in the same plot of unimproved chalk grassland, and eat preferably poor mineral species like Poaceae (Lamand, 1987).

25When biological conservation is the aim of management, this problem couldn't be resolve by the use of mineral fertilisation because of the decrease of plant richness and diversity (Bobbink, 1991; Willems et al, 1993). In the same way, the utilisation of a salt lick could enhance the apparition of undesirable ruderals species and trampling areas around the salt block (Chappel et al., 1971; Duvigneaud et al., 1990).

26Denudt & Lambert (1976) have shown that the maintenance of plant diversity in extensive hydrophilous meadows is favourable to the mineral balance of animals. Nevertheless, livestock must be able to eat mineral-rich species plants which are usually refused by improved animals. Therefore, the solution to restore species-rich chalk grasslands could be the realisation of an intermittent grazing system with unimproved animals, where periods of grazing and stocking-rate are in agreement with plant diversity and animal health (Wells, 1980; Willems, 1983).

27Sheep are the preferred grazers because they nibble close to the ground but without damaging the rootstocks or growing points of the plants (Grant, 1985). The choice of unimproved races of sheep (Mergelland in Dutch; Beulah in England) have shown that these animals could survive in chalk grassland without inwentering or depending on a large supply of extra food. In consequence, they consume coarse grasses and species-rich mineral plants forsaken by improved animals. They penetrate into scrub rather than staying in the open grassland as a flock. Therefore, the consumption of leaves, barks and shoots of the bushes doesn't require a high stocking rate in enclosed plots. Our study shows equally that mineral contents of barks are lower than herbaceous species. Nevertheless, they are more rich on Na, Zn and Cu than leaves. This could explain the wearing down of scrub barks in winter, when leaves and herbaceous species are rare.

CONCLUSION

28The long term solution for biological conservation of chalk grassland must involve grazing, but the management system to be employed will differ depending on, the objectives of conservation (animal or plant diversity, scrub or coarse grass decrease), the statute of the land (farmland, nature reserve, amenity area etc..), and the type of animals involved. If very good results have been achieved on England nature reserves by heavily grazing temporary paddocks of electrified “flexi-netting”, the use of an unimproved sheep livestock conducted by a shepherd could show also very good result without problems of unpalatable species and mineral deficiencies.

Table des illustrations

Légende Fig. 1: Map showing the study area in the Saint-Adrien nature reserve. [1] grazed plot. [2] ungrazed plot
URL http://books.openedition.org/purh/docannexe/image/8101/img-1.jpg
Fichier image/jpeg, 56k
Légende Tab. 3: Comparison of soils analysis (n=5) of the two plots (significance: **p<0.01; * p<0.05; NS Not Significant)
URL http://books.openedition.org/purh/docannexe/image/8101/img-2.jpg
Fichier image/jpeg, 32k
Légende Fig. 2: PCA carried out on the mean concentration (n=5) of plant mineral contents in the two plots (circle = grazed; square = ungrazed) of a chalk grassland. The code of herbaceous species is given in table 1. Keys for scrub species (I to 11 for leaves and A to K for barks) are given in tab. 2. See text for the definition of groups
URL http://books.openedition.org/purh/docannexe/image/8101/img-3.jpg
Fichier image/jpeg, 56k
Légende Fig. 3: Comparison of the mean concentration (n-5) of plant mineral contents in a grazed (plot 1) and an ungrazed (plot 2) chalk grassland. The code of herbaceous species is given in tab. 1
URL http://books.openedition.org/purh/docannexe/image/8101/img-4.jpg
Fichier image/jpeg, 88k
Légende Fig. 4: PCA carried out on mineral contents of leaves (1 to 11) and barks (A to K) of scrubs which invade chalk grasslands. Key for scrub species are given in tab. 2. See text for the definition of groups
URL http://books.openedition.org/purh/docannexe/image/8101/img-5.jpg
Fichier image/jpeg, 40k
Légende Fig. 5: Mineral contents of leaves (I to 11) and barkes (A to K) of scrubs which colonised chalk grasslands. Keys for scrub species are given in tab. 2
URL http://books.openedition.org/purh/docannexe/image/8101/img-6.jpg
Fichier image/jpeg, 76k

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search