Ecology, Management and Monitoring of Grey Dunes in Flanders

Journal of Coastal Conservation 10:33-42, 2004 © EU CC; Opulus Press Uppsala. 33

Ecology, management and monitoring of grey dunes in Flanders

Provoost, Sam1*; Ampe, Carole2; Bonte, Dries3; Cosyns,4 5 & Eric Hoffmann, Maurice14

1Institute of Nature Conservation, Kliniekstraat 25, B-1070 Brussels, Belgium; 1Ghent University, Laboratory of Soil Science, Krijgslaan 281, B-9000 Gent, Belgium; [email protected] ; 3Ghent University, Research group terrestrial Ecology, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium; E-mail [email protected]; 4Ghent University, Research group terrestrial Ecology, Krijgslaan 281, B-9000 Gent, Belgium; [email protected] ; * Corresponding author; E-mail sam.provoost@instnatbe

Abstract. Grey dunes are a priority habitat type of the Euro Introduction pean Union Habitats Directive and demand special attention for conservation and management. Knowledge of the ecology Grey dunes, defined as ‘fixed coastal dunes with of coastal grey dunes can contribute to this policy. Dune herbaceous vegetation’ in the CORINE biotope classifi grassland succession is initiated by fixation and driven by the cation, are considered a priority in the Annex I of the EU complex of soil formation (humus accumulation) and veg etation development. Leaching and mobilization of CaCO,, Habitats Directive. This status implies that grey dunes which are important in nutrient dynamics, complicate the deserve special conservation attention. Therefore it is picture. At present, grass- and scrub encroachment greatly useful to focus on the ecology of this landscape and its overrules these fine scaled soil processes and causes substan vegetation and to discuss the contribution of ecological tial loss of regional biodiversity. Belgium has an interna research to conservation strategy, management and moni tional responsibility in grey dune conservation because of toring. This paper summarizes the experience in Bel the limited range of its characteristic vegetation, flora and gian coastal dunes, based on several research projects. fauna. As biomass removal seems essential in grassland This research is to a large extent funded by the Nature preservation, grazing is an important management tool. Evalu division of the Ministry of the Flemish Community, ation of management measures focuses on biodiversity meas with contributions of LIFE. urements on the levels of landscape, community and species. Coastal grey dune is most readily defined using plant communities. Phytosociologically, grey dune veg

K eyw ords: Ammophila arenaria; Belgian coast; Biodiver etation includes moss dunes and dune grasslands (with a sity; Décalcification; Dune fixation: Habitats Directive. distinct A horizon), belonging to theCladonio-Koele- rietalia. In landscape-ecological terms grey dunes can be considered as the dry component of the ‘stressed dune landscape’ (Provoost 2004). In such a landscape, bio logical dynamics show a certain equilibrium between ‘top down’ regulating stress factors and ‘bottom up’ community succession. Dominance of top down regu lation, for instance by blowing sand, would lead to wards a ‘dynamic landscape’ - in casu marram dunes with Marram grass, Ammophila arenaria, mixing with the moss- and lichen-dominated Cladonio-Koele- rietalia vegetation, while dominance of bottom up organization would lead to the development of scrub and woodland with species such as Hippophae rham noides and Ligustrum vulgare. The main differentiating processes are related to dune fixation, soil formation, vegetation development and management. 34 Provoost, S. et al.

Processes in grey dunes additional effect of water repellence, caused by soil organic matter, is an underestimated factor of drought. Dune fixation and erosion This drought stress regime can strongly retard vegeta tion succession and soil development. Dune fixation implies a decrease in aeolian activity Despite these very harsh conditions, even in the and stabilization by vegetation (Jungerius 1990). Rhi early moss dune stages some biomass production can zome forming plants such asAmmophila arenaria and take place. It is the key element in soil development. Festuca juncifolia, which keep up with a certain sand Humus colloids, produced by decomposition of organic accumulation, play an important part and decaying tus matter, increase moisture retention and nutrient avail socks of Ammophila arenaria become a substantial ability. Accumulation of humus is reflected in species source of nitrogen in the early stages of grey dune composition and species’ competitive abilities; in Bel development. gian dunes, plant species such Sagina as nodosa and In Europe, mobility of dunes is generally affected by Linum catharticum occur beyond the reach of ground coastal defence works (stabilization) and human activi water only if soils are rich in organic matter. Since soil ties leading to erosion. Centuries of agricultural use of and vegetation processes are highly interwoven, early dunes have probably caused a great deal of secondary stages of dune grassland development represent a clas sand drift and as these activities ceased somewhere sic example of internal succession according to the during the 20th century, the dune landscape became resource-ratio model of Tilman (1985). Many of the fixed and vegetated. Other elements, such as the mili processes take place in the rhizosphere and are driven by tary activities during World War I, have been quite competition for water and nutrients. According to Slings devastating in a much shorter time span. More recently, (1994), grass roots are the main source of organic matter intensive recreation causes soil and vegetation degrada in grey dune soils. Veer & Kooijman (1997) found that tion, often resulting in local dune mobilization. root and moss biomass each are twice as large as above ground vascular plant biomass, so probably mosses play Phimus accumulation and decomposition an important role as well. Soil humidity is a main determining factor in bio The initial dune substrate is not very hospitable for mass production and soil development. The effect of the plant growth. Parent material of Belgian dune soils groundwater level on soil organic matter in Belgian consists of well drained sand (mean median grain size coastal dunes is illustrated in Fig. 2. Data are derived between 175 and 250 pm) with a low nutrient content. from 267 soil samples in the western part of the Belgian Drought stress is caused by climatological, geo- coast (Provoost & Ampe unpubl.; Waumans 2001). morphological and soil factors. Grey dune surface ex Organic matter was determined with loss on ignition posed to the sun can heat up to more than 60 °C (Fig. 1) (500 °C during 8 hr) and groundwater levels are interpo and soils can dry out down to 20 cm deep (Aggenbach & lated from nearby surrounding piezometers. Taking into Jalink 1999). Especially in moss-covered dunes, the account annual groundwater fluctuations of 0,5 to 1 m, the capillary effect and rooting depth, groundwater ex °C ceeding a depth of about 2 m under the soil surface is no longer accessible for most grassland plants. In this dry 60 zone, soil organic matter contents rarely exceeds about 50 - %OM

•45 - 30 - 40 -

- 20 35 -

10 - 30 -

25 -

20 - 0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00 15 -

10 - —♦-Air temperature on sunny day 5 - ■ Surface temperature (grey dune) on sunny day 0 - Air temperature on cloudy day - 0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0

F ig-1. Air and soil surface temperature on a sunny and cloudy Mean groundwater level (m below surface) day (17 August and 30 August 2001, respectively) on a south- Fig. 2.Soil organic matter content (loss on ignition) in the A- facing dune slope in De Westhoek, De Panne, Belgium. horizon in relation to groundwater level (see Text). - Ecology, management and monitoring of grey dunes in Flanders - 35

6 %. In moist dune grasslands, where the groundwater Acidity is an important parameter, affecting many table does not drop below 2 m but soils never get inun soil processes. Decomposition and mineralization of dated, maximum organic matter contents of 10 to 15 % organic matter is slowed down in acid soils (pH < 4,5 to are observed. These figures are comparable to the situa 5), resulting in an increase in the humus content tion in dunes on Goeree in the southwestern Netherlands (Aggenbach & Jalink 1999). During soil development, (Annema & Jansen 1998). Wet soils are permanently humic acids are released which enforce leaching of influenced by groundwater and show a spectacular in CaC03. This means that humus production and crease in (maximum) soil organic matter content. décarbonatation are involved in a positive feedback Humus accumulation is counteracted by several proc loop. esses. In young moss dunes, physical humus erosion can Décarbonatation of grey dune soils is slowed down be substantial, but in stable grassland, main losses of by mobilization of calcareous sand. Bioturbation is prob soil organic matter are due to biochemical decomposi ably an essential element. Ants (e.g.Formica cunicu tion. The ions formed by mineralization processes and laria) certainly increase the superficial lime content CaC03 are susceptible to leaching, caused by the pre (Bonte et al. 2003). Coprogenic fauna, becoming cipitation surplus. Bioturbation enforces decomposition more abundant after the introduction of large herbiv and mineralization. According to Bonte et al. (2003), ores in the grey dunes, bring calcareous sand towards ants affect the thickness of the A-horizon and the soil the surface when constructing their brood chambers penetration resistance (Fig. 3). These animals as well as (Ampe et al. 2002). Weeda (1992) also considers snails and other animals, are abundant in organic soil ‘sand spray’, a slight sand accretion, as an important layers. and characteristic natural cause of recalcification. In general, decomposition rates show an optimum- According to this author, this process prevents soil type response in relation to soil moisture. Within the acidification and enhances decomposition and miner humidity range of dune grassland, the net effect of alization of organic material inTaraxaco-Galietum biomass production and mineralization leads to the pat dune grasslands. tern in soil organic matter contents shown in Fig. 2. Nitrogen and phosphorus are key elements and partly co-limiting in the nutrient dynamics of grey dunes Soil décalcification and nutrient dynamics (Kooijman et al. 1998). Nitrogen is mainly supplied by decomposition of plant residues. The total nitrogen con Stabilized dune soils decalcify due to continuous tent is highly correlated with soil organic matter and carbonate leaching. At the time of deposition, the sand is generally increases during succession. Mineralization calcareous (up to 8 % CaC03) due to shell fragments in plant-available nitrogen shows a different response (Depuydt 1972). CaC03 is the main buffering substance and is rather correlated with standing crop biomass and and pH remains neutral to alkaline (6,5 up to 8) as long plant litter quality. Phosphorus availability is pH de as CaC03 concentration exceeds about 0,3 %. But be pendent and shows an optimum at pH 5.2 in calcareous low this point soil acidity shows an abrupt and steep dunes in The Netherlands (Kooijman & Besse 2002). At decline, which means that the gradual character of higher pH, phosphorus is largely fixed in calcium phos décarbonatation is not reflected in soil pH (Rozema et phate, limiting biomass production. Physiological adap al. 1985; Aggenbach & Jalink 1999). tations to this phosphorus limitation are characteristic of calciphilous plant species.

Thickness A horizon (cm) Vegetation development

20 In natural dune systems, fine-scaled soil processes as mentioned above, physical destruction by blowing 15 sand and invertebrate herbivory might be sufficient to 10 preserve grey dune vegetation and some its characteris tic species, at least within an ephemeral pattern. But at 5 present, substantial removal of plant biomass seems essential for sustainable dune grassland development. 0 Coastal dune vegetation in Belgium and many other L. psammophilus F. cunicularia European countries evolves unidirectionally towards Fig. 3.Effects of the presence of nests of the ants Lasius coarse grassland with Calamagrostis epigejos, Avenula psammophilus and Formica cunicularia on the thickness of pubescens or Arrhenatherum elatius, scrub with the A-horizon (Bonte et al. 2003). Hippophae rhamnoides, Ligustrum vulgare or Salix repens 36 Provoost, S. et al. and woodland (van Til et al. 2002; Provoost & Van with species suchas, Asperula cynanchica,Helianthemum Landuyt 2001). Grass encroachment is probably stimu nummularium and Thesium humifusum. Salisbury (1952) lated by atmospheric deposition but once established, described the same phenomenon of ‘species supply’ of grass-dominated vegetation is highly self-maintaining British dunes from the downs. Differentiation in inver due to increased nitrogen mineralization (Kooijman et tebrate communities is detected in the function of the al. 1998; Veer & Kooijman 1997). Scrub expansion is surrounding landscape: if woodland, heathland or chalk mainly caused by drastic changes in land use. Up to the grassland is present in the proximity, typical species middle of the 20th century, livestock grazing was com from these habitats also colonize grey dunes, probably mon in Belgian dunes and shrubs and trees were har within a source-sink context. vested for fuel. Because of these activities the grey dune On a landscape level, distribution of grey dunes is area covered an unnaturally large portion of the coastal limited to dry or moderately moist zones within fixed dunes. dunes. As most of the dune ridges along the Belgian Grazing and mowing can maintain vegetation struc coast were still mobile some decades ago, only young ture but do not stop succession (De Raeve 1989). Soil moss dunes are found in these zones (apart from scrub). and vegetation development continues through a differ Fixed dune grassland with a distinct soil profile is lim ent pathway, subject to new processes such as biomass ited to intermediate elevations within large dune slacks removal, trampling and dung deposition. The natural or to relief-poor inner dunes. Also patterns in soil acid processes described above can lead to a moder humus ity are largely revealed on a landscape level. Certainly form, whereas in situations with livestock grazing a within a particular moisture range, pH is mainly influ mullmoder develops (Slings 1994; Aggenbach & Jalink enced by décalcification, which in its turn is mainly time 1999). Such differences can be critical for the occur dependent. Thus general grey dune types appearing in a rence of (plant) species, which is shown for Viola dune area relate to the landscape’s ontogeny. rupestris in The Netherlands by Weeda (1992). Stress factors form another element in grey dune distribution. The ‘sand spray’ effect for example, will only take place in the vicinity of mobile dunes. Herbivory Spatial patterns in grey dunes pressure can be linked to rabbit density and present-day or historical livestock grazing. In Dutch dunes, an effect Regional and landscape level of past agricultural activities related to sea villages is reflected in vegetation and species composition (Slings Biogeographical patterns are related to climate, soil 1994) but this is not distinct along the Belgian coast. and history. Atlantic grey dunes are subject to a major Finally, fragmentation of grey dunes due to scrub stands gradient in temperature from southwest up to northeast decreases the number of typical invertebrate species and a substantial variety in precipitation (Dahl 1998). within each isolated patch (Bonte et al. 2002) because of Coastal dunes of Belgium and the adjacent areas take an smaller surfaces and metapopulation dynamics, where intermediate position with a mean temperature of the colonization and extinction depend on connectivity. coldest month of 2 °C and an annual precipitation of ca. 750 mm. During the months May till August a deficit Local level occurs of ca. 100 mm. Considering the importance of drought stress on grey dunes, climate is determinative Grey dune landscapes are usually characterized by for grey dune development and different vegetation an explicit microrelief, causing a fine-scale variability types can be distinguished according to the geographi in microclimate and edaphic factors. These elements cal position (Weeda & Schaminée 1990). Invertebrate primarily determine species composition and perform communities from North-Atlantic grey dunes respond ance. Superimposed, plant population dynamics, veg to these climatological differences as well and are clearly etation architecture and biotic interactions contribute to reflected in the declining presence of Mediterranean the development of complex vegetation patterns. As species at higher latitudes. many biological processes such as seed dispersal or Species pools are also affected by their geographical herbivory are highly stochastic, prediction of species position. North of Belgium, sand deposits and dunes occurrence becomes precarious (see e.g. Lichter 2000 along the rivers Rhine and Meuse act as an ecological for a dune example). corridor between the coast and inland habitats for species The response of several plant species to soil proper such asEuphorbia cyparissias and Eryngium campestre ties suggests that species assemblages change in func (Doing 1993). But Belgian dunes are more clearly influ tion of development time, which underlines the impor enced from the south, by the chalk-grasslands in north tance of the complex processes of plant growth, biotic ern France. Most probably, they enriched the dune flora interactions and soil formation. Terrestrial springtails - Ecology, management and monitoring of grey dunes in Flanders - 37 respond positively to this soil characteristic as well. species, for example Rosa pimpinellifolia, will hardly Spiders, which are their predators, aggregate in a similar affect the position in a vegetation classification but pattern, not because of direct responses but indirectly by might be of major importance for an invertebrate’s co-evolution for similar microhabitats. habitat. A local typology for management planning or In general, insect and spider communities largely evaluation or other ecological applications can be more depend on vegetation structure and assemblage stability useful if physiognomic properties and species domi is strongly influenced by the landscape patterns. Small nance is emphasized. Also, the use of fuzzy classifica patches are subject to significant edge effects, resulting tion techniques, with gradual transitions between dis in a high number of atypical species and a high spatial tinct types, might be helpfuk. community instability (Bonte et al. 2002). Communities in large patches are more stable due to higher internal microhabitat variation and the possibility of attaining Nature conservation and management viable population sizes. Biotic interaction also explains certain patterns in species composition: a number of Why preserve grey dunes! invertebrates (monophagous beetles, bugs and butter flies) are bound to typical host plants, while other inver In conservation policy, biodiversity - or rather the tebrates are bound to typical invertebrate hosts. Ants are loss of it - is generally used as a framework for nature peculiar in this respect since they host very rare evaluation (Reid & Miller 1989). In the ecosystem vi myrmecophilous spiders, bugs, woodlice and probably sion on coastal dunes of the Flemish government, this is many more invertebrate taxa. translated as ‘conservation or development of the char In the other direction, invertebrates can also influ acteristic habitat diversity by as natural as possible ence the presence of plant species. Ants again are of means’ (Provoost & Hoffmann 1996). major importance since their nests create suitable ger In an international context, the importance of grey mination conditions for several plant species (e.g.Thymus dunes is reflected in the designation as a priority habitat pulegioides). Additionally, they transport seeds to these in the EU Habitats Directive, mainly because of its nests. limited range. The Cladonio-Koelerietalia are consid ered a northern Atlantic vegetation type (Weeda & Schaminée 1990). Most of the characteristic grey dune Classification of grey dune vegetation plant species appearing in Belgium however, are not threatened, at least not in a European context. None of Phytosociological vegetation research contributed sub its characteristic species is included in the Bern Conven stantially to the present knowledge of dune grasslands tion list. A number of species, such asErodium lebelii (= (e.g. Rodwell 2000). According to Weeda et al. (1996), E. glutinosum), and probably some infraspecific mari the vegetation of most of the Belgian grey dunes belongs time varieties of Ononis repens, Koeleria albescens and to the phytosociological order Cladonio-Koelerietalia, others, are limited to Atlantic grey dunes (Doing 1993; an order of calcicole dune vegetation within the class Roisin 1969). This underlines the habitat specificity and Koelerio-Corynephoretea. Within the order two alliances the international responsibility for conservation. are distinguished: the Tortulo-Koelerion, including pio Asparagus prostratus is probably the best ‘flagship’ neer grasslands and moss-covered dunes and thePolygalo- species for Flanders, since it is strictly Atlantic and, at Koelerion containing closed grasslands on soils with a least regionally, rare and limited to grey dunes. Recent distinct organic layer. Within these alliances several research has confirmed its status as a separate species, associations have been distinguished. Their composi distinct from A. officinalis (Kay et al. 2001). tion of which depends largely on local circumstances; Within the Flemish region, grey dunes are important each segment of the European coast is after all charac for conservation of Red List species, including regionally terized by a unique combination of climate, geomor rare chalk grassland species, e.g.Thesium humifusum, phology, soil (e.g. reflected in CaC03 content) and Cirsium acaule and Asperula cynanchica, or species history. Belgian dune grasslands were described phyto- characteristic of grasslands of dry to moist nutrient-poor sociologically by Heinemann (in Lebrun 1949) as soils. An thy I lido-Thesietum. From an invertebrate fauna perspective, grey dunes Phytosociological classification, with an emphasis are the most endangered dune habitat, certainly within a on species composition, is quite abstract and often field Flemish context. Most characteristic dune invertebrates situations are hard to fit into the scheme because they however, are found in more dynamic habitats such as are degraded, in a subdeveloped stage, or are intermedi mobile dunes and young dune slacks. ate between ‘pure’ associations. The presence of one 38 Provoost, S. et al.

Threats and management epigejos, which succeeds in patches of decayed scrub, is successfully tackled and the impact of grazing on these Along the western part of the Belgian coast, the grey and other species-poor grasslands can be significant. In dune area decreased from 730 to 350 ha since the 1950s, most cases biomass of the dominant grass species col mainly due to urbanization and scrub encroachment. In lapses and after 3-4 yr several stress-tolerant dune grass addition, grey dunes are subject to internal degradation land species occur. due to the grass encroachment mentioned above. Until However, grazing does not simply influence scrub now, this evolution was not accompanied by a drastic dynamics but it also affects shrub seed germination and decline in the total number of grey dune species. How development via soil and vegetation structure. Probably ever, there is an apparent qualitative shift toward a less vegetation dynamics under grazing management fits in specific flora, which includes garden escapes and exotic a shifting mosaic pattern similar to the grazed wood species (Provoost & Van Landuyt 2001). lands as described by Olff et al. (1999). This biodiversity decline, at least regionally, urges Recruitment of favourable plant species seems to be managers to take nature management measures. Re promoted not only by grazing activity itself but also by moval of scrub and woodland, mowing and grazing, trampling, scratching or dust bathing (creating favour have proven to be appropriate tools for dune grassland able germination micro sites) and defecation. The latter restoration. Well documented examples of management shows the potential to bring in significant amounts of schemes are available for the Dutch dunes (e.g. Annema viable seeds of several plant species, including some of & Jansen 1998) and the LIFE initiative at the Sefton coast conservation interest (Cosyns et al. subm.). Inverte in the UK (Houston et al. 2001). In Belgian dunes, around brates react even faster on changes in the vegetation and 15 ha of scrub has been removed and currently nearly 350 landscape structure. ha are grazed (Herrier & Killemaes 1998). Mowing is Soil nutrients are redistributed through excretion. only very locally used for grey dune management. Most long-term observations show a decrease in C/N ratio. Simultaneously a large range of dung fauna ap pears in the soil system, increasing bioturbation and Preliminary evaluation of grazing enlarging the soil volume suitable for rooting.

Expectations of grazing as a management tool are often ample. Herbivores should maintain the open areas Monitoring after mechanical shrub removal and prevent scrub en croachment. They are expected to preserve biodiversity Which kind of monitoring? and to transform species-poor vegetation into species rich grasslands. Dune management surpassed the experimental, Vegetation composition responds to grazing in a small-scale stage, also in Belgium. One of the future rather slow manner - over decades rather than years. In challenges will be to elaborate sustainable and efficient order to evaluate short-term effects, the feeding ecology management programmes, with efficiency being meas of introduced ungulates (Konik and Shetland pony, ured mainly in terms of biodiversity gain related to cost. donkey and Scottish Highland cattle) is studied in sev This will require more detailed measurements in moni eral dune areas along the Belgian coast. Diet, habitat use toring and continuous feedback to management plan and behaviour are registered through all seasons and day ning (Fig. 4). periods. The first results show that all domestic herbi In a strict sense, monitoring implies repeated survey vores are mainly grass eaters (70 to 80% of their diet is and an assessment to what degree the results deviate composed of grasses and sedges) and they prefer grass from a predetermined norm (Goldsmith 1991). In prac land to graze in (Hoffmann et al. 2001; Cosyns et al. tice however, a lot of ‘monitoring’ does not meet this 2001). Diet composition is affected by seasonality: dur definition. The surveillance of the environmental con ing winter and early spring especially cattle and donkey text (climate, geomorphology and hydrology) for in consume a significant amount of browseLigustrum ( stance is meant as a multi-purpose frame of reference vulgare or Salix repens). Nonetheless browsing inten which must enable us to estimate the relative impor sity is not sufficient to push back or to prevent scrub tance of general (natural) and local, management driven encroachment (van Breukelen et al. 2002).Hippophae change. rhamnoides, the most prominent shrub species, is hardly Firstly, direct information can be gathered for many eaten by the herbivores. The animals do seem to respond taxonomic groups using various sampling techniques. to the spontaneous vegetation developments. However, there is a strong trade-off between detail and Rough vegetation, dominated byCalamagrostis geographical range. If the inventory effort seems to - Ecology, management and monitoring of grey dunes in Flanders - 39 large, indirect biodiversity measurement using habitat vegetation monitoring but research is still in a early models is more appropriate. This requires the selection stage. of a set of habitat parameters of a certain species or Detailed nature management evaluation requires high species group, which are easily mapped on a large spatial resolutions. At present, satellite sensors offer a 4 surface. And of course, before all, the species’ habitat m resolution multispectrally. Even higher resolutions characteristics must be thoroughly understood. (order of magnitude of dm) can only be achieved with digitized aerial photographs or airborne scanners. For Landscape level management evaluation in some Belgian coastal dunes, digital false colour orthophotos will be used with a Detailed information about topography, preferably spatial resolution of 15 cm. The spectral discernibility in combination with hydrology, is of vital importance in of several grey dune vegetation types on these images is dune research. In a GIS environment, abiotic habitat illustrated in Fig. 5. Vegetation data are derived from characteristics, such as surface slope and exposition, 270 precisely located 1-m2 plots in the Oostvoorduinen can easily be derived from a digital elevation model. (Koksijde). Using this information, highly detailed veg Vegetation structure is a basic biotic layer to be etation structure maps can be produced (Droesen 1998) added. Vegetation mapping using remotely sensed im which can be used in habitat models or for evaluation of, ages is a promising and powerful tool for planning and for instance, effects of grazing on grass encroachment. evaluation of management in spatially heterogeneous In digital grid files, a fuzzy information system can be environments such as coastal dunes. Most remote sens used, in which affinities for several vegetation types can ing techniques are based on detection of reflected solar be assigned to each pixel. radiation. Interaction with earth surface features alters the characteristics of the electromagnetic waves, bring Community level ing about a ‘spectral signature’ (Lillesand & Kiefer 2000). Sand, for example, reflects most radiation, while Evaluation of species composition requires a de green plants absorb red light (wave length 0.6 - 0.7 pm) tailed survey for which a variety of sampling techniques for photosynthesis and reflect near-infrared (NIR; 0.7 - is at our disposal. Plant communities can be mapped in 2 pm) due to properties related to plant tissue structure. classical vegetation maps, but these only contain general These characteristics allow us to estimate the amount of features. Plot measurements are considered on two levels. vegetation biomass using detected reflectance differ Typically, plots with a surface of at most several m2 are ences in red and NIR bands. It is the basic idea of used for vegetation research. They can represent homo- vegetation indices. Many ecological applications in re mote sensing are based on multispectral images, con sisting of several information bands. False colour pho RED reflectance tos for instance, with near infrared, red and green bands 160 are very useful for vegetation mapping. Hyperspectral 140 imagery, consisting of up to hundreds of narrow bands, offers very promising possibilities for more detailed 120 - •• M 100 -

Vision Monitoring EVALUATION 60 -

GOALS Sl 130 140 150 160 170 180 NIR reflectance « Open moss dune ° Closed moss dime with Tortula ruralis MANAGEMENT • Closed moss dime with llypnum cupressiforme ... » MANAGEMENT X Young dune grassland X Rough grassland • Grassland with Rosa pimpinellifolia (Prediction) + Salix repens dwarf scrub

Fig. 5. Spectral signatures of several grey dune vegetation types in red and near infrared bands, based on scanned false Fig. 4.Cyclic process of planning, management and monitor colour orthophotos of the Oostvoorduinen, Koksijde. ing in nature conservation. Reflectances are represented as 8 bit digital numbers. 40 Provoost, S. et al. geneous patches and are necessary to reveal detailed ruda cristata) or on the level of the patch occupancy for information on vegetation patterns and processes (Bakker typical and easily observable invertebrates such as et al. 1996). A stratified random set of some 100 perma Hipparchia semele, Oedipoda caerulescens, wolfspiders nent quadrats of 9 m2 for example, is used for monitor and the Sphecid Bembix rostrata. A follow-up of colo ing the effects of different management regimes at nization and extinction dynamics in particular, can re Belgian coastal dunes (Bonte et al. 1998). veal interesting data for species conservation. However, for management evaluation on the site level, using these small permanent quadrats would im ply a very large number of samples to be representative Conclusions of the whole area. Larger plots, up to several hundreds of m2 (20 X 20,50 X 50) could be more appropriate. An An entwined complex of vegetation and soil devel approximately equal time investment will provide infor opment is the driving force in grey dune succession. In mation on a much larger number of species, although natural situations, fine scaled dynamics of sand, CaC03 subtle changes in abundance will no longer be detect and nutrients are probably sufficient to maintain low able . The results will enable a broader quantitative evalu grey dune vegetation, at least locally. At present how ation, which has the extra advantage of being less sus ever, grey dunes spontaneously evolve towards scrub ceptible to annual fluctuations caused by weather condi and vegetation dominated by coarse grasses. Biomass tions. Plots will be spatially heterogeneous but this removal, e.g. by grazing is essential for preservation of should not be considered a disadvantage, since we aim dune grassland but significantly alters the natural sys at conclusions at a higher resolution level than the local tem’s characteristics. The response of several plants and plant community. Changes can be measured using scores Collembola to soil organic matter suggests the impor for several habitat types, based on species specificity tance of biological development processes beside or on and abundance. top of the underlying abiotic pattern. It forms a chal Standardized pitfall trapping is an effective method lenging area for future research. for monitoring changes in faunal (invertebrate) compo Atlantic grey dune conservation is of international sition. importance due to the limited range of the characteristic flora and vegetation. Syntaxonomical classification is a Species level valuable tool in conservation but shows some major limitations for vegetation mapping. Incorporation of Due to sampling, the collected data for rare species will vegetation structure and physiognomy is important. usually not include a sufficient number of replicas for Fuzzy classification techniques are promising, also in statistical analysis. But just these peculiarities and Red respect to vegetation maps derived from remotely sensed list species are of great interest to the conservationist. images. Especially for perennial vascular plants, detailed and Biodiversity monitoring requires surveys on differ site-covering inventories of a selection of ‘target species’ ent scale levels. A combination of detailed vegetation should therefore be considered. If possible, individuals structure maps and inventory of a selection of character are mapped, otherwise the abundance is estimated within istic and indicative species is a powerful tool for man a patch. This method largely excludes statistical prob agement evaluation. Distinguishing management effects lems since the entire population is considered. The from other changing parameters (natural as well as occurrence of mobile animals, annual plants and fungi is anthropogenic) remains a difficult but essential task in ephemeral and requires repeated extensive inventories. evaluation. An example of integral plant inventories in grey dunes is Grazing management has, at least on short notice, a given by Annema & Jansen (1998) for the dunes of strong impact on vegetation structure, composition and Goeree, The Netherlands. A program is set up for the dynamics and indirectly on faunal elements as well. Its Flemish coast, based on experience in the western re effects on the natural grey dune determinants, such as gion ‘Westkust’ (ca. 2000 ha). Grey dune plant species soil development, are complex and form the subject of included in this inventory are characteristic of moss further research. dunes (e.g.Viola curtisii ox Asparagus prostratus), chalk grasslands (Helianthemum nummularium, Silene nutans or Thesium humifusum) or other grassland types (Rhinanthus minor. Thymus pulegioides or Briza media). A program for detailed mapping of animal populations is worked out on the individual (territorial) level for birds ( Oenanthe oenanthe, Saxicola torquata or Galli- - Ecology, management and monitoring of grey dunes in Flanders - 41

References dune area in the first year after introduction. In: Houston, J.A..Edmondson, S.E. & Rooney,P.J. (eds.) Coastal dune Aggenbach, C.J.S. & Jalink, M.H. 1999.Indicatoren voor management, shared experience of European conserva verdroging, verzuring en eutrofiëring in droge duinen. tion practice, pp. 95-107. Liveipool University Press, Staatsbosbeheer, Driebergen, NL. Liveipool, UK. Ampe, C., Ngugi, N.M. & Langohr, R. 2002. Impact of Houston, J.A., Edmondson, S.E. & Rooney, P.J. (eds.) 1998. recently introduced large herbivores on soil properties of Coastal dune management, shared experience of Euro coastal dune soils of the ‘Westhoek' nature reserve, Bel pean conservation practice. Proceedings EUCC sympo gium. In: Veloso-Gomes, F., Taveira-Pinto, F. & das sium, Southport, September 1998. Liveipool University Neves, L. (eds.) Ute changing coast. Proceedings of the Press, Liveipool, UK. 6th international symposium Littoral 2002, Porto, 22-26 Jungerius, P.D. 1990. The characteristics of dune soils. In: September 2002, Vol III, pp. 433-438. Porto, PT. Bakker, T.M.W., Jungerius, P.D. & Klijn, J.A. (eds.) Annema, M. & Janssens, A.E.M. 1998. Het herstel van het Dunes o f the European coasts. Catena Stippi. 18:155-162. vroongrondgebied Midden- en Oostduinen op Goeree. Kay, Q.O.N., Davies, E.W. & Rich, T.C.G. 2001. Taxonomy Stratiotes 17: 20-60. of the western European endemicAsparagus prostratus Bakker, J„ Olff, H„ Willems, J. & Zobel, M. 1996. Why do we [A. officinalis subsp. prostratus) (Asparagaceae). Bot. J. need permanent plots of long-term vegetation dynamics? Linn. Soc. 137: 127-137. J. Veg. Sei. 7: 147-156. Kooijman, A.M. & Besse, M. 2001. The higher availability of B onte, D ., A m pe, C ., H offm ann, M ., Langohr, R ., Provoost, S . N and P in lim e-poor than in lim e-rich coastal dunes in the & Hemer, J.-L. 1998. Monitoring research in the Flemish Netherlands. J. Eco!. 90: 394-403. dunes: from a descriptive to an integrated approach. Pro Kooijman, A.M., Dopheide, J.C.R., Sevink, J., Takken, I. & ceedings European seminar on coastal dunes, manage Verstraten, J.M. 1998. Nutrient limitations and their im ment, protection and research. Skagen.pp. 39-79. Skagen, plications on the effects of atmospheric deposition in DK. coastal dunes; lime-poor and lime-rich sites in the Nether Bonte,D.,M aelfait, J.-P. & Baeri, L. 2002. Spider assemblage lands. J. Eco!. 86: 511-526. structure and stability in a heterogeneous coastal dune Lebrun, J., Noirfalise, P., Heinemann, P. & Vanden berghen, system (Belgium).J. Arachnol. 30: 331-343. C. 1949. Les associations végétales de Belgique.Bull. Bonte, D., Dekoninck, W., Provoost, S., Cosyns, E. & Soc. Roy. Bot. Belg. 82: 105-207. Hoffmann, M. 2003. Microgeographical distribution of Lichter, J. 2000. Colonisation constraints during primary suc ants (Hymenoptera: Formicidae) in a dune grassland and cession on coastal Lake Michigan sand dunes.J. Eco!. 88: their relation to the soil structure and vegetation.Anim . 825-839. B iol. 53: 367-377. Lillesand, T.M. & Kiefer, R.W. 2000. Remote sensing and C osyns, E ., D egezelle, T ., D em eulenaere, E . & H offm ann, M . image interpretation. 4th ed. John Wiley & Sons, New 2001. Feeding ecology of Konik horses and donkeys in York, NY, US. Belgian coastal dunes and its implications for nature man Olff, H., Vera, F., Bokdam, J., Gleichman, J., De Maeyer, K. agem ent.Belg. J. Zool. 131 Suppl. 2: 111-118. & Smit, R. 1999. Shifting mosaics in grazed woodlands Dahl, E. 1998.The phytogeography o f northern Europe {Brit driven by the alternation of plant facilitation and competi ish isles, Fennoscandia and adjacent areas). Cambridge tion. P lant Biol. 1: 127-137. University Press, Cambridge, UK. Provoost, S. 2004. Het kustecosysteem. In: Provoost, S. & Depuydt, F. 1972. De Belgische strand en duinfomiaties in het Bonte, D. (eds.) Levende duinen: een overzicht van de kader van de geomorfologie der zuidoostelijke Noordzee biodiversiteit aan de Vlaamse kust, pp. 10-45. Mede kust. Verh. K. Acad. Wet. Lett. Schone Kunsten Belg. Kl. delingen van het Instituut voor Natuurbehoud 22, Bins sei, Wet. 34-122: 1-214. BE. De Raeve, F. 1989. Sand dune vegetation and management Provoost, S. & Hoffmann, M. (eds.) 1996. Ecosysteemvisie dynamics. In: van der Meulen, F., Jungerius P.D. & Visser voor de Vlaamse kust. 2 Vols. Ghent University, Institute J.H. (eds.) Perspectives in coastal dune management, pp. of Nature Conservation & Minishy of the Flemish Com 99-109. SPB Academic Publishing, Den Haag, NL. munity, Nature division. Brussels, BE. Doing, H. 1993. Het Sileno-Tortuletum (ass. nov.), een Provoost, S. & Van Landuyt, W. 2001. The flora of the karakteristieke associatie van het zeedorpenlandschap. Flemish coastal dunes (Belgium) in a changing landscape. Stratiotes 6: 40-52. In: Houston, J.A., Edmondson, S.E . & Rooney, P.J. (eds.) D roesen, W . 1998.Spatial modelling and monitoring of natu Coastal dune management, shared experience of Euro ral landscapes with cases in the Amsterdam Waterwork pean conservation practice, pp. 393-401. Liveipool Uni D unes. Ponsen & Looijen, Wageningen, NL. versity Press, Liveipool, UK. Goldsmith, B. (ed.) 1991. Monitoring for conservation and Reid, W.V. & Miller, K.R. 1989. Keeping options alive: the ecology. Chapman and Haii, London, UK. scientific basis for conserving biodiversity. W orld R e Herrier, J.-L. & Killemaes.1.1998. Acquisition and protection sources Institute, Washington DC, US. of the coastal dunes of Flanders. Coastline 4: 10-15. R odw ell, J.S . 2000. British plant communities: maritime com Hoffmann, M., Cosyns, E., Deconinck, M., Lamoot, I. & munities and vegetation o f open habitats. Cam bridge U ni Zwaenepoel, A. 2001. Donkey diet in a Flemish coastal versity, Cambridge, UK. 42 Provoost, S. et al.

Roisin, P. 1969. Le domaine phytogeographique atlantique diy dunes.J. Coastal Conserv. 3: 41-48. d'E urope. Presses agronomiques Duculot,Gembloux,BE. Veer, M.A.C. & Kooijman, A.M. 1997. Effects of grass- Rozema, J.P., Laan, P., Broeckman, R., Ernst, W.H.O. & encroachment on vegetation and soil in Dutch dry dune Appelo, C.A.J. 1985. On the lime transition and décalcifi grasslands. P lant Soil 192: 119-128. cation in the coastal dunes of the province of North Hol Waumans, F. 2001. Vegetatie-ecologie van droge duingras- land and the island of Schiermonnikoog. Acta Bot. Neerl. landen aan de Westkust. Thesis, Ghent University, Gent, 34: 393-424. BE. Salisbury, E. 1952. Downs & dunes: their plant life and its W eeda, E.J. 1992. Zandviooltje ( Viola rupestris) in de duinen environment. Bell & Sons, London, UK. van Noord-Kennemerland: hoe een dwerg uit de steppe- Slings, Q.L. 1994. De kalkgraslanden van de duinen.Levende toendra standhoudt temidden van zand, zeewind en N at. 95(4): 120-130. konijnen. Wetenschappelijke mededelingen KNNV 206. Tilman, D. 1985. The resource-ratio hypothesis of plant suc KNNV Uitgeverij, Utrecht, NL. cession. Am . N at. 125: 827-852. Weeda, E.J., Doing, H. & Schaminée, J.H J. 1996. Koelerio- van Breukelen, L., Cosyns, E. & van Wieren, S. 2002. Wat Corynephoretea. In: Schaminée, J.H.J., Weeda, E.J. & weten we van terugdringen van duinstruwelen door her Westhoff, V. (red.) De vegetatie van Nederland 3. bivore zoogdieren? Levende Nat. 103(3): 101-105. Graslanden - Zomen - Droge heiden, pp. 61-144. Opulus van Til, M., Ketner, P. & Provoost, S. 2002. Duinstruwelen in Press, Uppsala, SE & Leiden, NL. opm ars. Levende Nat. 103(3): 74-77. Weeda, E J. & Schaminee, J.H.J. 1990. Nomenclatorische en Veer, M.A.C. 1997. Nitrogen availability in relation to vegeta systematische notities met betrekking tot de Koelerio- tion changes resulting from grass encroachment in Dutch Corynephoretea. Stratiotes 13: 77-97.

Received 3 July 2003; Revision received 15 January 2004; Accepted 1 March 2004. Co-ordinating Editor: F. Taveira Pinto.