Netherlands: Assessment of Fire as a Restoration Tool for Coastal Dunes

Project Stage:
Completed

Start Date:
1993-09-04

End Date:
2001-09-04

Primary Causes of Degradation

Degradation Description

The cover of mosses and tall graminoids has, since the 1970s, increased considerably on coastal and inland sand dune grasslands throughout The Netherlands (Veer & Kooijman 1997) and NW Germany (Biermann 1996, 1999). Terrestrial lichen diversity has previously been very high in these dunes, especially in those with calcium poor sand of Pleistocene origin. In the past, grazing by cattle and rabbits kept the coastal dune grasslands in a condition of short grasses with patches of bare sand, where succession was frequently restarted (van Dieren 1934). In these “˜grey dunes,’ the Corynephorus canescens vegetation on decalcified sand was rich in lichens (Westhoff 1947; Erichsen 1957; Oostra 1968). Since the 1970s the lichen-rich plant community Violo-Corynephoretum (VC), occurring on coastal dunes, (Weeda et al. 1996; Biermann 1999) was affected by a lush growth of tall graminoids such as Ammophila arenaria, Calamagrostis epigejos and Carex arenaria. Other grasses began to dominate the lichen-rich Spergulo-Corynephoretum on calcium-poor inland dunes, and both communities were invaded by the exotic moss species Campylopus introflexus (van der Meulen et al. 1987).

Increased grass dominance has been connected to increased deposition of atmospheric nitrogen (N), which, in the 1990s, amounted to 20 kg N.ha – 1.a – 1 on the West Frisian Islands, The Netherlands (Bleeker & Erisman 1996). Ammonia initially acts as a nutrient enrichment, but in the soil it also turns into nitric acid, which causes the acidification of the dune soils. The calcium poor dunes had a relatively high P availability but were N-limited, resulting in a large effect of this increased N input (Kooijman et al. 1998) and accelerated succession (Veer & Kooijman 1997). The vigorous growth of Ammophila arenaria was also due to acidification, which altered the plant-parasitic microbial community in the dune grasslands on Terschelling (de Boer et al. 1998). The decrease in the rabbit population due to viral diseases further accelerated the dominance of these graminoids (ten Harkel & van der Meulen 1996).

In 1990, the southern slope was dominated by tall graminoids with hardly any bare sand left. The cover of graminoids and herbs (living plus dead) was more than 85% and much litter had accumulated. The abundance of graminoids in relation to herbs varied considerably over the slope: from 100% graminoids at the top to 15% at the foot; with 2% herbs at the top and 5% at the foot; very few lichens remained.

Reference Ecosystem Description

In 1966 the vegetation composition on the southern slope was characterized by a combination of much bare sand (60%), vascular plants (28%), mosses (5%) and lichens (8%). Lichens from the pioneer stage of the VC, such as Cladonia foliacea, C. furcata and Cetraria aculeata, had moderate cover, while the humicole species of the later successional stages, such as Cladonia glauca and C. chlorophaea, were not as well developed on the southern slope as they were on the western slope.

In 1966, the western slope was covered by a well developed VC, very rich in lichens with high cover and much bare sand. Lichen species from the pioneer stage of the VC, such as Cladonia foliacea, C. furcata and Cetraria aculeata, had high cover and were growing in locations exposed to full sunlight. The humicole species of the later successional stages, such as C. glauca and C. chlorophaea, were growing abundantly on and between mosses (Hypnum cupressiforme and Dicranum scoparium), a substrate with a pH-H2O of 4.8 (Oostra 1968). They were growing in well lit places, but also in partial shade. At that time, only a few aero-hygrophytic species, such as the reindeer lichen Cladina portentosa, were present. The generally epiphytic lichens Hypogymnia physodes, Platismatia glauca and Pseudevernia furfuracea were growing on sand- and moss-covered dune soil (Ketner- Oostra & SÁ½kora 2004), a substrate with a pH-H2O of 5.0 (Oostra 1968).

Project Goals

The aim of the study was to assess the regeneration of burned dune grassland vegetation with possible restoration of the former open, dry-dune grassland rich in lichens (Oostra 1968). The hypothesis tested was that burning can be used to restore terricolous lichen diversity in former “˜grey dunes’ (VC) dominated by tall graminoids.

Monitoring

The project does not have a monitoring plan.

Description of Project Activities:
Fire has been frequently used to manage heathlands (Gimingham 1994) and is now considered a management tool in grass-dominated dune grasslands. Knowledge of the effect of fire on soil lichens and their re-establishment could provide basic information for future management measures to regain former biodiversity in coastal and inland dune grasslands. An opportunity for research presented itself in 1993 after a wildfire removed most of the vegetation on a dune body in the coastal dunes of Terschelling. This large dune had a monitoring history going back to 1966, when the lichen-rich vegetation was described both quantitatively, with large blocks in transects, and qualitatively with relevés (Oostra 1968). In 1990, just a few years before the fire, the dune was studied again to document the decline in species richness as result of increased dominance of Ammophila arenaria, Calamagrostis epigejos and Carex arenaria. After a very dry period in early spring 1993, the lichen-poor dune, dominated by tall graminoids and with much standing dead plant material and litter, was intensively burned by a wildfire. The field layer turned into ash, and the dwarfshrubs (Empetrum nigrum) on the upper part of the western slope burned deeply into their root system. The bottom layer turned partly into ash, and the high temperatures killed all cryptogams. Only a few blocks with patches of reindeer lichens at the bottom of the western slope were not burned. A total of 24 blocks (4 m x 4 m) were studied in a transect (three blocks wide by eight blocks long) on the southern slope of the RD dune, and 52 blocks in another transect (two blocks wide by 26 long) on the western slope. Cover of five major moss species and dwarf-shrubs (Empetrum nigrum) was estimated, along with the total cover of living and dead vascular plants (graminoids and herbs), litter, mosses, lichens, bare sand and ash, in 1966, 1990 and (after the fire in 1993) annually until 2001, except in 2000.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Immediately following the spring fire in 1993, both graminoids and rosette plants re-occurred. The cover of ash and bare sand was high. After two years, the cover of ash had decreased rapidly; after only two to four years, the cover of herbs, graminoids and litter had increased; and after three to eight years, the plant cover (including standing dead material) remained at a constant high level with much litter. The abundance of graminoids in relation to herbs still varied considerably over the slope during those years, especially the four years following the fire (1993-1997). This situation stabilized in the period 1998-2001: from 85% graminoids at the top to 20% at the foot; with 20% herbs both at the top and at the foot. The total moss cover increased from zero in the first year after the fire, to 40% in the second year and reached 65% after six years. Two lichen species were present three years after the fire (1996) and six species after five years. However, only a minor increase in number of lichen species and cover was found during the following years. Immediately after the fire on the western slope, a high proportion of the studied blocks were covered by ash with some bare sand. Empetrum nigrum on the upper part of the slope did not reappear after the fire. In all burned blocks, the cover of higher plants (including standing dead material) and mosses quickly increased after the fire, resulting in dominance by 2001, a build up of litter, and very low abundance of lichens. Only on the unburned lower slope (lichen-rich in 1990) did mosses and higher plants remain at about the same level, while lichen cover remained high. The moss increase was partly different compared to the south slope. The lichens were slow to colonize after the fire. Five species occurred after three years, with Cladonia foliacea and C. furcata as the main pioneers, while eight species occurred after five years. The humicole species C. chlorophaea was found on burned stumps of Ammophila arenaria. Colonization by some species typical of subneutral sand, such as C. rangiformis and C. humilis, led to 11 species in 1999 (probably the result of sand blowing in), while some increase in cover was apparent in 2001. However, the total lichen cover of the studied blocks was still low in 2001. The main difference on the southern slope between surveys conducted in 1990 (before the fire) and those from 2001 was the much higher moss cover found in 2001. Researchers observed the successive dominance of the three moss species Ceratodon purpureus, Campylopus introflexus and Polytrichum juniperinum. C. purpureus, a species strongly favoured by fire, reached a peak three years after the fire (southern slope 40%; western slope 55%). This moss is among the early colonizers in post-fire succession in heathland in Brittany (Gloaguen 1990) due to its wide diaspore dispersion (Clément & Touffet 1990). C. introflexus, a species characteristic of acidic, mineral-poor sand, gradually took over as succession proceeded. This species has been shown to have a high asexual reproductive rate that facilitates early successional dominance. Later in the succession, the habitat on the southern slope appeared to be particularly suitable for the xerophytic moss P. juniperinum. It increased from < 1% in 1995 to 45% in 1999, with a simultaneous decrease in cover of C. introflexus to 20%.

Factors limiting recovery of the ecosystem:
Annual surveys showed a slow and very limited re-establishment by lichens after the wildfire, especially on the southern slope. No living fragments were initially found, indicating that the wildfire must have been very intense. Some lichen species had settled on the western slope (11 species in total), but their total cover was very low (< 2 % after eight years, 2001). This can be compared with a mean cover of 42% in the original "˜grey dunes' of 1966. Lichens here are dependent on open vegetation with much bare sand. In the research area, lichen colonization was probably limited by the fast colonizer C. purpureus, soon followed by C. introflexus. Another important factor was the quick regrowth of vascular plants (graminoids and rosette plants), which has previously been shown to be caused by high N deposition in calcium-poor dune ecosystems (Kooijman et al. 1998). Thus, a similar vegetation as that found before the fire became established, partly from surviving perennial plant remains and partly from nearby areas of unburned vegetation, in accordance with the findings of Vestergaard & Alstrup (2001). The development of vegetation during later years was most probably a result of the same environmental factors that promoted dominance of graminoids before the fire, namely: (1) nitrogen deposition in combination with the available phosphates in the soil (Kooijman et al. 1998); (2) the impact of acidification on the soil micro-organisms which has been shown to influence the regrowth of Ammophila arenaria (de Boer et al. 1998) and (3) litter decomposition and soil formation during succession (Veer & Kooijman 1997; Kooijman & Besse 2002).

Socio-Economic & Community Outcomes Achieved

Key Lessons Learned

The results of this study show clearly that fire alone is not sufficient for management aimed at restoration of lichen-rich dune grasslands in calcium-poor coastal dunes, influenced by a relatively high N deposition. At the end of the study period (2001), the vegetation composition in the burned blocks was still quite different from the original composition in 1966. Even the Cladina mats, present at the foot of the western slope in 1990, did not recover after the fire.

Long-Term Management

Additional measures are recommended, such as large-scale clearing of the burned vegetation to the mineral soil layer and promoting sand to be blown in. The sand will physically harm and dry out the remaining mosses, and C. introflexus especially has been shown to be sensitive to such treatment (Ketner- Oostra & SÁ½kora 2000; Ketner-Oostra 2002). Neutral or sub-neutral sand might be blown in from foredunes, from reactivated blowouts, or artificially brought to the site. This is especially important for the colonization and maintenance of lichen vegetation in the first stages of the VC in these calcium-poor Terschelling dunes, based on earlier research (Ketner-Oostra & SÁ½kora 2000; Ketner-Oostra 2001).

Sources and Amounts of Funding

The State Forestry Service Frisia (Staatsbosbeheer Fryslán) supported this study financially.

Other Resources

Ketner-Oostra, R. et al., 2006. Restoration of lichen diversity in grass-dominated vegetation of coastal dunes after wildfire. Journal of Vegetation Science 17: 147-156.

Ketner-Oostra, Rita et al., 2004. Decline of lichen-diversity in calcium-poor coastal dune vegetation since the 1970s, related to grass and moss encroachment. Phytocoenologia 34(1): 521-549.

Rita Ketner-Oostra
Freelance ecologist
Algemeer 42, 6721 GD Bennekom, The Netherlands
E-mail: rita.ketner-oostra@wur.nl

Primary Contact

Organizational Contact