Factors affecting the declining Lacerta
agilis population in Voornes Duin
Abstract
Since 1994 the Sand Lizard (Lacerta agilis) shows an increase in the Netherlands. However, in Voornes Duin, an isolated nature reserve in Zuid-Holland, the population is declining. In this research we investigated different factors that could be the cause of this decline.
Since 2005 large parts of Voornes Duin have been opened up, mainly with the use of heavy machinery, to counteract encroachment of the dune landscape. Afterwards large grazers have been introduced to maintain this landscape. We hypothesize that the intensity of this grazing has been to high over the past years, leading to large areas with low vegetation height, an unsuitable habitat for L. agilis. to investigate this we divided our research in four parts, natural management, mechanical management, vegetation and geomorphology. With our main research question; Does management have an effect on vegetation and
geomorphology and does this affect the sand lizard population in Voornes Duin? We investigated this in Voornes Duin and in Duin & Kruidberg, which is known to house a healthy population of L.agilis. The number of L.agilis was monitored along routes with six sections, together with vegetation samples and height and raggedness measurements. We found that management can indirectly influence the population of L. agilis through its effect on vegetation and geomorphology. A high grazing intensity can cause a low raggedness of the vegetation, which has a negative effect on the L.agilis population density. We advise that in Voornes Duin the grazing intensity is lowered and locally the grazing is terminated
altogether. Thereby creating a vegetation structure with a higher mean raggedness and migration lanes where the raggedness is very high. This should help the L.agilis population to maintain a stable population and create the possibility for migration to new habitats.
Content
Introduction 3
Methods & Materials 12
Lizard monitoring 14
Routes and sections 14
Vegetation 15
Vegetation structure 16
Species diversity 16
Raggedness 16
Rabbits & Hares 17
Geomorphology 17
Grazing intensity 18
Statistics 19
Results 20
Relationship between L.agilis population density and vegetation characteristics (lead
author: Bart-Jan Akerboom) 22
Relationship between geomorphology and Lizards (Lead author: Gijs Bakker) 25
Mechanical management - Vegetation (Lead author: Aron Coffa) 27
Relationship between natural management and the L. agilis habitat suitability (Lead
author: Sarah Scholten) 28
Geomorphology (Lead author: Gijs Bakker) 30
Mechanical management (Lead author: Aron Coffa) 31
Relationship between natural management and the L. agilis habitat suitability (Lead
author: Sarah Scholten) 31
Overall conclusion and discussion 33
Acknowledgement 33 Appendix 35 Appendix I: 35 Appendix II: 35 Appendix III: 35 Appendix IV: 35 References 36
Introduction
Publications on environmental research during the second half of the 20th century
were responsible for increased public awareness. Hence, the European Union agreed on a EU-wide conservation network called Natura 2000. The main aim of this agreement is to preserve threatened species and valuable habitats. All member states must assign sites, according to specific criteria. Within these sites species and habitats specified under the Birds or Habitats Directive are protected. Once every six years, member states report the status on numbers and distribution of the species and areas under habitat directive protection to show their progress in protecting species and habitat types (Evans, 2012; Rijksoverheid, 2016; Sundseth, Mézard & Wegefelt, 2008). This network now covers 18% of land area and 6% of marine environment and is the largest of its kind in the world (European Commission, 2017).
More than 100 Habitat Directive species occur in the The Netherlands (Rijksoverheid, 2016). Among them the sand lizard (Lacerta agilis). Lacerta agilis is a widespread lizard species in Europe and parts of Asia (Hildebrand & Goslow, 2001). L. agilis occurs in two different habitat types: coastal sand dunes and heathlands (edgar et al. 2010). Figure 1 shows the distribution of L. agilis in the Netherlands.
Coastal sand dunes develop where there is an adequate supply of sand in the intertidal zone and where onshore winds are prevalent. Dry sand is blown landwards and deposited above the high water level, where it is trapped by specialised dune-building grasses, which grow up through
successive layers of deposited sand. Sand dune communities vary geographically, but a constant factor is the presence of marram grass (Ammophila arenaria) (BRIG, 2008).
A heathland is a dwarf-shrub habitat characterised by open, low-growing woody
vegetation, such as heath (Calluna vulgaris) blueberry (Vaccinium myrtillus), cowberry (Vaccinium vitis-idaea) and crowberry (Empetrum nigrum) (BRIG, 2008).Both habitats are present in the Netherlands and stable L. agilis populations occur in both (Overleg
duinhagedis, 1999). Despite this general increase in the Netherlands (Figure 2) a local decline in numbers is seen in Voornes Duin since 2005 (Figure 3).
Figure 1. Spread of Lacerta agilis in the Netherlands. The red dots being the present sightings (2015-2018) and the blue squares being the sightings before 2015 (Ravon verspreidingsatlas, 2018)
Figure 2. Sand lizard (Lacerta agilis) sightings in the Netherlands, the blue dots represent the sightings, the
blue line indicates the trend and the light blue area shows the uncertainty of this trend. From 1994 till 2017. NEM (RAVON, CBS), 05/2018.
Figure 3. Average number of L. agilis in Voornes Duin, the Netherlands. The average of the total L. agilis
sightings in one trajectory over a maximum of seven repetitions per year. Monitoring according to the RAVON guidelines.
Voornes Duin is an example of a sand dune habitat. The area is part of the Natura 2000 since 2008. Another example is Duin & Kruidberg. In this area the L. agilis population is slightly increasing over the past decades (RAVON, Figure 3). Because of the healthy L. agilis population this area is used as a reference area in this study.
Figure 4. Average number of L. agilis in Duin en Kruidberg, the Netherlands. The average of the total L.
agilis sightings in three trajectories over a maximum of seven repetitions per year. Monitoring according to the RAVON guidelines.
Since the L. agilis is a pilar species for Natura 2000 and Voornes Duin is a Natura 2000 area, the decline of the population is worrying. Sinervo et al. (2010), found that
increasing spring temperatures, due to global warming, have negative effects on many lizard populations in Mexico. However, Ollson et al. (2011), found the opposite effect on L. agilis populations in Sweden. The authors concluded that the increase in temperature allowed these ectotherms more possibilities for mating, thus resulting in more multiple paternity clutches. Multiple paternity clutches led to more eggs hatching and a greater survival of young.
Next to global warming, an increase in anthropogenic influences like agriculture and transport has caused more pollution over the last century. This pollution leads to so-called acidic deposition (De Vries, 2008). Acidic deposition, in the form of nitrogen compounds (N), can cause an increase in plant biomass, but only if enough phosphorus (P) is available (Kooijman et al., 1998). Acidification in the dunes is one of the causes of grass domination which results in a decrease in biodiversity (Veer & Kooijman, 1997). Management can be performed to counteract this encroachment and create a higher plant species diversity (Westerhoff, 1989). This nature conservation might be a cause of the decline of the sand lizard population in Voornes Duin.
The habitat requirements for L. agilis in coastal dunes can be summarized as follows: presence of cover against predators, bare sandy ground for the females to lay their eggs in, enough microtopography, the presence of prey and the presence of suitable places to bask (edgar et al., 2010, overleg duinhagedis, 1999). Furthermore, the suitable habitat must be large (at least 10 hectares) and well connected to allow a healthy population to develop (overleg duinhagedis, 1999). Whether these requirements are met is determined by a combination of the vegetation structure, species composition and the geomorphology of the habitat. The vegetation structure is defined as the the overall structure of a plant community, such as the vertical layers of plants of different heights. The species composition refers to the contribution of each plant species to the vegetation in a percentage. the geomorphology of the habitat is defined as a list of abiotic factors of the habitat.
According to Wouters et al. (2011) the vegetation structure is one of the most important factors influencing the habitat suitability for L. agilis. The species needs particular types of vegetation for different reasons. First of all, the vegetation can provide shelter against predators, such as birds of prey (Edgar et al., 2010, Overleg Duinhagedis, 1999). The presence of taller vegetation seems to serve this purpose the best (Overleg
Secondly, the vegetation must harbour enough prey species for L. agilis. These comprise a variety of invertebrates, especially spiders, grasshoppers, crickets, bugs, flies and insect larvae (Edgar et al., 2010). According to van Leeuwen & van de Hoef (1976) a high plant species richness can maintain a higher and more diverse population of insects, making a higher lizard population density possible. However, no study provides evidence that insect population density plays a significant role in the total lizard population density.
Thirdly, the presence of bare ground is an important aspect of the vegetation
structure, because the females to lay their eggs in bare, sandy, spots (Overleg Duinhagedis, 1999). Finally, and most importantly, the vegetation must allow for enough places for the lizards to bask in the sun, meaning that a very high coverage of taller vegetation is unfavourable (BIJ12, 2017)
Because of the necessity for places to bask and for places to find shelter and hunt a contradictory habitat requirement is present. While optimal places to bask are places with low vegetation, causing less shade, optimal places to hunt and find shelter are places with taller vegetation. The “amount of edge” in the vegetation is therefore of great importance for L.agilis habitat suitability (overleg duinhagedis, 1999). In this study, the “amount of edge” is defined as the ‘raggedness’ of the vegetation.
In addition to the structural characteristics of the vegetation, correlations have been observed between the presence of certain plant species and the number of lizards.
According to Edgar et al. (2010) areas with a luxuriant layer of bryophytes (Bryophyta) and lichens are particularly favoured. Correlations with other plant species have been found as well. Where variations occurred throughout the habitat, L. agilis was associated with long-standing heath communities or vegetation similar in structure, such as mixed grass species interspersed with shrubs and open ground (House & Spellerberg, 1983). According to the latter study in heathland, Calluna vulgaris is positively correlated with the presence of lizards. In coastal dune ecosystems C. vulgaris is a very rare species, however.
In coastal sand dunes, mainly Ammophila arenaria shows a strong positive
correlation with L. agilis presence (BIJ12, 2017). Furthermore, in the dune valleys and the middle dunes, L.agilis is often associated with sea buckthorn (Hippophae rhamnoides) (Overleg duinhagedis, 1999). It is, however, very likely that those plant species are only characteristic for a certain vegetation structure and therefore are, as a species, not directly influencing habitat suitability for L. agilis.
In short, two vegetation parameters can be expected to have an influence on the density of the L. agilis population: the structural characteristics and the species
composition of the vegetation. Of these two the first is expected to be of more influence than the latter, since the species composition of the vegetation is thought to be indicative for the vegetation structure. Of the vegetation structure we expect the raggedness of the vegetation will have the largest effect on the L.agilis population density.
Next to the biotic factors above, certain abiotic factors are of importance for the suitability of the habitat for L.agilis. One of the crucial demands for L. agilis is the presence of (micro)-topography of the soil; the presence and orientation of slopes (Overleg
duinhagedis 1999). Sand lizards have a preference for sunny slopes, particularly facing south to southwest on the leeward side[GO1] . These slopes receive the highest sun intensity,
which helps them with thermoregulation. The number of sun hours a slope can receive is not solely dependent on the exposition direction but also on the gradient of the slope (Overleg duinhagedis 1999). On slopes with high gradients sand lizards prefer the top of the slope while they prefer the bottom and middle of the slope at lower gradients, this is because the middle and bottom of the slopes are more protected from the wind.
The second factor of geomorphology can be divided in several aspects: the presence of stepping stones and the presence of migration corridors. Both aspects, however, need to suit the mobility of L. agilis to ensure migration. It has been shown that sand lizards don't move more than 500 meters through unsuitable habitat (Corbett, 1988); the distance
between suitable habitat patches must therefore be less than 500m to form part of the same metapopulation (van de Bund, 1991). Lizard populations that are located too far from each other (>500m) can be considered to be isolated. This could have severe consequences for
the survival of the lizard populations, because isolated populations do not exchange genetic material. In time this could lead to local extinction.
Two geomorphological aspects of an area can reduce this problem: the presence of stepping stones or migration corridors. Stepping stones are areas of around 1500 m2 that
management has opened up in forests or overgrown areas. Whilst these overgrown areas are not able to sustain stable populations of sand lizards they are able to sustain sink populations (Andrén et al, 1988). Corridors are longitudinal strips of suboptimal habitat connecting patches of suitable habitat, which allow sand lizards to migrate two to four kilometers a year (Klewen, 1988). These lanes of two to five meters wide and must provide enough shelter and protection from predators. Stepping stones and corridors can be very interesting aspects of an evaluation of optimal sand lizard habitat.
In order to preserve the above mentioned ecological and geomorphological
characteristics for a suitable L. agilis habitat, management can offer a solution. Management is in the aid of all protected animals and plants under the Natura 2000, so the sand lizard is no exception.
First of all, natural management can be conducted. In this study, this is defined as all types of management where people use natural processes to create desired habitat, mainly through the use of grazers (van Uchelen, 2006). In order to prevent ecosystems from changing into closed woodlands, grazing management can be a solution (van Uchelen, 2006; Rupprecht, Gillhaus & Hölzel, 2016). According to van Steenis (2006), grazers in Voornes Duin are one of the natural inhibitors on vegetation succession. Since L. agilis requires a specific habitat type, grazers might affect the lizard population as a result of changing habitat aspects.
Attum & Eason (2006) have shown that grazers havea positive correlation with L. agilis habitat suitability by shortening the vegetation and creating more bare patches. Their study suggested that lightly grazed sites are in favoured by lizards, because they can take
advantage of more open patches and can move more easily to other sites. Other studies indicated that multiple geomorphological aspects can be affected by trampling, grazing or rolling ofgrazing livestock. The main consequence is shrubs being destroyed by cattle trying to reach most palatable vegetation species (de Bonte, Boosten, van der Hagen & Sýkora, 1999). Grazing livestock can create an increased number on raggedness, because they prefer specific species (e.g. grasses) and lower these instead of less palatable species. These preferences incites a decreased plant species diversity in the environment (Olff & Ritchie, 1998). Thereby, according to Matsui, Khalil & Takeda (2009), livestock species roll in sandy spots to rid themselves of skin parasites. In this way, they create bare patches in areas that used to be covered with vegetation (Kooijman, Besse & Haak, 2005; Schonbach et al., 2011; Rupprecht, Gillhaus & Hölzel, 2016). However, Attum & Eason (2006) have shown that intensive grazing negatively affected the lizard population because of major changes in the vegetation structure. Areas with loss of tall vegetation also exhibit a decrease in food resources. In their studied area, a decrease in food resources costed the lizards more time to find enough food to survive.
Rabbits are also included by natural management in this research. They are selective grazers (Costin & Moore, 1960) and by grazing and digging burrows, rabbits create a heterogeneous habitat which positively affects L. agilis populations. Especially L. agilis females and juveniles were found more often at sites with burrows than in sites without. Since the conditions inside these burrows are more stable than outside, egg-laying sites can be found in burrows. Rabbits can also influence the soil by digging and create more patches of sand, which are favorable egg laying places (Kent, et al., 2001). This could result in a higher level of raggedness.
In addition to natural management mechanical management is an often used way of maintenance of a nature conservation. Mechanical management is defined as all types of management where maintenance of an area is conducted by human labour, often with the help of tools and machinery.
vascular vegetation with the choice of removing the residue. By removing the plant material, nutrients can be removed from the ecosystem, slowing down succession. Kooijman et al. found that mowing also leads to more plant species and light reaching the soil. “Spraggelen” is a variation on regular mowing, here the mowing residue or plant material is removed after mowing. In this way nutrients are removed from the environment, rather than returning them to the cycle.
To restore more humid areas, sod-cutting can be used to reach groundwater levels and remove organic matter to reduce acidification and encroachment (Grootjans et al. 2002). By removing the top layer of the soil in an encroached dune area, together with the
vegetation and even its roots, a bare landscape is created (Oosterveld, 1979). This offers suitable habitat for pioneer species a Kooijman et al. (2004). Bruin (2001) has found that sod-cutting increases numbers of rare plants, like Bog pimpernel (Anagallis tenella) and Fen pondweed (Potamogeton coloratus). Thus mechanical management can have an influence on the vegetation diversity in the Dutch dunes, which may in turn influence the L. agilis population (van Leeuwen & van de Hoef, 1976 en Spellenberg, 1988).
The four above mentioned factors will be investigated and afterwards combined into a management advice for the recovery of the L. agilis population in Voornes Duin. Figure 5 illustrates our approach to this research. The sand lizard population is dependent on a suitable habitat, which consists of ecological and geomorphological factors. These factors are in turn influenced by succession and to counteract that management can be adapted in two forms: natural and mechanical management. The following text will explain all
subsections in more details.
Figure 5: Flowchart of the variables investigated in this research influencing the habitat of L. agilis.
Mechanical and natural management influences both ecological and geomorphological factors. This influences the L. agilis habitat, which on its turn affects the L. agilis population. So, management indirectly influences the L.
agilis population.
We used Duin & Kruidberg, which comprises a large and viable population of L.agilis, as a reference area in our study. In this area the same variables will be measured as in Voornes Duin. Furthermore, the entire area is managed by Natuurmonumenten, just as the largest part of Voornes Duin (figure 4). Additionally, Duin & Kruidberg is also managed by a
combination of natural management and mechanical management. With 400 ha, the Duin & Kruidberg reserve is larger than Voornes Duin (190 ha), and it is connected to adjacent nature reserves. Because of the geographical location on an island, Voornes Duin is not connected to other nature reserves at all.
Our main research question is: Which ecological and geomorphological factors are correlated with the size of the L. agilis populations in Voornes Duin and Duin & Kruidberg, and how does management influences these factors?
In order to answer this main question, the study is divided into four parts: ecological characteristics, geomorphological characteristics, influence of natural management and influence of mechanical management. These four parts have been placed in a table, to provide an overview of their content, the complementary sub-questions and hypotheses (table 1).
Subsequent to this research, suggestions will be done for future management in Voornes Duin in order to create preferred habitat types of L. agilis, which may result in the recovery of L. agilis population.
Table 1: Overview of the research content and division. Parts of the research with an overview of their
content, sub-questions and sub-hypotheses. Sub-hypotheses are numbered for a quick reference.
Part Content Sub-question Sub-hypotheses
Ecological characteristics vegetation structure, species composition & raggedness Is L. agilis population density correlated with vegetation structure, species composition & raggedness? A diverse vegetation structure will be positively correlated with the L. agilis population density.e.1
Lichens, bryophytes, A. arenaria and H. rhamnoides will be positively correlated with L. agilis density.e.2 Plant species richness is positively correlated with L.agilis population density.e.3 Raggedness is positively correlated with L.agilis
population density.e.4
Geomorphological characteristics Distance, migration lanes, microrelief, exposition, wind shelters, sand patches and fire breaks
Are there any correlations between the geomorphological characteristics and the number of L.agilis in Voornes Duin? If the characteristics we checked have an effect on the presence of L. agilis, we will find a correlation between those characteristics and the presence of L. agilis.g.1
Natural management Grazing livestock intensity, rabbit
Does any form of natural
Grazers are able to lower vegetation and
population density & habitat management in Voornes Duin influence the ecological characteristics, the geomorphological characteristics or the L. agilis population? thereby decrease vegetation heights.n.1 Therefore, grazers can increase the raggedness within L. agilis habitat.n.2
Grazers can decrease the plant species diversity.n.3
Grazers can create sand patches by rolling and trampling.n.4
Rabbits are able to create egg laying places in favor of L. agilis, by digging activities and building burrows.n.5
Rabbits can
increase the level of raggedness.n.6 Mechanical management mowing, sod-cutting, logging, “spraggelen”, wood and vegetation removal
Does any form of Mechanical management in Voornes Duin influence the ecological characteristics, the geomorphological characteristics or the L. agilis population? Mechanical management in any form does influence ecological
characteristics.m.1
Mechanical
management in any form does influence ecological
characteristics.m.2
Methods & Materials
Study species
Lacerta agilis is a widespread species across Europe and parts of Asia. The species is ectotherm, that occupy two habitat types in the Netherlands: lowland heathland and sand dunes ( Edgar et al. 2010). In the Netherlands, Sand lizard populations can be found in the coastal dunes and on the inland sandy soils (Veluwe, Utrechtse Heuvelrug).
Males, in contrast to females, turn bright green during mating season in April and May. Both sexes are characterised by two dorsal stripes over the length of their bodies. Full grown specimens grow up to 21 cm in length including their tail. Females can lay up to 12 eggs per year, burying them in patches of open sand. The eggs hatch around August, depending on the incubation temperature.
As well as under the Habitat Directive annex 4, the Sand lizard it is indicated, on the Dutch Red list, as a Vulnerable species and is nationally protected under the Wet
(RAVON, 2018).
Within the sites where the species is found, the distribution of L. agilis is further restricted, often to relatively small areas, by its specific habitat requirements. This species requires warm, sheltered sites, with a varied topography, and especially south or south-west facing slopes (House & Spellerberg, 1983). In dune areas L. agilis occurs in two habitat types containing different key plant species. Closer to sea European marram grass
(Ammophila arenaria) provides protection against predators and wind. More inland common sea buckthorn (Hippophae rhamnoides), or other shrubs partly overgrown with grasses can provide shelter (Overleg Duinhagedis, 1999).
Study areas
Voornes Duin is a nature reserve on the island of Voorne in Zuid-Holland, it is approximately 1400 ha and most of the area is managed by Natuurmonumenten. A small part however is managed by the township of Westvoorne (Figure 6).
The management that is used by Natuurmonumenten consist of mostly natural management, namely year round grazing by Highland cattle and some local mechanical management in the form of mowing and sod cutting. In the northern most area that is managed by Natuurmonumenten they began with a Highland cattle herd of 18 on 100 Ha, which has been reduced to a herd of 12 Highland cattle in 2017. The Southern area of Natuurmonumenten started with a herd of 16 Highland cattle (Bos (primigenius) taurus) on 90 Ha, and has been reduced to a herd of 10 Highland cattle in 2017.
The central part of Voornes Duin which is under the management of the township of
Westvoorne, is solely managed with mechanical management, sowing, sod cutting and tree cutting. In 2008 the entire area under the management of the township was sod cutted.
Figure 6: The research area Voornes Duin. This figure shows the research area of Voornes Duin and the
ownership of all of its parts. Research was done in the brown and pink areas, managed by respectively Natuurmonumenten and the township of Westvoorne. Voornes Duin is located on the island of Voorne.
Duin & Kruidberg is a part of the national park Zuid-Kennemerland, it is located along the coast of Noord-Holland, next to Santpoort (Figure 7). The national park Zuid-Kennemerland is approximately 3800 ha, while Duin & Kruidberg is approximately 400 ha.
The entirety of Duin & Kruidberg is under management of Natuurmonumenten. The management they use consists of natural management in the form of grazing and localized mechanical management in the form of mowing, sod cutting and tree cutting. They have a herd of grazers consisting off 43 Konik horses (Equus ferus caballus),19 Shetland ponies (Equus caballus) and 50 Highland cattle. These grazers are closely monitored to keep the grazing at the desired intensity. While we only looked at Duin & Kruidberg these grazers are however able to roam throughout a major part of Zuid-Kennemerland of 2069 ha.
Figure 7: The reference area of Duin & Kruidberg. This figure shows the reference area Duin & Kruidberg,
which is a part of the national park South-Kennemerland. It is located in the coastside of North Holland, next to Santpoort. The entirety of the national park is under management of Natuurmonumenten.
Lizard monitoring
Lizard monitoring was done using a protocol based on the lizard monitoring guidelines from RAVON. For the protocol, see appendix I. The monitoring places were selected on the basis of an estimation of the habitat suitability of the area. These places were divided into routes which were divided into sections. The amount of repetition per route and section varied.
Routes and sections
L. agilis is only present in specific habitat types. Therefore, the monitoring routes were established in areas with habitat likely to be suitable for a L.agilis population. The sole purpose of a Route is that it is of a length which allows one researcher to perform all of his measurements in one day. This comes down to the length of all routes to be approximately 2 km. In Voornes Duin 5 routes were established (Figure 8) and in Duin & Kruidberg 4 (Figure 9) were done in those areas prefered by L. agilis.
Within each routes, 5 or 6 sections were established. A section is a part of a route between 60 and 120 meters in which the L.agilis monitoring measurements and the complementary measurements to ecological and geomorphological factors where performed. Since the sections are the primary unit of measurement they must meet the requirement that the vegetation and geomorphology are homogenous. An overview of the geographical distribution of the sections can be found in the appendix.
Figure 8: Routes in research area. These routes, in Voornes Duin, were set in order to monitor the data of the
research.
Figure 9: Routes in reference area. These routes, in Duin en Kruidberg, were set in order to monitor the
reference data of the research.
Vegetation
In each of the sections the vegetation structure and the plant species composition measurements were conducted in a sampling plot of 1 by 1 meter. The raggedness
measurement was done sectionwide by the use of a scale of 1 till 6. The plots were selected using a random number generator that determined the number of steps from the start of the section. These measurements were repeated by multiple researchers to increase the randomization and to exclude personal biases. For the protocols for the vegetation measurements see appendix II.
Vegetation structure
The vegetation structure was determined by using a classification system of the vegetation present in the vegetation plots. Vegetation heights are ranked in 5 classes (Table 2). The coverage of these classification was estimated in each of the vegetation plots. Coverage was estimated in percentage of the total plot.
Species diversity
In each plots all vascular plants were identified to the species level and the abundance of each species was written down using a DAFOR scale (Bell, 1993). This scale ranks abundance of plants according to D- dominant, A- abundant, F- frequent,
O-occasional and R- Rare. One of the greatest
weaknesses of this method of assessing the relative abundance of a species is that other factors will often unknowingly influence the judgement of the
observer. Small species tend to be rated lower than conspicuous ones, and similarly, a species in full flower receives a higher rating than one that is in a vegetative state at the time of sampling (Bell, 1993).
In some studies the different categories are
coupled with a coverage percentage of the species in a plot (Bertoli, 1996, Avila, 2003, Affre et al., 2010). These coverage percentage matched with the categories is not the same in every study, however. Every study approaches the following division: Dominant >= 75%, Abundant = 75 - 51%, Frequent = 50 -26 %, Occasional = 25 - 11% and Rare = 10 - 1%. In this study the categories are defined according to these coverage percentages.
Raggedness
The third parameter that was measured is the raggedness. This was done on a section wide scale instead of the 1 by 1 plots.
Measurements were done on a scale of 1 to 5. 1 stood for no vegetation height transitions at all and 5 for a very high number of
vegetation height transitions. Because this is hard to define, a standard area was defined as a scores of 5 (Figure 10).
Rabbits & Hares
Observations were made on the density of rabbits and hares in each of the sections. The density of these animals was based on the number of droppings and sightings of animals in each of the sections in all routes. This has been done on a scale from 0 (no
droppings/sightings) to 5 (many droppings/sightings) within each section for rabbits and hares separately.
Figure 10, Standard for a raggedness score of 5.
Geomorphology
The geomorphology was measured using a checklist of preferred factors for L.agilis habitat. The factors scored were derived from literature found in Overleg Duinhagedis (1999). The presence of the following factors was scored;
Table 3: Geomorphological factors checklist
Factors Present or not
Open sand spots (3-5 per ha) for the placement of eggs
(Micro) Relief
Exposition, slopes to the south/ southwest Wind sheltered (preferable), to help with thermoregulation
Distance to other populations within 500 meters, to be able to mix subpopulations Migration lanes present, lanes of possible habitat to migrate to other habitats/ mix with other subpopulations
Fire breaks accompanied by 2 to 5 meter of possible habitat, breaks in vegetation to prevent the spreading of wildfires
This checklist was scored for every section in each of the monitorings routes. Distance to other populations was measured using Google Maps and the Lizard data was obtained from our own monitoring in the field.
Grazing intensity
In order to measure the grazing intensity, we used the so-called livestock unit. This unit equalizes the grazing intensity between different species. The reference unit is one adult dairy cow producing 3000 kg of milk annually, without use of dietary supplements and is equal to 1.
Table 4: calculated livestock units for the investigated areas.
Area Number of grazing
livestock
LSU per species
Surface in ha Livestock unit per ha
Voornes Duin north 12 Highland cattle 1 90 0,12
Voornes Duin south 10 Highland cattle 1 100 0,1
Gemeente Duin 0 0
Duin & Kruidberg 50 Highland cattle 43 Konik horses
1 0.8
19 Shetland ponies 0.8 Parts of both
Voornes Duin and Duin & Kruidberg
Statistics
All data analysis and statistics were done using the program R Studio with p < 0.05 being significant.
Table 5: Performed wilcox.tests for the vegetation heights.
Voornes duin Duin & Kruidberg
L.agilis numbers
vegetation height: 0 cm vegetation height: > 150 cm Raggedness
Table 6: Performed cor.tests between vegetation composition and diversity parameters and the number of L.agilis.
Voornes Duin
raggedness L.agilis numbers
H. rhamnoides
Bryophytes lichens Marram grass vegetation diversity
Duin & Kruidberg
raggedness L.agilis numbers
H. rhamnoides
Bryophytes Lichens Marram grass vegetation diversity
Table 7: Performed t.tests for the vegetation height classes.
Voornes duin Duin & Kruidberg
vegetation height classes 0-5 cm
5 - 20 cm 20-150 cm
Table 8: Preformed Wilcox.tests between the geomorphological factors and the L.agilis numbers. L.agilis numbers were corrected for the difference in section lengths, these test were done for the data of Voornes Duin and Duin & Kruidberg.
Voornes Duin
Sand L.agilis numbers
Relief Exposition Wind Distance Migration Firebreaks Duin & Kruidberg
Sand L.agilis numbers
Relief Exposition Wind Distance Migration Firebreaks
Results
This study has five separate segments in which five different aspects of Voornes Duin are researched. First, an observational study was done in which the population of L. agilis in Voornes Duin was determined. Secondly, multiple aspects of the flora in Voornes Duin were investigated and their relation to the population of L. agilis was determined. Thirdly, the geomorphological characteristics of Voornes Duin were investigated and their relation to the ecological aspects and the L. agilis distribution. Fourthly, the ways in which natural
management was conducted, where geographically analysed, and compared to the L. agilis distribution. Lastly, the ways in which mechanical management was conducted, where geographically analysed, and compared to the ecological aspects, the geomorphological aspects and the L. agilis distribution.
Lizard counts
about the L.agilis population in Voornes Duin two results will be presented: the density in comparison to the reference area (figure 11) and the geographical distribution of L.agilis in both Voornes Duin and Duin & Kruidberg (figure 12 & 13).
Firstly, the L.agilis density was significantly lower in Voornes Duin than in Duin & Kruidberg (Figure 11).
Figure 11: Mean L. agilis number/counting round per Area. This figure shows that there is a significant
difference (Wilcoxon rank sum test, W = 9013.5, p-value < 0.001) between the Mean L.agilis number/counting round of Voornes Duin and Duin & Kruidberg.
Secondly, Large differences in L. agilis density between the different sections within one area were present. mapping the sections geographically shows that the highest L. agilis density is present in the Northern part of Voornes Duin managed by Natuurmonumenten (figure 8). The same mapping in the reference area shows that the highest L.agilis density in this area is situated in the centre (figure 9).
Figure 12: The average number of L. agilis per counting round per section in Voornes Duin. The highest
L.agilis numbers are present in the most Northern part of Voornes Duin managed by Natuurmonumenten. In the
Figure 13: The average number of L. agilis per counting round per section in Duin & Kruidberg. The
highest L.agilis numbers are present in the centre of the area: in the mid dune. The sections in the beach ridge and the more forested inland dunes have lower L.agilis numbers.
Relationship between L.agilis population density and vegetation
characteristics
(lead author: Bart-Jan Akerboom)
Possible explanatory factors for the density of lizard populations are the vegetation characteristics: the vegetation structure, the species composition and the raggedness.
The percent cover of vegetation height categories did not differ significantly between areas, except for the category 0 cm (figure 14). The pairs of height categories 0-5 cm, 5-20
cm, 20-150 cm and >150 cm no significant differences were found.
Figure 14: Vegetation structure of the monitoring Routes in Voornes Duin and Duin & Kruidberg. The
coverage percentage of bare ground in Voornes Duin and Duin & Kruidberg differ significantly (Wilcoxon rank sum test, W = 177.5, p-value = 0.004). In the other categories no significant differences were present between the areas.
L.agilis population density was not significantly correlated with the cover of vegetation in any of the height classes in Duin & Kruidberg as well as in Voornes Duin. Additionally, the vegetation height diversity did not correlate with L.agilis density.
Also the four plant species that we hypothesized to be positively associated with L.agilis: lichens, bryophytes, H.rhamnoides and A. arenaria did not correlate significantly with L.agilis population density in Voornes duin nor in Duin & Kruidberg.
Plant species diversity was significantly positively correlated with L.agilis population density in Duin & Kruidberg (Figure 15). No correlation was observed in Voornes Duin.
Figure 15: The correlation between the L.agilis population density and the mean plant species diversity in Duin & Kruidberg. A positive correlation exists between the mean L.agilis number per section and the plant species diversity per section in Duin & Kruidberg. (Spearman's rank correlation, S = 1214.5, p-value = 0.020, rho = 0.47).
Plant species diversity in Duin & Kruidberg is significantly lower than in Voornes Duin (Figure 16).
In both areas, raggedness of the individual trajectories is positively correlated with the mean number of L.agilis observed per section (Figures 19 & 20).
Figure 18: Correlation between lizard number per section and raggedness in Duin & Kruidberg. A positive correlation exists between the mean L.agilis number per section and the raggedness score in Duin & Kruidberg. (Spearman's rank correlation, S = 792.24, p-value = 0.002, rho = 0.61).
Figure 17: raggedness means of Voornes Duin and Duin & Kruidberg. No significant difference was found in Raggedness between Voornes Duin and Duin & Kruidberg (Wilcoxon rank sum test, W = 291.5, n = 24, p-value = 0.31).
Figure 16: The mean plant species diversity per section of both areas. The plant species diversity in Duin & Kruidberg is significantly lower compared to the diversity in Voornes Duin (Wilcoxon signed rank test, V = 300, p-value < 0.001)
Figure 19: Correlation between lizard number per section and raggedness in Voornes duin. A positive correlation exists between the mean L.agilis number per section and the raggedness score in Voornes Duin. (Spearman's rank correlation, S = 2069.2, p-value = 0.007, rho = 0.49).
Relationship between geomorphology and Lizards
(Lead author: Gijs Bakker)
To test the relationship between the geomorphilogical factors and the number of lizards in the sections of the routes, we conducted several Wilcox tests. With these tests, we tested for a difference in the average amount of L.agilis between the presence and absence of the geomorphilogical factors. The number of L.agilis was recalculated to number of lizards per kilometer, to correct for the different length of each section. From these tests only two significant results were found. We found a significant difference in the average number of L. agilis between the absence and presence of the factor distance and the factor of wind. These results were only found in Voornes Duin, in Duin & Kruidberg no significant
differences were found between the presence and absence of the geomorphilogical factors. The first significant difference we found was between the geomorphilogical factor of wind and the number of L.agilis in Voornes Duin (figure 20). We found that the number of L. agilis present in the sections was significantly higher when the geomorphilogical factor wind was present in the sections, for the factor wind the presence and absence of wind shelter was counted, resulting in a higher number of L.agilis at windshelterred areas( P = 0.009).
The second significant difference we found was between the geomorphilogical factor of distance and the number of L.agilis in Voornes Duin (figure 21). We found that the number of L.agilis was significantly higher when other populations of L.agilis were within 500 meter, which we called the factor distance (P = 0.001).
Figure 20: Histogram of the average number of L.agilis per kilometer and the presence of the geomorphilogical factor wind. For this factor the presence and absence of wind shelter was counted. This resulted in wind sheltered at wind factor present and open area at wind factor absence. A significant difference can be seen, between the number of lizards in wind sheltered and wind open areas, a higher number of L.agilis is found in wind sheltered areas (P = 0.009).
Figure 21: Histogram of the average number of L.agilis per kilometer and the presence of the geomorphilogical factor distance. For this factor the presence of populations of L.agilis within 500 meter was scored. Resulting in populations present at distance present and populations absent at distance absent. A significant difference was found between populations present and absent, resulting in higher amounts of L.agilis when other populations were within 500 meter (P = 0.001).
Mechanical management - Vegetation
(Lead author: Aron
Coffa)
Using a CCA analysis influence of management types on plant species can be visualised, an ANOVA can than be used to determine the significance. For vegetation no data was
excluded so for Voornes Duin n=211 and for Duin en Kruidberg n=120. A significant influence of management on plant species was found in Voornes Duin (p=0.034). Yet this significance explains only 3.6%. 96.4% of the differences in vegetation cannot be explained by management. Figure 23 shows that vegetation removal and mowing are correlated with each other. The same goes for spraggelen and logging.
Figure 22: CCA of management versus plant species arrows indicate influence and correlation of
management types (blue arrows). Red crosses visualize different plant species and their position indicates correlation with different management types.
No further significant correlation between management and vegetation was found for Duin en Kruidberg. Also no significant results were found in the CCA of management against geomorphological factors.
Table 9: Eigenvalues CCA, the eigenvalues of different axes in the CCA analyses including the explained and cumulative proportions
CCA1 CCA2 CCA3 CCA4 CCA5
Eigenvalue 0.1104 0.08021 0.05448 0.0438 0.0368
Proportion explained 0.339 0.24626 0.16727 0.1345 0.113
Cumulative proportion 0.339 0.58525 0.75252 0.887 1
Relationship between natural management and the L. agilis
habitat suitability
(Lead author: Sarah Scholten)
The grazing intensity between the research area and the reference area is not significantly correlated. The LSU of 0, 0.0481, 0.1 and 0.12 per ha (table 10) are classified as
respectively no, low, middle and high grazing intensity. The majority of the vegetation reaches a height up to 5 - 20 cm (figure 24). Furthermore, the vegetation heights in areas
without grazers are almost equally divided over 5 - 20, 20 - 50 and 50 - 150 cm. There is significant difference between the middle grazing intensity with the vegetation height of 20 - 50 cm.
Figure 24: Vegetation height compared with grazing intensity. This figure shows the influence of grazers on the vegetation height. Significant correlations between vegetation heights and grazing intensity have been analyzed with an ANOVA (*: p = 0.007), after there has been concluded that it is normally distributed with a Shapiro-Wilk test.
There cannot be found any significant values between topography within the landscape and egg-laying sites with grazing intensity.
Since there are no significant relations found between vegetation heights and the number of rabbits there could be stated that rabbits have no influence on the vegetation heights. No significant differences were found between between egg laying places and relief with rabbits. Rabbits are not significantly correlated with the raggedness.
Another significant correlation can be found with Spearman rank correlation test between the number of rabbits and the average number of L. agilis per kilometer (rho = 0.431, p = 0.001). Lastly, there is a significant negative rank correlation between the raggedness and the grazing intensity (rho = -0.223, p < 0.001). Grazing intensity was negatively correlated with the number of L. agilis (rho = -0.292, p < 0.001). Also, a negative correlation exists between the mean plant species diversity per section and the grazing intensity in the research area (rho = -0.269, p < 0.001) and in the reference area (cor = -0.18, p < 0.001).
Discussion & conclusion
Relationship between L.agilis population density and vegetation
characteristics
(lead author: Bart-Jan Akerboom)
A diverse vegetation structure was expected to be positively correlated with the L. agilis population density (e.1). However, in both areas the vegetation structure diversity is not correlated to the L.agilis population density. Hypotheses e.1 is therefore rejected.
Furthermore no correlations between any of the vegetation height classes and the L.agilis population density were found in Voornes Duin as well as Duin & Kruidberg. Therefore, can be said that the vegetation height composition did not have an impact on the L.agilis
population density in Voornes Duin.
Secondly, the plant species composition was not correlated to the L.agilis population density in both areas, thereby rejecting hypothesis e.2.
In Duin & Kruidberg a positive correlation was observed between plant species richness and L. agilis population density (figure 15) (e.3). This correlation was not found in Voornes Duin, however. Furthermore, Duin & Kruidberg had a significantly lower mean plant species diversity per transect than Voornes Duin (figure 16). In Duin & kruidberg the plant species diversity of certain sections could be so low that it dropped beneath a threshold for L.agilis habitat requirements and hereby, Hereby, making the amount of plant species a more important factor in Duin & Kruidberg than in Voornes Duin. This could be caused, as van Van Leeuwen & Van de Hoef (1976) suggested, by a lowered insect population density in less diverse vegetation, creating food scarcity. Another possibility is that the species diversity creates a more complex vegetation structure, not measured for in our vegetation structure measurements. In Voornes duin the plant species diversity per transect is high enough that it does not influence the L.agilis population density that much. There must, however, also be taken into account that the overall L.agilis population density in Voornes Duin is much lower than in Duin & kruidberg, making issues such as food scarcity less problematic.
The results were completely in line with hypothesis e.4. In both areas a significant correlation was present between the raggedness of the sections and the L.agilis population density (fig 18, fig 19).However, the raggedness of Duin & kruidberg did not differ
significantly to the raggedness of Voornes Duin (fig 17). Thus, the raggedness of Voornes Duin cannot be assumed to be a cause for the lower L.agilis density in this area compared to Duin & Kruidberg.
Taking all the results from the vegetation analysis into account, can be concluded that the combination of ecological characteristics of the habitat does has an influence on the L.agilis population density. The results show a strong indication that plant species diversity is correlated with the L.agilis population density. The coverage percentage of the vegetation height categories does not influence the L.agilis population. The raggedness is strongly correlated with the L.agilis population density.
Geomorphology
(Lead author: Gijs Bakker)
My main goal was to find out if the geomorphilogical factors we selected had any effect on the population of L.agilis in Voornes Duin and Duin & Kruidberg. Following the results we got, we can conclude that the geomorphilogical factors we chose have no significant effect on the population of L.agilis in Duin & Kruidberg. However we can conclude that they do have a significant effect in Voornes Duin.
In Voornes Duin we found that the geomorphilogical factor of distance and wind affect the number of L.agilis that we found in the sections. For the factor of distance we measured if other populations of L.agilis were within 500 meter, so they could possibly mix and replenish each others subpopulation. Our results show that if these subpopulations are within 500 meter, the average number of L.agilis we found went up significantly. This shows that the presence of these subpopulations does indeed play an important role and that so called isolation of these subpopulations can be a negative influence on the populations overall. For the factor of wind we measured if the section was wind sheltered or not. We
hypothesized that shelter from the wind could benefit the thermoregulation of L.agilis and would therefore have a positive effect on the number of L.agilis. Our results confirmed this hypothesis, we found that wind sheltered sections had a significant higher number of L.agilis than non sheltered sections.
The relations we found between the geomorphilogical factors and the number of L.agilis, were only significant in Voornes Duin and not significant or even present in Duin &
Kruidberg. We think this is because of their location, because Voornes Duin is located on an island it is already isolated for natural influx from L.agilis populations outside of the island. Therefore we believe they have to rely on influx from populations on the island itself to counter sudden declines in the other populations in Voornes Duin. As we discovered these populations in Voornes Duin are also isolated from each other making it hard, if not
impossible to have influx between them. Next to this we believe that wind is more present on an island than on land, making it more influential in Voornes Duin than in Duin & Kruidberg.
Mechanical management (
Lead author: Aron Coffa)
From the anova done on the CCA in figure 22 we can conclude that mechanical
management does have a small, but significant, influence on the vegetation diversity in Voornes Duin. This means we do not reject the m.1 hypothesis A couple of plant species only occur once, this gives some disturbance in our data. It turned out that the mechanical management in Duin en Kruidberg was not as extensive as we thought it would be. No extensive papers were available with past and present management, but the ranger could provide us with some general information. This might have resulted in a less accurate analysis.
Overall we can conclude that through our methods only a small influence of management can be found. This might be the case because when looking at suitable routes and sections we did not entirely focus on management, but key vegetation structures for L. agilis. This might in fact be in areas with less management or less recent management, like Duin and Kruidberg. Future research can focus on following a recently managed area, to see how this develops over time in both floral and faunal ways.
Relationship between natural management and the L. agilis
habitat suitability (
Lead author: Sarah Scholten
)
The main question in this part of the study was “To what extent does a grazer population affect L. agilis habitat suitability?”
Against expectations, our study has made clear that grazing intensity hardly influences the cover percentages of different vegetation height classes.n.1 Only one one significant
difference on the vegetation height was observed in the research area (figure 24). In
contradiction to previous research, no significant correlations were found between vegetation height and grazing intensity in the reference area Duin & Kruidberg.
woody plants to complement their diet of grasses, horses forage on sedges and herbs (Cromsigt, Kemp, Rodriguez & Kivit, 2017; Menard, Duncan, Fleurance, Georges & Lila, 2002).
Voornes Duin is exclusively grazed by Highland cattle, while Duin & Kruidberg contains also horses. The difference in diet and the difference of bite might be another explaining factor on the difference in vegetation pattern (Hongo & Akimoto, 2003). According to de Bonte, Boosten, van der Hagen & Sýkora (1999) grazers are able to create different levels in vegetation heights, because they prefer some plants above other plant species. In line with their research, this analysis indicates a negative correlation between the raggedness and the grazing intensity.n.2 Additionally, as was expected, grazing intensity decreases the mean
plant species diversity.n.3 Grazers prefer the most palatable and nutritious species (Olff &
Ritchie, 1998; Bakker, et al., 2006).
The geomorphological aspects egg laying places and relief were not significantly correlated with grazing intensity. Apparently, grazing livestock does not influence the geomorphological aspects of both areas.n.4 The grazing intensities in the study areas might
be relatively low compared to other studies in which grazers do have influence on relief and sand patches. However, this is only speculation since other studies does not give detailed information about the intensities.
Rabbits have no significant effect on vegetation heights, even though it is a grazing species (de Beer, 2012). This can be explained by the fact that rabbits only occur on short grasses, because they do not lower grasses (Iason, Manso, Sim & Hartley, 2002). The relief and egg laying places in the landscape are also not influenced by rabbits (de Beer, 2012; Gómez-Brandón, Lores, & Domínguez, 2013).n.5 With their digging, rabbits create patches of
bare sand or very open vegetation (de Beer, 2012). No correlations between rabbits and the raggedness are found.n.6 Similar to grazing livestock, rabbits are specific grazers, which
could be a reason for different levels in the vegetation (Gómez-Brandón, Lores, & Domínguez, 2013). So, overall, natural management barely influences L. agilis habitat suitability.
There was no significant correlation between the total number of lizards and grazing intensity. The only significant factor is the raggedness. This component of variation in vegetation structure might be associated with grazing livestock, which therefore indirectly affects the L. agilis population. Striking is that rabbits are correlated with the number of L. agilis. This might be explained by the fact that the research area had until very recently almost no rabbits. Nowadays, rabbits are being released in the research area. Besides, the numbers of grazers had just been reduced in 2018. Even though grazers barely affect L. agilis habitat, vegetation probably needs time to recover from the intensive grazing and become more suitable for L. agilis. Because research proofs intensive grazing to lower the vegetation. As time to recover increases, the vegetation growth rate increases as well (Oesterheld, & McNaughton, 1991). Hence, vegetation heights in the research area might change in the (near) future.
This research has used several monitoring areas within the research environment. The majority of both dunes are being grazed, it was therefore difficult to obtain data from sites with a grazing intensity of 0. An analysis consisting of more samples without grazers might be more reliable on the results about the influence of grazers on the L. agilis habitat suitability. Besides, grazing livestock and rabbits might affect other factors on the habitat suitability of L. agilis. For example, the reptiles need insects in order to feed themselves. There is a chance that there is a correlation between the number of insects available for L. agilis to consume and the number of grazers or rabbits. This has not been analysed within this research. Moreover, roe deer have not been taken into account in this study, even though they might be able to affect the L. agilis habitat. They consume grasses and keep thicket vegetation low (Abbas et al., 2013). Information and details about the numbers of this species in the areas were not detailed enough to use in our study. Following research should take these aspects into account.
Overall conclusion and discussion
The main research question was: Does management have an effect on vegetation and geomorphology and does this affect the sand lizard population in Voornes Duin? Answering this question required adressing four topics: ecological and geomorphological characteristics and natural and mechanical management.
From the results, we can conclude that natural management has an effect on the vegetation as well as on the geomorphology. Grazing intensity is negatively correlated with raggedness. Mechanical management on the other hand, appears only to have a small effect on the vegetation structure.
Subsequently the question arises what the effects of vegetation and geomorphology were on the population of L. agilis in Voornes Duin. It can be concluded that especially the raggedness of the vegetation has a positive effect on the L. agilis population. The
geomorphological factor distance influences the population negatively.
Based on these results, we can conclude that management can indirectly influence the population of L. agilis through its effect on vegetation and geomorphology. However, the decline in L. agilis population of Voornes Duin cannot be attributed to management alone, since this is very similar in Voornes Duin and Duin & Kruidberg. However, other factors could have caused the decline in the L.agilis population on Voorne.
Since a positive correlation has been found between the distance between
populations and the population density of L. agilis, it can be expected that this factor can be of impact on a bigger geographical scale. Two important differences between Voornes Duin and Duin & Kruidberg can play a role in this: the size of the area and the degree of isolation. The smaller size of Voornes Duin (190 ha) could impede good placement of core
populations. Secondly, the geographical isolation of the area prevents influx from core populations from other areas.
Therefore we advise that in future management extra attention is given to the creation of suitable habitat for L. agilis in central places in Voornes Duin with the aim of creating a core population in a central place. From this core population L. agilis can colonize other habitats themselves, potentially with help of migration lanes or stepping stones. The management parameters for this habitat should be: high raggedness and enough bare ground. According to the results, this can be accomplished by the use of the grazers that are already present in the area: Highland cattle, but with a decreased grazing intensity.
Acknowledgement
We would like to thank all researchers and staff of the national parks: Voornes Duin and Duin & Kruidberg for making this research possible, and special thanks to Ingo Janssen, Gerard Oostermeijer, Han Meerman, Ruud Luntz, Mathijs Broere, Lisette de Ruiter and Marcel Westerhout for their contribution to this study.
Appendix
Appendix I:
Lizard monitoring protocol: Before the monitoring:
● Preferably monitor L.agilis between 09:00 and 12:00 at right conditions ○ Between 20 and 30 degrees Celsius
○ Sunny with a light cloud cover ○ No rain, hail or snow
○ Wind Speed below 5 Beaufort
● If not, wait for a moment in the day that the conditions are favourable. During monitoring:
○ Do not walk faster than 4 km/h and stop now and then
○ Don’t walk in a straight line, but look for suitable patches along your Route. ○ Males, females and subadults are scored
○ Subadults: smaller individuals which lack adult skin pattern.
○ If a Lizard has found cover before the sex could be determined score as unknown.
Appendix II:
Vegetation measurements protocols:
● Score the raggedness of the section using the standard picture.
● Determine a plot location by counting your steps given by a random number generator.
● Lay out a 1x1m plot using rope.
● Make an inventorisation about which plant species are present. ● Estimate their coverage on the DAFOR scale.
● Estimate height within 5 categories: ○ Bare ○ 0-5 cm ○ 5-20 cm ○ 20-150 cm ○ >150 cm
Appendix III:
Protocol rabbits● asses the size of the rabbit population by the droppings density per section ● score the size on a scale of 1 to 5.
Appendix IV:
Figure 1: Locations of the sections in route 1 of Voornes Duin.
Figure 3: Locations of the sections in route 3 of Voornes Duin.
Figure 5: Locations of the sections in route 5 of Voornes Duin.
Figure 7: Locations of the sections in route 2 of Duin & Kruidberg.
Figure 8: Locations of the sections in route 3 of Duin & Kruidberg.
References
Abbas, F., Picot, D., Merlet, J., Cargnelutti, B., Lourtet, B., Angibault, J., Verheyden, H. (2013). A typical browser, the roe deer, may consume substantial quantities of grasses in open landscapes.
European Journal of Wildlife Research, 59(1), 69-75.
Adler, P., Raff, D., & Lauenroth, W. (2001). The effect of grazing on the spatial heterogeneity of vegetation. Oecologia, 128(4), 465-479.
Affre, L., Suehs, C. M., Charpentier, S., Vila, M., Brundu, G., Lambdon, P., ... & Hulme, P. E. (2010). Consistency in the habitat degree of invasion for three invasive plant species across Mediterranean islands. Biological Invasions, 12(8), 2537-2548.
Andrén, C., S.A. Berglind & G. Nilson, 1988. Distribution and conservation of the northernmost populations of the Sand Lizard Lacerta agilis. Mertensiella, supplement zu Salamandra, 1: 84-85.
Abbas, F., Picot, D., Merlet, J., Cargnelutti, B., Lourtet, B., Angibault, J., Verheyden, H. (2013). A typical browser, the roe deer, may consume substantial quantities of grasses in open landscapes.
European Journal of Wildlife Research, 59(1), 69-75. doi:10.1007/s10344-012-0648-9
Avila, S. P. (2003). The littoral molluscs (Gastropoda, Bivalvia and Polyplacophora) of Sao Vicente, Capelas (Sao Miguel island, Azores): ecology and biological associations to algae. Iberus, 21(1), 11-33.
Attum, O. A., & Eason, P. K. (2006). Effects of vegetation loss on a sand dune lizard. Journal of Wildlife Management, 70(1), 27-30.
