Small mammals, Ixodes ricinus populations and vegetation structure in different habitats in the Netherlands

Transcription

1 WAGENINGEN UNIVERSITEIT/ WAGENINGEN UNIVERSITY LABORATORIUM VOOR ENTOMOLOGIE/ LABORATORY OF ENTOMOLOGY Small mammals, Ixodes ricinus populations and vegetation structure in different habitats in the Netherlands No.: Name: Arno Hoetmer Period: March August e Examinator: Leo van Overbeek 2e Examinator: Willem Takken

2 Content Abstract... 2 Introduction...3 Tick Habitat, life cycle and ecology of ticks... 3 Borrelia burgdorferi sensu lato... 5 Hosts... 6 Previous research in study areas... 7 Aim of the research... 8 Material and Methods... 9 Study Area... 9 Meteorological data Collection of ticks Rodent trapping Statistical Analysis Results Meteorological data Ticks questing in the vegetation Tick hosts Discussion Conclusion Future research Improvements for this study Acknowledgements References Appendix A: photographs of the plots Appendix B: Weighted mean method Appendix C: activity of the rodents in each subplot Appendix D: Summary of the captured rodents

3 Abstract Ticks are important vectors for Lyme-borreliosis. Different variables influence tick abundance in an area, for example the microclimate, the type of vegetation and the presence of hosts like mammals and birds. Different life stages of ticks mostly use different hosts. Adult ticks usually use large mammals as hosts, nymphs use mostly medium sized mammals and larvae mostly use small mammals such as rodents. This research studied the influence of rodent populations on tick abundance. During this study 3 different areas were used, the Kwade Hoek (KH), the Amsterdamse Waterleidingduinen (AWD) and the Nationaal Park De Hoge Veluwe (HV). In each area two different plots were selected, which consisted for the AWD and the HV of a pine forest and an oak forest. In each plot rodents were captured and each individual rodent was analysed for the presence of ticks. In each plot ticks on the vegetation were collected and counted. The highest number of nymphs in the plots was found in the two HV plots and the AWD2 plot. In the pine dominated forests (HV2 and AWD2) more nymphs were collected than in their oak dominated forest counterparts (HV1 and AWD2). The number of questing adults and nymphs was lower in the morning (collecting started at 9:00) than in the evening (collecting started at 21:00). The wood mouse (Apodemus flavicollis) and the bank vole (Clethrionomys glareolus) were the most commonly caught rodents. In the KH plots only wood mice were caught. The KH1 (51 wood mice), followed by the HV2 (43 wood mice) and the HV1 (36 wood mice) had the largest wood mice populations. The AWD1 (44 bank voles, 8 wood mice) was the only plot with a higher population of bank voles than wood mice. The AWD1 had the largest bank vole population followed by the HV1 (21 bank voles), and these two plots were also the two oak forest plots. The rodents in the HV2 carried the highest number of larvae, followed by the HV1 and the AWD2. The AWD1 had the lowest number of larvae on the rodents, and also in the vegetation, even though it did have a large rodent population which consists mostly of bank voles. Irrespective of the plot, bank voles were parasitized by fewer ticks than wood mice. The AWD2 had a high number of nymphs in the vegetation, but only few rodents and also only few ticks on the rodents. Therefore we expect that another host, possibly one or more bird species, is mostly responsible for the abundance of nymphs in the vegetation of the AWD2. The KH1 had one of the largest rodent populations but only a few nymphs were collected in the vegetation, and relative lower number of ticks were found on the rodents. This suggests that another factor, possibly vegetation type, is also important for tick abundance, other than the rodent population. The wood mouse seems to be more important in tick development than the bank vole. Bank vole populations do not seem to have a positive effect on tick populations. 2

4 Introduction After mosquitoes, ticks are the most important vectors of infectious diseases in humans (de Boer et al. 1990). Lyme-borreliosis, one of the main tick-borne diseases, is becoming more frequent in the Netherlands (den Boon et al. 2003, Smit et al. 2004). The number of tick bites might even have tripled in some areas from 1994 to 2001 (den Boon et al. 2003). The main vector of Lyme-borreliosis is Ixodes ricinus, a very common tick in Dutch woodlands (de Boer et al. 1990). I. ricinus is the most important vector of human diseases in Europe. It is also associated with diseases such as Tick-borne encephalitis virus, Tularaemia, Babesiosis and Ehrlichiosis (Lindström et al. 2003). In Europe two different Ixodes species are significant Lyme-borreliosis vectors. These species are I. persulcatus and I. ricinus (Gray 1998), and both species are three-hosts ticks. Tick Habitat, life cycle and ecology of ticks Ticks feed on the blood of vertebrates and use this blood for their development. A full life cycle of ticks usually takes about three years (Randolph et al. 2002) (Figure 1), but can take up to six years. During a ticks life it has four developmental stages: egg, larva, nymph and adult (Figure 1). Ticks in all developmental stages only feed on a host for a couple of days. Thereafter they detach from the host and hide in the vegetation to develop to the next developmental stage, which takes approximately one year (Gray 1998). After moulting the tick emerges as a new developmental stage and starts questing for a new host. During the immature stages (larva and nymph) the ticks mostly parasitize small to medium sized mammals and birds. During the adult stage the ticks mostly parasitize large mammals (Gray 1998). Questing adult and larval ticks have a maximum life expectancy of 3 to 4 months if they do not feed (Walker 2001). Tick activity starts when the weekly averaged daily maximal temperature is above 7 ºC (Perret et al. 2000). Nymphs and adults can quest at all times during the year as long as the temperature and humidity are suitable, although they usually are quiescent in the winter (Walker 2001). 3

5 Figure 1: life cycle of ticks, the relative size of the host approximates their significance as host for the different life stages in a typical woodland habitat (Courtesy of J. Grey and B. Kaye). Different variables can influence the tick density in certain areas. These variables include: Microclimate (Michalik et al. 2003) Type of vegetation (Gray 1998, Lindström et al. 2003, Michalik et al. 2003) Presence of hosts, for example large mammals (Gray et al. 1999, Michalik et al. 2003, Pichon et al. 2006, Wielenga 2006), medium sized mammals (Michalik et al. 2003) and small mammals (Michalik et al. 2003). Ticks are most successful in an area with a diverse mix of fauna (Gray 1998). The number of tick bites has been associated with the proportion of area covered by woods, sandy soil, dry uncultivated land and the number of sheep and humans (de Mik et al. 1997). Ticks seem to be mostly present in woodlands (Gray 1998, Lindström et al. 2003, Michalik et al. 2003), particularly woodlands containing beech (Lindström et al. 2003) and oak (Gray 1998, Wielenga 2006). There is a very low abundance of ticks in open areas such as meadows (Lindström et al. 2003), however if the vegetation is rough and there is a high rainfall ticks are able to survive in open areas (Gray 1998). Some studies suggest that ticks are more abundant in pine forest than in oak forests (Smit et al. 2004), other studies suggest the opposite (Gray 1998, Lindström et al. 2003, Michalik et al. 2003). So there is no general agreement on the most favourable habitat for tick survival found in literature. One factor that could influence the difference in the results in the previous studies could be the undergrowth. Smit and coworkers suggest that in their study the pine forest had a more favourable undergrowth (grass 4

6 vegetation) than the oak forests (small bush vegetation) (Smit et al. 2004). Most studies do show that a woodland is favourable over an area without trees (Gray 1998, Lindström et al. 2003, Wielenga 2006, Wielinga et al. 2006). It is known that ticks are very sensitive to desiccation (Randolph et al. 1999), the relative humidity should be 80% throughout the year (Gray 1998). Especially in open areas the microclimate is important for tick survival (Lindström et al. 2003). However, the different tick densities in different vegetation types might not only be related to the difference in vegetation, soil or microclimate. It could also be caused by a difference in density of host animals (Lindström et al. 2003), since host animals also prefer certain vegetation types. Ticks either become active and feed immediately in early autumn or wait until the following spring. Most unfed ticks start questing in April and May. The group that feeds in early autumn suffers a morphogenetic diapause before moulting, and the group that feeds in spring suffered a behavioural diapause before feeding (Walker 2001). Both groups start moulting to the next stage in the following autumn (Randolph et al. 2002). The timing of questing activity is caused by a temperature threshold. This threshold varies for ticks with different sizes; larger ticks become active before the smaller ones (Randolph et al. 2002). In the Netherlands the highest number of ticks is found in the months June, July and August (Wielinga et al. 2006). When the tick community starts questing they move to a higher place in the vegetation. How high they climb is correlated to the height of the vegetation. In general larvae quest at a lower height then nymphs and adults (Mejlon et al. 1997). The part of the tick population that quests is determined by seasonal patterns, host contact and mortality rates (Randolph et al. 1999). Questing activity of ticks could be influenced by microclimate (Randolph et al. 1999, Michalik et al. 2003). The proportion of ticks questing in the vegetation depends on the humidity. In dry conditions the nymphs stay closer to the ground. Larvae seem to avoid dry conditions by becoming quiescent and only quest when the humidity is higher (Randolph et al. 1999). Larger ticks can quest more continuously than the smaller one who are forced to temporal inactivity caused by water stress during high summer (Randolph et al. 2002). Borrelia burgdorferi sensu lato Lyme-borreliosis is caused by Borrelia burgdorferi sensu lato spirochaetes that are transmitted to human by tick bites (Wielinga et al. 2006). Ticks get infected by the Borrelia species when they feed on a vertebrate who is also infected with Borrelia spp and when the infected tick feeds on another vertebrate, that vertebrate will also be infected. So for Borrelia to be present in an area both ticks and vertebrate hosts are necessary. The Borrelia spirochaetes inhabit the midgut of the ticks. When the tick feeds on a host they penetrate the midgut wall and translocate to the salivary glands via the haemolymph. After that they pass into the feeding lesion with the saliva. This migration of the spirochaetes means that most infections do not occur for at least two days after attachment (Gray 1998). There are different genospecies of Borrelia burgdorferi s.l.. All genospecies transfer Lymeborreliosis (Smit et al. 2004). In Europe most common species are Borrelia afzelii (Le Fleche et al. 1997, Pichon et al. 1999, Christova et al. 2003, Hanincová et al. 2003, Rauter et al. 2005, Wielinga et al. 2006, Pecchioli et al. 2007), Borrelia garinii (Le Fleche et al. 1997, Pichon et al. 1999, Hanincová et al. 2003, Rauter et al. 2005, Wielinga et al. 2006, Pecchioli et al. 2007), Borrelia burgdorferi sensu stricto (Le Fleche et al. 1997, Pichon et al. 1999, Wielinga et al. 2006, Pecchioli et al. 2007), Borrelia lustiniae (Le Fleche et al. 1997) and 5

7 Borrelia valaisiana (Le Fleche et al. 1997, Hanincová et al. 2003, Wielinga et al. 2006, Pecchioli et al. 2007) In Europe there are about 5 major genospecies of Borrelia burgdorferi s.l., but only 1 major tick species. Most Borrelia burgdorferi s.l. genospecies are linked to certain vertebrate hosts. The species of Borrelia burgdorferi s.l. can be divided into three ecological groups: 1 species that are adapted to small mammals, 2 species that are adapted to birds, 3 species that are not specialized (Kurtenbach et al. 2002). Since all these groups are transmitted by one tick species, and this tick parasitizes different vertebrate hosts, it can be concluded that the niche of the Borrelia burgdorferi s.l. genospecies is determined by the vertebrate host species and not by the tick (Kurtenbach et al. 2002). B. afzelii (Humair et al. 1999, Pichon et al. 1999, Hanincová et al. 2003, Khanakah et al. 2006) and B. lustiniae are both associated with mammals (Kurtenbach et al. 2002). B. garinii and B. valaisiana are associated with birds (Hanincová et al. 2003, Pichon et al. 2006), although in some studies it was shown that B. garinii can be present in mammals (Khanakah et al. 2006), and that the different ribotypes of B. garinii might have different vertebrate hosts. B. burgdorferi s.s is not specific (Pichon et al. 1999, Kurtenbach et al. 2002), so it can occur in both mammals (Khanakah et al. 2006) and birds. Ticks can also contain mixed infections of more then one genospecies of Borrelia burgdorferi s.l. (Pichon et al. 1999, Khanakah et al. 2006, Wielinga et al. 2006). Overall Borrelia infection in tick populations is highly variable ranging from 0% to over 50% (Gray 1998, Christova et al. 2003, Wielinga et al. 2006). Different factors influence the infection rate of tick population. There is a correlation between ticks density and density of infected ticks (Jouda et al. 2004). Also the tick infection rate can be different in different vegetations (Wielinga et al. 2006). Another Dutch study found higher tick infection rates in oak forest compared to pine forest (Smit et al. 2004). Adult ticks have a higher infection prevalence then nymphs (Jouda et al. 2004, Rauter et al. 2005). There is no clear relation in Borrelia infection between tick sex, some studies found higher proportion of infection among males (Wielinga et al. 2006), others found a higher proportion among females (Michalik et al. 2003). The transmission of Borrelia is usually trans-stadial, which means from stage to stage, and not transovarial, which means that an infected female transmits the Borrelia to her eggs. Therefore reservoir hosts such as rodents are important for new infections of larvae and nymphs (Gray 1998). However even though most transmission is trans-stadial, it is suggested that the transovarial transmission might have a considerable role in maintaining the circulation of spirochaete (Gray 1998). Hosts There are two type of hosts for ticks (Gray 1998). Reproduction hosts are mainly parasitized by adult ticks. These reproduction hosts have little influence in the maintenance of Borrelia infected ticks (Gray 1998), but the presence of these reproduction hosts does increase the tick abundance (Pichon et al. 1999, Wielenga 2006). Larger mammals such as deer, cattle and sheep are reproduction hosts. Smaller mammals such as rodents, squirrels and hedgehogs are called reservoir hosts. They are mostly infested with nymphs and larvae, and therefore, if they are infected with Borrelia, are able to infect a tick in a sub adult life stage. This infected tick is able to spread the obtained Borrelia to other vertebrates. Therefore these reservoir hosts 6

8 heavily influence the maintenance of Borrelia infested ticks (Gray 1998, Michalik et al. 2003). Birds could also be a major reservoir host for Borrelia (Gray et al. 1999, Estrada-peña et al. 2005). However, birds and mammals usually are hosts to different genotypes of Borrelia (Le Fleche et al. 1997, Kurtenbach et al. 2002), so they will also be reservoir hosts for different Borrelia genospecies. Field mice (Apodemus spp.) and voles (Clethrionomys spp.) are abundant rodent genera in Dutch nature areas (Dijk 2003, Smit et al. 2004, Gassner 2005). Together with other rodent species these species are examples of reservoir hosts (Gray 1998, Khanakah et al. 2006). Rodents are mostly infested by larvae of I. ricinus and much less by nymphs (Humair et al. 1999, Michalik et al. 2003). Different factors influence the number of attached ticks to the rodents. The number of ticks in the area is a factor determining the number of ticks attached to rodents. The infection rate of rodents can also be influenced by the vegetation type. A study in the Netherlands found a higher infection rate of rodents in deciduous woodland over evergreen woodland (Smit et al. 2004). The number of ticks attached to voles can differ among sex and age, with more ticks on males than females, and more ticks on adults then on juveniles (Tälleklint et al. 1997). Among different species of rodents there can also be a difference in tick infestations. In most studies field mice were more heavily infested with ticks than voles (Tälleklint et al. 1997, Humair et al. 1999, Dijk 2003, Hanincová et al. 2003, Michalik et al. 2003). Only in one study in Spain voles contained more ticks than field mice (Estrada-peña et al. 2005). The reason for the different result by Estrada-peña and co-workers is unclear. The proportion of Borrelia infected field mice and voles was different in different studies. Some found a higher Borrelia infection in voles than in field mice (Humair et al. 1999). However most studies found the opposite, a higher proportion of field mice was infected with Borrelia spp. than voles (Hanincová et al. 2003, Smit et al. 2004, Khanakah et al. 2006)). Some studies suggest that mice and voles might have different immunological strategies to cope with tick-borne pathogens. Mice are more focused on the micro parasite, whereas voles are more focused on reducing the vector (Humair et al. 1999, Hanincová et al. 2003). This different immunological strategy theory is supported by the fact that Clethrionomys glareolus (bank vole) reduced the percentage of fully engorged ticks after repeated infestations. This way the time ticks feed on C. glareolus is reduced. Apodemus flavicollis does not show this resistance to repeated infestations of I. ricinus (Dizij et al. 1995). Furthermore field mice show higher levels of B. burgdorferi-specific antibodies than voles do (Kurtenbach et al. 1994). Voles are therefore a more effective reservoir host in transmission of Borrelia compared to field mice (Michalik et al. 2003), but since field mice are mostly more abundant (Gray et al. 1999, Estrada-peña et al. 2005, Khanakah et al. 2006) and heavily infested by ticks, field mice might be a more important vector host for Lyme-borreliosis (Dijk 2003). Previous research in study areas Different studies on tick populations, Borrelia infections and rodent populations have been done in Dutch nature areas such as the National Park De Hoge Veluwe (NPHV) and the Amsterdamse Waterleidingduinen (AWD) (Dijk 2003, Smit et al. 2004). However in the Kwade Hoek (KH) no such studies have been done before. The NPHV has a higher rodent population than the AWD (Dijk 2003, Smit et al. 2004). A. sylvaticus (wood mouse) was the most common rodent in both the AWD and the NPHV, followed by C. glareolus (Dijk 2003, Smit et al. 2004). The area and habitat did not have an 7

9 influence on the number of rodents caught (Dijk 2003), only with C. glareolus some influence of habitat was found, namely more C. glareolus were found in a pine woodland compared to oak woodland (Dijk 2003). A higher proportion of A. sylvaticus was infested with Borrelia than C. glareolus (Smit et al. 2004). In a pine woodland the Borrelia infection rate of rodents was higher in the AWD, but in the oak woodland it was higher in the NPHV (Smit et al. 2004). In general the Borrelia infection in rodents was higher in oak woodland compared to pine woodland (Smit et al. 2004). The tick density was higher in the AWD than in the NPHV (Smit et al. 2004). This could be caused by a more favourable microclimate in the AWD; the average temperature and the humidity was higher in the AWD (Smit et al. 2004). Another reason for the higher abundance of ticks in the AWD is the higher abundance of reproduction hosts in the AWD (Smit et al. 2004). The Borrelia infections in ticks population was higher in oak woodland than in pine woodland (Smit et al. 2004). B. afzelii was the most abundant Borrelia genospecies in both areas. The other genospecies that were present in the AWD and the NPHV were B. garinii, B. valaisiana and B. burgdorferi s.s. (Smit et al. 2004). Aim of the research The aim of this research was to gain more knowledge on the interaction between rodent populations, vegetation structure and Ixodes ricinus populations in the Netherlands. This research took place in three different nature areas in the Netherlands, respectively the National Park De Hoge Veluwe (NPHV), the Amsterdamse Waterleidingduinen (AWD) and the Kwade hoek (KH) (Figure 2). Figure 2: Map of the Netherlands by Google earth with the 3 different areas 8

10 Different rodents were caught and were examined for tick attachment and B. burgdorferi s.l. infections. Furthermore ticks were collected and will be examined on B. burgdorferi s.l. infections. These infestations were linked to different factors, among them rodent species, rodent population densities and (in co-operations with another MSc student) vegetation type. The B. burgdorferi s.l. analyses were not performed in this study, but will be performed in a future study. The main research question is: What is the effect of rodent populations on the I. ricinus populations? In order to answer the main research question the following secondary research questions were formulated: a) What are the most dominant rodent species in all areas and do they differ per area? b) Are there differences in rodent populations between different locations within the same area? c) Are all rodents of the same species parasitized equally? Is there a difference between areas and plots? d) Are there differences in the number of ticks that parasitize rodents between rodent species? e) Is there a relation between number of ticks on rodents and number of ticks on the vegetation in a location? f) Is there a relation between number of rodents and the number of ticks on the vegetation in a location? Hypothesis (H1): There is a relation between rodent populations and tick populations in natural areas. H0: There is no relation between rodents populations and tick populations. Borrelia infestations of rodents were not analysed during this internship, but blood and ears were collected and will be analysed by another student in this project. Material and Methods Study Area Three study areas were selected: the KH, the HV and the AWD. In all three areas two plots of 30 by 50 meter were chosen. These two plots differed in vegetation type. Both plots were subdivided into four different subplots. The research sites in all areas were selected to represent different vegetation types, if possible two types of forests vegetation. National park Hoge Veluwe (HV) is an area in the east of the Netherlands. The size of the area is about 5000 hectare. The area consists of mixed landscapes with areas of drifting sands, heath land and almost 50 % of the area is woodland. Most woodland consists of evergreen forests, but there are also some oak dominated areas. Most of the area consists of nutrient poor sand soils. The large herbivores inhabiting this area consist of about 200 red deer (Cervus elaphus), 200 roe deer (Capreolus capreolus), 50 wild boar (Sus scrofa) and 150 mouflon (Ovis mousiman). Two plots have been selected, respectively HV1 and HV2 (Figure 3). HV1 (N52º ; E005º ) is a Sessile oak (Quercus petraea) forest with an undergrowth dominated by Blueberry (Vaccinicium myrtillus). HV2 (N52º ; 9

11 E005º ) is a Scots pine (Pinus sylvestris) dominated forest and the undergrowth is also dominated by Blueberry (Table 1). Figure 3: National park Hoge Veluwe from Google earth. The two plots (HV1 and HV2) are shown. The Amsterdamse Waterleidingduinen (AWD) is a nature area on the border of Noord- and Zuid-Holland, just south of Zandvoort. The size of the area is about 3400 hectare. Besides being a nature area, the area is also used for drinking water production and water purification. The landscape is varied. The area consists mostly of different dune vegetations and at the edges some deciduous woodlands, mostly oak dominated. The large herbivores inhabiting this area consist of fallow deer (Dama dama), roe deer (C. capreolus) and also cattle and sheep. The number of large herbivores in the AWD is higher than in the NPHV (Smit et al. 2004). Two plots have been selected, respectively AWD1 and AWD2 (Figure 4). AWD1 (N52º ; E004º ) is a forest dominated by Sessile oak with some Sycamore maple (Acer pseudoplatanus), the undergrowth is dominated by Bracken (Pteridium aquilinum). AWD2 (N52º ; E004º ) is a forest consisting of mainly Scots pine, the undergrowth is dominated by grass and by European honeysuckle (Lonicera periclymenum) (Table 1). 10

12 Figure 4: The Amsterdamse Waterleidingduinen Amsterdamse from Google earth. The two plots (AWD1 and AWD2) are shown. The Kwade Hoek (KH) is a nature area on the north side of the Zuid-Holland island Goeree- Overflakkee close to Goedereede. The size of the area is about 380 hectare. The KH is unique since the hydrology is a mix of saline and fresh water, with part of the area under inundation during spring tide. The vegetation is mostly coast vegetation and dune vegetation. The large herbivores inhabiting this area consist of about 30 roe deer (C. capreolus) and in some areas also cattle. In the Kwade Hoek also two plots have been selected, respectively KH1 and KH2 (Figure 5). KH1 (N51º ; E003º ) is a wetland on the south side of the vegetation dominated by Sea-buckthorn (Hippophae rhamnoides) and Blackberry (Rubus fructosus). Other plants that were dominant were Common reed (Phragmites australis) and Yorkshire fog (Holcus lanatus). KH2 (N51º ; E003º is a grass land on the north side of a dune. It is also dominated by Sea-buckthorn, but also by grass (Table 1). 11

13 Figure 5: The Kwade Hoek from Google earth. The two plots (KH1 and KH2) are shown. Table 1: Ground and tree vegetation of the different plots. Meteorological data In each area the relative humidity and temperature (maximum, minimum and average) were recorded continuously (every 30 minutes) during the field study period (May and June) using data loggers (type Gemini Tinytag Plus TPG 1500). The data loggers were protected from damage in a metal cage. A 5 cm diameter PVC drainage tube cap suspended over the data loggers prevented direct rain water impact. Collection of ticks Ticks were collected at three different occasions (Table 2). In May ticks were collected in the different areas once, two weeks before the fieldwork period. The second time ticks were collected was at the beginning of the pre-bait, between 21:00 and 1:00. Three days later the third sampling was done at the same transect between 9:00 and 13:00. 12

14 Table 2: Dates when the ticks were collected and the date of the start and end of the fieldwork period per area. Questing ticks were collected by dragging a 1 square meter white cotton blanket with a rope attached to a PVC pipe. With blanket dragging only questing ticks are captured and these captured ticks may be considered representative for the overall tick population (Pichon et al. 1999). In every subplot the blanket was dragged up and down over a transect of 25 meters (Figure 6). After 25 m the blanket was examined and the attached ticks were counted and collected. The collected ticks were placed in an eppendorf tube using a fine forceps. Captured ticks were stored at 4 C in 70% ethanol for better DNA extraction. Figure 6: Sampling transect. The dotted line is the dragging route of the blanket. A, B, C and D are the different subplots. Every 25 m the blanket is checked for ticks. 13

15 Rodent trapping In May and June rodents were caught during one period for each location. Every trapping period consisted of three days of pre-bait and after that five days of actual trapping during which on the last day also blood and ears of the collected rodents were collected, and one day of only blood collecting. During the pre-bait the traps were placed in the area, but they were not set. This way the rodents could get used to the traps. In every area two plots of traps were used. Every plot consisted of 48 Longworth lifetraps with 6 rows consisting of 8 traps, every trap was placed 5 metres from each other (Figure 7). In the Kwade Hoek the traps were placed in 2 rows of 24 because the vegetation did not allow the normal dimensions. The bait used in this experiment consisted of peanut butter, oatmeal and a mealworm. The bait was refreshed after every catch and at least once a day. The traps were set from 6 PM to 6 AM and they were checked at 12 PM and 6 AM. The decision for this 6 hour interval was a result of the advice of the Ethical Animal Experimentation Committee in order to reduce the stress of the rodents. Every caught rodent, except shrews, was inspected for age, sex, weight and tick density and a sample of the ticks attached to the rodents was also collected (if nymphs were found they were collected) with a maximum of 5 per rodent individual. During the capture period blood samples (200 microlitre) of different rodent individuals were collected on two different days per area, with a maximum of 30 bank voles and 30 wood mice per plot. From the rodents of which blood was collected as many ticks as possible were collected. Figure 7: Traps placed in 6 rows of 8. Every trap is at least 5 meter apart from the next one. Every square with a different shade of grey is a different subplot (for vegetation measurements) 14

16 To estimate the population size of the different rodent species the Weighted mean method was used (Lange et al. 1986). The individually caught rodents were marked with a specific number by clipping some of the fur (Figure 8). Caught shrews were immediately released to minimize the stress of these rodents. Figure 8: Rodent marking places for the mark-recapture technique. Statistical Analysis All statistical tests were performed using SPSS 12.0 for Windows. All data were analysed on normality by the Kolmogorov-Smirnov test for normality. The differences in collected ticks between periods were analysed with a Wilcoxon paired sample test. The difference between collected ticks per plots and ticks parasitizing rodents per plot was analysed with Kruskal Wallis, followed by a Scheffe multiple comparison. The difference between larvae parasitizing bank vole and wood mice was analysed with a Mann-Whitney U test. The comparison of the proportion of larvae in the vegetation and the proportion of larvae on the rodents was done by a Chi-square test. A correlation between collected nymphs and rodent activity was analysed with a Spearman-r test. Results Meteorological data In every area the temperature and relative humidity was recorded at 5 cm of the litter layer. Because the data logger in the KH got inhabited by a spider the relative humidity data became unreliable, therefore the relative humidity and the saturation deficit were not calculated in the KH. In case of rain and usually during early morning when water vapour condensates on the vegetation, relative humidity could reach 100 %. This could influence the average relative humidity. The relative humidity data was corrected for unrealistic outliers. The relative humidity in the AWD and the HV was corrected for values below 15% and also manually smoothened if necessary. Because of these corrections the relative humidity data should be carefully used. The average temperature was calculated over the period from May 5 th until May 30 th (Figure 9). The average temperature of each area was for the HV: 13.4 ºC ± 0.08; the AWD 12.8 ºC ± 0.06 and the KH 14.4 ºC ± In the month of May the KH was the warmest area, followed by the HV and the coldest was the AWD. 15

17 Figure 9: Temperature measured at 5 cm of the litter layer from collecting period one until the end of the fieldwork period in the different plots. Every day has a night (from 21:00 until 9:00) and day (from 9:00 until 21:00) average. The average relative humidity was calculated from May 5 th until June 10 th. The relative humidity was the highest in the HV (92% ± 0.23) and the driest area was the AWD (90% ± 0.18). Note that the relative humidity is mostly above 80% (Figure 10). Since the data loggers in the KH gave unrealistic data, we used the result per day of the KNMI station in Vlissingen for the relative humidity (81 ± 1.1). Figure 10: Relative humidity (RH) measured at 5 cm of the litter layer from collecting period one until the end of the fieldwork period in the different plots. Every day has a night (from 21:00 until 9:00) and day (from 9:00 until 21:00) average. 16

18 The mentioned microclimate factors are important for tick activity. Ticks respond strongly to a combination of temperature (T) and relative humidity (RH) called the saturation deficit (SD) (Randolph et al. 2002). The SD (Figure 11) was calculated by the following equation: SD = (1-RH/100) * * e ( * T). The HV had a higher SD (1.58 mmhg ±0.05) than the AWD (1.44 mmhg ± 0.03) from May 5 th until May 16 th. Tick activity has an optimum at an SD of 4.4 mmhg. When the SD is higher than 4.4 mmhg the SD and tick activity seemed to be negatively correlated (Perret et al. 2000). Figure 11: Saturation deficit (SD) calculated from the temperature and relative humidity measured at 5 cm of the litter layer from collecting period one until the end of the fieldwork period in the different plots. Every day has a night (from 21:00 until 9:00) and day (from 9:00 until 21:00) average. Ticks become less active during rain. Therefore the precipitation data in the different areas was collected from the recorded data in nearby weather stations. For the KH the data of the KNMI station in Vlissingen was used, for the HV the data of the Haarweg meteorological station in Wageningen was used and for the AWD the data of the KNMI station in Amsterdam was used (Figure 12). During the fieldwork period in the AWD more rain fell compared to the fieldwork period in the other areas. Especially was a wet day, this was also the day on which the night collecting of ticks in the AWD1 and AWD2 was performed. This should be considered during the analyses of tick collecting data. 17

19 Figure 12: The daily precipitation obtained from the data of three different weather stations near the three different areas (KH: Vlissingen; HV: Wageningen: AWD: Amsterdam). A difference in microclimate was determined between the areas in the same period. However the ticks were not collected at the same days. Therefore an average of the temperature, RH and the SD of the days the ticks were collected was calculated. During their respective collecting days the KH (15.9 ºC ± 0.25) had the highest average temperature and the HV (15.4 ºC ± 0.24) had a warmer climate than the AWD (14.0 ºC ± 0.16). The average RH humidity was higher in the AWD (98.6 % ± 0.16) than in the HV (93.3 % ± 0.55). As a result of this the average SD was higher in the HV (1.2 mmhg ± 0.11) than in the AWD (0.2 mmhg ± 0.02). This was probably caused by the high amount of rainfall during the collecting period 2 of the AWD. Ticks questing in the vegetation For the analysis of the daily fluctuation and temporal variation in ticks questing in the vegetation, the data of a research done at the same time by a fellow student was added to the data of this research. This data consisted of two additional plots in the KH, two in the AWD and three additional plots in the HV. For the rest of the results only the data gathered during this research was used. In this research 3642 larvae, 1212 nymphs and 61 adults were collected. The data was not normally distributed, even after square root and logarithmic transformation. Each subplot was used as a replicate within a plot. This was under the assumption that ticks do not travel far in a short period of time without rodent hosts. Furthermore, the vegetation was more or less homogeneous between subplots, which means the most dominant plant species were similar. 18

20 Since rodents are not equally active during the entire day (Macdonald et al. 1993), it is important to see if there is also a fluctuation in tick daily activity. Therefore the number of ticks collected in the morning during the collecting period two was compared with the number of ticks collected in the evening. First all areas were pooled together to see an overall change in active tick populations, and after that every area was analysed separately. The number of questing ticks collected in the vegetation did differ during the day, although this was not the case for all life stages of the tick (Figure 13). Overall the number of larvae collected during the morning (65 ± 14) was not significantly different from the number of larvae collected during the evening (65 ± 14) (Wilcoxon paired sample test - Z = ; P = 0.731). Specifically in the KH (Wilcoxon paired sample test - Z = 0.000; P = 1.00; Figure 14) and the HV (Wilcoxon paired sample test - Z = ; P = 0.411; Figure 15) no difference occurred in larvae questing during the day. However in the AWD significantly more larvae quested during the night (Wilcoxon paired sample test - Z = ; P = 0.011; Figure 16). The number of nymphs (morning: 16 ± 3; night: 24 ± 3) (Wilcoxon paired sample test - Z = ; P = 0.002) collected overall during the evening was significantly higher than during the morning. When we look at the areas individually we see that only in the KH (Wilcoxon paired sample test - Z = ; P = 0.003; Figure 14) this difference was observed, in both the HV (Wilcoxon paired sample test - Z = ; P = 0.185; Figure 15) and the AWD (Wilcoxon paired sample test - Z = ; P = 0.093; Figure 16) no difference occurred. The number of adults (morning: 0.1 ± 0.13; night: 1.5 ± 0.3) (Wilcoxon paired sample test - Z = ; P = 0.001) collected overall during the evening was also significantly higher than during the morning. Both in the KH (Wilcoxon paired sample test - Z = ; P = 0.011; Figure 14) as in the HV (Wilcoxon paired sample test - Z = ; P = 0.021; Figure 15) this difference was also observed. However in the AWD (Wilcoxon paired sample test - Z = ; P = 0.437; Figure 16) no difference was found. Overall the number of nymphs and adults questing in the vegetation was higher from 21:00 until 0:00 than from 9:00 until 12:00. This difference was most obvious in the KH. Figure 13: Comparing the daily fluctuation and temporal variation in larvae (A), nymphs (B) and adults (C) questing in the vegetation (N=52 per period). The y-axis represents the number of ticks collected. The x-axis represents the different periods. Collecting period one is during a day at the beginning of May and the collecting period two lasts from the end of May until June 23 rd. The error bars represent the ± 1SE of the mean. The periods were tested using a Wilcoxon paired sample test. NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<

21 The date of the start of the collecting period two was different for every area. There was a two week period between the end of the first area (KH) and the beginning of the last area (AWD). This difference in time could influence the collected ticks in the vegetation. Literature shows that the activity of larvae and nymphs could change over time (Gassner 2005). To examine this difference in activity, the ticks collected during collecting period one (during the day in the beginning of May) was compared with the ticks collected during the day of the collecting period two in the beginning of June. The difference in time between these periods did have some influence on the activity of ticks (Figure 13). Figure 14: Comparing the daily fluctuation and temporal variation in larvae (A), nymphs (B) and adults (C) questing in the vegetation in the KH (N=16 per period). The y-axis represents the number of ticks collected. The x-axis represents the different periods. Collecting period one is during a day at the beginning of May and the collecting period two lasts from May 22 nd until May 31 st.. The error bars represent the ± 1SE of the mean. The periods were tested using a Wilcoxon paired sample test. NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<0.001 However, only the number of larvae questing changed: at collecting period one less larvae (10 ± 3) were questing than at the collecting period two (65 ± 14) (Wilcoxon paired sample test - Z = ; P < 0.001). This difference was not observed in the KH (Wilcoxon paired sample test - Z = 0.000; P = 1.000; Figure 14), which is logical since no larvae were collected in the KH plots. Both the HV (Wilcoxon paired sample test - Z = ; P < 0.001; Figure 15) and the AWD (Wilcoxon paired sample test - Z = ; P = 0.018; Figure 16) had a significant higher number of larvae on the vegetation in collecting period two. The number of nymphs overall (collecting period one: 21 ± 3; collecting period two: 16 ± 3) stayed the same (Wilcoxon paired sample test - Z = ; P = 0.104). This was also observed in both the KH (Wilcoxon paired sample test - Z = ; P = 0.345; Figure 14) and the HV (Wilcoxon paired sample test - Z = ; P = 0.398; Figure 15), however the number of nymphs questing in the AWD was lower in collecting period two (Wilcoxon paired sample test - Z = ; P = 0.005; Figure 16). 20

22 Figure 15: Comparing the daily fluctuation and temporal variation in larvae (A), nymphs (B) and adults (C) questing in the vegetation in the HV (N=20 per period). The y-axis represents the number of ticks collected. The x-axis represents the different periods. Collecting period one is during a day at the beginning of May and the collecting period two lasts from June 3 rd until June 12 th. The error bars represent the ± 1SE of the mean. The periods were tested using a Wilcoxon paired sample test. NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<0.001 The number of questing adults overall also stayed the same (collecting period one: 0.4 ± 0.08; collecting period two: 0.1 ± 0.13) (Wilcoxon paired sample test - Z = ; P = 0.088). This was also observed in every area separately, the number of adults collected in the KH (Wilcoxon paired sample test - Z = ; P = 0.564; Figure 14), the HV (Wilcoxon paired sample test - Z = ; P = 0.244; Figure 15) and the AWD (Wilcoxon paired sample test - Z = ; P = 0.272; Figure 16) stayed the same. Therefore, it seems that during this study the number of collected larvae could be influenced by the order the areas are visited. This will be discussed during the discussion. Overall the amount of nymphs and adults was not different in the different collecting period. However the amount of larvae was higher in the second colleting period. This was observed both in the HV as in the AWD. Figure 16: Comparing the daily fluctuation and temporal variation in larvae (A), nymphs (B) and adults (C) questing in the vegetation in the AWD (N=16 per period). The y-axis represents the number of ticks collected. The x-axis represents the different periods. Collecting period one is during a day at the beginning of May and the collecting period two lasts from June 14 th until June 23 rd. The error bars represent the ± 1SE of the mean. The periods were tested using a Wilcoxon paired sample test. NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<

23 Both the rodent population and the vegetation could influence the tick population. The influence of the vegetation will be analysed more closely by the research of a student of the Resource Ecology Group (Wolfs 2007). The collected questing ticks, during the three collecting periods, in the different plots were compared with each other. The number of collected larvae in a subplot ranged between 0 and 292, the number of nymphs between 0 and 49 and the number of adults between 0 and 7. In the HV1 and the HV2 significantly more larvae were captured than in the other four plots (Kruskal wallis, Chi square= ; df = 5; P<0.001; Figure 17). The number of nymphs questing on the vegetation also varied among the plots (Kruskal wallis, Chi square = ; df=5; P<0.001; Figure 17), with the highest number of nymphs in the AWD2 and the two HV plots. The lowest number of nymphs was collected in the AWD1. The number of adults was not significantly different between the different plots (Kruskal wallis, Chi square = 5.433; df=5; P=0.356; Figure 17). So the HV1, HV2 and the AWD2 had the highest amount of nymphs questing in the vegetation. The AWD1 had the lowest. Figure 17: The number of captured larvae (A), nymphs (B) and adults (C) questing in the vegetation in each plot (N=12 per plot). The y-axis represents the number of ticks captured per subplot. The x-axis represents the different plots. The error bars represent the ± 1SE of the mean. Letters show the statistically different groups as calculated from the Scheffe multiple comparison test. Tick hosts Some observations were done on wildlife other than rodents in the different plots. However no quantified data were gathered. The HV plots had high number of tracks of wildlife, mostly of red deer, roe deer and wild boar. Wild boar, roe deer and badger were observed near or in the HV plots. In the KH1 a roe deer was observed. On different occasions Phasianidae were also observed in and near the KH plots, in the KH1 a nest of Phasianidae with two young was found. In the AWD many fallow deer were observed in and near the plots, the number of tracks found in these plots was much less than in the HV plots. These large hosts are important for tick abundance (Gray 1998, Pichon et al. 1999). During this research 247 rodents were captured and analysed for ticks, of which 160 wood mice, 83 bank voles and 4 common voles. Besides that, a couple of greater white-toothed shrews (Crocidura russula) and common shrews (Sorex araneus) were captured, but not used for any analysis. The rodents captured during blood collections were also not included in the 22

24 tick analyses since the effect of ethanol on ticks feeding on rodents is unknown. In the plots in the AWD and the HV both the wood mouse and the bank vole are the most dominant species (Table 2). In the KH only the wood mouse and the common vole occurred. The populations of wood mouse and bank vole in every plot were estimated using the weighted mean method (Lange et al. 1986). The population of common vole in the KH1 was 2 according to the weight mean method. However, because of this low number of common voles they will be left out of any further analyses of differences between species, they will be used in all analyses of ticks on rodents in general per plots. The AWD1 was the only plot in which the bank vole was the most common rodent captured. In the other five plots the wood mouse was the most common rodent captured, even tough the HV1 also had a large bank vole population (Table 3). Overall the AWD2 and the KH2 had small rodent populations. Table 3: Population of rodents, the mean ± SE of ticks questing on the vegetation and feeding on individual rodents per subplot in every plot. N represents the number of rodents used for the calculation of ticks on rodents. The maximal distance between traps in which the same rodent was captured was calculated for rodent captured more than 5 times. In the HV1 the maximal distance for a rodent captured more than 5 times was 29m, the minimal was 10m. In the HV2 the maximal distance was 25m and the minimal distance 5m. In the AWD1 the maximal was 35m and the minimal distance was 10m. In the AWD2 the maximal was 35 m and the minimal distance was 14m. In these two areas there was not a large difference in the distance between bank voles (maximum of 25m, minimum of 5m) and wood mice (maximum of 35m, minimum of 5m). The KH had different dimensions in the plot setup which makes in difficult to compare this area with the other two areas. In the KH1 the maximal distance was 110m and the minimal was 15m, in the KH2 the maximal distance was 115m and the minimal distance was 15m. The number of parasitizing ticks per rodent species in the different plots was compared. Does a wood mouse in a certain plot have more ticks compared to a wood mouse in other plots? The number of larvae that parasitized a wood mouse were significantly different for some of the different plots (Kruskal wallis, Chi square = ; df=5; P<0.001; Figure 17; Table 3), but between most plots there was no significant difference. The KH plots had the lowest number of larvae per wood mouse and significantly less larvae than the wood mice in the HV2. Between the other plots there was no significant difference. There was no significant difference between the nymphs parasitizing wood mice in the different plots (Kruskal wallis, Chi square = 0.513; df=5; P=0.992; Figure 18). 23

25 Figure 18: Mean number of larvae (A) and nymphs (B) parasitizing an individual rodent (y-axis). The x-axis represents the different plots. The error bars represent the ± 1SE of the mean. The grey bars represent the ticks parasitizing on a wood mouse and the white bars ticks parasitizing on a bank vole. Numbers under bars represent the number of rodents (N). Letters (Italic: wood mouse; Bold: bank vole) show the statistically different groups as calculated from the Scheffe multiple comparison test. Differences between species (for every plot) were tested using a Mann whitney-u test NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<0.001 The same analysis was done for the bank vole. Bank voles were not captured in the two KH plots, but between the other four plots the bank voles in the HV1 carried the highest number of parasitizing larvae and this number of parasitizing larvae was significantly higher than on bank voles in the AWD1 (Kruskal wallis, Chi square = ; df=3; P=0.008; Figure 18). Just as with the wood mouse, with the bank vole there was no difference in the number of 24

26 parasitizing nymphs between plots (Kruskal wallis, Chi square = 7.757; df=3; P=0.051; Figure 17). To see how large the influence of the different rodent species is on the tick population, the number of ticks parasitizing wood mice and bank voles overall was compared. This comparison showed that the wood mice are parasitized by an average of 36 ± 3.2 larvae (with a minimum of 0 and a maximum of 241 larvae). Of these 36 larvae per wood mouse 17 ± 1.6 were found on the ears of the mice, 5 ± 0.5 were found on the nose, 9 ± 1.1 were found on the neck and 5 ±0.8 were found on other body parts of the mice. Only 1.3% of the wood mice were not infected with any larvae and 87.5% was parasitized by more than 3 larvae. The bank voles are parasitized by an average of 11 ± 3.6 larvae per individual (with a minimum of 0 and a maximum of 235 larvae). Of these 11 larvae per bank vole 5 ± 1.1 were found on the ears, 1 ± 0.5 were found on the nose of the rodent, 2 ± 0.9 were found on the neck and 3 ± 1.4 were found on another body part % of the bank voles were not infected with any larvae and only 32.5% was parasitized by more than 3 larvae. In general wood mice were parasitized by more larvae than bank voles (Mann-Whitney U test, Z = ; P < 0.001; N = 243; Figure 19.A). This difference was present in all the four plots in which both species were present. For every plot the number of ticks parasitizing wood mice and bank voles was compared. In every plot the wood mouse were parasitized by significantly more larvae than the bank voles (HV1: Mann-Whitney U test, Z = ; P = 0.001; N = 61; HV2: Mann- Whitney U test, Z = ; P < 0.001; N = 50; AWD1: Mann-Whitney U test, Z = ; P = 0.001; N = 56; AWD2: Mann-Whitney U test, Z = ; P = 0.035; N = 21; Figure 18 ). Wood mice were also parasitized with significantly more nymphs (0.33 per individual with a minimum of 0 and a maximum of 12 nymphs) compared to bank voles (0.18 per individual with a minimum of 0 and a maximum of 8 nymphs) (Mann-Whitney U test, Z = ; P = 0.012; N = 243; Figure 19.B). However, when analysed per individual plot, only in one of the four plots (AWD1) in which both species occurred this difference could be shown (Mann- Whitney U test, Z = ; P = 0.014; N = 56), in the other 3 there was no significant difference (HV1: Mann-Whitney U test, Z = ; P = 0.980; N = 61; HV2: Mann-Whitney U test, Z = ; P = N = 50; AWD2: Mann-Whitney U test, Z = ; P = 0.25; N = 21; Figure 18 ). So overall there is a difference in the number of nymphs found on the two rodent species, but this difference was not found in every plot individually. 25

27 Figure 19: Mean number of larvae (A) and nymphs (B) parasitizing individual rodents per species. The y-axis represents the number of ticks, the x-axis represents the two different rodent species. The numbers under the s species name represent the number of rodents (N). The error bars represent the ± 1SE of the mean. Species are tested using a Mann-Whitney U test: NS: P >0.05; *: P < 0.05; **: P<0.01; ***: P<0.001 Since the population size of wood mouse, bank vole and common vole is different in each plot, it is useful to look at the number of ticks parasitizing rodents in general in each plot, because this combined number shows how many ticks are actually on the rodents. So by combining the different species, the number of ticks infesting the individual rodents captured in each plot can be compared. The number of larvae feeding on rodents in general is different per plot with the highest number of larvae per rodent for both areas in the HV and also a high number of larvae on the rodents in the AWD2. Again the AWD1 had a low number of larvae on the rodents. This seems to be in relation to the large bank vole population. In the KH there were also a low number of larvae on the rodents To see the importance of rodents in the development of ticks the proportions of the tick stages on the rodent and on the vegetation can be compared. If these proportions differ than it might be an indication that certain tick stages use rodents more commonly than others. No adults were found on the rodents. If we disregard the adults on the vegetation (so larvae + nymphs represents 100%), and we look at the proportion of larvae on the rodents and on the vegetation we see that the proportion of larvae is much higher on the rodents compared to the proportion of larvae on the vegetation (Table 4). The number of rodents that had a higher proportion of larvae on them than expected (with the expectation being the proportion of larvae in the vegetation) was significantly higher than the number of rodents that had a lower proportion than expected (Chi-square= 228; df=1; P<0.001). Most of the plots had an average proportion of 0.99 larvae per rodent. However the KH2 had a proportion of 0.91 this is still much higher than the proportion of larvae on the vegetation, but it is lower than the other plots. However this could also be influenced by the low number of rodents captured in the KH2. Table 4: The average proportion of larvae (larvae / (larvae + nymphs)) on the vegetation and on the rodents in every plot. The distribution of rodents captured in the different plots was analysed. The traps were checked 10 times during the fieldwork. The number of times the traps were checked and a rodent was captured was summed. This way a rough estimation of rodent activity in the different parts of the plots can be made. In most plots the distribution of the captured rodents is not homogenic over the plots. In both the KH plots most rodents are captured next to the Sea-buckthorn vegetation (Appendix C), which is for the KH1 the row on the north and for the KH2 the row on the south-west. In the HV2 plot in one subplot few rodents were captured (Appendix C) and in the subplot east of that subplot many rodents were captured. In the AWD2 (Appendix C) in the north and south subplot very few rodents were captured. The captured rodents in every subplot were compared to the collected nymphs in those subplots during the day sampling in the collecting period two. This result suggests that an increase in rodent activity does not necessarily lead to an increase in collected nymphs within a plot. For 26

28 example in the AWD2 in the two subplots with the lowest number of captured rodents the highest number of nymphs were collected. To determine if the different rodent species have a correlation with the nymphs collected from the vegetation, the number of the different rodent species in the different subplots is compared with the amount of nymphs collected in the different subplots. This is done by analysing if there is a correlation between captured rodents per subplot (like calculated in the previous paragraph) and the total number of collected ticks in that specific subplot. There was no correlation between wood mice and nymphs (Spearman r = ; P = 0.992; N = 24). There was a negative trend between bank voles and nymphs, however this was not significant (Spearman r = ; P = 0.085; N = 16). Discussion The result of the KNMI station in Vlissingen is difficult to compare with the data from a data logger, since the data loggers are closer to the ground, and usually have a higher relative humidity than the data from a weather station. For example the relative humidity in Amsterdam by a KNMI weather station (79 % ± 1.1) was about 10 % lower than the data from the data logger in the AWD. Therefore for the SD and the RH we only compare the AWD and the HV. Overall the average temperature never drops below 7 ºC, which is suggested to be a lower threshold for tick activity in Switzerland and England (Perret et al. 2000). A study in the AWD and HV by Smit and co-workers showed that ticks became active at a temperature above ± 5 ºC (Smit et al. 2004). The KH had the highest temperature and this was probably caused by the absence of trees to provide shade. We expect that this absence of trees also caused the RH of the KH to be lower than in the other areas. Over the entire fieldwork period the highest RH was found in the HV. However these numbers are calculated until the end of the field period in the HV, and during the collecting period 2 the AWD had a much higher rainfall than the HV. This high rainfall during tick collecting could influence the activity of ticks. During this study there seemed to be a fluctuation in activity of ticks during the day. During this study the nymphs and adults were more active during the evening, whereas this was not seen for larvae. A reason for the fluctuation in tick activity during the day could be that the hosts of the tick are also more active during the evening and night. Fallow deer, Roe deer, Red deer, Wild boar and Hedgehog and most rodent species are all mostly nocturnal or crepuscular in humanly disturbed areas (Macdonald et al. 1993). Another factor that could influence the tick daily activity is the microclimate (Randolph et al. 1999, Michalik et al. 2003). This microclimate is different during the day than during the night. It is known that nymphs stay close to the ground during dry conditions and when the suns sets the conditions get more humid, which could cause the nymphs and adults to become more active. However both these factors should also cause the larvae to become more active during the evening and night, and that was not the case during this study. It could be possible that the larvae were more influenced by rain during night collecting than nymphs and adults, this is supported by a study that found that questing larvae can be positively correlated by temperature and negatively correlated by relative humidity (Walker 2001). Another factor could be the collecting method, which will be discussed later in this discussion. The difference in nymphs questing during the day and during the night was the most striking in the KH, this could also be caused by the absence of trees in the KH plots. Without trees the microclimate could be especially unsuitable for ticks during the day, because with the absence of shade the microclimate could become very dry during the day. So the ticks could be forced to quest more during the night when the RH is higher and desiccation is less likely. 27

29 The activity of the larvae did increase from the tick collecting period one (beginning of May) to the collecting period two (end of May/beginning of June). However, that was not the case for the nymphs and adults. This could be explained by the adults and nymphs becoming active earlier in spring than the larvae (Randolph et al. 2002, Gassner 2005). The majority of the nymphs and adults could already be active during the collecting period one (beginning of May), and stayed active during the collecting period two. The majority of the larvae on the other hand only became active after the tick collecting period one and before the collecting period two. Therefore the larvae are also influenced by time of year during this study. This should be included during the plot comparison, since there is a factor of time of year between the plots. Hence one could expect that most larvae are active in the plots collected last (AWD1 and AWD2). However larvae can also have an optimum activity peak (Estrada-peña et al. 2005). The length and time of this peak can differ (Randolph et al. 2002), some studies found a peak of about two weeks (Randolph et al. 2002, Gassner 2005). This study also showed that within an area there can be large differences in number of ticks questing on the vegetation between plots. With the adults no significant difference was found, even tough the AWD2 had the highest number of adults and the AWD1 had the lowest number of adults of all the plots. With the nymphs some significantly different numbers of nymphs per subplot were found. The AWD2 had the highest number of nymphs questing on the vegetation, and again the AWD1 together with the KH1 had the lowest number of nymphs. The KH2 also had a low number of nymphs. Both the HV plots had a high number of nymphs. This shows that there are differences within areas in the number of nymphs questing. Both the HV plots had a higher number of larvae than all the other 4 plots. The only surprise about this is that the AWD2 had the highest number of adults and the highest number of nymphs, but only a few larvae. This could be caused by the higher amount of rainfall during the collecting period two in the AWD compared to the fieldwork periods of the HV and the KH. It is also possible that during the collecting period two in the HV the activity peak of the larvae took place. This could influence the outcome of the number of ticks collected between the different areas, and could be partially responsible for the highest number of larvae found in the HV plots. However in collecting period one when the ticks were collected in all areas in three successive days the HV also had the highest number of larvae. Therefore we expect that the HV did have the highest number of larvae, but that the difference between the number of larva in HV plots and in the other plots is probably not as large as it seems during this study. Another factor that could influence the number of collected ticks is the collecting method, since with this method the length of the vegetation influences the chance of collecting a tick (Wielinga et al. 2006). Overall it seems that by this collecting method the larvae population in the vegetation will be underestimated. So it might be better to use nymphs collected in the vegetation as a parameter for number of ticks present in the vegetation If we look at every plot separately a few comments can be made on these results. Both transects of the KH1 and the KH2 were situated in grass next to the higher vegetation. No larvae and few nymphs were collected in these two plots. Since ticks are less abundant in open areas (Lindström et al. 2003) and because they are very sensitive for desiccation (Randolph et al. 1999) the larvae are probably situated in the more abundant bush like vegetation of the Sea-buckthorn. However no ticks could be collected in this more dense vegetation because the thorns made blanket dragging impossible. So there could be more ticks present in the vegetation than can be concluded from the results. However since no trees 28

30 occurred in both the KH plots (Wolfs 2007) we expect that these plots do have low amount of ticks. Both the HV plots and the AWD2 plot had a large number of ticks in the vegetation. Both in the HV and in the AWD the plot in the pine forest (HV2 and AWD2) had more nymphs and larvae than their counterpart in the oak forest (HV1 and AWD1). However, different studies showed that pine forest are inhabited by small tick populations compared to deciduous forests because of the unfavourable microclimate (Gray 1998, Lindström et al. 2003, Michalik et al. 2003). This difference in result could be influenced by the pine forests in these studies usually having a minimal developed shrub layer (Michalik et al. 2003). Ticks are able to survive in a pine forest if the litter layer is sufficiently thick (Gray 1998) and can act as a buffer for extreme environmental factors (Michalik et al. 2003). Another study also showed a higher number of ticks in pine forest over oak forest (Smit et al. 2004). Smit and co-workers suggested that this result could be caused by the vegetation of the herb layer, their study had a herb layer dominated by grass in the pine dominated forest and a herb layer dominated by bracken in the oak dominated forest. The shrub and herb layer during this study were not less abundant in the pine forest than in the oak forest (Wolfs 2007). This result again suggests the importance of the vegetation type. Especially the difference in occurrence of nymphs between the two AWD plots was striking. The AWD2 did have a larger herb and shrub layer (Wolfs 2007). Also opposing some studies (Smit et al. 2004) the dune areas (AWD and KH) did not have higher number of ticks than the HV. The wood mouse was the most common rodent captured. The highest population of wood mouse was in the KH1. However this number should be used carefully since the traps in the KH plot were not set in the same dimension as the traps in the plots in the other areas. So instead of the, for population research, commonly used 6x8 traps plot the KH had 2x24 traps. This way there are more traps at the edge of the plot, which can lead to a higher probability of trapping rodents, especially since in the KH most rodents were trapped in the side of the plot that was close to the Sea-buckthorn. On the other hand since most rodents were captured next to the Sea-buckthorn and almost none in the row 5 metre from the vegetation, it can also be expected that if the traps were placed in the Sea-buckthorn vegetation that more rodent would be captured. So it is difficult to determine what the impact of the differently placed trap grid is. Since it is unclear what the impact of the different dimensions of the trap plot is we assume that the calculated population is a good representation of the rodent population in the plot. In both the KH1 and the KH2 no bank vole population was found. There were more wood mice in the KH1 than in the KH2. Both HV plots had large wood mouse populations. The HV1 had a larger bank vole population than the HV2. When the rodent populations in the different plots were compared one of the things that was notable is that the two largest bank vole populations were in oak forests (HV1 and AWD1). Other studies showed the opposite, they found more bank voles in pine forests (Dijk 2003, Smit et al. 2004), however they stated that this was caused by the dense undergrowth. In the study areas used in this research the undergrowth in the oak forest were not more dense than the pine forests (Wolfs 2007). So that could explain the difference in the results between both studies. Both the AWD plots did not have large wood mouse populations. The mean number of larvae per rodent was in the same range as a previous study done in the AWD and the HV (Smit et al. 2004). The rodents in both the HV plots had a high number of ticks parasitizing them. The AWD1 had the rodents with the lowest number of ticks. This was mainly caused by the high number of bank voles in that plot, because the bank voles in that plot had a low number of ticks on them, especially compared to the wood mouse. The wood mouse in the AWD1 did have a high number of ticks on them, comparable to the ticks on the wood mouse in the HV plots. Overall the bank voles were less infected with larvae than wood 29

31 mouse in every plot they occurred in. Different studies also showed that the bank vole is less infested with ticks than the Apodemus mice (Tälleklint et al. 1997, Gray et al. 1999, Humair et al. 1999, Dijk 2003, Hanincová et al. 2003, Michalik et al. 2003, Sinski et al. 2006). Some studies suggest that bank voles develop a resistance against ticks (Dizij et al. 1995, Gray et al. 1999, Hanincová et al. 2003). This is supported by Tälleklint and co-workers which found that engorged ticks detaching from bank vole weigh less than engorged ticks detaching from wood mice (Tälleklint et al. 1997). These results suggest that the bank vole is less important in the development of the ticks than the wood mouse. When the number of larvae on the vegetation is compared to the number of larvae on the rodents the HV plots have the highest number of larvae in both the vegetation as well as parasitizing the rodents. Between the HV plots the HV2 has the highest number of larvae both on the vegetation as on the rodents. The AWD2 also has a high number of larvae both on the rodents and on the vegetation. In the KH no larvae were found on the vegetation, and also a low number of larvae were found on the rodents. In the KH no bank voles occurred, and the wood mice were parasitized by low numbers of ticks. We discussed the absence of larvae in the vegetation and suggested that this could be caused by the difficulties of the collecting method. However, this low number of larvae on the wood mouse could indicate that the number of larvae questing in the vegetation of the KH plots is low. The absence of larvae in the vegetation during tick collecting is probably caused by a combination of difficulties in tick collecting method, and a low number of ticks in the vegetation and possible unsuitable weather. There seems to be a relation between larvae on the vegetation and larvae on the rodents. The three plots with the rodents with the most larvae on them (HV1, HV2 and AWD2) were also the three plots with the highest number of larvae in the vegetation. However it seems that the number of larvae in the vegetation might be underestimated with the dragging method. The number of larvae on the rodents should influence the number of nymphs in the vegetation. The top three plots in number of larvae on the rodents (HV1, HV2 and AWD2) also had the highest number of nymphs in the vegetation. Furthermore, the plot with the lowest number of larvae on the rodents also had the lowest number of nymphs in the vegetation (AWD1). There seems to be a relation between these two variables. The AWD2 had the highest number of nymphs in the vegetation. However, a lower number of larvae per rodent than the two HV plots in combination with the fact that the AWD2 also had a much smaller rodent population than the two HV plots, could indicate that in the AWD2 another animal than rodents (for example birds) might be an important host for the ticks. Other studies also found that birds could be an important vector host (Estrada-peña et al. 2005). When the number of rodents living in the area is also taken into account some interesting conclusions can be drawn. The HV plots both had a lot of wood mouse, and also a large number of ticks both in the vegetation and on the rodents. However the plot in the HV with the fewest wood mouse en the most bank voles had the lowest number of ticks (both larvae and nymphs) in the vegetation. Again this result could indicate that bank vole is less important for the development of ticks than the wood mouse. The KH plots are more difficult to explain. Especially KH1 had a large population of wood mouse, but they did not have a lot of ticks on the vegetation. This could be caused by an unsuitable microclimate caused by the absence of trees. The AWD is especially interesting. When the two plots are compared with one another we see that the AWD1 has a large bank vole population. The total rodent population is the second largest of all the plots (wood mouse and bank vole combined 52). The AWD2 has a small bank vole population, and a below average wood mouse population. 30

32 However the AWD1 had the lowest number of ticks on the vegetation and on the rodents. This again is a good indication that a larger bank vole population does not necessarily increase the number of ticks. It might even be possible that bank voles have competition with wood mice over food sources. If that is the case a large bank vole population could reduce the wood mouse population, and thereby reduce the availability of suitable hosts for ticks. Therefore it is possible that a large bank vole population could lead to fewer ticks instead of more ticks. However this is speculation and more research is needed to verify if this really is the case. The rodents had a higher proportion of larvae on them than the proportion of larvae on the vegetation. More specifically every caught rodent that was infected by ticks had a higher proportion of larvae on them than the proportion of larvae in the vegetation of the matching plot. This shows that the rodent is mainly used by larvae, and much less by nymphs. This is in agreement with other studies that suggests that nymphs generally parasitize medium sized mammals (Talleklint et al. 1994, Mejlon et al. 1997) and rarely rodents (Talleklint et al. 1994, Humair et al. 1999) especially if the conditions are humid enough (Randolph et al. 1999). This is probably caused by the fact that nymphs quest higher in the vegetation than the rodents usually occur (Mejlon et al. 1997). When the nymphs stay closer to the ground because the condition are too dry the number of nymphs on the rodents increases (Randolph et al. 1999). On the rodents the proportion of larvae relative to other tick life stages is around 0.99, whereas in the vegetation this proportion is much less. For the Borrelia to infect the rodents, nymphs are needed. However, only few nymphs parasitize rodents. If a rodent is infected with Borrelia by a single nymph, it is able to transfer the Borrelia to many larvae, since rodents can be parasitized by up to 240 larvae. Some studies suggest that transovarial transmission of Borrelia is possible (Gray 1998, Nefedova et al. 2004), However the percentage of transovarial infected larvae is probably very low (Nefedova et al. 2004). However since a single rodent can be infected with many larvae one single Borrelia infected larva can infect many other larvae trough a rodent. The rodent activity is not homogenic over a plot. Some traps caught more rodents than other traps, and this could be caused by the vegetation, since in the traps in the KH close to the Seabuckthorn more rodents were caught. This leads to believe that in the KH the rodents mostly stay in the Sea-buckthorn vegetation. In the other plots it seems that high rodent activity does not have to lead to a tick hotspot. For example in the AWD2 the two subplots with the lowest rodent activity did have the higher nymphs density. There was no correlation between wood mice and nymphs, which was partly caused by the KH in which many wood mice occurred but only few nymphs. There was a negative trend between bank voles and nymphs, however this was not significant, and the KH was not used in this correlation because no bank voles occurred in those plots. Furthermore some question marks should be placed with this test, since the subplots were used as a replication in the plot. This is not a problem with the ticks, since they do not travel large distances, however using subplots separately for mouse collecting leads to problems. Rodents do travel large distances and this causes the subplots to be influenced by other subplots, and this leads to pseudoreplication. So overall no correlation between nymphs and a rodent species could be shown. 31

33 Conclusion To answer the main research question (what is the effect of rodent populations on the composition of I. ricinus?) we first look at the secondary research questions. What are the most dominant rodent species in all areas and do they differ per area? In all the areas the wood mouse and bank vole were the most dominant species. In the AWD1 the bank vole was even the most common rodent species. The HV1 also had a large bank vole population. In the KH1 some common voles and common shrews were caught, and in the KH2 some greater white-toothed shrews were captured. In the AWD2 some common shrews were caught Are there differences in rodent populations between different locations within the same area? The HV1 had a large bank vole population, whereas the HV2 only had a small bank vole population. Both had a large wood mouse population. The bank vole was the most common rodent in the AWD1, but only a few bank voles occurred in the AWD2. Both had a relative small wood mouse population, with more wood mouse in the AWD2. The KH1 had a much larger wood mouse population than the KH2. Are all rodents of the same species parasitized equally by ticks? Is there a difference in location? There is a difference in number of larvae parasitizing rodents per plot. Among wood mice the HV plots had the highest number of larvae per wood mouse, closely followed by the wood mice in the AWD plots. The wood mice in the KH had the fewest number of larvae on them. Among bank voles the voles in the HV1 had the highest number of larvae on them. The bank voles in the HV2 and the AWD plots had fewer larvae on them. Most of these differences are trends and not significant. Are there differences in the number of ticks that parasitize rodents between rodent species? Wood mice were parasitized by significantly more larvae than bank voles. This difference was shown in each plot in which both species occurred. Overall the wood mice also had significantly more nymphs on them than the bank voles. Is there a relation between number of ticks on rodents and number of ticks on the vegetation in a location? The 3 plots with the highest number of larvae on the rodents also had the highest number of larvae and nymphs on vegetation. So there seems to be a relation. However the AWD2 had the highest number of nymphs in the vegetation but less larvae on the rodents than the two HV plots, this could be caused by another host for tick larva such as birds. It seems that more larvae on the rodents lead to more nymphs on the vegetation. Is there a relation between density of rodents and the number of ticks in the vegetation in a location? This relation is not clear. One of the plots with the highest density of rodents (KH1) did not have a lot of ticks in the vegetation. The plots with high bank vole populations had lower number of ticks in the vegetation than their counterparts with more wood mouse. So apparently other factors also influence the tick density. Is the larvae : nymph ratio comparable on the vegetation and on the rodents? The proportion of larvae on the rodents was significantly higher than the proportion of larvae in the vegetation. Overall this research suggests that the vegetation type is a very important factor influencing the tick populations. Both the KH plots are a good indication of this effect since the KH plots had a large population of rodents, but not that many ticks on the vegetation or on the rodents. 32

34 This influence of vegetation type on ticks is in accordance with previous studies (Gray 1998, Lindström et al. 2003, Michalik et al. 2003). Pine forests seem to contain more ticks than oak forests, which is different from most other studies (Gray 1998, Lindström et al. 2003, Michalik et al. 2003). However in this study the bank vole populations were largest in oak forests. Also the dune areas do no necessarily have more ticks than non-dune areas. This opposes the general hypothesis that dune areas contain more ticks (Smit et al. 2004). However since the variation within an area can be large it might be difficult to make a comparison between areas. The rodent populations can also influence the tick populations. However in this study the plots with high rodent populations did not necessarily have large number of ticks in the population. Furthermore, the AWD2 had the highest number of nymphs in the vegetation, but the population of rodents was small compared to other plots. Furthermore the number of larvae on the rodents was smaller than on the rodents in the HV plots. Other animals such as birds can also increase the tick populations (Estrada-peña et al. 2005). The rodents are important for the development of larvae to nymphs. The wood mouse seems to be more important for tick development than the bank vole, or at least more favourable as a blood host, which is supported by the study from Tälleklint and co-workers which found that engorged ticks detaching from wood mice weigh more than engorged ticks detaching from bank voles (Tälleklint et al. 1997). A large bank vole population does not seem to have a positive effect on tick populations. So overall the vegetation type seems more important than rodent populations for tick occurrence. If the vegetation is not suitable, few ticks will inhabit the area. If the vegetation is suitable than a large wood mouse population is most favourable for tick populations. Future research The most important aspect for future research is analysing the blood of the captured rodents, and the ticks collected from the rodents. Especially the comparison of the Borrelia infections between wood mouse and bank vole is interesting. The wood mouse carries more ticks than the bank vole, so you should expect that they also have a higher infection rate (Hanincová et al. 2003, Smit et al. 2004, Khanakah et al. 2006). However some literature suggest that mice are more suited to minimize the bacterial infections, and voles concentrate more on limiting the vector (Humair et al. 1999, Hanincová et al. 2003). Comparing the infection rates of the rodents and the ticks in the vegetation should also give a good idea of how large the influence of rodents in Borrelia transmission is. Some studies suggest that bank vole is an efficient host for the spread of B. burgdorferi s.l. (Michalik et al. 2003). If the infection percentage of bank vole is comparable to the infection percentage of wood mouse than this suggestion is supported since wood mice are parasitized by more ticks. The nymphs collected in the AWD2 should be analysed for the different Borrelia genospecies, we suggest in this study that birds might be an important host for the tick, and therefore we expect more bird associated Borrelia species (B. garinii and B. valaisiana) in the AWD2 nymphs compared to the HV1 and the HV2 ticks. New studies could be done in more different areas. This study showed that bank vole are more common in oak forest, it can be useful to make the comparison of oak forest and pine forest in other areas. This study suggested that in oak forest fewer ticks are present than in pine forests and more bank vole are active in oak forest than in pine forests. With more studies in different areas with these two types of forests these conclusions can be supported. 33

35 Improvements for this study If this research is done again some points could be improved. Tick collecting could be done every two weeks in every area. This way more data points for ticks in the vegetation are collected, which makes it easier to have statistically significant differences. The vegetation of the dragging samples could also be recorded differently. 25 metre per subplot is dragged. Per 25 metre the type of vegetation that is dragged should be recorded. So every plant that is touched by the blanket should be recorded as well as other differences between dragging strips such as tracks of vertebrates, and the percentage per plant per transect should be calculated. This way it is easier to determine if certain plant species/grow styles contain more ticks than others, and if there are tick hotspots in the vegetation. If enough people are available the rodent trapping in the different areas could be done at the same time, to decrease the influence of time. However logistically this might be difficult to arrange. Acknowledgements I would like to thank my supervisors Willem Takken and Leo van Overbeek for their valuable comments on the work plan and report. I would also like to thank the managers of the three areas, Leo van Breukelen (Amsterdamse Waterleiding Duinen), Jacob Leidekker (Nationaal Park De Hoge Veluwe) and Natuurmonumenten specifically Han Meerman (Kwade Hoek) Furthermore I would like to thank several people for their help in the practical part of this study. Marianne Wolfs for her help during rodent capturing. Fedor Gassner for his advice on the proposal and practical tips for the field study, Maarten Holdinga for his help with the blood collecting and Peter Kastelein, Francoise Kaminker and Frans Jacobs for their advice on tick collecting. I would also like to thank Sip van Wieren for his comments on the field study and the people of CKP for their help with the preparation of the blood collecting. 34

36 References Christova I. I., van de Pol J., Yazar S., Velo E. and Schouls L. (2003) Identification of Borrelia burgdorferi sensu lato, Anaplasma and Ehrlichia species, and spotted fever group Rickettsiae in ticks from southeastern Europe. European journal of clinical microbiology & infectious diseases 22 (9), de Boer R., Nohlmans M. K. E. and van den Bogaard A. E. J. M. (1990) Dutch ticks as transmitters of infectious diseases Nederlandse teken als overbrengers van infectieziekten. Nederlands tijdschrift voor geneeskunde 134 (27), de Mik E. L., van Pelt W., Docters-van Leeuwen B., van der Veen A., Schellekens J. F. P. and Borgdorff M. W. (1997) The geographical distribution of tick bites and erythema migrans in general practice in the Netherlands. International journal of epidemiology 26 (2), den Boon S. and van Pelt W. (2003) Verdubbeling consulten voor tekenbeten en ziekte van Lyme. Infectieziekten bulletin 14 (5), Dijk F. (2003) Rodents as host for ticks. Thesis at Nature Conservation and Plant Ecology Group, Wageningen University and Research centre. Dizij A. and Kurtenbach K. (1995) Clethrionomys glareolus, but not Apodemus flavicollis acquires resistance to Ixodes ricinus L., the main European vector of Borrelia burgdorferi. Parasite immunology 17 (4), Estrada-peña A., Osácar J. J., Pichon B. and Gray J. S. (2005) Hosts and pathogen detection for immature stages of Ixodes ricinus (Acari: Ixodidae) in North-Central Spain. Experimental & applied acarology 37 (3-4), Gassner F. (2005) Does grazing affect Lyme disease risk? The impast of grazign livestock on tick populations and Borrelia infections in a natural woodland area in the Netherlands. Wageningen UR, Laboratory of Entomology, MSc-thesis No Gray J. S. (1998) The ecology of ticks transmitting Lyme borreliosis. Experimental & applied acarology 22 (5), Gray J. S., Kirstein F., Robertson J. N., Stein J. and Kahl O. (1999) Borrelia burgdorferi sensu lato in Ixodes ricinus ticks and rodents in a recreational park in south-western Ireland. Experimental & applied acarology 23 (9), Hanincová K., Schäfer S. M., Etti S., Sewell H.-S., Taragelová V., Ziak D., Labuda M. and Kurtenbach K. (2003) Association of Borrelia afzelii with rodents in Europe. Parasitology 126 (1), Humair P. F., Rais O. and Gern L. (1999) Transmission of Borrelia afzelii from Apodemus mice and Clethrionomys voles to Ixodes ricinus ticks: Differential transmission pattern and overwintering maintenance. Parasitology 118 (1), Jouda F., Perret J.-L. and Gern L. (2004) Density of questing Ixodes ricinus nymphs and adults infected by Borrelia burgdorferi sensu lato in Switzerland: Spatio-temporal pattern at a regional scale. Vector borne and zoonotic diseases 4 (1), Khanakah G., Kocianová E., Vyrosteková V., Řeháček J., Kundi M. and Stanek G. (2006) Seasonal variations in detecting Borrelia burgdorferi sensu lato in rodents from north eastern Austria. Wiener klinische Wochenschrift 118 (23-24), Kurtenbach K., de Michelis S., Etti S., Schäfer S. M., Sewell H.-S., Brade V. and Kraiczy P. (2002) Host association of Borrelia burgdorferi sensu lato - The key role of host complement. Trends in microbiology 10 (2), Kurtenbach K., Dizij A., Seitz H. S., Margos G., Moter S. E., Kramer M. D., Wallich R., Schaible U. E. and Simon M. M. (1994) Differential immune responses to Borrelia burgdorferi in European wild rodent species influence spirochete transmission to Ixodes ricinus L. (Acari: Ixodidae). Infection and immunity 62 (12),

37 Lange R., Winden A., van Twisk P., de Laender J. and Speer C. (1986) Zoogdieren van de Benelux. Herkenning en onderzoek Le Fleche A., Postic D., Girardet K., Peter O. and Baranton G. (1997) Characterization of Borrelia lusitaniae sp. nov. by 16S ribosomal DNA sequence analysis. International journal of systematic bacteriology 47 (4), Lindström A. and Jaenson T. G. T. (2003) Distribution of the common tick, Ixodes ricinus (Acari: Ixodidae), in different vegetation types in southern Sweden. Journal of medical entomology 40 (4), Macdonald D. W. and Barret P., Collins field guide Mammal of Brittain and Europe. (Harpercollins, London, 1993). Mejlon H. A. and Jaenson T. G. T. (1997) Questing behaviour of Ixodes ricinus ticks (Acari: Ixodidae). Experimental & applied acarology 21 (12), Michalik J., Hofman T., Buczek A., Skoracki M. and Sikora B. (2003) Borrelia burgdorferi s.l. in Ixodes ricinus (Acari: Ixodidae) Ticks Collected from Vegetation and Small Rodents in Recreational Areas of the City of Poznan. Journal of medical entomology 40 (5), Nefedova V. V., Korenberg E. I., Gorelova N. B. and Kovalevskii Y. V. (2004) Studies on the transovarial transmission of Borrelia burgdorferi sensu lato in the taiga tick Ixodes persulcatus. Folia parasitologica 51 (1), Pecchioli E., Hauffea H. C., Tagliapietraa V., Bandib C., Genchib C. and Rizzoli A. (2007) Genospecies of Borrelia burgdorferi sensu lato in Ixodes ricinus ticks from the Autonomous Province of Trento, Italy. International journal of medical microbiology 297 (1), Perret J. L., Guigoz E., Rais O. and Gern L. (2000) Influence of saturation deficit and temperature on Ixodes ricinus tick questing activity in a Lyme borreliosis-endemic area (Switzerland). Parasitology research 86 (7), Pichon B., Kahl O., Hammer B. and Gray J. S. (2006) Pathogens and host DNA in ixodes ricinus nymphal ticks from a German forest. Vector borne and zoonotic diseases 6 (4), Pichon B., Mousson L., Figureau C., Rodhain F. and Perez-Eid C. (1999) Density of deer in relation to the prevalence of Borrelia burgdorferi s.l. in Ixodes ricinus nymphs in Rambouillet forest, France. Experimental & applied acarology 23 (3), Randolph S. E., Green R. M., Hoodless A. N. and Peacey M. F. (2002) An empirical quantitative framework for the seasonal population dynamics of the tick Ixodes ricinus. International journal for parasitology 32 (8), Randolph S. E. and Storey K. (1999) Impact of microclimate on immature tick-rodent host interactions (Acari: Ixodidae): Implications for parasite transmission. Journal of medical entomology 36 (6), Rauter C. and Hartung T. (2005) Prevalence of Borrelia burgdorferi sensu lato genospecies in Ixodes ricinus ticks in Europe: A metaanalysis. Applied and environmental microbiology 71 (11), Sinski E., Pawelczyk A., Bajer A. and Behnke J. M. (2006) Abundance of wild rodents, ticks and environmental risk of Lyme borreliosis: A longitudinal study in an area of Mazury Lakes district of Poland. Annals of agricultural and environmental medicine 13 (2), Smit R. and Takken W. (2004) Gevolgen voor de transmissiedynamiek van de ziekte van Lyme. Wageningen university and research center (Report). Talleklint L. and Jaenson T. G. T. (1994) Transmission of Borrelia burgdorferi s.l. from mammal reservoirs to the primary vector of lyme borreliosis, Ixodes ricinus (Acari: Ixodidae), in Sweden. Journal of Medical Entomology 31 (6),

38 Tälleklint L. and Jaenson T. G. T. (1997) Infestation of mammals by Ixodes ricinus ticks (Acari: Ixodidae) in south-central Sweden. Experimental & applied acarology 21 (12), Walker A. R. (2001) Age structure of a population of Ixodes ricinus (Acari: Ixodidae) in relation to its seasonal questing. Bulletin of entomological research 91 (1), Wielenga S. (2006) The effect of cattle presence and roe deer density on the number of questing ticks in oak forest, pine forest and a meadow. Wageningen UR, Resource Ecology Group, MSc-thesis No. H Wielinga P. R., Gaasenbeek C., Fonville M., de Boer A., de Vries A., Dimmers W., Akkerhuis Op Jagers G., Schouls L. M., Borgsteede F. and van der Giessen J. W. B. (2006) Longitudinal analysis of tick densities and Borrelia, Anaplasma, and Ehrlichia infections of Ixodes ricinus ticks in different habitat areas in the Netherlands. Applied and environmental microbiology 72 (12), Wolfs M. (2007) The effect of vegetation on small rodent and tick abundance. Wageningen UR, Resource ecology Group, Msc thesis 37

39 Appendix A: photographs of the plots. Photographs taken from the corner of the different plots, taken around the fieldwork period (KH pictures were taken 1.5 month after fieldwork period). 38

40 Appendix B: Weighted mean method Example of a calculation scheme in accordance with the weighted mean methods (Lange et al. 1986) in the AWD1 for wood mouse (A) and bank vole (B) A Population size: 230/(27+1)= B Population size: 3100/(70+1)=

41 Appendix C: activity of the rodents in each subplot Number of times a rodent was captured in the different traps in the different plots. Every number represents a trap. The surrounding of every trap is coloured according to the number of times a rodent was captured in each plot: 0-3 = white; 4-6 = light grey; 7-9 = medium grey; 10 = dark grey. The bold numbers in the circle represent the number of larvae collected in the day sample of collecting period two. Each number represents 25 metre dragged, which is half a subplot. The north is shown by the compass. 40