Beaver territories and waterfowl breeding in beaver dams in Klosterheden State Forest, Denmark

by Aurélie Charrin

directed by Sten Asbirk & Thomas Borup Svendsen

2004 BEAVER TERRITORIES IN KLOSTERHEDEN STATE FOREST AND SURRONDINGS ...... 4

A. INTRODUCTION...... 4 1. PRESENTATION OF THE SPECIES : EUROPEAN BEAVER, CASTOR FIBER LINNAEUS 1758...... 4 1.1. Taxonomy ...... 4 1.2. Body description and characteristics...... 4 1.3. Behaviour and ecology...... 5 2. BEAVERS IN DENMARK ...... 8 B. BEAVER STUDY...... 9 1. IN KLOSTERHEDEN FOREST DISTRICT ...... 9 1.1. Definition...... 9 1.2. Møllesøen/Flynder Å...... 9 1.3. Ællebæk/Hestbæk...... 11 1.4. Discussion on the methodology...... 12 2. BEAVER POPULATION ESTIMATIONS ...... 13 2.1. Introduction...... 13 2.2. Material and methods...... 13 2.3. Results ...... 14 2.4. Discussion ...... 15 3. OUTSIDE KLOSTERHEDEN STATE FOREST ...... 20 3.1. Investigation along Storå river ...... 20 3.2. Search for the lodge in Flobæk ...... 21

MONITORING OF WATER BIRDS IN THE BEAVER PONDS ...... 22

A. INTRODUCTION...... 22

B. MATERIAL AND METHODS ...... 22 1. STUDY AREA ...... 22 Description of the vegetation ...... 22 2. OBSERVATIONS ...... 25 3. GRAPHICAL METHODOLOGY...... 25 C. RESULTS...... 26 1. RESULTS BY SPECIES...... 26 1.1. Mallard (Anas Platyrhychos)...... 26 1.2. Muet swan (Cygnus olor) ...... 26 1.3. Moorhen (Gallinula chloropus) ...... 27 1.4. Pied wagtail (Motacilla alba) ...... 27 1.5. Grey wagtail (Motacilla cinera) ...... 27 1.6. Little grebe (Tachybaptus ruficollis)...... 28 1.7. Grey heron (Ardea cinerea) ...... 28 1.8. Cormorant(Phalacrocorax carbo) ...... 28 1.9. Kingfisher (Alcedo atthis) ...... 28 1.10. Teal (Anas crecca) ...... 28 1.11. Tufted duck (Aythya fuligula)...... 28 1.12. Coot (Fulica atra) ...... 29 1.13. Sandpiper ...... 29 1.14. Egyptian goose (Alopochen aegyptiacus) ...... 29 1.15. Water rail (Rallus aquaticus)...... 29 1.16. Black stork (Ciconia nigra)...... 29 1.17. Shelduck (Tadorna tadorna) ...... 29 1.18. Black headed Gull (Larus ridibundus) and other gull (Larus sp.)...... 29 1.19. Unindentificated duck ...... 29 2. RESULTS BY TERRITORY...... 29 D. DISCUSSION...... 30 1. RECORDS OF THE SPECIES...... 30 1.1. Mallard...... 30 1.2. Muet swan ...... 31 1.3. Moorhen ...... 31 1.4. Little grebe ...... 31 1.5. Grey heron...... 32 1.6. Wagtail ...... 32 1.7. Cormorant ...... 32 1.8. Sandpiper ...... 32 1.9. Unindenticated duck...... 33 2. RESULTS BY TERRITORY...... 33 3. BEAVERS’INFLUENCE...... 34 E. PASSERINES OBSERVATIONS...... 35 1. RING-MARKING IN FRUERBÆK ...... 35 2. OTHER OBSERVED SPECIES IN KLOSTERHEDEN ...... 35 3. OTHER OBSERVATIONS...... 37

ACKNOWLEDGEMENTS...... 38

BIBLIOGRAPHY ...... 39 Beaver territories in Klosterheden State forest and surrondings

A. Introduction

1. Presentation of the species : European beaver, Castor fiber Linnaeus 1758.

1.1. Taxonomy Beavers are semiaquatic mammals, they belong to the order Rotentia, the family Castoridae. They are divided in two species : Castor fiber Linnaeus 1758, the European beaver and Castor canadensis Kull 1820, the north-American beaver. North-American beavers have got 40 pairs of chromosomes (following Robertsonian fusion of eight pairs of chromosomes) while European beavers have got 48 (Lavrov & Orlow, 1973 in Rosell & Sun, 1999 and in Hartman, 1994). Rosell & Sun (1999) found that colour and viscosity of the anal gland secretion (AGS) could also be used for discrimination between the two species. Nevertheless, these two species are similar both morphologically and behaviourally. However the North-American beaver has a greater ecological flexibility, survives better under severe climatic conditions, adapts more easily to substitute foods, regulates hydrological regimes of water body better and is more fecund than the Eurasian beaver (Danilov & Kan’shiev, 1983 and Hill, 1982 both in Rosell & Sun, 1999).

Castor fiber may also be divided in eight sub-species according to relict populations and morphometric characteristics.

1.2. Body description and characteristics Weight: 18-20 kg for 75-100 cm length. Short plump legs support the body. Its hind and front feet have 5 digits with claws. The hind feet are webbed and the 4th claws form a comb, used for grooming (Richard, 1980; Bau, 2001; Laanetu, 2001). On land, beavers walk on all four legs but can also walk on their hind feet in a upright position with the tail stabilising the posture (Carlson & Welker, 1976 in Bau, 2001; Richard, 1980). They usually don’t walk very far from the shore.

As a semiaquatic animal, beavers have special feature; the most remarkable being their large tail covered by scalelike skin with small bristle hair (Carlon & Welker, 1979 in Bau, 2001; Richard, 1980). The tail is used as driving power, manoeuvring, diving while swimming (Bau, 2001). It also assumes a function in regulation of body temperature, body energy reserve and defense against enemies (“tail slapping”) (Richard, 1980; Bau, 2001; Laanetu, 2001). The fur is thick, formed by two layers which function as isolation when in water (Richard, 1980; Bau, 2001; Laanetu, 2001). The nostrils are closeable as well as the opening of the ears; and eyes are covered by transparent eyelids when beaver dives (Richard, 1980; Bau, 2001; Laanetu,2001). Beavers can also shift the position of the epiglottis in order to prevent water going into breathing ducts during chewing and transportation of timber under water (Richard, 1980; Bau, 2001; Laanetu,2001).

Bearver territories in Klosterheden State Forest, Denmark 4 Diving time is 2-5 minutes normally but can extend up to 15 minutes (Rouland & al., 1984; Richard, 1980; Bau, 2001).

Beaver dental formula is 1-0-1-3 / 1-0-1-3 (Rouland & al., 1984). As a rodent, the incisors are constantly growing. The outside surface is covered by hard orange enamel.

Beavers present a pseudocloack where the urethra, the anus, the reproductive organs and two pair of glands end up into. It is situated underneath the base of the tail. Castor glands produce the castoreum. Mixed with urine, it is used to scent-mark on mounds of mud or plant material. Secretion of anal glands serve on waterproofing the fur and also for scent-marking (Richard, 1980).

No external criteria could be use to identify male or female. However males have a bacculum (bone) which could be felt by touching the reproductive organ inside the pseudocloack or seen by radiography (Richard, 1980; Rouland & al., 1984). Sucked females are more easily recognizable, as teats could be seen passing from the fur (Richard, 1980).

Average longevity is 7-8 years but beavers can live until around twenty years.

1.3. Behaviour and ecology Beavers have an active period during the night but in quiet places, they can be active during daytime (Richard, 1980; Bau, 2001; J-C Gaudin, pers.comm.). Doboszyńska & Żurowski (1983) reported that new born young are active in the lodge during daytime as their mother is with them.

1.3.1. Reproduction and social group Beavers are monogamous and live in family with an average of 2-7 members concerning European beavers : an adult reproducing pair, young from last years and young of the years.

Sexual maturity is reached at the age of 1½ -2 years. Beavers are polyoestrus from 1 to 5 oestrus during one single mating season (Richard, 1980). The heat lasted from 12-24 hours. Mating usually occurs in January or February in the water (Richard, 1980, Doboszyńska & Żurowski, 1983). However mating could be observed in December, October or November (Richard, 1980). Pregnancy lasts for 105-107 days and young beavers bear between mid-April to mid-August (Doboszyńska & Żurowski, 1983). Lactation in beavers lasts for 90 days but they begin to eat plant material as soon as they are 10 days old. The kids remain in the lodge the first 3-6 weeks (Richard, 1980, Lancia & Hodgdon, 1983 in Bau, 2001). All the family members take care of the new borns (Patenaude, 1983). Dispersal of the young usually occurs at two years old and in spring. However, dispersal could be delayed (e.g. Svendsen, 1989).

As an adult, beavers are totally under water when swimming, and only the head is up. Rosell & Pedersen (1999) propose to evaluate age class according to the shape of the beaver body in water.

Bearver territories in Klosterheden State Forest, Denmark 5 1.3.2. Foraging Beavers are herbivorous animals. The main source of food is deciduous trees: bark, leaves, shoots, buds and twigs. They also forage on herbaceous plants. Beaver diet has been studied by many authors (Lapiński & Staliński, 2001; Borglykke, 2002; Svendsen,1980 and Belovsky, 1984 in Borglykke, 2002). Populous (Populus sp.), willow (Salix sp.) and birch (Betula sp.) are preferred.

1.3.3. Territory and building activities Beaver territories are highly differentiated.

a. The shelters The shelters are ground burrows or lodges but some intermediaries between these two states could be found (Rouland & al., 1984). All have immersed entry from where an oblique gallery opens out onto at least one room where the layer is. This room is often ventilated by an open whole in the ground (Rouland & al., 1984). - Ground burrows are dug by beavers where the banks are high and the soil soft enough (Rouland, 1984; Bau, 2001). They may be overbuilt with branches especially when the roof is collapsed (Rouland & al., 1984). - Lodges are built where the cliff shaped banks are far away from the wateredge as well as when beavers can’t dig in the bank (Rouland & al., 1984; Bau, 2001). The lodges are built by accumulation of wood branches from feeding rest. In springtime and summer time, beavers can use nests. These are depressions in the ground located under bushes, a felled tree or tree roots (Bau, 2001).

Beaver colonies can use more than one lodge, especially in summer time (Richard, 1980, Bau, 2001). A family inhabiting the same territory for a long time likely has several shelters (Richard, 1980).

b. The dams Dams are built with branches, grass and mud especially in stream habitats. They are used to maintain a high water level around the lodge for keeping the entrance under water (and therefore protected from predators), making easier swimming and transport of building material (Rouland, 1984; Rosell & Parker, 1996 in Bau, 2001). A waterlevel of at least 50 cm is necessary to provide a quick route for escape (Macdonald & al., 1995 in Bau, 2001). Beavers can also build dams to increase the waterlevel of streams and channels up to feeding sites (e.g. Flynder Å; pers.obs).

c. The refectories and the feeding site Refectories are places used by beavers to eat bark from the branches they have just cut. These sites are located near the wateredge and the branches are accumulated on the bank or under the water (pers.obs; Bau, 2001, Richard, 1980). The feeding sites are more scattered in the territory and easily recognizable: cut trees are shaped as a pencil tip and cuts have got a characteristic oblong section (Fig.1). The different feeding sites are used continuously along the year according to season changes in the availability of food (e.g. bark, leaves, shoots, buds and twigs).

Bearver territories in Klosterheden State Forest, Denmark 6 d. The food caches They are a collection of branches and twigs non-barked located usually under water and sometimes on the wateredge. These stocks are used during winter. e. The passage way tracks As beavers use the same way to go to the different feeding sites, they leave in the ground or/and grass some tracks. These tracks are also found where beavers cross roads to go from one water area to another (e.g. in Ællebæk, in Risbæk territories; pers.obs.).

Channels are dug in order to link to water area separated by a little way of ground, go across river meandering and make easier swimming in swamp area or accessing feeding sites far-off (Richard, 1980; Rouland & al., 1984).

1.3.4. Territory and scent-mark European and North-American beavers are highly territorial and use scent-marking as a means to territory defense and delineating family territories (Rosell & Nolet, 1997, Rosell, 1994, Müller-Schwarze & Heckman, 1980; Rosell & Bergan, 2000, Nolet & Rosell, 1994) Scent-marks are mounds of debris material and mud or sand where beavers deposit castoreum and/or AGS (Svendsen, 1980, Müller-Schwarze & Heckman, 1980, Rosell & Bergan, 1998). Both sexes and all age classes (except kids under 5 months old) participate in scent marking (Butler & Butler, 1979). However few differences exist as males scent-mark more than females in summertime (Thomsen, 2001).

Scent-marking activity varies with seasons (Rosell & Bergan, 2000; Rosell & al., 1998; Svendsen, 1980) with peaks during spring when dispersal of the 2-year-olds may occur (Svendsen, 1980; Rosell & al., 1998). Rosell & Bergan (2000) found that the median number of scent-marking is higher when near and during the oestrous than the non-breeding portion of winter. Rosell & Nolet (1997) didn’t observe a significant difference in the number of scent-markings during their study.

Beaver families with close neighbours scent mark more often than isolated families (Rosell, 1994; Müller-Schwarze & Heckman, 1980, Rosell & Nolet, 1997). Scent-markings may concentrate near the territory limits (Rosell & Nolet, 1997; Rosell & al., 1998; Rosell, 1994). Some marks appear to be maintained frequently whereas some others don’t (Rosell, 1994).

Beaver respond to intruder scent-markings with aggressive behaviour (Rosell & al., 2000, Müller-Schwarze & Heckman, 1980; Müller-Schwarze & al., 1983), with however some difference in observed behaviour (Butler & Butler, 1979; Rosell & al., 2000). After the response, the beavers were often observed starting to patrol the territory (Rosell & al., 2000).

Nolet & Rosell (1994) observed that beavers swim faster through a foreign territory than in their own territory as well as they don’t spend time on river banks. Transient beavers are often 2-years old, but they could be males temporarily leaving their family lodge or adult females without young (Townsend, 1953 in Müller-Schwarze & Heckman, 1980).

Bearver territories in Klosterheden State Forest, Denmark 7 Bau (2001) described scent-marking behaviour in Castor fiber as including pawing with the front feet, placing the cloacae above the pile scraped together and depositing scent. Stamping the hind legs on the pile for a couple of times could follow this scent-marking sequence (Bau, 2001). Scent-marking behaviour lasted between 6 and 18 seconds (Bau, 2001).

2. Beavers in Denmark Beavers used to live in Denmark but disappeared from the country for more than two thousand years (Bau; 2001). In October 1999, 18 Eurasian beaver (Castor fiber) - 12 adults, 4 juveniles and 2 kids - from Germany were reintroduced in Klosterheden State Forest District (north-western part of Denmark) in two successive releases (Asbirk, 2001; Bau, 2001). The reintroduction area covers the upstream parts of the Flynder Å stream catchment area inside the Klosterheden State Forest District (KLS). Beavers were released at six lakes where artificial lodges had been previously built (Asbirk, 2001; Bau, 2001). Rather quickly, the beavers have left four of the release sites and spread out in KLS and surroundings (Asbirk, 2001; Bau, 2001; Elmeros & al., 2004). They have been breeding successfully and the first young had been observed in August 2000.

The beavers have actually colonized the most part of the Flynder Å stream catchment area: five territories are in KLS and eight on private owner lands outside KLS (Fig.2). The beavers have settled in lake habitat as well as stream habitat. Beaver tracks are also recorded in the river Storå (Elmeros & al., 2004; see also part B.4.2). According to the last counting (April, 29th and 30th & May, 1st 2004), the population is estimated at 52 beavers and 9 young are observed for 2003 (O.Olsen, pers. comm.).

In addition to an extensive monitoring program coordinated by the forest staff for the reintroduction, two deeper studies were made in order to increase the knowledge of the novel beaver population. Bau (2001) studied beaver behaviour during the settling phase using infrared video recording equipment and beaver tracks recording on field. Borglykke (2002) studied beaver diet using feces analyses.

Bearver territories in Klosterheden State Forest, Denmark 8 B. Beaver study

The aim of this work is to clarify some knowledge on the actual beaver territories and bring out some more accurate information.

1. In Klosterheden forest district

1.1. Definition A Territory is delimited between two tracks assessing a continuous beaver presence : scent- mark, lodge and refectory in activity, dams. Natural limits as a main road were also used. A freshly marked scent mound with a scent detectable by a human nose at 2 cm or more, was termed scent mark. This also included marks directly on the ground or on tussocks (Rosell & al., 1998).

In this part, Møllesøen, Ællebæk, Flynder Å and Hestbæk refer to a beaver activity area unless when stated (Fig.3).

1.2. Møllesøen/Flynder Å Møllesøen lake is the one chosen for reintroduction and therefore one of the oldest beaver territories of Klosterheden plantation. At present, the beaver territory stretches from upstream Abogård to Øvej. Crossed by Flynder Å River, it is roughly composed of 3 parts: - Reed mace swamp upstream Møllesøen; - Møllesøen lake boarded by coniferous forest; - Grass meadows with willow area downstream Møllesøen lake = Flynder Å. With the artificial lodge built for reintroduction, two other lodges was found in this territory: one on the third island of Møllesøen ; one downstream at the intersection of Flynder Å river and a channel from the lake (Fig.3). According to observations made during the annual count, the forest staff discusses the possibility of having two beaver families and therefore two territories instead of one. Boundaries would be roughly near the Gl. Landevej.

1.2.1. Material and methods Territories were drawn during daytime by walking along the riverside of all territories waterways and recording all signs of activity on a map. Cartography of Møllesøen territory was made in late April. However, scent-marks were recorded around once a month for the part between Gl. Landevej and Øvej and for the lake once again in August. As the waterfowl monitoring (see part 2) took place on the easthern riverbank of the lake, the scent-mark found there was summed up on the map (Fig.4).

Boundaries were studied in relation with scent-mark as beavers clumps scent-mark near territorial borders (Rosell & al., 1998) for territories with close neighbours.

Observation point was carried out opposite to the first fresh supposed scent-mark i.e the supposed boundaries (Point A & B - Fig.3).

Bearver territories in Klosterheden State Forest, Denmark 9 From mid-May to mid-June, observations were made from sunset and ended either when bad weather conditions caused to be very low or when the equipment was insufficient. No camouflage equipment was used. One observer carried out all observations. Military night vision glasses were used (Photonic type NS-B 3.3-80) as well as infrared light . However the infrared light system was abandoned, as it was inefficient.

No beaver was marked for individual recognition (Bau, 2001). Therefore beaver families were determined according to their arrival way. Age class, duration, behaviour and swimming speed were also recorded.

1.2.2. Results No beaver was observed at the point A during all the observation time: 53h15 of observation enclosed 3 entire nights. However a possible tail slapping was heard once around 1 o’clock downstream in the Flynder Å. No beaver was observed at the point B during all the observation time : 30h of observation enclosed 3 entire nights plus 2h made by a other observator.

1.2.3. Discussion Brady & Svendsen (1981) noted during their study on American beaver that the male patrol the entire perimetre of the pond at the beginning of each activity period. If distrubed, it would return to the lodge (for up to one half hour). When it would reemerge it would carry on the parol of the pond from the disturbance site. I did not observe this behaviour during my watching. During the entire night, it blew sometimes and this may have brought the odour to the beavers. Furthermore, I may have been too noisy as the shelter was moving with the wind. However there has be some watching without wind and without shelter. I did not observe a beaver neither. The observation point was chosen as beeing a possible boundaries. We expect beaver to scent- mark each night on their territory boundaries according to an abundant litterature. Therefore, we assume that they scent-mark in all the scent-mark place of the boundary. They didn’t. They seem to scent-mark once in a couple of days at a given place. As the scent-mark making is very quick (from 6 to 18 second (Bau, 2001)), it is lucky to see the beaver scent-marking in a given point on the boundary. This, indeed don’t exclude that beaver do spend a lot of time on their boundaries. If they do patrol the territory each night, and if Flynder Å and Møllesøen had been one territory, I must have seen a beaver pass before the point A. I have not. Therefore, Flynder Å and Møllesøen must be two different territories. The scent-mark pattern shows that the place chosen for the watching have got a permanent scent-mark except for August when the water has flotted the place. However they managed to scent at the other place downstream (Fig.4). In Øvej, they uphold the scent-mark during all the study period. When the water level increased in August they managed to scent-mark little heigher with sand or grass as the riverbank lend itself to. For the same reason than above, Fynder Å and Ællebæk must be two different territories. To conclude, two family inhabit there: Møllesøen territory is reduced to the lake and Fynder Å territory stretches from Gl.Landvej to Øvej (Fig.5).

Bearver territories in Klosterheden State Forest, Denmark 10 1.3. Ællebæk/Hestbæk Hestbæk territory stretches from Øvej to upstream the fishery and covers both Ællebæk and Hestbæk. Only one lodge has been recorded in this area (Fig.3). According to observations made during the annual counts, the forest staff discusses the possibility of having two beaver families and therefore two territories instead of one i.e Ællebæk and Hestbæk. Boundaries would be on the river Flynder Å.

1.3.1. Material and methods Territories were drawn during daytime by walking along the riverside of all territories waterways and recording all signs of activity on a map. Cartography of Ællebæk was made in late April. Searches of lodges in Ællebæk were made in April and July. Area between Gl.Landvej and Jagtvej were prospected in June and between Jagtvej and Tårnvej in July by following the streams and the bank of the lake. Scent-marks along the river Flynder Å were recorded in April-May, July and August for the part downstream Øvej. Boundaries were studied in relation with scent-marks as beavers clumps scent-mark near territorial borders (Rosell & al., 1998) for territories with close neighbours.

Observation point was carried out in different parts of the area (Fig.3). No camouflage equipment was used. Military night vision glasses were used (Photonic type NS-B 3.3-80).

1.3.2. Results Until June 20th at least seven beavers inhabited this territory with at least 1 young. According to the observation made at the point C (enclosing two entire nights), some beavers sleep and spend daytime in Ællebæk but no resting place has been found yet. However some beavers sleep and spend daytime somewhere else. Some beavers were seen crossing the road from Ællebæk to Flynder Å river late in the morning (around 6 o’clock) and didn’t come back (at least in the observation time). During Summer, the first beaver has been seen more often crossing the road from Ællebæk to the river Flynder Å. A collapsed burrow was found along the Flynder Å just downstream the point D. Observations made in point E confirm that the beavers use the lodge in Hestbæk. Beavers were observed alone, swimming upstream to me when I was watching from point D. They could have stopped and disappeared from view. Sometimes, I heard some eating noise. I found eating tracks in site 1 while a dam flotted the area behind. When beavers passed before me, they disappeared from view.

1.3.3. Discussion As said above, a boundaries is assessed on Øvej between FlynderÅ and Ællebæk/Hestbæk. According to the data, Ællebæk could not be discrimated from Hestbæk as a territory for itself (Fig.5). A beaver family can use more than one lodge or burrows and especially in summer time (Richard, 1980; Bau, 2001). The fact that beavers have a resting place in Ællebæk is not enough satisfactory. Beavers seen swimming alone upstream to Ællebæk (at point D) must be members of Hestbæk as the closer family downstreams lives far away after the fishery (U2). Therefore, Hestbæk family use both sides of the river (Fig.5).

Bearver territories in Klosterheden State Forest, Denmark 11 The important scent-marking on the riverbank downstream Øvej (Fig.4) could be explained as this part is a necessary downstream dispersion way for the 2-years old from the two family upstream. Indeed most of the scent-marks were found in April-May just before the 2-years old dispersion. Ællebæk is a bush area with many willows while Hestbæk is a common reed swamp more or less flotted where the bushes are inside the vegetation and therefore less easy to reach.

As I have never seen a beaver at the bundary on Øvej, it is impossible to tell which family (Flynder Å or Hestbæk) uses this area. Therefore I stopped the territories before the bridge. In early setpembre, I found a new cutting tree and fresh eating tracks there.

The beavers visit sometimes Ællebæk Sø but have not setlled any territory there. Indeed just a few fresh cuts were recorded near Ællebæk Sø and close to Jagvej. However no cuts or tracks were found between Gl.Landvej and Ællebæk Sø. Most of the tracks near the lake were old and may date from the reintroduction. Therefore I limited the beaver territory at Ællebæk where the last beaver tracks have been found (i.e scent-mark, dams and refectory).

1.4. Discussion on the methodology

1.4.1. Record of scent-mark A scent-mark is easy to recognize as the smell is very characteristical. Beavers can scent mark also with AGS (Svenden, 1980, Müller-Schwarze & Heckman, 1980, Rosell & Bergan, 1998). I have considered only the scent-mark with castoreum as we didn’t know the smell of the AGS. The April-May record was made with a low level of water and a low vegetation. Afterwards the water level increased with a maximum for the August prospection. The vegetation have grown up on the riverbank first, and inside the river in a second time (especially for burreed (Sparganium erectum)). The August prospection has been made from inside the river because the vegetation was too abundant. I may have missed some scent-mark especially where the riverbank is covered by trees and bushes (i.e. western part of Møllesøen, Ællebæk). Some scent-mounds were also found without odor anymore and therefore non-recorded on the map. Some scent-marks are uphold along the time whereas the others are only made once or twice (Fig.4). This is in accord with Rosell (1994). Scent-mark recording is the most important process with radio-telemetry when studying beaver territory (Rosell & al., 2000; Nolet & Rosell, 1994 (for the old territories), Rosell & Nolet, 1997; Rosell, 1994; Borgykke, 2002). The beaver territory study focusing on observation only is possible. It needs a lot of time and an efficient material. With the use of infrared video, this work could have been made more quickly and without doubt. Furthermore, this method would have avoided the problem of the odour and noise from the observer. Indeed it is more expensive and not necessary in this case. However the next study could be thinked of under this way. The markings of the individual have not been chosen because of the death factor. However the popualtion has grown and telemetry study would be envisaged in a few years. By marking the 2-years old, dispersion could be studied as well as territory boundaries when the 2-years old would settle their own territory. Pair formation and duration would be also studied this way.

Bearver territories in Klosterheden State Forest, Denmark 12 This could also bring a solution for the dead recorded, because many of the transmittors could have different signal when the animal is dead (Ouellet, 1999). Furthermore, an indice of population growth could be settled between the marked individual and the non-marked individual so the population size would be known too.

1.4.2. Night vision glasses observation With a night vision glasses system, all is seen in a green color more or less bright. The vegetation and ground is quickly mixed up when dark and the point of reference more or less lost. The best use was at point C where the road offers a good contrast. Like common glasses, there is no use possible with fog. Difference between grown-ups and young are difficult to make and have been made satisfactory only when together.

2. Beaver population estimations

2.1. Introduction Population estimations for beavers has not been found yet: neither lodge building (Ouellet, 1999) nor food caches (Osmundson & Buskirk, 1993) seem be usable. Furthermore the abundance of scent marks varies along the year (Rosell & Bergan, 2000; Svendsen, 1980; Rosell & al., 1998) and with the number of close neighbouring (Butler & Butler, 1979; Müller-Schwarze & Heckman, 1980) and therefore can’t be used. Beavers inhabit the same territory for several years (Richard, 1980). For this reason, eating tracks should increase with years even if the size of the colony doesn’t change. Furthermore eating tracks vary according to seasonal food abundance and beaver activity. Therefore, eating tracks seems to be a bad choice for population estimation. In order to estimate the beaver population in Klosterheden and surroundings two yearly beaver counts were organised with 30-40 participants observing beavers during two evenings and two mornings in April and October since the reintroduction. Data from all beaver counts were analysed. The aim of this work is to bring out a pattern in beaver observation and reduce personnel and/or observation time for the beaver count.

In the following part, the words “observation” refers to a sight of beaver, the words “watch” to a watching of a part of a beaver territory during a period of time when an observer may have seen none, one, or several beavers.

2.2. Material and methods

2.2.1. Characteristics of the observation during beaver counts. Data were sorted out according to observation characteristics as age class of the beaver observed (i.e. adults or juvenile) and period when observation was made (i.e evening or morning) (Tab.1; Fig.6 & 7). The number of observation points has increased with years according to the increase in the beaver population and its dispersal. Furthermore, all the observation points were not supplied for all the observation periods through a given count. The number of made watches was calculated for each beaver count and inside each observation period (Tab.1).

Bearver territories in Klosterheden State Forest, Denmark 13 An index “observation number/watch number” was calculated in order to compare the different periods of counting (Fig.7). For the kid seen in October 2000, no period has been noted. Therefore we divided the number of juveniles by two and assess the result to each period.

In a second time, data were sorted out according to the hours of observation (before/after sunrise/sunset) (Tab.2; Fig.8 & 9). An observation beginning before sunrise/sunset and finishing after, counts for one observation before and one after. This is the only case where duration of the observation was considered. If the observation is noted as made when the sun goes down/up, the observation counts for one observation before and one after. Two individuals observed in the same time count for two observations. Only the data with an hour noted could be use.

Lots of parameters are unsettled (numbers of territories, posts, observers...). Statistical analysis could not be applied with efficiency and therefore was not used. Data were analysed graphically with the Microsoft Excel® software.

2.2.2. Hypothesis on different methodologies for beavers count Actually beaver counts took place twice a year, in spring and autumn. Two days are used to observe the beavers during 1 hour in sunset (30min before, 30min after) and 1 hour in sunrise (30min before and 30min after). The beavers are counted at counting zone where only some of them fit with the beaver territories. This work is almost entirely made by volunteers. The young are considered according to information on the breeding in each counting zone. This knowledge corresponds to the beaver kids seen out of the counting period. The maximum number of juveniles per territory is known when several kids are seen together. Dead beavers subtracted. Dead beavers are all the beavers which dead bodies have been found.

Data covers counting from April 2000 to April 2004. Different kinds of counting were simulated according to the data: - One day of counting twice a year (i.e spring and autumn); - Two days of counting and 1 counting season (i.e spring or autumn); - One day of counting once a year (i.e spring or autumn). Any possible combining between data were analysed. Data obtained with the present methodology (i.e two days of counting and two counting seasons) are took as basis (100% of information) for the percentage calculation.

2.3. Results

2.3.1. Characteristics of the observation during beavers count. Beavers are observed as often in the morning as in the evening. However, more beavers have been observed in the morning until now (Fig.6). Age class may play a role in beavers observation: adults were observed especially during morning counting whereas juveniles were during evening counting.

However no pattern can settle as the trends could go in the opposite direction according to the beaver count (e.g. April 2003, October 2003) (Fig.7).

Bearver territories in Klosterheden State Forest, Denmark 14 Sun hours may play a role in beaver observation. Evening observations (µ=0,67) are made mainly after sunset (Fig.8). Morning observations (µ=0,61) have been made before sunrise and especially during the four last count (Fig.9).

2.3.2. Hypothesis of other methodology The results are presented without subtracting the dead beavers unless when stated (Tab.3)

- One day / 2 counting seasons 128 combinations could be found. As they are redundant, only 10 solutions exist finally for the estimation of the beaver population in this case. They spread out from 43 to 52 beavers but each possibility doesn’t have the same chance to be found.

- One day / October counting 6 solutions could be found. As they are redundant, the beaver population could be estimate at 37, 38 or 39 beavers with respectively 33,33%, 50% and 16,67% to be found.

- One day/ April counting Fore solution could be found with the same chance: 41, 42, 43, 44.

The percentage of conserved/lost information varies with the kind of information (i.e. beavers). However, they follow the same pattern (Fig.10).

2.3.3. Personnel featuring Number of observation points and number of watches increased with the growth of beaver population. This went with an increase in personnel featuring. However, all the observation points were not supplied for all the observation periods through a given count. In average; 86% of the observation points are supplied during the four counting periods of a beavers count. The ratio “number of personnel/number of observation points” is more than 1 (µ=1,38). Therefore, all the personnel did not participate during all the beavers count.

2.4. Discussion

2.4.1. Characteristics of the observation during beaver count. Bau (2001) found that beavers in Klosterheden showed no pattern for their first appearances. However the first sighting had an average of 78 minutes after sundown (Bau, 2001). Even if the study was carried out at the beginning of the reintroduction time and only in few families (as well just one family could be monitored with the help of video recording), data from beaver count follows a similar pattern as the beavers are more often seen once the sun had disappeared. However, in the morning, beavers could spend a lot of time after sunrise outside their lodge (personal observation), this could explain the observation made in April 2001 and April 2002.

Beavers show no significant preference for one of the periods (i.e morning or evening). The tendency showing in Fig.6 should be considered with reservations as they could vary (Fig.7). Furthermore, a coincidence inside the data could not be remoted because most of the parameters are not settled around the counting period (from 2001 to 2004). One of them could have an influence.

Bearver territories in Klosterheden State Forest, Denmark 15 2.4.2. Hypothesis of other methodology Changing the counting method will cause loss of information. However two methodologies could be seriously envisaged as the loss is less than 15 % on the estimate of the population: one day with two counting season (A method) and two days of counting made in April (B method) (Tab.3). The A method is more efficient but the B method would be more economic. Indeed, lots of participants don’t work in the forest and several of them come from Copenhagen. It will be less expensive to come here for two days once a year than one day twice a year. Furthermore, the efficiency of the A method could be dued to an effect on the data combination. Indeed, most of the juveniles have been observed just once during counting days (Fig.11).

April counts are more efficient than October counts even if there is more people per observation point for October counts (for all the counting season). This is due to the fact that the beavers seen in October are often seen once more in April. For all the counting season, sights of beaver juveniles are more important in April (µkids sight in April >µkids sight in October) There is less vegetation and more light in April than in October. Therefore the conditions of observation are better. Other explanation could be in relation with the weather as the temperature is generally lower in October than in April. Beavers spend less time outside the lodge in winter, as the water temperature is cold. An another explanation could be found in the young beaver’s behaviour. In early October, young beavers have only a 2-month experience outside the lodge. The first young are mainly seen in early August. In April, they are older and more experienced. Therefore, they may spend more time outside the lodge and will be seen more easily.

2.4.3. Analysis of the actual methodology of counting

a. Overestimation argument: Recording of dead beavers. Dead beavers are taken away from the population only when a body has been found. For the moment, only two dead beavers have been taken away. In the two cases, private individuals transmitted the information to the staff forest. The forest staff was called for dead beavers many times but it was rarely a beaver. Mixing up with otters is the general case but it can be with all sorts of animals, even a roe deer (O.Olsen, pers. comm.). Not all the dead beavers could be found, as all the wateredges cannot be checked all the time. Furthermore, many beaver territories are on private owned land and some without any close inhabitant around. Therefore, the dead beavers cannot be recorded easily. The body would drift but the possibility that it will be found is also low if it drifts in an uninhabitated portion of river.

Average longevity is 7-8 years but beavers can live until around twenty years. Five of the reintroduced beavers were more than 51/2 years-old when released in October 1999 (Bau, 2001). Four years and a half later, these individuals have passed over the average longevity and may have died. Seven reintroduced beavers were between 2 and 51/2 years-old and some of them have passed over or have just reached this average longevity too. Therefore, the population may count at least 3 beavers less.

Bearver territories in Klosterheden State Forest, Denmark 16 Consideration of the death factors. The predation must be very low in Denmark as neither wolves, bears, nor lynxes live in the surroundings of the release site. Otters have been reported to prey on beavers whereas red foxes, minks and fishers (Martes pennanti) are known to prey on American beavers (i.e. Collen & Gibson, 2001). However, they are not considered as regular predators. Furthermore only the otter is well represented in the area but it eats preferentially fish.

Humans play an important role in beaver populations as they destroy their habitats. In Denmark where the beavers have been especially reintroduced for creating wetland habitats, it is more the lack of suitable habitats which should be considered. This will be a limiting factor for the population increase more than a death factor.

Destruction of dams and lodges especially during winter could bring out beavers to death if they don’t find a new suitable place. For the moment, only a few dams have been removed on private owned land by the forest staff. In the forest, a pipe has been put under Øvej to avoid the flooding of the road but the beavers are now used to cross the road. Hopefully the road is forbidden to public circulation and only used by the forest staff. Road kills are recorded as a death factor in beaver population (examples referred to in Bau, 2001, Rouland & al., 1984). This factor must be considered in Denmark as a car hurt a beaver, a few days after the reintroduction. In this case the beaver was able to be cured and released in nature after. For the moment, no poaching has been recorded in Denmark but it is a death cause in other countries. With the growth of the population, hunting has already be suggested and poaching may come too.

Beavers suffer of intestinal tract disease, pseudotuberculosis (Rouland & al., 1984, Bau, 2001) and parasitism (Rosell & al., 1996 in Bau, 2001). Only rabies and tularaemia has not be found yet in Denmark and blood samples of the reintroduced beavers were analyed before the reintroduction (Asbirk, 2001; Bau, 2001).

Extremely cold winter and heigh swelling have not yet be recorded in Danemark.

At last, interspecific competition and fights between males can cause death especially in dense population. In Denmark, the population could not yet be considered to be dense but fights may occur anyway between dispersed individual and residents.

Furthermore, it seems luckiest if all the kids reached the adult size even if the kids mortality in beavers is low.

b. Underestimation argument: Recording of the new-born kids The risk of overestimating the breeding is low as the additional beavers are considered only if several kids are seen together. Moreover, kids are seen often just once through a counting period (Fig.11) and mixing up with kids is rare. It can happen especially in Hestbæk where the territories enclose three counting zones. In this case, details of the observation helps to assess the number of young.

Bearver territories in Klosterheden State Forest, Denmark 17 The risk of underestimation is more important. The litter is 2-4 kids in European beavers with an average of 2,7 (Salvesen, 1927 and Danilov & Kan’shiev, 1983 in Bau, 2001). In Denmark, most of the litter have been observed with only 1 kid (47% of case) and some families have never breeded yet (Tab.4) In 2002, four 1 year-olds have been seen and did not correspond to the data on the breeding in 2001. Therefore, they haven’t been seen as kids and were summed afterwards. This proves that all the kids are seen neither during the counting nor during another period of the year. As an other example, 3 adults were recorded in Fruerbæk in September 2003 but neither kids nor 2 year-olds were recorded there.

The breeding must be underestimated.

Conclusion: Considering one and the other, it may find a balance and therefore the counting gives a good idea of the population estimation.

2.4.3. Index on number of territory. When the 2 years-old leave the lodge, they go find their own territory and pair mate to build a family. The number of territories should follow the variation effectif inside the population with a delay of roughly two years.

Assessments: All the 2 year-olds disperse during their 2 years. Two beavers will live in a territory and form a pair (therefore the sex ratio is assumed to be equal to 1). The number of territories is known according to the place where the counting has been taken place and information reported in Elmeros & al. (2004).

Only 2 kids under 1 year-old were release in October 1999 (at Møllesøen) but four individuals between 1-2 year-olds were release as well (Bau, 2001). The first dispersion should begin during 2000 whereas the kids released in Møllesøen may have disperse in 2001. Until 2001, seven territories were settled: Hestbæk, Risbæk, U1, U2, U3, U4, U5 while five of the release site were left: Stenbæk Sø, Rørkær Sø, Ellebæksøern, Rishøj Sø, Neder Sø. Møllesøen is the only release site where the beavers have settled a territory. This is also the only site where a family unit was conserved with a pair and kids from this year. During all this reshapes, two new family unit appeared. But with 6 beavers in dispersion, it misses at least one territory. In fact, a beaver in dispersion can find a place in a previously settled territory where the reproductive pair is broken. Svendsen (1988) found that mate-turnover happens frequently during a beaver life. Death of the mate is the main cause but desertions of the family have been also recorded (4% of cases). The territory could also not be recorded as it was not settled yet and the individuals always in dispersion or not be recorded because no found. The young at Møllesøen could have also delayed their dispersion for some reason. Svendsen (1989) found that the dispersion is delayed in dense population area. It was obviously not the case but the reintroduction and settlement in a new place could have perturbated the kids.

In 2002, two new territories were found (Fruerbæk and U7). However, U6 and U8, found in early 2003 (winter), could descend from the dispersion during 2002 as 2 years-old dispersion occurs normally in spring time.

Bearver territories in Klosterheden State Forest, Denmark 18 In 2002, the young from the first breeding in Denmark (in 2000) reached the time of dispersion and 6 kids have been recorded in 2000. By making pairs with individual, there is one territory too many. It could correspond to the territory not found for 2001. However, not all family units are composed by a reproductive pair with kids (Richard, 1980). This is reinforced by a sex ratio which may not be equal to 1.

After March 2003, no new territory was recorded but ten kids were noted for 2001 and must be in dispersion. Several tracks were recorded along the Storå: cut and fallen trees and sights of beavers, but neither lodge nor dams were found. In fact, nothing permited to know if it was the same beaver, which was swiming in all places or if several beavers made the different tracks. In 2003, the second lodge in Møllesøen territory was built (Elmeros & al.; 2004) downstreams the lake. As we assess now that it is a new territory (i.e. Flynder Å see part 1.2), therefore it may be a territory descending from the dispersion during 2003.

The stagnation in the number of territories until 2003 could therefore have two principal explanations : - the replacement of a lost mate inside a settled family unit (and therefore a underestimated death record); - the not-yet recorded new territories.

However, the variation in the number of territories seems to follow the variation of the population (Fig.12).

To use such an index, it will be important to decide which tracks are significant of a territory settlement: active lodge/burrows, dams and scent marks can be chosen. However, dams are not built if the water is enough high. Burrows are difficult to find if they are not overbuilt by branches and the entrance kept under the water surface, as it must be. Scent-mark is the most important tool for drawing beaver territories (Rosell & al., 2000; Nolet & Rosell, 1994 (for the old territories), Rosell & Nolet, 1997; Rosell, 1994; Borgykke, 2002) but their record needs to spend a lot of time along the wateredge. Walking along the wateredge is not always possible as most of the area is a private owned land and/or full of grass. Regular observations of beavers in an area can also serve to assess the presence of a territory.

Now that the beavers disperse outside the forest, all the information will depend essentially on private people. As for dead beaver recordings, not all the area can be checked and some of it is uninhabited. Until now, it works but such a dependence could be a problem in a few years, when people will be used to the beavers.

Another problem will come when the beavers are leaving the territory forever. The recorded of abandoned territories could be delayed in time if the lodge/dam is not destroyed quickly. Furthermore, this does not correspond all the time to an abandonation as a family could use more than one lodge, or build/dig a another shelter. Broken dams are not always a sign of deserted area (e.g. Flynder å). To rebuild a dam could take a lot of time (more than one month) (Flynder Å, pers. obs.). Obviously, the scent-mark will disappear as well as regular sight of beaver in the area but they are both time-costing methods.

Bearver territories in Klosterheden State Forest, Denmark 19 In the next two years, it will be interesting to see if the number of territories will follow the pattern of the kid records (Fig.12).

Recommendation. After the age of ten, beavers could be taken away from the population. However if the non-recorded deaths correspond to non-recorded kids, this will cause an underestimation of the population. A survey on the reproduction is needed, to know the potential breeding of the population and a reliable fecund rate, as they are very low until now. (See part 1.5 for argument on radio-marking).

In order to change the counting method, April count (A method) should be tested. As we bring it on data from all the previous countings, we cannot assess that the lost information will not grow up with the population increase. If such a relation is not settled, April count should be applied for counting. In Germany, Dr. Heidecke used a method of mapping territories/lodges and multiplied by 4 individuals (S.Asbirk, pers.comm.). As the index of territories follows the variation in population estimation (part 2.3.4) and as doing the same in Klosterheden, we get the same population estimates as with the beavers counts (S.Asbirk, pers.comm), I think this method could be considered for the future. However, with marked beavers it would be easier to map the territory without any doubt.

3. Outside Klosterheden state forest

3.1. Investigation along Storå river The Storå river runs south from the Flynder Å catchment area. Beaver signs were found on Storå : several branches were cut at the intersection between Storå and Ellebaek and upstream from Vemb. Direct observation of beavers took place in Holstebro and on October, 1st upstream from Vemb. However, the forest staff wasnn’t be sure about the first observation. Investigations made until now have not established a permanent beaver presence on Storå: beavers evidence being less significant. Therefore, beaver territories on Storå remain to be proved.

3.1.1. Material and methods Investigations were made by canoë from Holstebro to Vemb. Each riverside was observed simultaneously and randomly by two observers. In this way, one investigation was made on April 6th. From Vemb to about 100m before the river mouth, investigation was made using a little motorized boat. Three observers looked at one bank during the way down and the other bank during the way up. In this case, one investigation was made on April 7th.

3.1.2. Results From Holstebro to Vemb, no beaver signs were found. Four beaver signs were found near the mouth in a section of roughly 1km long. It was cuts of few little branches and two barked branches lied on the ground. Signs were recent, in regard of their light and unoxydized colour.

Bearver territories in Klosterheden State Forest, Denmark 20 3.1.3. Discussion Investigations of April 6th and 7th were made after a raining period. Therefore, the water was brown owing to streaming and the water level was high. This bad visibility condition didn’t permit to bring to the fore of possible burrows. Even high bank could be a good place to dig burrows, habitat is favoured to beavers especially after Vemb, where deciduous trees areas are more important. This could explain the lack of tracks from Holstebro to Vemb. Beaver signs founded are in few number and not significant of a territory settlement (unlike burrows, lodge, dams and scent-mark). On the other hand, it proves way of beavers and a fortiori Storå could be seen as a colonization way for beavers. Furthermore, dispersal of the 2-year-olds beavers took place during spring, and few founded signs could mean that dispersal has just begun.

Tributaries and streams have not investigated yet. The possibility of a beaver territory somewhere in this region persists.

3.2. Search for the lodge in Flobæk The area is situated on a private owner land, and cows feed on the fields on both sides of a river. The river sides are sloppy and therefore protected by a fence. The burrow the beavers have built there collapsed. No other shelter has been recorded for this time but eating tracks still spread along the area.

3.2.1. Method Search of the lodge has been made by walking along the riverside. The investigation took place in late July.

3.2.2. Result No lodge has been found. Two ventilated wholes have been found in the ground.

3.2.3. Discussion The ventilated wholes could be assigned to many things. I was not able to see under the water if an entrance had been dug. The water was too dark as it was after a raining period. The eating tracks were few and concentrated along the first field (i.e. closer form the house). However the cover of tree was low and therefore the tracks less easier to find. Not all the riverbank could have been checked.

Bearver territories in Klosterheden State Forest, Denmark 21 Monitoring of water birds in the beaver ponds

A. Introduction After the reintroduction of European beavers Castor fiber in Klosterheden state forest district in Denmark, waterlands areas have increased thanks to the beavers building activity and the flooding of their territories. Bird breeding was monitored from 2000 to 2003, especially for the sparrow species (Elmeros & al., 2004). However waterfowl breeding was less studied. The aim of this work was to monitor a breeding season on beaver territory and therefore to assess the impact of beaver introduction made three years before.

B. Material and methods

1. Study area Klosterheden State Forest (KLS) is situated in the western part of Juland (56°28’N, 8°20’E), Denmark. It is a mixture with coniferous plantation and heath area, covering approximately 7000 hectares. Forest raisings, meadows, beaches and small forests are also present on the low area (Borglykke, 2002). Coniferous trees species as sikta spruce (Picea sitchensis), scotch pine (Pinus sylvestris), common spruce (Picea abies) and noble fire (Abies procera) dominate the forest area (Borglykke, 2002). Deciduous trees species as birch (Betula sp.), alder (Alnus sp.), willow (Salix sp.) and aspen (Populus tremula) are present only in few percentage (Borglykke, 2002). Flynder Å water system crosses the area with a low stream but before the reintroduction, some of its have been dammed to create small lakes or ponds (rapport de reintroduction).

12 areas in or surroundings Klosterheden plantation were studied (Tab.5). This enclose all the beaver territories present in this areas (i.e. Møllesøen, Flynder Å, Hestbæk, Ællebæk, Fruerbæk, Risbæk, U1 and U2) and four control zone (i.e. Øvre Sø, Rishøje Sø, Ællebæk Sø and Tårnsøen) where the beavers are not recorded regularly and therefore supposed to represent the general breeding pattern for waterfowl in KLS (Fig.13). Since May 30th, Stenbæk Sø was added to the monitoring.

Description of the vegetation Algae have not been identified but they are present in all places.

Møllesøen : North-east/east river bank is a mixture of water plants (Acrocus calamus, Iris pseudacorus, Sparganium erectum, Typha latifolia), sedges (Carex paniculata, C. rostrata), various grass (Epilobium sp., Rumex hydrolapathum, Filipendula ulmaria), wild angelica, nettles (Urtica dioica), lotus, water Forget-Me-Nots (Myosotis scorpioides), vetches (Vicia sp.), joncus (Joncus effusus, Joncus conglomeratus) and a few willows. In the northern part, it becomes a reed swamp with reed maces (Typha latifolia), common reeds (Phragmite australis), great water docks (Rumex hydrolapathum) and sedges (Carex sp.) where the Flynder Å runs with meandering. The western river bank is invaded by reed maces and various decidious trees. The sounthern river bank presents less diversified vegetation than the

Waterflow breeding in Klosterheden State Forest, Denmark 22 northeast part and leads to some resting places. The Flynder Å leaves the lake there in two running streams which meet a few metres downstream. A third stream run under the resting- place through a pipe and lead to a stagnant place invaded by the water vegetation (Sparganium erectum and Lemna minor). The water of the lake is largely covered by aquatic plants (Persicaria amphibia and Potamogeon natans) and dispersed burreeds (Sparganium erectum). There are also some “forest” of horstails (Equisetum fluviatile), patch of bulrushes (Scirpus lacustris), and/or Acrocus calamus. Water plantain (Alisma plantago-aquatica) is also recorded as well as Lemna minor. The first island is covered with reed maces, Epilobium, great water docks and burreeds but the other Møllesøen’s islands have got more diversified grass patterns with deciduous trees.

Flynder Å (monitored part) : The iver bed is invaded by burreeds (Sparganium erectum), with iris (Iris pseudacorus) and reed maces. The sloping river banks are enclosed (Lotus penduculus), nettles (Urtica dioica), water Forget-Me-Nots (Myosotis scorpioides), Epilobium spp, Gallium spp. and rare trees (Salix sp). On both sides spread out wet meadows with joncus species (Joncus effusus). Lesser duckweeds (Lemna minor) spread out along the water. Canadian pond weed (Elode canadensis) forms some underwater carpet especially before and after the monitored portion.

Ællebæk : the dam have flooded forest swamp dominated by willow with bog myrtle (Myrica gale), purple moor-grass (Molinia caerulea), common reed (Phragmite australis) and sedge. This vegetation spread out in the western side before lead on the coniferous forest. The water plants are represented by Lesser duckweed (Lemna minor).

Hestbæk : The dam is situated on a reed swamp area dominated by common reed (Phragmite australis) with bog myrtle (Myrica gale), willow (Salix sp.) and meadowsweet (Filipendula ulmaria). Great water dock (Rumex hydrolapathum), sedge (Carex paniculata) and Joncus effusus are also present. In the water, lesser duckweed is the main species.

Fruerbæk : The water has spread out in a moor scrubland of willow (Salix sp.), bog myrtle (Myrica gale) with bog blueberry (Vaccinium uliginosum), joncus species and very few trees (Popolus tremula, coniferous sp.). On the south-eastern part, it is followed by a grass meadow (Deschampsia flexuosa, Molinia caerulea) with heather (Calluna vulgaris, Erica tetralix), and common crowberry (Empetrum nigrum) whereas in the north-eastern part, the grass meadows are more diversified. The coniferous forest is just behind. Lesser duckweed (Lemna minor) is scattered over the water surface.

Risbæk : The first dam is dominated on both sides by bog myrtles (Myrica gale), purple moor-grass (Molinia caerulea) with more or less deciduous trees (willow especially in the eastern part, and birches especially in the western part). The wateredge is bordered with Joncus effusus and grass. In the water, European frog-bits (Hydrocharis morus-ranae) and cutleaf waterparsnips (Berula erecta) are found as well as a few burreeds.

The second and third dam are surrounded by wet meadows dominated by Joncus effusus. There, burreed is well represented with some European frog-bits (Hydrocharis morus-ranae) and Lesser duckweeds (Lemna minor).

Waterflow breeding in Klosterheden State Forest, Denmark 23 U1 : The monitored dam is a flooded area between forest and swamping meadow with Joncus effusus and various grass, bordered by mace reeds (Typha latifolia) and burreeds (Sparganium erectum) which scattered also in the dam. Lesser duckweed (Lemna minor) is well represented especially in the swamp meandering. It belongs for a part to a private owner area.

U2 : The lake is situated on a private land and a fishery spreads out on the Flynder Å from the southern-east to the northern-east side of the lake. The eastern and the western bank belong to a wet meadow more and less flooded with highly diversified vegetation. It is dominated by grass (Phelum pratense), meadowsweet (Filipendula ulmaria), Joncus effusus, Joncus articulatus and Rhinanthus serotines. Closer to the water, reed maces and sparganiums (Sparganium erectum) scattered out. The eastern bank is bordered by birch (Populus sp.) and few willow (Salix sp.) whereas in the western, common reed domines. An island in the eastern water area is covered by reed maces and sedges. European frog-bits (Hydrocharis morus-ranae) and Potamogeon natans cover most of the water surface. Inside the closer dam, Lesser duckweeds (Lemna minor) and cutleaf waterparsnips (Berula erecta) have been found.

Øvre Sø : The south bank is separated on Øvrevej by a small willow which branches fell down in the water. They carry on to the eastern by mixing with bog myrtle (Myrica gale), sedge (Carex sp.), joncus species and few reed mace in front of an open oak area (Quercus sp.) with grass underwood (Deschampsia flexuosa). The northern part becomes a reed swamp with reed mace, sedge (Carex sp.) and mosses. The western part is an islet of grass swamp dominated by cottongrass (Eriophorum angustifolium). The water is covered by Potamogeon natans.

Rishøj Sø : The south bank is bordered by willow which mixed in the western part with birch before leading immediately to the coniferous forest. The north part is a reed swamp dominated by reed mace. In the western, the coniferous forest turns in a scrubland meadow with heather (Calluna vulgaris and Erica tetralix) and common crowberry (Empetrum nigrum). It gets more various with sedges (Carex paniculata), joncus species , purple moor-grass (Molinia caerulea), cottongrass (Eriophorum angustifolium) when going down. Finally, the vegetation stretches out over moss carpets covered by sedges (Carex paniculata, C. rostrata), purple moor-grass (Molinia caerulea), cottongrass (Eriophorum angustifolium) and bog-myrtle and turn in island and penninsula which go meanderings in the western and north-western part of the lake. In the water, bulrushes (Scripus lacustris) form some patch while Potamogeton natans land on the water surface.

Ællebæk Sø : the biggest lake is largely covered by common reed (Phragmite australis) with few reed maces (Typha latifolia), a patch of horsetail and an islet of bulrushes (Scripus lacustris). Potamogeton natans, Hydrochrius morus-ranae, Urticularia australis and Sparganium sp. made up the aquatic vegetation. From the western to the south-eastern, a coniferous forest (Pinus sp., Larix sp.) borders the common area encolsed birch (Betula pubescens) and willow . The rest of the bank lead to an open area of a mixture with bushes (Myrica gale, Salix sp., Vaccinium uliginosum), sedges (Carex paniculata, Carex echinata), joncus and cottongrass (Eriophorum vaginatum, Eriophorum angustifolium) on thick moss carpet which turns on a grass meadow (Deschampsia flexuosa, Molinia caerulea) with heather (Calluna vulgaris and Erica tetralix) and common crowberry (Empetrum nigrum).

Waterflow breeding in Klosterheden State Forest, Denmark 24 In the water of the smallest lake, yellow pond-lily (Nymphaea lutea) shares the surface with Potamogeton natans.

Tårnsøen : Bushes (Myrica gale, Salix sp.) spread over the bank of the biggest lake in an arrengement of joncs, sedges (Carex sp), and grass (Molinia caerulea, Deschampsia flexuosa). The most part of the bank lead to an open area of heather (Calluna vulgaris) and common crowberry (Empetrum nigrum) which becomes a grass meadow (Deschampsia flexuosa). The rest of the bank leads to a coniferous forest. Only a few reed maces (Typha latifolia) are present, and especially in the northern part. The aquatic vegetation is poor with small covers of Potamogeton natans. A small island on the southern part is covered by trees. The smallest lake is situated between two walls of coniferous forest. The moss bank encloses bush cover (Myrica gale, Salix sp.) with horsetail (Equisetum sylvaticum). Reed maces and a few common reeds (Phragmite australis) spread over on both sides. Burreeds (Sparganium sp.), Potamogeon natans and marestail (Hippuris vulgaris) compose the aquatic vegetation.

Stenbæk Sø : a scrubland meadow with heather (Calluna vulgaris) and common crowberry (Empetrum nigrum) spread out over the north/north-eastern side and vary with moss areas cover by bog asphodel (Narthecium ossifragum) and bog blueberry (Vaccinium uliginosum). Close from the water, the vegetation is dominated by Eriophorum angustifolium growing up on moss carpets with sedges and joncus species. Typha latifolia et Alisma plantago-aquatica spread over on the bankside, the water and the island. The south part is composed of bushes and a common reed.

2. Observations The twelve areas were plotted in order to permit a monitoring with a 4-days interval for each territory, therefore twice in 8 days. The study was carried out form May 5th to August 24th 2004. The observation took place in the morning, on a sample road and/or high observation point when possible. Binoculars (8x21) were used. The water level was recorded as on a 3 steps scale (normal, high and low level). Weather conditions were noted for rain, wind and fog. The different broods were descriminated according to the growth of the young.

3. Graphical methodology All data were analysed by graphical method (using Microsoft Excel® software). In order to be able to compare the different areas, data were plotted by week (Fig.17 to 29). When two observations from a same week, the higher number of birds seen per observation is taken as equal to the number of birds for the week.

Waterflow breeding in Klosterheden State Forest, Denmark 25 C. Results The breeding patterns have been summed up on Tab.6 & 7 and schemes 15 & 16. The average number of kids by pair was calculated on the total number of kids on the total number of broods (Tab.6).

1. Results by Species

1.1. Mallard (Anas Platyrhychos) Recorded since March 29th, pair formation was observed during April especially in Møllesøen. Thirteen breeding pair was estimated for KLS (Tab. 6). The first young was observed in Møllesøen on May 6th with two broods (7 and 10 young respectively). In the same week, one brood was observed at Ællebæk Sø (9 young). The next week, 7 young were recorded in U1 (May 9th) and 6 in Fruerbæk (May 11th). On June 3rd , in Møllesøen, a third brood was recorded with 7 kids, a forth (10 young) on June 7th. On July, 2 new broods were recorded with 5 young each (Fig.14).

In Hestbæk, 3 young were recorded on June 3rd. On June 15th, a second brood (at least 6 young) was observed running away. However a young has been seen in Flynder Å in May. In Flynder Å, 2 broods were observed, on June 23rd (7 young) and July 1st (4 young) respectively. During beavers observations, 6 juveniles were recorded in Ællebæk on June 20th and one brood was also observed in the stream from Ællebæk on June 4th. Three grown-up young were also recorded in July the 3rd on the Flynder Å.

On August 1st , a female with 5 young around 1-2 weeks were recorded in Risbæk. Juvenile feathers was observed since June.

1.2. Muet swan (Cygnus olor) A pair of swans was recorded regulary for the end of March in Møllesøen. A nest had been built on the western bank but the swan was never seen hatching. Indeed absence of pair was noted during May (10th & 27th) but they carry on to defend the area : a second pair was recorded once but chased by the resident pair when it came back. They have left Møllesøen after the 22nd week (Fig.17). Another pair of swans ramdomly occuppied Tårnsøen during spring.

2 breeding pairs were estimated for KLS (Tab.6). A swan was observed hatching since May 5th in Rishøj Sø and since May 14th in U2. Rishøj Sø brood hached between May 17th and 21th (6 young : 5 grey, 1white). The same brood with the pair was observed in Luskestien Sø in May 25th (Fig.22) and after in Stensbæk Sø (no-monitored lake at first) since May 30th (Fig.28). Since June 2nd, 2 children of the initial Rishøj Sø brood have disappeared. Another one was observed floating on the lake on June 18th. On July 20th, an other young have disappeared. Therefore, at the end, only 2 young were alive. This is the only brood which could be monitored all the time.The swans have chosen the first island for a resting place.

U2 brood hached between June 2nd and 6th (4 young). After June 22nd, the family left the area (Fig.24). The young were around 2-3 weeks-old.

Waterflow breeding in Klosterheden State Forest, Denmark 26 However, a family with 2 young around this age was observed in U1 on July the 3rd . A pair with at least 3 young were also recorded in U1 (on the big dam) on the time of the 24th week.

An adult single swan was regulary observed in Ællebæk Sø from April to August (Fig.29).

1.3. Moorhen (Gallinula chloropus) Adults were recorded on this study area since March 29th. Five breeding pairs were estimated for KLS (Tab.6).

First young of moorhen are present in Risbæk since May 13th (4 young). Afterwards, they have been observed until the end of May. Observation in this area in June and July are sporadic (Fig. 22).

In Møllesøen, adults were regulary recorded since April and during the study (Fig.17). On July the 4th, 2 young were recorded in Møllesøen but the brood may count 3 young (obs. July 19th & 26th). Since July 29th, at least 2 young have got a juvenile feather.

In U2, records began in June on the lake as well as on the little dam. A brood is recorded on July 20th with 2 young seen. However it may enclose 3 young (obs. July 24th).

Moorhen were regulary recorded in U1 with at least 3 individual present on the area. On July 24th, a brood was observed (4 young).

Ællebæk Sø present a regular record pattern (Fig.29). No kids have been seen but a juvenile have been well observed on August 7th.

The species have been noted in Tårnsøen (1 obs.), in Hestbæk (3 obs. in spring time), Rishøj Sø (4 obs. in spring time) and Stenbæk Sø (Tab.7; Fig.16).

1.4. Pied wagtail (Motacilla alba) Adults were observed in Klosterheden since April 2004. During the monitoring, this species was regulary recorded only in Møllesøen, Flynder Å and U2 (Fig. 17, 18 & 24). However, few observations were made in Ællebæk. Males were seen more often than females. Juveniles were observed in Klosterheden in August.

1.5. Grey wagtail (Motacilla cinera) This species was recorded regulary between the 21st and 25th week in Møllesøen and more randomly during the rest of the study. Juvenile observation was made on June 7th in Møllesøen. Adults were also regulary seen in Flynder Å (Fig.18). They have bred in a nest under the bridge in øvej (A.Hartelius, pers comm).

Waterflow breeding in Klosterheden State Forest, Denmark 27 1.6. Little grebe (Tachybaptus ruficollis) The little grebe was recorded in Ællebæk Sø since April (Fig.29). One pair at least was observed and males were recorded singing during April and May. One kid was observed with an adult since August the 7th.

Presence of this species was confirmed also in Rørkær Sø (no-monitored) at the end of May and beginning of June (A.Hartelius, pers. comm; Pers. obs.) and 2 young were observed since June 9th. Another brood may have hatched afterwards as kids have been regulary seen until August.

An adult was observed several time in Mølesøen, always diving inside the water vegetation- covered surface. Once it was heard singing. Therefore, 2 breeding pairs were estimated for KLS (Tab. 6).

1.7. Grey heron (Ardea cinerea) Grey herons were recorded in Fruerbæk, Rishøj Sø, Ællebæk Sø, U2 and Hestbæk. Flyingindividuals are common and especially above U2. During summer time, and after the juveniles have left the nest, observations were made also in Flynder Å, Mollesøen, Tårnsøen. Great observations on juvenile were made in Ællebæk Sø. Observations were also made on individual landing in the tree-peak close to Møllesøn and Ællebæk Sø.

1.8. Cormorant(Phalacrocorax carbo) Cormorants were often observed flying since April and were also recorded landing in Møllesøen (April 12th & 14th , May 21st) and Tårnsøen (May 21st, June 4th).

1.9. Kingfisher (Alcedo atthis) Kingfisher brood regularly in Flynder Å and in U2 (T. Borup Svendsen, A. Hartelius, pers.comm.). Individuals werre recorded flying in Flynder Å and Møllesøen since May 6th but the nest was built on Flynder Å near U2. An adult was also recorded in U2 landing on a branch above the lake. In August, flying individuals were seen above Ællebæk and Ællebæk Sø as well as a non-monitored part in U1.

1.10. Teal (Anas crecca) Males were recorded on different lakes and dams during May. Females were only noted always accompanied by their male : on Øvre Sø (May 5th), on Møllesøen (May 19th), in a pond near Rishøjvej (May 21st) and Risbæk (on the stream).

1.11. Tufted duck (Aythya fuligula) One pair of tufted ducks was chased away form Møllesøen by two mallard males on May 10th. However, one pair plus one male were recoreded on May 19th. One pair was observed in Rishøj Sø on May 17th.

Waterflow breeding in Klosterheden State Forest, Denmark 28 1.12. Coot (Fulica atra) Presence of coots was only recorded in April in Møllesøen. However, one adult was recorded in Stensbæk Sø on June 2nd.

1.13. Sandpiper One individual was recorded in U1 on June 22th. Twice an individual was recorded in Fruerbæk in August.

1.14. Egyptian goose (Alopochen aegyptiacus) Two individuals of these geese were observed swimming in Tårnsøen in May 27th .

1.15. Water rail (Rallus aquaticus) A water rail was heard singing in Hestbæk three times from May to June. On August 23rd, an individual was recorded in Risbæk (A.Hartelius, pers. comm.).

1.16. Black stork (Ciconia nigra) The Black stork was seen around KLS (T.B.Svendsen, pers. comm.). It is the 4th year that it is recorded in the area.

1.17. Shelduck (Tadorna tadorna) Two individuals were observed flying above Møllesøen in April 8th at sundown. They made two passages.

1.18. Black headed Gull (Larus ridibundus) and other gull (Larus sp.) This species have been seen near U2 wheeling round above the fisherie. No accurate record of these species have been made as their presence is linking up only to the fisheries presence.

1.19. Unindentificated duck Six ducks were seen in Ællebæk Sø at the last observation. They were half than a mallard with a general spotted grey color. The observation was made by rain with dark light.

2. Results by territory Species records for all the area is presented on from schemes 17 to 29. Møllesøen present the best record of observed species (12 species) followed by U1 (7 species); Ællebæk Sø and U2 (6 species); Tårnsøen, Flynder Å, Risbæk area and Rishøj Sø area (5 species), Ællebæk, Hestbæk and Stenbæk Sø (4 species), Fruerbæk (3 species) and at last Øvre Sø (2 species) (Fig.16). All the species are not recorded in all places which may coorespond with habitat and space preferences.

Waterflow breeding in Klosterheden State Forest, Denmark 29 Apparently there seems to be no clear correlation between the size of the monitored area and the number of brood inside (Tab.7; Fig.15).

D. Discussion

1. Records of the species

1.1. Mallard Broods were discriminated on the young size. Confusion between the different broods could happen. The broods of May were seen on different areas far away from each other and without clear water way between them. In Møllesøen, the two broods were seen in the same time with no possibility for mistake. Therefore, five broods have hatched in early May (Møllesøen (2); Ællebæk Sø (1); U1 (1); Fuerbæk (1)). In June, the brood recorded in the stream linking up Ællebæk to Fynder Å on June 4th may mix up with the brood seen on June 3rd. Mallard grown-ups as well as juveniles and broods have been observed several times swimming along the Fynder Å and on the stream to Ællebæk. As the dates of observation are too close, the brood must be the same seen twice. The brood recorded in Ællebæk on June 20th was seen on sunset during beaver observation, the size could be identified as a young (not juvenile). Five days discrimate this brood with the brood seen in Hestbæk on June 15th (6 young). The brood seen in Flynder Å on June 23rd (7 young) could be mixing up with the brood from Ællebæk. Broods in Møllesøen were recorded with a four days interval. However, it is only with one week interval that the difference in the young size are quite satisfactory to discrimate the brood. Therefore, in June a maximum of six brood (Møllesøen (2); Fynder Å/Hestbæk/Ællebæk (4) were recorded and a minimum of four brood (Møllesøen (1); Fynder Å/Hestbæk/Ællebæk (3)). In July, two brood were seen in Møllesøen and on Flynder Å on the same day. As the observations were made close to each other, the birds are considered to not move from one area to an other, especially the young one. Therefore, three broods are recorded in July (Møllesøen (2); Fynder Å (1)). Mallard may not breed in the river Fynder Å. The width of the river is small with no secure place to hatch. They may use it to move between Hestbæk, Ællebæk and Møllesøen, the abundant vegetation protecting their move. A female with its young was seen getting under cover above the bridge on Øvejen.

The young recorded in August in Risbæk count for a new brood as no young were recorded previously in this area. The young were not able to fly at this time. They could walk from one of the close lakes in the surroundings of Risbæk. However no brood have been seen there at this time too. The brood is considered to have hatched a week before according to the size of the young : around July 25th.

All the brood were seen accompanied by a female except in Fruerbæk which was accompanied by a male. This observation in Fuerbæk is made against the litterature as the female hatch alone the kids (Géroudet,1999). Considering the whole brood, a total of 97 young have been observed (Tab.5). In August, the juvenile get together especially in Møllesøen (44 juveniles on August 6th) and Ællebæk (18 juveniles on August 23rd). However some of this juveniles could come from breeding places outside KLS.

Waterflow breeding in Klosterheden State Forest, Denmark 30 Mallards have been recorded at least once in each area. In July and August, the juveniles were dominant. They could mix up with the females as they share quite the same feather. Some of the observation were made against the sunlight and juveniles were identified in doubt case.

1.2. Muet swan Stenbæk Sø brood must be the Rishøj Sø brood. Five young were found as Stenbæk Sø, having roughly the same age as the young in Rishøj Sø. The missed young could have died during their walk from lake to lake as it happens frequently (predation by foxes). Stenbæk Sø was not monitored at the beginning of the study. We can exclude the possibility that Stenbæk Sø has got its own family of swans as there used to have breeding swans on this place. We didn’t see a nest in this area but the nest in Rishøj Sø, U2 and Møllesøen have been removed during the summer. In Møllesøen it was because of an increased water level. Indeed the brood in Rishøj Sø have got a white young. This is dued to a mutation called “false albinism” that followed the heredity rules. The broods in Stenbæk Sø count also a white kid. Therefore, I assume that the broods in Stenbæk Sø come from Rishøj Sø. A pair may have hatched in U1 on the big dam. At least three young were seen there in June 10th. Therefore between four and two pairs have hatched this year in Klosterheden. In Ællebæk Sø only one adult was recorded during the whole period of the study. However its absence was also recorded. It has landed at the same place during a week but left the lake after. Four days afterwards, it had returned. However, I have never seen a mate with him. I think it didn’t form a pair and hatch. The reasons are not known.

1.3. Moorhen Moorhen have usually two to three successive broods but I did not observe the second brood. The broods have got a average of 3.5 young. This is lower than the litterature (5 – 11 eggs). Predation on eggs could explain this low number. It is easy to discriminate the different broods as the new-born are black but the juvenile (called ”pullus”) have a brown olive color with a white throat (Heinzel & al., 1996; Beaman & Madge, 1998). Moorhen have been often recorded singing especially in U1. The singing record may underestimate the population as only two singing heard in different places could be discriminated.

1.4. Little grebe Little grebes are a small waterfowl, diving all the time to find his food (insect, larva, tadpole...). These two points made it difficult to observe. Furthermore, the little grebe found in water plant cover surface some good camouflage. Its singing is characteristic and easily recognizable.

Little grebe have two to three hatch per year (Géroudet, 1999). Adults hatch four to six eggs during 21 days. In Rørkær Sø, at least two kids were observed on June 9th but several brood may have hatched as kids have been observed until August. As the lake was not monitored, the number of broods and kids are not precisely known. In Ællebæk Sø, only one young was recorded in August. This is much later than in Rørkær Sø. Predation on eggs and kids could explain the low record of young. The vegatation

Waterflow breeding in Klosterheden State Forest, Denmark 31 is abundant in Ællebæk Sø with common reed and important cover of water plants and I may have miss some individuals.

1.5. Grey heron Grey heron breed in colony at the tree-peak. No such place have been recorded close to the monitored area. Grey heron must breed near U2 as lots of individuals were often seen flying and wheeling round above the forest (i.e. Klosterheden). This is supported by the spread of the juveniles in Klosterheden since July (observation made on juveniles in Hestbæk, Ællebæk Sø, Tårnsøen). Adults were identificated in doubt case. Therefore, there could have more juveniles than seen. A breeding place is known close to a beavers territory (A.Hartelius, pers comm).

1.6. Wagtail Confusion between pied wagtail and grey wagtail is possible only with flying individuals. However a confusion between female of grey wagtail and yellow wagtail (Motacilla flava) is possible. Yellow wagtail was recorded in 2000 but not afterwards (Elmeros & al., 2004). Grey wagtail have been identificated without mistake : the nuptial male have got a black throat which makes it clearly different than a yellow wagtail (Heinzel & al., 1996). During the record, I did not make a difference between male, female and juvenile as the observation time was too short. However, grey wagtail juveniles have been identificated in June on some observations and pied wagtail juveniles have been identificated on early August in Fruerhøj. This last species may have bred earlier but the identification of juvenile was reliable only on the observation in August.

1.7. Cormorant This seabird could be found in lake, river, estuary and coastal water. They breed on cliffs, rocks or trees (Heinzel & al., 1996). Seen more often flying, only three records were made on landing bird. All of them were on big lake (Møllesøen 3,3 ha and Tårnsøen 1,3 ha). The species did not breed in KLS on the monitored area but it may breed in the surrondings closer to the sea.

1.8. Sandpiper In U1, I observed a sandpiper feeding itself on the mud. It had got a white stomach, dark grey wings, roughly neck side and face light grey, thin and quite long beak, long light legs (yellow or green). After some minute, it flew to the peak on dead wood and stayed not moving for a while. Afterwards, I identificated it as a common sandpiper (Actitis hypoleucos) or a green sandpiper (Tringa ochropus) or a wood sandpiper (Tringa glareola). The habitat corresponds better to the green sandpiper and the wood sandpiper. The observed individual did not move the tail as it is written the common sandpiper do, but I well remembered a white sickle above the wings. Therefore I don’t assess more precisely name to this sandpiper. Common sandpiper have been observed in KLS and in very few number some wood sandpiper (A.Hartelius, pers. comm.).

In Fuerbæk, I observed twice a sandpiper with four days intervalle. The first fed itself on the mud and also find shelter in a bush. It walked in and out the vegetation making the observation difficult. It had got a rather orange beak, a black band on

Waterflow breeding in Klosterheden State Forest, Denmark 32 the eye. The feather looked like the one of the common redshank but I was not able to see the color of the legs. The stomach and the anus are white and keep moving. The habitat is not the one of the common redshank and the observed individual is smaller than the individual of this species. The behaviour corresponds to the one of the common sandpiper. The second individual was seen against the sunlight, standing up on a grass mound before jumping in the water and disappearing from view.

All the observations were made with bad conditions of light as they were in the shadow or against the sunlight. Furthermore the glasses used were not efficient for this kind of observation condition as they are not enough powerful.

1.9. Unindenticated duck The ducks seen in Ællebæk on August 23rd were accompanied by mallard juveniles. Their size was half of the mallard size. With the bad light conditions, nor the mirror color nor the beak color could be determined. However a white area near the end of the tail has been clearly seen. This could correspond to teal juveniles which feathers are close from the teal female.

2. Results by territory Møllesøen present the best diversity (12 species) as well as the best record of breeding for mallards (5-6 brood). This pattern could not be reliably linked to the beaver presence as it is the biggest lake in KLS. However, the beavers don’t trouble the waterfowl breeding in this place. The typically beaver dam as Ællebæk, U1, Fruerbæk, Risbæk present an interesting waterfowl breeding pattern with more or less one mallard brood each and two moorhen brood at the end.

Øvre Sø presents a bad record as well as the lack around Risbæk (Nedre Sø, Døjbæk Sø) due to the presence of predator fish.

The bad record noted on the end of the monitoring at Risbæk may be linked to forest work and noise disturbance in this area during august. The dams have also been drained quite completely (waterlevel decreased of more than 50cm).

Tårnsøen is on a windy area and just few vegetation covered the water surface. This could explain the low waterfowl presence.

Waterflow breeding in Klosterheden State Forest, Denmark 33 3. Beavers’influence. With increasing the water area in Klosterheden, the beavers have favoured the presence of waterfowl species (Moorhen, Sandpiper, Mallard, Wagtail) as well as wetland sparrow species (Emberiza schoeniclus). The moorhen have profited on the beavers presence with four broods in beavers territories (which two in beavers dams) against one in the control zone. The mallards use all the water place created by the beavers as breeding place and resting place. The swans have bred in a control zone and in a beaver territory and haven’t been troubled by the beaver presence. The main factor seems to be the area rather than their inhabitants. Swans breed in lake with abundant vegetation but have avoided Møllesøen and Tårnsøen for undetermined reason. The breeding pattern this year is lower than it used to be : swans have been breeding in Møllesøen, Tårnsøen, Risbæk sector, Stenbæk Sø and U2. However no breeding season have been monitored until now and the breeding may have been overestimated. Indeed, a pair of swans has been observed at the beginning of the breeding season in all of this place but they didn’t choose to breed. As they are territorial during the breeding period (Géroudet, 1999) the presence of one swan in Ællebæk Sø prevent the coming of a breeding pair. The sandpiper presence seems linked to the beaver activities as they have been observed only in beaver dams who provide an adequate habitat. Grey herons have profited of beavers area as eating and resting place whereas they could use the forest for breeding. Little grebes were not found yet in beaver dams but they are known to use all the place suitable to breed, even the smallest piece of water (Géroudet, 1999). Therefore, the increase of water place in KLS could favour this species on the long-term. The swallow (Hirundo rustica) and house martin (Delichion urbica) have been observed all over the forest but they use the water area to drink and catch their food (insects). Lots of observations have been made above beaver territories. The diversity in insect population have increased thanks to the beaver reintroduction (Elmeros & al., 2004) but the insect population may have incresed as well by the increase of the wetland area. Therefore, swallows and house martins find more food in the area and the population may increase. Furthermore, houses and garages inside the forest provide lot of possible breeding places.

During springtime, teals and tufted ducks have been recorded using beaver territories and/or control zones to rest. They breed in other parts of Juland. The possible juvenile of teal observed in Ællebæk Sø on August 23rd could come from one of this places. Their presence is promising and they could breed in KLS in a few years. The black stork recorded since the beaver reintroduction would perhaps breed in a few years here.

Globally this year was a poor year for bird observation (A.Hartelius, pers comm.) as it used to. The study made should be considered to a starting point for evaluating the waterfowl breeding in KLS. However the beavers presence have been favorable to the waterfowl species by increasing the potentiality of resting and breeding places in the forest.

Waterflow breeding in Klosterheden State Forest, Denmark 34 E. Passerines observations

1. Ring-marking in Fruerbæk Passerines species were recorded occassionnaly during this study. However, a ring-marking study was carried out near Fruerbæk by an other authority (P.Flansmose for the museum of Zoology). Three field works were organized between May 2nd and May 31st. Data of this study give a good knowledge of birds diversity in Klosterheden and near beaver dams. Catchings were more important near the stream where areas have been floated by beaver activity than in the forest, even for forest specific species.

Recorded species Acanthis cannabina (Linnet) Acanthis flammea (Redpoll) Anthus trivialis* (Tree pipit) Carduelis spinus (Siskin) Dendrocops major* (Lesser-spotted woodpecker) Emberiza schoeniclus* (Common reed bunting) Erithacus rubecula* (Robin) Fringilla coelebs* (Chaffinch) Garrulus glandarius (Jay) Hirundo rustica* (Swallow) Loxia curvirostra (Red crossbill) Loxia pytyopsittacus (Parrot crossbill) Parus ater* (Coal tit) Parus cristatus* (Creasted tit) Parus major* (Great tit) Phylloscopus collybita* (Chiffchaff) Phylloscopus trochilus* (Willow warbler) Prunella modularis* (Dunnock) Pyrruhula pyrruhula* Regulus regulus* (Goldcrest) Sylvia atricapilla* (Blackcap) Sylvia communis* (common whitethroat) Turdus merdula* (Blackbird)

2. Other observed species In Klosterheden Most of this species was recorded during water bird monitoring or beaver observations. However little ringed plovers (Charadrius dubius) and lapwings (Vanellus vanellus) were recorded on private owners area in the limits of Klosterheden near the aerodrome. Lapwings were also recorded in Stenbæk Sø on July. This two species have bred with young observed since the end of May (A.Hartelius, pers, comm.).

Alauda arvensis* (Skylark) Buteo Buteo (Common buzzard) Carduelis chloris* (Green finch)

Waterflow breeding in Klosterheden State Forest, Denmark 35 Certhia familiaris (Common treecreeper) (A.Hartelius, pers.comm) Charadrius dubius (Little ringed plover) Coccothraustes coccothraustes (Hawfinch) Columba palumbus* (Wood pigeon) Corvus corax ( Raven) Corvus corone cornix (Hooded crow) Corvus monedula (Jackdaw) Cuculus canorus (Cuckoo) Delichion urbica (House martin) Emberiza citrinella (Yellow bunting) Falco sp. (Falcon) Lanius collurio* (Red-backed shrike) Miliaria calandra* (Corn bunting) Parus caeruleus* (Blue tit) Parus palustris* (Marsh tit) Passer montanus (Tree sparrow) Phasianus colchicus torquatus (phasian) Pica pica (Magpie) Picus viridis (Green woodpecker) Streptopelia decaocto (Collared dove) Strix aluco (Tawny owl) Sturnus vulgaris* (Straling) Sylvia borin* (Garden warbler) Sylvia carruca* (Lesser whitethroat) Troglodytes troglodytes* (Winter wren) Turdus philomelos (Song thrush) Turdus viscivorus (Mistle thrush) Vanellus vanellus (Lapwings)

Young were observed and/or heard for the species in blue ink. Species with a star were previously noted in the area (Elmeros & al., 2004). Three species recorded in this report have not been seen : Muscicapa striata (Spotted flycatcher), Lullula arborea (Woodlark), and Carduelis carduelis (Goldfinch). The last two species have not been recorded since 2000 and their disappearance was assessed to natural variation in small population (Elmeros & al., 2004).

In winter time, a Jack snipe (Lymnocryptes minimus) has been observed in Hestbæk (T.B.Svendsen, pers.comm.). This species is in Danemark during winter only (Heinzel & al., 1996).

Waterflow breeding in Klosterheden State Forest, Denmark 36 3. Other observations Odonata: In Klosterheden, from May to August : - Pyrrhosoma nymphula - Ischnura elegans (Blue-tailed damselfly) - Coenagrion pulchellum (Variable damselfly) - Calopteryx virgo (Beautiful demoiselle) - Calopteryx splendens (Banded demoiselle) - Libellula quadrimaculata ( Four-spotted skimmer) - Brachytron pratense (Hairy dragonfly) - Corduliidae sp - Aeshnidae sp - Libellulidae sp

Lepidoptera In Klosterheden, from April to August : - Anthocharis cardamines (IV-V) (Orange Tip) - Inachis io (IV-Vand VIII) (Peacock) - Euthrix potatoria (caterpillar of Drinker moth) - Aphantopus hyperantus (VI-VII) (Ringlet) - Heodes virgaureae (VII) - Pieris sp.

Waterflow breeding in Klosterheden State Forest, Denmark 37 Acknowledgements

I would like to thank Sten Asbirk, Thomas Borup Svendsen, Ole Olsen for supervision and advices, Karstern Jensen and Arne Hartelius for assistance in the field, Jérôme Lhullier for correcting the manuscript. I would like to thank Peder Flensmose and Henry Thørgersen for inviting me to their ring-marking. I also thanks all the Klosterheden’s staff for help and precious informations.

38 Bibliography

Asbirk, S. 2001. Reintroduction of the European beaver (Castor fiber) in Denmark. Pages 25-28 in : Czech, A. & G. Schwab (eds): The European beaver in a new millennium. Proceedings of 2nd European Beaver Symposium, 27-30 Sept. 2003, Balowieża, Poland. Carpatian Heritage Society, Kraków.

Brady, C.A. & G.E Svendsen. 1981. Social behaviour in a family of beaver, Castor canadensis. Biology of Bihaviour 6: 99-114.

Bau, L. 2001. Behavioural ecology of reintroducted beavers (Castor fiber) in Klosterheden State Forest, Denmark. MSc Thesis. Departement of Animal Behaviour, University of Copenhagen, Denmark.

Beaman, M. & S. Madge (K. M. Olsen). 1998. Fuglene: Europa, Nordafrika og Mellemøsten. Gods Forlag København. 867 pp.

Borglykke, M. 2002. Bæverens (Castor fiber) fødevalg i Klosterheden Statsskovdistrikt, Danmark. Specialerapport i Biologi, Zoologisk Institut, Københavns Universitet.

Butler, R.G. & L.A. Butler. 1979. Toward a functional interpretation of scent marking in the beaver (Castor canadensis). Behavioural and Neural Biology 26: 442-454.

Collen, P. & R.J. Gibson. 2001. The general ecology of beavers (Castor spp.), as related to their influence on stream ecosystems and riparian habitats, and the subsequent effects on fish – a review. Reviews in Fish Biology and Fisheries 10:439-461.

Doboszyńska, T. & W. Żurowski. 1983. Reproduction of European beaver. Acta Zool. Fennica 174: 123-126.

Elmeros, M.; Berthelsen, J.P. & A.B. Madsen. 2004. Overvågning af bæver Castor fiber i Fylnder Å, 1999-2003. Faglig rapport fra DMU, nr xxx 2004. Danmarks Miljøundersøgelser, Miljøministeriet. 76 pp.

Géroudet, P. 1999. Les Palmipédes d’Europe. Edition mise à jour par Michel Cuisin. Quatrième edition. Delachaux et niestlé. 510 pp.

Hartman, G. 1994. Ecological studies of a reintroduced beaver (Castor fiber) population. Dissertation, departement of Wildlife Ecology. Swedish University of Agricultural Sciences.

Heinzel, H., Fitter R. & J. Parslow. 1995. Oiseaux d’Europe, d’Afrique du Nord et du Moyen-Orient. Les Compagnons du Nauturalistes. Delachaux et niestlé. 384 p.

Laanetu, N. 2001. Action plan for the protection and management of the european beaver (Castor fiber L.) in Estonia. In the frame of the project “ Banat-Estonia”; project funded by the german federal Ministry of the Environnement, Nature Conservation and Nuclear Safety. TARTU 2001.

Lapiński, S. & J. Staliński. 2001. Utilization of woody species by beaver (Castor fiber) in different habitats. Pages 142 in : Czech, A. & G. Schwab (eds): The European beaver in a new millennium. Proceedings of 2nd European Beaver Symposium, 27-30 Sept. 2003, Balowieża, Poland. Carpatian Heritage Society, Kraków.

39 Müller-Schwarze, D. & S. Heckman. 1980. The social role of scent marking in beavers (Castor canadensis). Journal of Chemical Ecology 6(1): 81-95.

Müller-Schwarze, D.; Heckman, S. & B. Stagge. 1983. Behavior of free-ranging beaver (Castor fiber) at scent-marks. Acta Zool. Fennica 174: 111-113

Nolet, B. & F. Rosell. 1994. Territoriality and time budgets in beavers during sequential settlement. Canadian Journal of Zoology 72: 1227-1237.

Ouellet, J.P. 1999. Techniques d’aménagement de la Faune. Notes de Cours. Université du Quèbec à Rimouski. 58pp.

Osmundson, C. L. & S. W. Buskirk. 1993. Size of food caches as a predictor of beaver colony size. Wildl. Soc. Bull. 21: 64-69.

Patenaude, F. 1983. Care of the young in a family of wild beavers, Castor canadensis. Acta Zool. Fennica 174: 121-122.

Richard, B. 1980. Les castors. Baland,Paris. 168pp.

Rosell, F. 1994. Factors affecting territory size and scent marking behavior in the European beaver (Castor fiber). Hovedfagsoppgave (cand. scient.) i etologi. Zoologisk Institutt, AVH. Universitetet i Trondheim.

Rosell, F. & F. Bergan. 1998. Free-ranging Eurasian beavers, Castor fiber, deposit anal gland secretion when scent marking. The Canadian Field-naturalist 112(3): 532-535.

Rosell, F. & F. Bergan. 2000. Scent marking in Eurasian beavers Castor fiber during winter. Acta Theriologica 45(2): 281-287.

Rosell, F.; Bergan, F. & H. Parker. 1998. Scent-marking in the Eurasian beaver (Castor fiber) as a means of territory defense. Journal of Chemical Ecology 24(2): 207-219.

Rosell, F.; Johansen, G. & H. Parker. 2000. Eurasian beavers (Castor fiber) behavioral response to simulated territorial intruders. Can. J. Zool. 78: 931-935.

Rosell, F. & B. Nolet. 1997. Factors affecting scent-marking behavior in Eurasian beaver (Castor fiber). Journal of Chemical Ecology 23(3): 673-689.

Rosell, F. & K. V. Pedersen. 1999. Bever. Landbruksforlaget. 272p.

Rosell, F. & L. Sun. 1999. Use of anal gland secretion to distinguish the two species Castor canadensis and Castor fiber. Wildlife Biology 5(2): 119-123.

Rouland, P. ; Perraud, P. & R. Nozerand. 1984. Le Castor castor fiber L. Bull. Mens. Off. Natl. Chasse. 76: 26- 32.

Svendsen, G. E. 1980. Patterns of scent-mounding in a population of beaver (Castor canadensis). Journal of Chemical Ecology 6(1): 133-148.

Svendsen, G. E. 1989. Pair formation, duration of pair-bonds, and mate replacement in a population of beavers (Castor canadensis). Canadian journal of Zoology 67(2): 336-340.

Thomsen, L.R. 2001. Sex differences in scent marking in a monogamous mammal, the eurasian beaver (Castor fiber). MSc Thesis. Departement of Zoology, University of Aarhus, Denmark.

40 Fig.1. Tree and branches cut by beavers (Flynder Å, 2004)

41 Fig.2. Beaver territories in Klosterheden Forest District and surroundings in 2003 (Elmeros & al., 2004). The empty circles show the release sites and the black circles the known lodges. I1 : Møllesøen, I2 : Hestbæk, I3 : Flynder Å, I4 : Risbæk, I5 : Fruerbæk.

42 Møllesøen

Flynder Å

Ællebæk

Hestbæk

Fig.3. Beaver activity area in Møllesøen and Hestbæk beaver territories.

43 Fig.4. Map of the beaver scent-mark.

44 Møllesøen

Fynder Å

Hestbæk

Fig.5. Beaver territories according to the study results.

45 Tab.1. Beaver observation characteristics during the beaver count from October 2000 to April 2004 Evening Morning Beaver count Adults Juveniles Watches Adultes Juveniles Watches 10.11.12 Oct 2000 31 0,5 20 35 0,5 20 24.25.26 Ap 2001 247 4929354 1.2.3 Oct 2001 180 4233343 16.17.18 Ap 2002 359 5637450 9.10.11 Oct 2002 16 10 39 23 3 38 9.10.11 Ap 2003 44 18 66 38 9 64 28.29.30 Sept 2003 23 14 53 18 7 53 29.30 Ap & 1 May 2004 386 6036860

46 Beavers seen on evening 0,9 Beavers seen on morning Adults seen on evening 0,8 Adults seen on morning 0,7 Juveniles seen on evening Juveniles seen on morning 0,6

0,5

0,4

0,3

0,2 Average of observations/watches of Average 0,1

0,0

Fig.6. Distribution of the beaver observations according to the age of the observed beaver during beaver counts from 2000 to 2004

47 2,0 1,8 1,6 1,4 1,2 1,0 0,8 0,6 observations/watches 0,4 0,2 0,0 oct. 2000 april 2001 oct 2001 april 2002 oct 2002 april 2003 oct 2003 ap. & may Beavers Counting 2004 Beavers seen on evening Beavers seen on morning Adults seen on evening Adults seen on morning Juveniles seen on evening Juveniles seen on morning

Fig.7.Variation of beaver observations according to the age of the observed beaver during beaver counts from 2000 to 2004

48 Tab.2. Hours of the beaver observation according to avaible data from beaver counts from 2001 to 2004 Beaver Counts Before sunset After sunset Before sunrise After sunrise 24.25.26 Ap 2001 13 15 16 19 1.2.3 Oct 2001 5122213 16.17.18 Ap 2002 15 31 14 17 9.10.11 Oct 2002 423186 9.10.11 Ap 2003 26 38 26 15 28.29.30 Sept 2003 432218 29.30 Ap & 1 May 2004 21 24 27 16

49 1,00 g 0,90 0,80 0,70 0,60 0,50 0,40 0,30 0,20 0,10 observations / watches made on evenin watches / observations 0,00 ap. 2001 oct. 2001 ap. 2002 oct. 2002 ap. 2003 oct. 2003 ap. 2004 Beaver count

obs. before sunset obs.after sunset

Fig.8. Distribution of the beaver observations made on evening during beaver counts.

50 1,00

0,90

0,80

0,70

0,60

0,50

0,40

0,30

observations / watches made on morning watches / observations 0,20

0,10

0,00 ap. 2001 oct. 2001 ap. 2002 oct. 2002 ap. 2003 oct. 2003 ap. 2004 Beaver count obs. before sunrise obs. after sunrise

Fig.9. Distribution of beaver observations made on morning during beaver counts.

51 Tab.3. Beaver population estimation obtained with different methods of counting according to data of the beaver counts from April 2000 to April 2004 Method 2 days / 2 1 day/ 2 2 days/ 1 days/ October 2 days / April 1 day / April counting counting October counting counting counting counting Population in 1999 18 beavers 18 beavers 18 beavers 18 beavers 18 beavers 18 beavers Population in 2004 54 beavers 47,5 ± 0,7 40 beavers 37,67 ± 1,15 47 beavers 42,5 ± 1,29 (without taking out the beavers beavers beavers dead beaver) Information conserved 100 % 87,96 % 74,07 % 69,75 % 87,04 % 78,70 % Information lost 0 % 12,04 % 25,93 % 30,67 % 12,96 % 21,30 % Minimal 43 beavers in 37 beavers in 41 beavers in 1,6 % of case 50 % of case 25 % of case Maximal 52 beavers in 39 beavers in 44 beavers in 1,6 % of case 16,67 % of case 25 % of case Population in 2004 52 beavers 45,5 ± 0,7 38 beavers 35,67 ± 1,15 45 beavers 40,5 ± 1,29 beavers beavers beavers Information conserved 100 % 87,5 % 73,08 % 68,60 % 86,54 % 77,88 % Information lost 0 % 12,5 % 26,92 % 31,40 % 13,46 % 22,12 % Population growth 36 beavers 29,5 ± 0,7 22 beavers 19,67 ± 1,15 29 beavers 24,5 ± 1,29 beavers beavers beavers Information conserved 100 % 81,94 % 61,11 % 54,64 % 80,56 % 68,06 % Information lost 0 % 18,06 % 38,89 % 45,36 % 19,44 % 31,94 %

52 50 45 40 35 30 25 20

Information lost (%) lost Information 15 10 5 0 2days/ 2 countings 1 day/ 2 count 2 day/ October 1 day/ October 2 day/ April count 1 day/ April count count count

Population growth Population estimation without taking out the dead beavers Population estimation (with taking out the dead beavers)

Fig.10. Variation of the loss of information according to the considered information.

53 100%

90%

80%

70%

60%

50%

40% Sight of juveniles of Sight 30%

20%

10%

0% ap. 2001 oct. 2001 ap. 2002 oct. 2002 ap. 2003 oct. 2003 ap. 2004 total Beavers count

juvenile seen once juvenile seen twice juvenile seen three times juveniles seen four times

Fig.11. Sight of juveniles during the beaver count from April 2OO1 to April 2004

54 Tab.4. Breeding in beaver territoies from 2000 to 2003 Territories Kids born in 2000 Kids born in 2001 Kids born in 2002 Kids born in 2003 Møllesøen 1 1 1 Hestbæk1223 Risbæk 2 1 1 1 Kurts Dam (U2) 1 1 0 1 Bækmarksbro (U4) 3 3 Keldseng (U1)1121 Fruerbæk Høkær (U8) U6 1 U3 1 U5 1 2 1 Grønkær bæk (U7) Kids Total 6 10 11 9 Population24344552 fecond rate (‰) 0,25 0,29 0,24 0,17

55 12

10

8

6

4

2

0 1999 2000 2001 2002 2003 years

beaver kids

Fig.12. Beaver breeding in Danmark since the reintroduction of 18 individuals on October 1999. The kids for 1999 are reintroducted kids

56 Tab.5. Areas monitored for waterfowl breeding Study area Number Surface Habitat Beavers presence Møllesøen 1 ca.3,3 ha. lake present Fynder Å 2 stream present Ællebæk 3 0,2 ha. floded pond present Hestbæk 4 1 ha. floded pond present Fruerbæk 5 0,2 ha. folded pond present Risbæk 6 1 ha. floded stream/pond present U1 7 2 ha. floded pond present U2 8 lake/pond present Øvre Sø 9lakeabsent Rishøj Sø 10 ca. 1,1 ha. lake absent Ællebæk Sø 11 ca. 1,4 ha lake absent Tårnsøen 12 ca. 1,3 ha lake absent Stenbæk Sø 13 ca. 0,7 ha lake absent

1 2 11 10 3 9 4 12 6 13

7 8

5

57 Fig.13. Areas monitored for waterfowl breeding

45

40

35 7/06

30 3/06 6/05 25 J 20 1/7 Y Individuals F 15 M 10 9/7 5

0

Fig.14. Mallard presence and observation of the brood in Møllesøen from May 2004 to August 2004

58 Tab.6. Breeding pattern in KLS during the summer of 2004 Breeding pair Average of young by pair Mallard 13 7,2 Muet swan 25 Moorhen 53 Little grebe 21.5

59 Tab.7. Number of observations made on waterfowl in Klosterheden State Forest between April and August 2004.

The dates are the first and the last observation for the species in the given area. Observations made on juveniles are written in bold. The date of each brood correspond to the date of the first young observation, they are written in bold with the number of young in bracket. Abreviation: A=April ; M=May ; J=June ; j=July ; a=August

Tachybaptus Gallinula Ardea Cinera Cygnus Olor Anas platyrhychos ruficollis chloropus Møllesøen 19/6-18/8 19/6-22/8 29/3-30/5 29/3-22/8 29/3-22/8 A=8; M=7; J=7; J=2, j=1; A=1 J=1 ; j=2; A=2 A=8 ; M=5 A=9; M=7+7, J=7+5, A=6; M=7; J=4; j=8; A=6 j=3+6; A=1+6 j=6+5; A=3+4 6/5(7); 6/5(10); 3/6(7); 4/7(3) 7/6(10); 1/7(5); 9/7(5) Fynder Å 9/7: j=1 17/4 –13/7 A=6; M=7; J=7; A=4 ;M=2 ; J=2+1; j=8; A=6 j=1+3 23/6(7); 1/7(4) Ællebæk 9/4-4/7 A=1; M=7; J=7; A=1 ; M=1 ; J=3+1; j=8; A=6 j=1+1 4/6 (x); 20/6 (6) Hestbæk 7/7-22/8 10/5-29/7 6/5-18/6 A=1; M=7; J=7; J=1 ; j=2+1; A=1 M=3; J=2+2; j=1 M=6; J=1; j=8; A=6 3/6(3); 15/6(6) Fruerbæk 7/5-22/7 11/5-11/8 A=1; M=7; J=7; M=2 ; J=2 ; j=1 M=4+1; J=0; j=1; A=1 j=8; A=6 11/5(6) Risbæk 8/7: j=1 2/6-1/8 5/5-18/6 A=1; M=7; J=7; J=1+1, j=1; A=1+1 M=5+3; J=4 j=8; A=6 25/7 (5) 13/5(4) Nedre Sø 8/4 : A=1 A=1; M=7; J=7; j=8; A=6 Luskestein Sø 8/7 : j=1 25/5 M=7; J=7; j=8; A=6 M=1+1; Dojbæk Sø 25/5 : M=1 25/5-8/7 A=1; M=7; J=7; 5/8: A=1 M=1; J=1; j=1 j=8; A=6 U1 4/7: j=1+1 5/5-28/7 5/5-21/8 M=7; J=7; j=8; A=6 M=6+2; J=3+2; j=6 M=7; J=7; j=6+1; 9/5(7) A=3+1 28/7(4) U2 17/4-8/7: M=1; J=1; 17/4-8/7 17/4-22/8 6/6-21/8 A=1; M=7; J=7; j=2 A=1; M=6; J= 6+5; A=1; M=3; J=2; j=0; J=5; j=3+3 ; A=3+1 j=8; A=6 individual were often j=1 A=2 20/7(3) seen flying 6/6 (4) Øvre Sø 13/5-14/6 A=1; M=7; J=7; M=2; J=1+1 j=8; A=6 Rishøj Sø 9/5: M=1 5/5-21/5 5/5-2/6 13/5-14/6 A=1; M=7; J=7; M=5+1 M=4; J=1 M=2; J=3 j=8; A=6 21/5 (6) Ællebæk Sø 19/5-19/8: M=3; J=1; 3/4-7/8 12/4-23/8 3/4-23/8 11/5-7/8 A=5; M=7; J=7; j=2+2; A=1 A=5; M=5; J=2 ; j=5; A=1; M=5; J=5; j=7; A=4; M=3+4; J=1+3; M=6; J=4; j=4+1; A=1 j=8; A=6 A=4+3 A=4 j=2+3, A=4 18/7(1) 7/8(1)? 7/4(9) Tårnsøen 20/6-13/8: J=1; j=1+1; 19/5-31/5 6/5-27/5 6/5 : M=1 M=8; J=7; j=8; A=6 A=1 M=4 M=4; Stensbæk Sø 18/6: J=1 30/5-1/9 10/6-13/8 M=1+1; J=7+7; J=1; j=2; A=1 J=8+8; A=5+5 Rørkæk Sø 9/6(2)

60 Moorhen

Mallard

Swan

Little grebe

Fig.15. Waterfowl broods according to the monitoring made in KLS during the summer of 2004

61 Moorhen Pied wagtail Teal

Mallard grey wagtail Tufted duck

Muet swan Cormorant Coot

Little grebe Sandpiper Egyptian goose

Grey heron Kingfisher Water rail

Fig.16. Waterfowl species observed in KLS according to the monitored area during the summer of 2004.

62 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Juvenile Swan Male Mallard Female Mallard Juvenile Mallard Adult Moorhen

Pied wagtail Grey wagtail Little grebe Grey heron Cormorant

Tufted duck Sand piper King fisher Egyptian gouse Water rail

Fig.17. Waterfowl records in Møllesøen from May to August 2004 12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.18. Waterfowl records on Flynder Å from May to August 2004. (For the legend, see Fig. 17)

10

9

8

7

6

5

4

3

2

1

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.19. Waterfowl records on Hestbæk from May to August 2004. (For the legend, see Fig. 17)

65 4,5

4

3,5

3

2,5

2

1,5

1

0,5

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.20. Waterfowl records on Ællebæk from May to August 2004. (For the legend, see Fig.17)

12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.21. Waterfowl records on Fruerbæk from May to August 2004. (For the legend, see Fig. 17)

66 18

16

14

12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.22. Waterfowl records on Risbæk from May to August 2004. (For the legend, see Fig. 17)

18

16

14

12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.23. Waterfowl records on U1 from May to August 2004. (For the legend, see Fig. 17)

67 14

12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.24. Waterfowl records on U2 from May to August 2004. (For the legend, see Fig. 17)

12

10

8

6

4

2

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.25. Waterfowl records on Rishøj Sø from May to August 2004. (For the legend, see Fig. 17)

68 9

8

7

6

5

4

3

2

1

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.26. Waterfowl records on Øvre Sø from May to August 2004. (For the legend, see Fig. 17)

7

6

5

4

3

2

1

0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.27. Waterfowl records on Tårnsøen from May to August 2004. (For the legend, see Fig. 17)

69 14

12

10

8

6

4

2

0 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Adult Swan Juvenile Swan Male Mallard Female Mallard Juvenile Mallard Adult Moorhen Coot Lapwing

Fig.28. Waterfowl records on Stenbæk Sø from May to August 2004.

70 22 23 24 25 26 27 28 29 30 31 32 33 34 Weeks

Fig.29. Waterfowl records on Ællebæk Sø from May to August 2004. (For the legend, see Fig. 17) Bibliography