Influence of Dermanyssus gallinae and Ascaridia galli infections on behaviour and health of laying hens (Gallus gallus domesticus)

Transcription

1 British Poultry Science Volume 46, Number 1 (February 2005), pp Influence of Dermanyssus gallinae and Ascaridia galli infections on behaviour and health of laying hens (Gallus gallus domesticus) O. KILPINEN, A. ROEPSTORFF 1, A. PERMIN 1, G. NØRGAARD-NIELSEN 2, L.G. LAWSON 2 AND H.B. SIMONSEN 2 Danish Institute of Agricultural Sciences, Danish Pest Infestation Laboratory, Lyngby, 1 Centre for Experimental Parasitology and 2 Department of Animal Science and Animal Health, The Royal Veterinary and Agricultural University, Copenhagen, Denmark Abstract 1. The effect of infections with Dermanyssus gallinae (poultry red mite or chicken mite) and Ascaridia galli (roundworm) on the behaviour and health of laying hens was investigated. 2. Six groups of 15 pullets (Isa Brown) were kept in indoor pens from 18 weeks of age. Two groups were artificially infected with D. gallinae, two groups with A. galli and two groups were kept as uninfected controls. The hens were observed for behavioural reactions and physiological changes (weight gain and various blood variables) to the parasitic infections. 3. Infections with D. gallinae resulted in reduced weight gain, anaemia and even death of some of the hens. Behavioural changes were also observed, as the mite-infected hens showed higher self-grooming and head scratching both during the day and night. 4. A. galli resulted in a lower weight gain but no significant changes were seen in blood variables or behavioural activities. INTRODUCTION Parasitic infections may have serious consequences for wild as well as domestic animals; infections of the latter are strongly associated with husbandry systems because transmission between hosts generally depends on environmental factors. In particular, the newly reintroduced trend of keeping production animals in extensive husbandry systems with access to the outdoor environment may include a risk of increased parasite burden and thereby reduced welfare (Appleby and Hughes, 1997). Dermanyssus gallinae, the poultry red mite or chicken mite, is the most important ectoparasite affecting egg layers in many European countries (Maurer et al., 1993a; Höglund et al., 1995; Kilpinen, 2000). D. gallinae is a temporary ectoparasite, spending most of the time hidden in cracks and crevices, only coming out to feed for short periods during the night. With high reproduction potential, a population doubling in 6 d under optimal conditions (Maurer and Baumgärtner, 1992), small numbers of mites may rapidly develop into serious mite infestations. In addition to its role as a potential vector for several pathogens of medical and veterinary importance (Zeman et al., 1982; Durden et al., 1992; Chirico et al., 2003), D. gallinae can cause anaemia and death of affected birds, increased food consumption, lowered egg production and blood stained eggs with a lower commercial value (Arends, 1991). However, little experimental evidence for these hypotheses exists. In an experimental infestation study, Maurer et al. (1993b) found that hens abandoned their favourite resting and sleeping places on perches and spent the night in the litter area, presumably because the mite density in the litter remained relatively low as compared to the perches. The grooming activities of the birds may well be affected by blood-sucking mites, but this has never been investigated. According to Spruijt et al. (1992), behaviour related to care of the body surface, such as preening and sand bathing, plays a crucial role in the behaviour repertoire of birds. Also removal of parasites (Brown, 1974) is given much attention and birds may spend considerable time taking care of their feathers. Ascaridia galli is a widespread nematode in poultry (Ackert, 1931) frequently occurring in non-intensive systems (Permin et al., 1999). Correspondence to: Ole Kilpinen, Danish Institute of Agricultural Sciences, Danish Pest Infestation Laboratory, Skovbrynet 14, DK-2800 Kgs. Lyngby, Denmark. Tel: þ Fax: þ Ole.Kilpinen@agrsci.dk Accepted for publication 28 September ISSN (print)/ISSN (online) ß 2005 British Poultry Science Ltd DOI: /

2 EFFECTS OF ENDO- AND ECTORARASITES 27 Perches above dropping pit Nest boxes in two levels Food Water Door Figure 1. Schematic drawing of the pens. Hens in organic systems are by law obliged to have straw-bedded houses and access to outdoor runs while prophylactic anthelmintic treatment is prohibited, and therefore it is predictable that a Danish survey has shown that 64% of organic layers are infected with A. galli (Permin et al., 1999). Although hens may regulate the worm burdens to some extent (Permin et al., 1997a), infections may result in production losses due to loss of appetite, weight, reduced egg production and death (Ackert and Herrick, 1928; Ikeme, 1971; Permin et al., 1997a). Observations indicate that A. galli may interfere with host behaviour, causing an increase in cannibalism (Ikeme, 1971), while a behavioural response to mild infections seems less likely. The present experiment was designed to investigate the influence of infections with D. gallinae and A. galli on behaviour and health of laying hens, with special focus on plumage maintenance activities during day and night and on blood variables related to health. MATERIALS AND METHODS Animals, housing and care taking Ninety parasite-free hybrid pullets (Isa Brown) were purchased from a commercial breeder at 18 weeks of age. The pullets were randomly allocated to 6 groups of 15 birds and placed in pens each measuring 3.4m1.5m. In one room, two control groups (C1 and C2) were placed nearest to the entrance, while two groups infected with A. galli (A1 and A2) were placed further back. The two groups infected with D. gallinae (D1 and D2) were placed in a separate room. All pens had concrete floors, covered with wood shavings, which was supplemented regularly but not replaced during the experiment. All groups were visually and physically isolated from each other by plywood walls 130 cm high with wire netting extending to the ceiling. The front of the pens was 65 cm high to allow for behavioural observations. The back wall was of concrete, on which nests were hung, with 6 nest boxes in two levels (Figure 1). Next to the nests were perches with a droppings pit underneath. The 4 perches were wooden measuring 5 cm 5 cm 60 cm and placed in a ladder-like arrangement, 35 cm above each other and separated 30 cm horizontally. At the beginning of the experiment, before the first mite introduction, the hens were not using the perches during the night. Therefore, all hens were manually placed on the perches three times a week, immediately following lights off, to ensure a standard exposure for mites in the Dermanyssus groups. Room temperature was maintained at approximately 23 C by a combination of forced ventilation and thermostatically regulated radiators. Light was provided by 40 watt light bulbs above the partition walls, one for every two pens, providing a light intensity of approximately 15 lux at floor level for 16 h per d. The hens were fed ad libitum a commercial layer mash containing 180 g protein/kg from a round food trough. Water was provided ad libitum from a drinking bowl. Parasites and parasitological techniques Dermanyssus gallinae The two Dermanyssus groups (D1 and D2) were infected twice. The first infection was on d 33, by placing one container with mites at the ends of each of the 4 perches in both pens (in total 8 containers per pen). Approximately mites were introduced to each room in the first infection. Probably due to starvation, most of the mites were dead at the day of introduction. Because mite numbers in the rooms were quite low in the following 3 weeks (Figure 2) a second infection was given on d 56. All mites were taken from a laboratory culture, originally collected in an organic egg production facility in 1997 and since then kept at the Danish Pest Infestation Laboratory. Hiding places for the mites were made out of two pieces of plywood (20 cm

3 28 O. KILPINEN ET AL. Mites/100 ml litter Day Mites in hiding places (mill.) Figure 2. Populations of D. gallinae found in rooms D1 (squares) and D2 (circles). The shaded symbols show the number of mites in the hiding places (right Y-axis). The black symbols show the number of mites per 100 ml litter sample (left Y-axis) next to the perches and the open symbols in litter samples in the front of the room. Mites were first introduced on d 33 and on d 56 additional mites were added (white arrows). On d 75, chemical control was applied in nearby areas as explained in the text (black arrow). 20 cm) with 5 pieces of corrugated cardboard in between. The mite hiding places were covered by wire netting to prevent the birds from damaging them. One hiding place was placed at each end of all perches in all pens. Estimation of mite population In order to determine the total weight (w T )of the mites in the hiding places they were weighed once a week. Three to 5 samples of the mites in the hiding places were collected in Pasteur pipettes and transferred to lactic acid with lignin pink which makes counting the number of different life stages easier. The ratio between the 4 different life stages (d 1 : eggs and unfed stage 1 nymphs; d 2 : fed stage 1 nymphs; d 3 : fed stage 2 nymphs and males; d 4 : adult females) was determined in these samples. Beforehand, the average weight of the same 4 life stages had been determined in the laboratory (w 1 ¼ 0.01 mg; w 2 ¼ 0.04 mg; w 3 ¼ 0.12 mg; w 4 ¼ 0.27 mg). The estimated total number of mites (n) in all hiding places together was determined from the following formula: n ¼ w T =ðd 1 w 1 þ d 2 w 2 þ d 3 w 3 þ d 4 w 4 Þ: In all pens, two litter samples (50 or 150 ml) were collected once a week from two areas: underneath the perches and close to the door in the front of the pen. The mites in these samples were isolated by means of a modified Tullgren funnel, transferred to lactic acid with lignin pink, and counted. When more than 2000 mites were present, smaller subsamples were taken, weighed and counted. This rough estimate of the total mite number was used as a relative measure of the mite burden in the mite groups. In the C- and A-pens the hiding places were opened and checked for mites to ensure that no mites were accidentally introduced into the control and Ascaridia groups. Chemical control From d 75, chemical control methods were applied in an attempt to restrict the movement of mites out of the Dermanyssus room. A barrier (10 cm wide) of alpha-cypermethrin (0.3% in water) was applied with a brush to the walls and the ceiling outside the mite-infested room. Furthermore, a pyrethrin spray (0.36%) was used in adjacent rooms not involved in the experiments, in order to kill mites migrating through the ventilation system. Ascaridia galli The hens of the Ascaridia groups (A1 and A2) were individually inoculated by oral administration of 500 embryonated A. galli eggs in 0.1 N H 2 SO 4 once a week, starting when the hens had been in the experimental pens for 21 d. The hens of the C- and D-groups were sham inoculated with 0.1 N H 2 SO 4. The Ascaridia strain was obtained from a naturally infected farm 20 km south of Copenhagen. Non-embryonated eggs were isolated from the uteri of adult female worms and cultured for 30 d in 0.1 N H 2 SO 4 at 20 C to obtain full infectivity, as described by Permin et al. (1997b).

4 EFFECTS OF ENDO- AND ECTORARASITES 29 Estimation of helminth burden Five faecal samples were collected from each of the 6 experimental groups on the day of the first inoculation and weekly from week 6 after introduction to the pen. Ten weeks after the first inoculation, all hens were killed by decapitation, a faecal sample was collected and the content of the intestine was washed on a 100 mm sieve and transferred to Petri dishes. The Petri dishes were examined for the presence of immature and mature A. galli using a stereo microscope at 40 and 100 times magnification. All faecal samples were analysed for helminth eggs by a McMaster concentration technique (Permin and Hansen, 1999) with a lower detection limit of 20 eggs per g faeces (EPG). Clinical examinations All hens were weighed once a week throughout the experimental period. Three blood samples for haematological analyses were collected from the vena ulnaris of all hens on d 21, 70 and 96. The blood samples were analysed for haemoglobin concentration (HGB), packed cell volume (PCV), number of red blood cells (RBC) and number of white blood cells (WBC). The mean cell haemoglobin (MCH) was calculated from these values as MCH ¼ HGB/RBC, the mean cell haemoglobin concentration (MCHC) as MCHC ¼ HGB/PCV and the mean cell volume (MCV) as MCV ¼ PCV/RBC. A differential count was carried out at the end of the experiment. Egg production Hens of all groups began to lay eggs at 20 weeks of age (2 weeks after arrival), and continued throughout the study. Eggs were collected daily. However, in all groups a number of birds were laying their eggs on the floor and an unknown number of eggs were damaged or eaten. The data on egg production were therefore excluded. Behavioural observations Direct observations during the daytime were by focal animal recordings of selected behaviours according to the definitions of Table 1. The behavioural elements were measured as the number of bouts, that is, the number of times an element appeared, and the amount of time spent on the behaviour in 10 min sessions. Six hens from each group, selected at random at the beginning of the experiment, were observed in each session. A total of 16 h of observations was made on each group. The first observations were 18 d after the hens arrived in the pens and the last were 70 d later. The observations were distributed evenly over early and late afternoons on alternating days, so that the C- and A-groups were observed on the same day, while the D-groups were observed on alternative days. Observations were restricted to the afternoon to exclude interference from nest seeking during egg laying behaviour as far as possible. Both the number of bouts (the number of times a behavioural element appeared) and the amount of time spent on a behavioural element were recorded when possible. Video recordings were made at night with infrared equipment in selected periods. Of these recordings three nights from each group were analysed for nocturnal activity of the hens. Every 15 min the hens in the group under observation were scanned for the following activities: preening (defined as the beak actively in contact with Table 1. Description of elements in the behavioural analysis during the daytime Key Element Description Agg Aggression Pecks directed at the head or threats leading to an avoidance reaction in an opponent. Number recorded. BoSh Body shaking Shaking around the body axis or flapping the wings. DuB Dust bathing Sitting/lying performing activities associated with dust bathing behaviour. Number of vertical wing shakes recorded. FpG Gentle feather pecks Pecks directed at the surface or tip of the feathers which were not seen to move, including pecking at particles on the plumage. Number recorded. FpS Severe feather pecks Pecks by which the feathers were moved or bent as the hen with the beak grasped the feathers and possibly pulled them. Number recorded. HeSc Head scratching Scratching head with one foot. HeSh Head shaking Turning the head quickly from side to side, covering an angle of approximately 180 or more, while keeping the head above the height of the shoulders. Pree Preening Preening own plumage. SoPr Social preening Pecking or nipping at feathers on the neck region approaching frontally and from beneath. Number recorded. Yawn Yawning Opening the beak while head and neck are stretched forward and upwards. Number recorded. InAc Inactive Standing or sitting with the head in the plumage or under the wing possibly with eyes closed. OtBe Other behaviour Any behaviour not covered by the above-mentioned categories.

5 30 O. KILPINEN ET AL. the plumage), sitting or standing on the perch, or showing any kind of movement at all. Termination of the experiment On d 97 all animals were killed by neck dislocation followed by decapitation in accordance with Danish welfare regulations and were subsequently examined for active egg production, pathological changes and parasites. Statistics Blood variables were tested with a General Linear Model and Tukey s multiple comparison ( ¼ 0.05) of the means in each group on separate days. The weight of the hens in the D- and A-groups were tested against the weight of the hens in the C-groups with a repeated measures analysis of variance test. A descriptive analysis of the daytime behavioural data was carried out for number of events and duration of individual behavioural elements within the three experimental groups on the unmodified data. All statistical tests were performed using SAS statistical software. Mixed linear models (SAS, 1997) were used for analysis and to test whether the residuals were normally distributed. RESULTS Dermanyssus gallinae Mites were first introduced to the birds of the D-groups on d 33, 12 d after the Ascaridia inoculations of the A-groups started. However, after 3 weeks, the numbers of mites found in the hiding places and in the litter were low and therefore additional mites were introduced on d 56. After another 2 weeks, the mite population increased rapidly (Figure 2). As the hens avoided sleeping on the perches at the beginning of the experiment, mites were found elsewhere than in the hiding places on the perches. We estimated that approximately 50% of the mites were resting on the ceiling and in other places within the room. Because of the uncertainties regarding the precise mite numbers the values should be taken mostly as a relative measure of the mite burden in each pen. In the C- and A-pens no mites were found until d 77 when one mite was found in the litter samples from pen C2. At the final examination (d 96), 13 mites were found in the hiding places and one mite in the litter of pen C2, and two mites were found in the litter samples of pen A2. Ascaridia galli During the entire experimental period no A. galli eggs were found in the faecal samples from the hens of the control and Dermanyssus groups, while eggs were excreted in the Ascaridia groups from week 10. At slaughter, the mean faecal egg counts (SD) of the hens in A1 and A2 were respectively and eggs per g of faeces. The mean number of worms (SD) of the hens in A1 and A2 were respectively and of which 47% were adult worms. None of the C- and D-group hens were found to have intestinal worms at slaughter. Blood variables The blood samples taken on d 70, just one week before the maximal mite infestation in D2, showed several significant differences between the D-groups and the other groups. By d 70, there was a significant reduction in PCV in the D-groups compared with the C- and A-groups (Figure 3A). Towards the end of the experiment, at d 96, the PCV in the D-groups had increased but remained significantly lower than in the other groups. Similarly, the D-groups showed a significantly lower MCH concentration (Figure 3B) and a significantly larger MCV (Figure 3C) by d 70. But by d 96 these differences had decreased and were no longer significant. The differential count at the end of the experiment showed no significant differences between groups. Mortality In the control group C2 one bird died relatively early on d 42 (no specific pathological findings at post-mortem); in the A-groups no birds died before the experiment was terminated. However, around the peak in the mite infection two (13%) and 5 (30%) acute deaths occurred in groups D1 and D2, respectively. No signs were observed prior to the deaths which could have indicated that those particular birds were under more severe stress than the rest of the birds; all birds were moving around normally. The sudden increase in mortality led to an increased surveillance of the birds with the purpose of terminating the experiment if more deaths occurred. However, the simultaneous reduction in the mite population apparently reduced the pressure on the birds and no more deaths occurred. In the post-mortem examinations performed on all dead birds, the only pathological signs were related to anaemia. A relatively large number of blood-filled mites were found in the trachea on the dead birds in pens D1 and D2. Similar observations are known from veterinarians

6 EFFECTS OF ENDO- AND ECTORARASITES 31 Packed Cell Volume Mean Cell Haemoglobin mmoles/l Mean Cell Volume femto litre a a a a a a bb c c C 21 D A C 70 D A C 96 D A c e a b ab ab e a ab c c c e e d d e e 21 D A C 70 D A C 96 D A Day examining dead hens in mite-infested poultry houses (unpublished observations). Therefore, it was decided to examine the tracheae of all birds at the post-mortem examination at the end of the experiment. However, no mites were found in the trachea of the hens in the mite groups at the final post-mortem examination. This indicates that the mites had moved into the trachea of the hens after they had died, probably in the search for hiding places. Body weights Because there were no significant differences between the replicates within each treatment b b d d d d e e C 21 D A C 70 D A C 96 D A 21 c c e a a a de aa a bb b b de de de d C Figure 3. Average blood variables of the 6 experimental groups, showing packed cell volume, the concentration of haemoglobin in an average blood cell and the volume of an average blood cell. Dark grey bars show the results from groups C1 and C2, white bars from D1 and D2 and light grey bars from A1 and A2. Error bars show standard deviations. Similar letters indicate groups not significantly different (Tukey multiple comparison test, ¼ 0.05). (C1 vs C2, D1 vs D2 and A1 vs A2) the results were pooled and handled as three treatment groups. Figure 4 shows the average weight gain in each treatment group. Both parasites resulted in a significantly lower body weight of the hens (D-groups vs C-groups: P ¼ 0.003; A-groups vs C-groups: P ¼ 0.014). Around the peak of the mite population there was a depression of the average weight of the hens in the D-groups. Some individuals lost more than 100 g during one week, but it was not necessarily those that died around that period. From d 30, approximately one week after the rooms were infected with mites, the average weight of the hens did not increase for the rest of the experiment, whereas the control birds on average gained 159 (81) g in the same period. During the period from d 50 to d 75, the average weight of the hens in the A-groups was 94 g lower than the control groups. Towards the end of the experiment this difference diminished. Behaviour during the daytime Descriptive analysis for the different behavioural elements expressed as number of bouts and minutes of activity per 10 min of observation is shown in Table 2. The mite-infected hens showed higher activity in preening (Pree), head scratching (HeSc) and gentle feather pecking (FpG) behavioural categories, and they performed severe feather pecking (FpS) for longer periods than the control hens and the Ascaridia hens. The control group performed less dust bathing (DuB) as compared with the other two groups and the mite-infected hens performed less other behaviour (OtBe) as compared with the two other groups. Behaviour at night Preening was the only night-time activity where there was a difference between the treatment groups (Figure 5). At the beginning of the experiment, an average of 5 to 29% of the birds in each group was engaged in preening during the night. In the final recording, close to the peak in the mite infestation in the D2-group, the preening activity in the C- and A-groups was reduced to 1 to 5%, whereas the incidence of preening was 18% in D1 and 50% in D2. However, the low number of measurements and the large variation calls for caution in interpretation of the results. DISCUSSION This study has demonstrated that D. gallinae and A. galli can have a serious impact on the welfare

7 32 O. KILPINEN ET AL. Average weight (g) C D A Day Figure 4. Average weight of the hens in each treatment group (circles: C-groups, squares: D-groups and triangles: A-groups). Error bars show the standard error. Shaded arrow indicates the introduction of A. galli to the A-groups and the open arrows indicate the introduction of D. gallinae to the D-groups. Finally, the black arrow indicates the application of chemical control around the D-groups. The levels of significance for the differences between the C-groups and the D- and A-groups were and 0.014, respectively (SAS GLM procedure, repeated measures). Table 2. Behaviour patterns in the three experimental groups of birds measured in number of bouts per 10 min and in minutes of activity per 10 min. Averages (and standard deviations) measured over the entire experimental period. Similar letters indicate groups which are not significantly different (General Linear Model with Tukey multiple comparison test, ¼ 0.05) Behaviour No. of bouts Minutes of activity/10 min Control Dermanyssus Ascaridia Control Dermanyssus Ascaridia Agg 1.03 (0.23) 1.10 (0.12) 1.15 (0.25) 0.02 (0.01) 0.03 (0.01) 0.03 (0.01) BoSh 0.56 (0.08) 0.65 (0.09) 0.60 (0.08) 0.02 (0.00) 0.02 (0.00) 0.02 (0.00) DuB FpG a (0. 00) 6.70 (3.19) b (0. 23) (5.03) b (0. 23) 1.34 (0.40) (0. 03) 0.14 (0.06) (0. 15) 0.27 (0.09) (0. 14) 0.04 (0.01) FpS 2.11 (1.12) 2.42 (0.98) 0.73 (0.46) cd 0.04 (0.02) c 0.07 (0.03) d 0.02 (0.01) HeSc HeSh b 1.01 (0.11) 7.93 (l.06) a 2.27 (0.29) 9.05 (l.00) b 1.02 (0.13) 7.34 (l.14) d 0.05 (0.01) 0.15 (0.02) c 0.10 (0.02) 0.17 (0.02) d 0.04 (0.01) 0.14 (0.02) Pree b 1.95 (0.23) a 3.15 (0.35) b 1.93 (0.25) cd 1.22 (0.20) c 1.74 (0.21) d 1.05 (0.21) SoPr 2.95 (l.11) 2.61 (l.20) 4.48 (3.70) 0.09 (0.03) 0.05 (0.02) 0.05 (0.03) Yawn (0. 04) (0. 02) (0. 19) InAc (0. 00) 0.37 (0.14) (0. 00) 0.08 (0.04) (0. 00) 0.17 (0.06) OtBe d 7.86 (0.23) c 7.15 (0.29) d 8.13 (0.29) Total and health of poultry, most clearly seen in the increased mortality in the mite-infected hens. The lower PCV, the lower concentration of haemoglobin in the red blood cells and the increased volume of the blood cells indicate that the Dermanyssus-infected hens were suffering from regenerative anaemia (Kerr, 1989) with the birds losing blood faster than the rate of haematopoesis (young immature RBC are large and contain little haemoglobin). Thus it seems likely that the deaths were caused by the severe anaemia from large numbers of blood-sucking mites. A. galli in contrast, had no significant impact on the blood variables. Even the initial low mite infestation depressed the weight gain of the young birds, whereas the birds in the control groups gained 159 g (81 g) on average during the last 9 weeks (from d 35) of the experiment. A significant effect of A. galli on weight gain was also observed. After the initial infection, the weight gain was reduced and remained 80 to 100 g below the control groups for the rest of the experimental period. The increase in self-grooming as well as in social feather pecking activity are important as indicators for the affected welfare of the hens. It is possible that the disturbance of normal behaviour during both day and night may have contributed to the reduced weight gain. The mite-infected birds were in general less inactive than the two other groups. Although this difference was not statistically significant it may have played a role in relation to the body

8 EFFECTS OF ENDO- AND ECTORARASITES 33 Active birds 60% 50% 40% 30% 20% C1 D1 A1 C2 D2 A2 10% 0% Day Figure 5. The percentage of birds engaged in light feather pecking activity during the night (circles: C-groups, squares: D-groups and triangles: A-groups, open symbols show the results from C1, D1 and A1, and solid symbols from C2, D2 and A2). weight. The reduced incidence of Other Behaviour in mite-infected birds as compared to the two other groups is probably a direct effect of the increased preening and feather pecking. Dust bathing activity was more pronounced in both mite- and worm-infected birds as compared to the control birds. Removal of ectoparasites has been suggested as a primary function of dust bathing (Simmons, 1964). This however has never been experimentally proved. The increased dust bathing activity in the worm-infected hens is more difficult to explain. It may, however, be hypothesised that worm-infected birds for some reasons have a less perfect plumage, thereby stimulating dust bathing behaviour, because dust bathing has been demonstrated to improve feather structure (Healy and Thomas, 1973). Despite the uncertainty regarding mite numbers in this study, we believe that it is realistic that the serious health problems appeared when the infection level reached to mites per bird. It is clear that when fatal cases appeared the mite populations were increasing rapidly (Figure 2). The data on the blood variables show that the hens were also increasing their production of new red blood cells. However, it is possible that the hens died because the mite populations and thereby the blood loss were increasing faster than blood production could increase. Furthermore, the video recordings demonstrated that the birds were severely disturbed during the night by the attacking mites. On average, the birds spent half the night in light feather pecking activity. Altogether, these observations indicate that major problems occur when birds are under continuous stress, leaving no time for rest and when the blood loss due to blood-sucking mites exceeds the increased blood production. The drop in the mite population observed towards the end of the experiment is probably due to the chemical control applied in the neighbouring rooms. Despite the great care taken to avoid spread of the pyrethrin spray applied in the adjacent rooms, it might have been transported through the ventilation system and diffused from the installation passages, which were all placed in pen D2. This may explain the major drop in the mite population in the litter of this particular pen. The mites in the hiding places were probably more protected from the volatile chemicals and therefore continued to increase for longer. At present, no information is available on the number of mites present in commercial egg production units, however, our subjective judgement was that the infestation level in the present experiment seemed comparable to a poultry house under severe attack. Under ideal conditions, D. gallinae feed every 2 to 3 d, but this depends on factors such as temperature and availability of hosts (Maurer and Baumgärtner, 1992). Furthermore, variation in the infestation of individual birds seems likely, due to the clumped distribution of mites in the pens and because of differences in the attractiveness of individual birds to mites. Therefore, even with a more precise measure of the mite population, it would be difficult to estimate the mite burden of individual birds. Previous studies on the effect of D. gallinae on laying hens have produced variable results. Kirkwood (1967) showed that exposing adult cockerels to around 1 million mites each resulted in almost 100% death. In contrast, Maurer et al. (1993b) found no negative influence on either blood variables or egg production of 40 adult hens in a room with around 4 million poultry red mites. A reason for this could be that the mite population was increasing slowly so that the hens were able to compensate by increasing blood production. In a field study in Sweden,

9 34 O. KILPINEN ET AL. Höglund et al. (1997) compared various blood and morphological variables of hens on two farms, one of them infested with D. gallinae. On both farms hens were kept in cages and in a freerange system. Packed cell volume was 14% lower among the hens in cages on the farm with poultry red mites, but between the hens in the free-range systems no differences were found. The major problem with such field studies is that differences in the management systems between farms can have an important influence on the physiological conditions of the hens. The major concerns arising from the present work are the reduced weight gain due to the presence of D. gallinae and A. galli as well as the anaemia and increased mortality due to high numbers of blood-sucking mites. The reduced weight gain particularly highlights the importance of avoiding major infections with poultry red mites and A. galli during the growth period of chickens and both effects call for effective treatments against parasites. At present, some European countries have no acaricides at all registered for application against D. gallinae, and in general the number of registered anti-parasitic compounds is decreasing all over Europe. The growing number of organic farming systems presents a particular problem since they are not allowed to use many of the remaining antiparasitic products and they are not allowed to use preventive treatments. ACKNOWLEDGEMENT Financial support by a grant from the Danish Environmental Research Programme is highly appreciated. We would like to thank Nicolai Hansen, Niels Midtgaard, Jørgen Olesen, Margrethe Pearman and Rikke Frahm Lundvig for technical assistance. These experiments comply with the current Danish laws of animal care. REFERENCES ACKERT, J.E. (1931) The morphology and life history of the fowl nematode Ascaridia lineata (Schneider). Parasitology, 23: ACKERT, J.E. & HERRICK, C.A. (1928) Effects of the nematode Ascaridia lineata (Schneider) on growing chickens. The Journal of Parasitology, 15: APPLEBY, M.C. & HUGHES, B.O. (1997) Animal Welfare (Wallingford, CAB International). ARENDS, J.J. (1991) External parasites and poultry pests, in: CALNEK, B.W. (Ed) Diseases of Poultry, pp (London, Wolfe Publishing). BROWN, N.S. (1974) The effect of louse infestation, wet feathers and relative humidity on the grooming behaviour of the domestic chicken. Poultry Science, 53: CHIRICO, J., ERIKSSON, H., FOSSUM, O.& JANSSON, D.(2003) The poultry red mite, Dermanyssus gallinae, a potential vector of Erysipelothrix rhusiopathiae causing erysipelas in hens. Medical and Veterinary Entomology, 17: DURDEN, L.A., LINTHICUM, K.J. & TURELL, M.J. (1992) Mechanical transmission of Venezuelan equine encephalomyelitis virus by hematophagous mites (Acari). Journal of Medical Entomology, 29: HEALY, W.M. & THOMAS, J.W. (1973) Effects of dusting on plumage of Japanese quail. Wilson Bulletin, 85: HÖGLUND, J., NORDENFORS, H. & UGGLA, A. (1995) Prevalence of the poultry red mite, Dermanyssus gallinae, in different types of production systems for egg layers in Sweden. Poultry Science, 74: HÖGLUND, J., NORDENFORS, H. & UGGLA, A. (1997) Infektionsvägar, skadeverkningar och bekämpning av blodsugande kvalster i svenska värphönsanläggningar. Meddelanden från parasitologen, SVA, Sweden. IKEME, M.M. (1971) Observations on the pathogenicity and pathology of Ascaridia galli. Parasitology, 63: KERR, M.G. (1989) Veterinary Laboratory Medicine (Oxford, Blackwell Scientific). KILPINEN, O. (2000) Problems caused by the poultry red mite, Dermanyssus gallinae, in the Danish egg production, in: GOOD, M., HALL, M.J., LOSSON, B., O BRIEN, D., PITHAN, K.& SOL, J. (Eds) Mange and Myiasis of Livestock, pp (Brussels, Cost 833, Commission of the European Communities). KIRKWOOD, A.C. (1967) Anaemia in poultry infected with red mite Dermanyssus gallinae. Veterinary Record, 80: MAURER, V. & BAUMGÄRTNER, J. (1992) Temperature influence on life table statistics of the chicken mite Dermanyssus gallinae (Acari: Dermanyssidae). Experimental & Applied Acarology, 15: MAURER, V., BAUMGÄRTNER, J., BIERI, M.& FÖLSCH, D.W. (1993a) The occurrence of the chicken mite Dermanyssus gallinae (Acari: Dermanyssidae) in Swiss poultry houses. Mitteilungen der Schweiziche Entomologishe Geselschaft, 66: MAURER, V., FÖLSCH, D.W. & BAUMGÄRTNER, J. (1993b) Untersuchung der Populationsdynamik der Roten Vogelmilbe im Hinblick auf die Entwicklung von nachaltigen Bekämpfungsmethoden. Landwirtschaft Schweitz, 6: PERMIN, A.& HANSEN, J.W. (1999) The Epidemiology, Diagnosis and Control of Parasites in Poultry (Rome, FAO). PERMIN, A., BOJESEN, M., NANSEN, P., BISGAARD, M., FRANDSEN, F. & PEARMAN, M. (1997a) Ascaridia galli populations in chickens following single infections with different dose levels. Parasitology Research, 83: PERMIN, A., PEARMAN, M., NANSEN, P., BISGAARD, M. & FRANDSEN, F. (1997b) An investigation on different media for embryonation of Ascaridia galli eggs. Helminthologia, 34: PERMIN, A., BISGAARD, M., FRANDSEN, F., PEARMAN, M., NANSEN, P.& KOLD, J. (1999) The prevalence of gastrointestinal helminths in different poultry production systems. British Poultry Science, 40: SAS (1997) SAS/STAT Software: Changes and Enhancement through Release 6.12, pp (Cary, NC, SAS Institute Inc.). SIMMONS, K.E.L. (1964) Feather maintenance, in: THOMPSON, A.L. (Ed) A New Dictionary of Birds, pp (New York, McGraw-Hill). SPRUIJT, B.M., VAN HOOFF, J.A.R.A.M. & GISPEN, W.H. (1992) Ethology and neurobiology of grooming behavior. Physiological Review, 72: ZEMAN, P., STIKA, V., SKALKA, B., BÁRTÍK, M., DUSBÁNEK, F. & LÁVICKOVA, M. (1982) Potential role of Dermanyssus gallinae de Geer, 1778 in the circulation of the agent of pullurosis-typhus in hens. Folia Parasitologica (Praha), 29: