EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES IN SWEDISH MOOSE Margareta Stéen1, Ing-Marie Olsson Ressner2, Bodil Olsson3, and Erik Petersson4 1Department of Anatomy, Physiology and Biochemistry, Swedish University of Agricultural Sciences, P. O. Box 7090, SE-750 07 Uppsala, Sweden; 2Swedish Chemicals Agency (KemI), P. O. Box 2, SE-172 13 Sundbyberg, Sweden; 3TNS Sifo, P.O. Box 115 00, SE-404 30 Gothenburg, Sweden; 4Department of Aquatic Resources, Swedish University of Agricultural Sciences, SE-178 93 Drottningholm, Sweden ABSTRACT: A total of 961 harvested and 241 unharvested moose (Alces alces) carcasses and parts from throughout Sweden were examined for Elaphostrongylus alces from 1985 to 1989. When avail- able, the central nervous system and skeletal muscles were searched for adult nematodes, and lungs and feces were examined for first-stage larvae. The parasite was distributed throughout Sweden with highest prevalence (56%) in the central region and lowest in the south (13%). Prevalence was highest in calves and old moose (>9 years) and lowest in middle-aged animals (5–9 years), with no statistical difference between sexes, although prevalence trended higher in young males. Body condition and abundance of Elaphostrongylus alces were negatively correlated, and condition was poorer in unhar- vested than harvested moose. A short (39–73 days) prepatent period was documented, and calves as young as 1.5 months were infected. These results indicate the importance of continued surveillance of Elaphostrongylus alces, particularly because a warming climate will likely increase abundance of intermediate mollusk hosts and possibly cause increased infection of moose. ALCES VOL. 52: 13–28 (2016) Key words: Alces alces, climate, body condition, Elaphostrongylus alces, intermediate host, gastropods, moose, prepatent period, protostrongylidae, Sweden The moose (Alces alces) population in Scandinavia began to rise in the 1970s, peaking in the mid-1980s in Sweden. With few large predators at that time, it was not unusual to find dead or sick animals (Hörnberg 2001, Stéen et al. 2005), and in the 1980–1990s, high mortality was noted in both Swedish and Norwegian moose, as well as in semi- domestic reindeer (Rangifer tarandus). A pre- viously unknown disease, elaphostrongylosis (Stéen and Rehbinder 1986, Stuve 1986), was reported in the 1980s and sick animals were characterized by locomotive abnormalities such as ataxia, incoordination, swaying of the hindquarters, broad and stamping gait, and a certain way of hypermetria that sug- gested paralysis of ascending proprioceptive nerve fibers (Stéen and Roepstorff 1990). A previously undescribed species of elaphostron- gyline nematode with a dorsal-spine larva, Ela- phostrongylus alces (Stéen et al. 1989) was invariably associated with sick and dead moose (Stéen and Rehbinder 1986). Parasites of the genera Parelaphostron- gylus and Elaphostrongylus belong to the subfamily Elaphostrongylinae (Protostrongy- lidae, Metastrongyloidea, Nematoda). Species of the genus Parelaphostrongylus (P. tenuis, P. odocoilei, P andersoni) affect the central nervous system (CNS) and skeletal muscle Corresponding author: Margareta Stéen, Department of Anatomy, Physiology and Biochemistry, Swedish University of Agricultural Sciences, PO. Box 7068, SE-750 07 Uppsala, Sweden, margareta.steen@slu.se 13 mailto:margareta.steen@slu.se fasciae of Nearctic cervids in North America including white-tailed deer (Odocoileus vir- ginianus), black-tailed deer (O. hemonious hemonious), mule deer (O. h. columbianus), and occasionally wapiti (Cervus canadensis) and moose (Alces alces spp.). Species of Elaphostrongylus (E. alces, E. cervi, E. panticola, E. rangiferi) affect the CNS, the peripheral nerve system (PNS), and the skel- etal muscle fasciae in Eurasian cervids in- cluding moose, red deer (Cervus elaphus), maral deer (C. e. sibiricus), roe deer (Capreo- lus capreolus), and reindeer (Lankester 2001). In the New World, as in the Old World, central nervous disorders and mortal- ity occur in wild cervids infected with ela- phostrongyline nematodes (Anderson 1964, Lankester 2001, 2010). Representatives of the genera Elaphostrongylus and Parela- phostrongylus are also harmful to domestic ruminants (Lankester 2001). Both Elaphostrongylus spp. and Parela- phostrongylus spp. develop from the first to third larval stage (L1–L3) in their gastropod intermediate host, and develop from the L3 to the adult (L5) stage in their cervid (final) host (Olsson et al. 1998, Olsson 2001). Specific identification of adult protostrongylids and first-stage larvae (L1) in feces of Swedish moose was a result of multiple studies. The morphology of E. alces was initially described by Stéen et al. (1989) and (Stéen and Johansson 1990), and subsequent com- parison of specific proteins in protostrongylid L1 indicated that L1 and adult E. alces had the same protein pattern in moose, but dif- fered from the L1s and adult protostrongylid parasites in other wild ruminants (Stéen et al. 1993). Experimental infection of captive moose indicated that L1 collected from wild moose caused elaphostrongylosis, and L1 excreted from infected and sick moose and transmitted to terrestrial snails (Arianta arbustorum) in which larvae develop (Lankester et al. 1998), were identified as E. alces using genomic DNA (Gajadhar et al. 2000) and single-strand conformation poly- morphism (SSCP) analysis (Chilton et al. 2005, Huby-Chilton et al. 2006). Collective- ly, these studies indicate that protostrongylid larvae in Swedish moose are E. alces. Given the prevalence and deleterious effect of this disease, our objective was to determine if E. alces is related to age, sex, condition, and geographic distribution of moose in Sweden. STUDY AREA Sweden was divided into 6 regions from the far north (69°03′36″N 20°32′55″E) to the far south (55°20′13″N 13°21′34″E) to de- termine the distribution and prevalence of elaphostrongylosis (Fig. 1). Sweden is shel- tered by the Scandinavian mountains and has a continental climate with large differ- ences in temperature and precipitation be- tween summer and winter, and a relatively small amount of precipitation (Swedish Meteorological and Hydrological Institute [SMHI]). Summer temperatures are similar to those in North America and Asia at similar latitude, although due to the Gulf Stream, winter in Sweden is typically milder (SMHI 2015). METHODS Hunting begins on the first Monday of September in northern Sweden, on the second Monday of October in the south, and seasons end in December or January (Swedish Asso- ciation for Hunting and Wildlife Manage- ment 2015). Prior to the hunting seasons (1986 and 1987), we sent hunters report cards (hunting site, sex, and approximate age) and wrapping materials to pack body parts (i.e., lungs, feces, spinal cords, and mandibles). Moose carcasses and parts (n = 1137) were examined in 5 consecutive years (1985– 1989); 896 (79%) were associated with har- vested moose (1986, 1987, 1989) and 241 (21%) were from non-harvested animals (i.e., euthanized or found dead; 1985–1989). 14 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 We used 1020 lungs and 1084 fecal samples to identify presence of elaphostrongylid L1, 655 spinal cords (membranes) to identify presence of adult worms, and 636 mandibles to measure fat content (Table 1). Age was determined by dental wear (Gasaway et al. 1978) or from information provided by hunters; 151 animals were not aged due to lack of information. Five age classes were established: 1) calves were ≤12 months), 2) yearlings were >12 and ≤24 months, 3) young animals were >24 months and ≤5 years, 4) middle-aged animals were >5 and ≤9 years, and 5) old animals were >9 years. Sex was determined from the whole carcass or hunter information. Evaluation of body condition was by vis- ual inspection, location, and appearance of body fat (n = 948), and/or by measuring fat content (%) in mandible bone marrow (n = 636; Engelsen Etterlin et al. 2009). Three categories of condition were established: BD AC Z Y X W S T U C B D EO M H F Arc�c Circle 66°33'39" N Treriksröset 69°03'36''N, 20°32'55''E Smygehuk 55 °20'13''N, 13°21'34'' Region 1: n= 135 Region 2: n=558 Region 3: n=37 Region 4: n=100 Region 5: n=40 Region 6: n=26 N K G Fig. 1. Map of Sweden with county codes and 6 regions: Region 1 = The Laplandic counties (AC, Z, BD, and Y), Region 2 = southern part of Norrland (W, S, and X county), Region 3 = northern Svealand (C, U, T, and B county), Region 4 = southern Svealand and Northern Götaland (D, P, R, E, and O county), Region 5 = Småland (F, H, and G county), and Region 6 = southern Götaland (N, L, K, and M county). Dots represent locations of moose infected with Elaphostrongylus alces; n = harvested moose. ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 15 normal, poor (below normal), and emaciated (lack of adipose tissue). The fat content in bone marrow was measured with standard techniques under specified assay conditions and techniques (NMKL No 131, Nordic Committee on Food Analysis 1989) and also used to assign condition: normal = 75– 94%, poor = 16–<75%, and emaciated = 0.4–<16% fat content. Assigning condition from visual inspection (without measuring fat content) was considered reliable because of the strong correlation between the condi- tion category assigned from fat measure- ments and visual inspection of the same animals (Rs = −0.801, P < 0.001, n = 592). Bodies/parts were inspected for adult E. alces worms and L1 with necropsy proce- dures described previously (Stéen and Rehbinder 1986, Stéen et al. 1997, 1998) and included examination of muscle fasciae, the cranial cavity, brain, and spinal cord membranes and epidural space of the spinal cord (Stéen and Rehbinder 1986, Stéen et al. 1997). Lungs from all animals were palpated and inspected for nodules, and 20 g samples of minced lungs and feces were processed to detect L1 (Baermann 1917). L1s were identified as protostrongylids quantified in a counting chamber under a stereo micro- scope, expressed as larvae per gram of wet feces (lpg), and classified into 7 levels of relative abundance ranging from none (0) to heavy (6) (National Veterinary Institute, Sweden). There were 4 categories of infec- tion: 0 = uninfected, 1 = in the epidural space but not in lungs or feces, 2 = in lungs but not feces, and 3 = in feces. Animals were categorized as either infected or unin- fected (presence or absence of L1 and/or E. alces worms) for certain statistical com- parisons (e.g., sex or age groups, prevalence in population or region), Data management and statistics Data were tested for normal distribution and seasonal variation, and if not normally distributed, normality was achieved with log-transformation. A peak function analysis was used to identify the best fit to the rela- tionship between bone marrow fat and sea- son (TableCurve software, Systat 2002). A mean value was calculated for harvested ani- mals and this value was applied together with the individual values for remaining Table 1. Total number of moose (harvested/unharvested) and sample location/type – epidural space of the spinal cord (epidural), lungs, feces, mandibles – used to study Elaphostrongylus alces in Sweden, 1985–1989. Moose Epidural Lungs Feces Mandibles Sex Males 386/87 173/85 343/84 369/85 260/36 Females 456/147 223/144 404/145 427/145 262/57 Unknown 54/7 23/7 37/7 49/7 21/– Age group Calves 457/107 187/103 392/102 434/103 270/48 Yearlings 227/36 113/36 52/12 220/36 182/11 Young 40/27 23/27 39/27 38/27 36/10 Middle aged 31/19 27/19 28/19 26/19 30/5 Old 6/37 4/37 6/37 6/37 5/16 Unknown 135/14 65/13 110/14 123/14 20/1 Total 896/241 419/236 784/236 847/237 543/93 16 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 animals. The residuals for all animals were calculated (harvested moose were not com- bined as above), and adjusted values were calculated by adding the residual to the com- mon mean. This produced a few values >100% that were not further corrected in subsequent analyses. Bone marrow fat (adjusted for seasonal variation) was subse- quently analysed using generalized linear models. Body condition was also analysed with generalized linear models, modeling the probability of being in normal condition (see above) assuming a binary distribution of the response variable. The total parasite infection or parasites found in either the lungs, feces, or in the epidural space were similarly corrected, and the probability of being infected was tested with respect to 3 predictors (age, sex, region). The age when calves were infected was estimated with birth date information from each county. Comprehensive data were available from 5 counties: Västerbotten (AC in Region 1), Västra Götaland (O in Re- gion 4), Kalmar (H in Region 5), Kronoberg (G in Region 5), and Södermanland (D in Region 4) (Fig. 1). In 3 counties (H, G, and D) the mean value + SD (Malmsten 2014) was used as the birth date, and in 2 counties (AC and O) the mean value + SD was esti- mated (Broberg 2004). Birth dates for the counties without data were estimated using a multiple imputation (PROC MI in SAS statistical software, SAS 2014) with a Markov chain Monte Carlo method in which longi- tude and latitude of resident cities were used with the number of imputations set to 60. Other than 3 counties with a minor inconsist- ency (3–4 days), the approach produced an acceptable trend of earlier birth dates in southern Sweden, and the dates corresponded well with the span of birth dates reported by a national hunting organization (Swedish Asso- ciation for Hunting and Wildlife Management 2015) (Table 2). The mean category of infection (0–3) in each age group was calculated to illustrate the relationships among age (mean age of group), category of infection, and body condition. These values were used to de- velop a contour graph using SigmaPlot soft- ware (Systat 2008) where body condition, age group, and infection category were interpolated. RESULTS Infection, age and sex Age of moose was skewed towards young animals (Table 1), and age in the two groups (harvested and unharvested) was not distributed evenly. Unharvested moose were older than those harvested for combined age classes, calves, and by sex (Table 3). The average age of harvested ani- mals (n = 761) was 10.4 months (95% CI = 9.6 – 10.4; range = 0–15 years), and 22.3 months (CI = 17.4–28.4; range = 0–20 years) for unharvested animals (n = 227). Females were older in the yearling, middle-aged, and combined age groups. A slight majority (57%) of the harvested sample (n = 896) was infected with L1 and/or adult E. alces worms. The prevalence was similar between sexes in each age class for L1 in lungs, L1 in feces, and adult worms in the epidural space of the spinal cord (Fig. 2). There was a tendency (P = 0.074) toward higher prevalence in males than females in the young age class. Worms were found in the epidural space of the spinal cord in animals 3 months to 2 years old, but not in animals 3 to 9 years old; worms were found in a single 10-year old moose. The abun- dance of L1 in lungs (n = 784) was high in calves and yearlings, lower at 3–4 years of age, and minimal in adults. Nearly the entire sample (98%) of unhar- vested moose (n = 241) was infected with E. alces (Fig. 3). Worms were found in the epidural space of the spinal cord in 3 month to 4 year-old animals. The average age of ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 17 infected calves was 4.8 months (95% CI = 4.7–4.9). No worms were found in the epi- dural space of the spinal cord in 5–9 year- old moose, but worms reappeared at 10–16 years of age. The prevalence of adult worms in the epidural space of the spinal cord was 36% in the combined data (harvested and unhar- vested, n = 655); the prevalence of L1 in lungs (n = 1020) and feces (n = 1084) was 64 and 53%, respectively. The prevalence (worms/ L1) was 66% overall; 88% in old moose, 74% in yearlings, 67% in calves, 55% in young, and 48% in middle-aged animals. There were differences (P < 0.001) in fre- quency of infection among age groups; the oldest animals had the highest frequency of infection (L1) and the middle-aged the lowest. The frequency of worms in the epidural space of the spinal cord was high in calves/ yearlings, leveled out at 4 years, and then was not identified until 10–16 years at low frequency. The abundance of L1 in lungs of old animals was at the highest level (6). Body condition Body condition of harvested animals (n = 981) was either normal (40% overall, 24% calves) or poor (59%, 75% calves). In unhar- vested moose (n = 227), body condition was normal in 38% overall, with calves and old animals lower; 25% calves, 45% young, and 29% old animals were in normal condition. For all moose, body condition and cat- egory of infection were correlated (Rs = 0.215, P < 0.001). In separate age classes, this Table 2. Prevalence of Elaphostrongylus alces (adjusted for Julian date and age of the sampled moose) and birth date of moose in Swedish counties. Birth dates marked with an asterisk are observed values; others are estimated (see Data management and statistics). Region County Mean prevalence (%) N Birth date (Julian date) Birth date 1 AC Västerbotten 43.8 82 167* 16 June 1 BD Norrbotten 51.4 13 168 17 June 1 Z Jämtland 53.5 8 171 20 June 1 Y Västernorrland 58.2 21 164 13 June 2 W Dalarna 63.2 32 160 9 June 2 X Gävleborg 67.7 457 157 6 June 2 S Värmland 49.4 36 159 8 June 3 B Stockholm 61.7 2 149 29 May 3 C Uppsala 100.0 3 152 1 June 3 T Örebro 67.4 17 156 5 June 3 U Västmanland 79.3 15 154 3 June 4 D Södermanland 100.0 5 148* 28 May 4 E Östergötland 27.6 12 150 30 May 4 O Västra Götaland 53.9 12 153* 2 June 5 F Jönköping 39.1 16 151 31 May 5 G Kronobergs 28.2 10 144* 24 May 5 H Kalmar 32.2 11 143* 23 May 6 K Blekinge 37.4 5 140 20 May 6 M Skåne 0 8 144 24 May 6 N Halland 41.5 12 146 26 May 18 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 correlation was found in yearlings (Rs = 0.262, P < 0.001, n = 239), young (Rs = 0.463, P < 0.001, n = 67), middle-aged (Rs = 0.441, P = 0.002, n = 47), and old animals (Rs = 0.456, P = 0.003, n = 40), but not in calves (Rs = 0.054, P = 0.239, n = 471). For all moose, body condition was correlated inversely with category of infection (Rs = 0.084, P = 0.025, n = 721); separate correlations were found in yearlings (Rs = 0.213, P = 0.002, n = 227) and young animals (Rs = 0.398, P = 0.011, n = 40). Figure 4 illustrates the prob- ability of normal body condition relative to age and category of infection, indicating that calves have poor body condition regardless of category of infection, and that some middle-aged animals have normal body condition despite high abundance of L1 in feces. Old individuals were generally in nor- mal body condition if not infected, although few were without infection. Bone marrow fat content (n = 615) var- ied annually (Table 4, Fig. 5). On average, harvested animals had higher fat content (93%, 95% CI = 91 – 96) than unharvested animals (70%, CI = 66 – 75) with values cor- rected for time of year, sex, and age class (Table 4). In a combined sample, a negative correlation was found between bone marrow fat content and category of infection (Rs = −0.212, P < 0.001, n = 635). This negative correlation was found in calves (Rs = −0.131, P = 0.020, n = 319), yearlings (Rs = −0.223, P = 0.002, n = 193), young (Rs = Table 3. Age in months (mean and 95% CI) of Swedish moose examined for Elaphostrongylus alces, 1985–1989. The column to the far right gives level of significance between harvested and euthanized + dead moose (for the last three rows the t-tests are performed on log transformed data; the data presented in the table are back-transformed values). If all sexed animals are combined, the sexes differed in age (P < 0.05). Age class Sex Harvested Unharvested t-value Calves Females 4.3 (4.1–4.4) 8.5 (8.1–8.8) 22.3, P < 0.001 Males 4.2 (4.0–4.4) 8.1 (7.8–8.4) 21.1, P < 0.001 All calves‡ 4.2 (3.5–5.0) 8.3 (6.6–9.9) 4.28, P < 0.001 Yearlings* Females 18.4 (17.7–19.0) 19.2 (17.5–20.9) 0.91, P = 0.362 Males 17.7 (17.0–18.4) 17.3 (15.6–19.0) 0.36, P = 0.716 Combined 18.0 (16.9–19.1) 18.3 (15.4–21.1) 0.17, P = 0.863 Young Females 46.5 (42.8–50.4) 47.4 (43.0–51.7) 0.28, P = 0.782 Males 42.0 (36.9–47.1) 44.6 (37.4–51.8) 0.58, P = 0.562 Combined‡ 44.7 (42.0–47.4) 46.2 (42.9–49.5) 0.70, P = 0.481 Middle-aged* Females 90.3 (85.0–95.5) 93.4 (86.7–100.2) 0.74, P = 0.461 Males 84.0 (74.4–93.6) 76.8 (65.5–88.1) 0.98, P = 0.333 Combined‡ 89.0 (86.0–92.1) 89.1 (85.1–93.0) 0.01, P = 0.994 Old Females 144.0 (106–181.7) 154.2 (141.3–167.2) 0.52, P = 0.606 Males 120.0 (144.6–195.4) 144.0 (100.5–187.5) 0.56, P = 0.580 Combined‡ 146.0 (139.1–152.9) 153.4 (150.6–156.2) 1.94, P = 0.053 All females§ 12.3 (10.7–14.1) 29.6 (21.0–41.6) 4.73, P < 0.001 All males§ 9.0 (8.0–10.0) 14.2 (10.7–18.7) 3.92, P = 0.004 All moose 10.4 (9.6–11.4) 22.3 (17.4–28.4) 5.74, P < 0.001 *Sexes differ by age class (all causes of death included). ‡Includes individuals not sexed. §Sexes differ with all age classes combined. ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 19 −0.618, P < 0.001, n = 46), and old (Rs = −0.736, P < 0.001, n = 21), but not middle- aged moose (Rs = −0.319, P = 0.062, n = 35). Time of infection The earliest identification of a calf diag- nosed with elaphostrongylosis was at ~1.5 months on 21 July in Region 3, County of Calves Yearlings Young Mid. age Old 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Epidural space Fr eq u en cy 0.0 0.2 0.4 0.6 0.8 1.0 Females Males Lungs 0.0 0.2 0.4 0.6 0.8 1.0 Feces Total E. alces Fig. 2. The mean abundance (frequency) of Elaphostrongylus alces measured in adult worms in the epidural space of the spinal cord, and larvae in lungs and feces of harvested moose, Sweden, 1985– 1989. The values are least-squared means and 95% confidence limits from a generalized linear model. The interaction between age groups and sex was a categorical predictor and Julian date was a continuous predictor. 20 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 Uppsala (Table 5). The abundance of L1 was category 6 in the lungs and 4 in feces, and the calf was in normal body condition. The earliest calf death where worms were found in the epidural in the spinal cord was on 10 October (~4 months old) in Region 2, County of Värmland; the abundance of L1 was category 6 in the lungs and 3 in feces. Worms were first found in harvested calves on 15 September (~3 months old). Fig. 4. Contour plot of the probability of normal body condition in moose (n = 613) versus age and 3 categories of Elaphostrongylus alces infection intensity, Sweden, 1985–1989. Ca lve s Ye arl ing s Yo un g Mid dle ag ed Old P ro po rt io n (% ) 0 20 40 60 80 100 Uninfected E. alces worms in epidural space only E. alces L1 in lungs E. alces L1 in feces H HH H HD D D D D Fig. 3. The 4 categories of Elaphostrongylus alces infection within 5 age groups of Swedish moose, 1985–1989. H bars represent harvested moose (n = 761) and D bars represent unharvested moose (euthanized or found dead; n = 227). ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 21 Table 4. Analysis of bone marrow fat (%) and body condition of harvested and unharvested moose, Sweden, 1985–1989. Values are mean ± SE with sample size in parentheses. Pair-wise comparisons (t-values) of harvested and unharvested animals are provided in each age group (raw). Means denoted by the same letter in each column (percent fat and body condition separately) were not different (P < 0.05). The values for percent fat are adjusted for time of year causing certain values to be >100% (see Data management and Statistics). Variable Age class Harvested Euthanized or dead Statistic Bone marrow fat (%) Calves 83.6 ± 0.8a (270) 60.5 ± 1.8a (48) t = 11.6 P < 0.001 Yearlings 97.1 ± 0.9b (182) 69.9 ± 3.8b (11) t = 6.88 P < 0.001 Young 98.7 ± 2.1b (36) 83.6 ± 4.0c (10) t = 3.32 P < 0.001 Middle aged 99.6 ± 2.3b (30) 77.4 ± 5.7bc (5) t = 3.62 P < 0.001 Old 104.6 ± 5.7b (5) 63.3 ± 3.2ab (16) t = 6.34 P < 0.001 All animals 96.7 ± 1.3 (523) 70.9 ± 1.7 (92) t =11.6 P < 0.001 Probability of normal condition Calves 0.17 ± 0.03a (368) 0.25 ± 0.06a (102) z = 0.81 P = 0.420 Yearlings 0.48 ± 0.04b (204) 0.30 ± 0.12c (35) z = 1.19 P = 0.235 Young 0.50 ± 0.11bc (40) 0.47 ± 0.15b (27) z = 0.25 P = 0.800 Middle aged 0.74 ± 0.12c (30) 0.64 ± 0.21b (17) z = 0.34 P = 0.734 Old 0.84 ± 0.17abc (6) 0.12 ± 0.06ab (34) z = 2.53 P = 0.012 All animals 0.33 ± 0.02 (648) 0.29 ± 0.05 (215) z = 0.56 P = 0.578 Julian Date 25 50 75 100 125 150 175 200 225 250 275 300 325 350 B od y co nd iti on (b on e m ar ro w fa t % ) 0 20 40 60 80 100 Fig. 5. The dependence of moose body condition (expressed as percent fat) on day of the year (Julian date). In this analysis harvested moose are represented by a single value (Julian date = 292.80; percent fat = 66.05). The line in the figure is the estimated peak function; a Lorentzian peak function; y = 45.58 + [− 34.65/(1 + (((X – 115.05)/62.57))^2)]; F3,89 = 6.23, r2 = 0.174, P < 0.001. The peak are estimate to Julian date = 115.05, which is 25 April. 22 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 The abundance of L1 was category 0 in the lungs and 6 in feces. First stage larvae (infec- tion intensity = 6) were found in lungs from 14 August (~2 months old) to 4 June the fol- lowing year (~12 months old). Abundance of L1 in calves ranged from categories 1–6 by 2 months old, and the lung infection remained high; 81% had an L1 abundance category of 4–6 in the first year. The excretion of larvae began at a low level (2) on 14 August, and calves continued to excrete larvae through- out the first year at all levels of abundance (1–6). The prevalence of infection in harvested moose (n = 896) differed among regions (Fig. 6), ranging from 13% in southernmost Region 6 to 56% in Region 3 (Fig. 1 and Table 2). Infection was most prevalent in central Sweden, least prevalent in southern Sweden, and similar (P < 0.05) in southern and northern Sweden. DISCUSSION Although parasites at low abundance are generally less harmful to their host, when the host population increases rapidly, as with Swedish moose in the 1970–80s (Hörnberg 2001, Stéen et al. 2005), an increasing risk to the individual and host population is pos- sible (Toft 1991). The proportion of elaphos- trongylosis (symptoms of nervous disorder and/or emaciation) varies among age-classes in moose, with young animals more prone to illness (Stéen et al. 2005). Similarly, we found that E. alces worms located in the epi- dural space of the spinal cord were more prevalent in calves and yearlings, and only occasionally found in adults. The high abun- dance measured in young animals may simply reflect that the Swedish moose population is skewed towards young animals (Sand et al. 2011). Conversely, abundance of L1 in lungs and feces was highest in old moose, and low- est in young and middle-aged moose. Both Stuve (1986) and Stéen et al. (2005) suggested that E. alces most frequently infects males and young animals; however, we found no difference in the abundance within the epidural space, lungs, or feces be- tween sexes or age groups of harvested moose, only a tendency toward males in the young age group. Similarly, male reindeer calves with dominant mothers had higher abundance of E. rangiferi than female calves, and it was suggested that because these calves had better access to forage, they were at greater risk of ingesting infected gastro- pods (Halvorsen 1986a). Calf weight is de- pendent on summer browse availability in a Table 5. Age (in days) of moose calves infected by Elaphostrongylus alces, Sweden, 1985–1989. Parasite location N Mean age ± SD Min age Max age Epidural 281 140.8 ± 20.0 50 215 Lung 348 140.2 ± 20.8 50 215 Feces 416 141.0 ± 17.6 101 215 Region 1 2 3 4 5 6 P re va le nc e 0.1 0.2 0.3 0.4 0.5 0.6 0.7 a ab a b b b Fig. 6. The prevalence of Elaphostrongylus alces in regions of Sweden. The values are calcu- lated with logistic regression, Regions was a categorical predictor and Julian date a contin- uous predictor. The values are mean ± SE; means with the same letter are not different (P > 0.05). Only harvested moose were used in the analysis. ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 23 cow’s home range, with access to and quality of forage related to its relative status (Saether and Heim 1993). Stuve (1986) attributed the difference in infection rate between sexes in older moose to physiological changes asso- ciated with the rut, as suggested with reindeer (Halvorsen 1986b). A novel finding of our study was that L1 were found in lungs and feces of calves by 21 July, and adult worms in the epidural space by 15 September, or ~50–100 days after birth (Broberg 2004, Malmsten 2014). This prepatent period aligns with experimen- tal infections of E. alces in moose in which patent infection was realized 39–73 days post-infection (Stéen et al. 1997). Because calves sample vegetation in the first days of life to promote development of rumen microbes (Syroechkovsky et al. 1989), their potential to exposure to E. alces L3 is almost immediate. Not surprisingly, adult E. alces were identified in the epidural space of the caudal vertebral canal in 2 other calves har- vested in September (Handeland and Gibbons 2001). Further, calves and yearlings were most frequently infected in the epidural space of the spinal cord which seemingly corroborates that moose shed most E. alces L1 during their early years, after which a sharp drop in larval shedding and low num- bers of adult worms in older animals occur (Stuve 1986, Stéen et al. 2005). In both harvested and unharvested moose, E. alces worms were found in the epidural space of the spinal cord of animals aged 3–4 months to 4 years, not in middle- aged animals, and again at 10–16 years. Conversely, high levels of larvae were found in lungs and feces irrespective of age. We believe that the low frequency of worms in older animals, despite having L1 in lungs and shed larva, is due to migration from the CNS/PNS into the muscle fasciae, as with some other elaphostrongylins (Lankester 2001). The pattern of E. alces adults migrating out to the muscle fasciae, presumably due to an immune response in the epidural space (Stéen et al. 1997, 1998), differs somewhat from that of E. rangiferi, E. cervi, and P. tenuis. The latter are believed to remain in the CNS as adult worms during their entire life (in the subdural or subarachnoid space, inside the meninges), although E. rangiferi also migrates to the muscle fasciae (Hemmingsen et al. 1993). E. rangiferi, E. cervi, and P. tenuis may realize an immuno- logical harbor within the CNS, as might P. andersoni that is associated with blood ves- sels and connective tissues where females de- posit eggs (Lankester 2001). We hypothesize that E. alces worms are attacked by the im- mune system in the epidural space, and they migrate to the muscle fasciae where, with lower immunological defense, they deliver most of their larvae. After ingestion, L3 migrate from the gastrointestinal (GI) tract to the perineal cav- ity along the mesenchyme nerves, and into the abdominal wall associated with the more posterior lateral nerves. It is likely that E. alces does not need to enter the CNS par- enchyma to develop to the 5th stage (adult), as other Elaphostrongylus spp., but remains epidurally-associated with lateral nerves of the PNS and finally migrates to the muscle fasciae (Olsson et al. 1998). The lack of worms in the epidural space of the spinal cord in moose during their prime could be explained by this migration; however, it could also reflect an immune response to pre- vent reinfection as described for P. andersoni that realizes declining larval output as deer age with few adult worms in deer >1 year old. Further, repeated infection in white- tailed deer resulted in sharp decline in larval numbers and a strong cellular response to adult worms (Lankester 2001). Worms in the epidural space of older moose could sim- ply be a reinfection associated with a weaker immune system, or an initial infection. Whether some L3s migrate directly to muscle 24 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 fasciae without being associated with neural tissue is unknown. Infected animals, on average, had lower body condition than uninfected animals ex- cept for middle-aged animals in their prime. Calves were in poorer condition regardless of category of infection (as expected for young, growing animals), middle-aged were likely in normal condition despite high shed- ding rate of L1 in feces, and old individuals were in normal condition if uninfected. Thus, infection, not age per se, seemed to re- flect relative body condition. However, indi- vidual variation of immunological response to the parasite presumably exists because some individuals die young, others remain in normal condition through prime, and old animals are increasingly susceptible. In contrast with E. alces, no protostron- gylid L1 of E. cervi were recovered from Iberian red deer fawns (Cervus elaphus hispanicus) (Vicente and Gortázar 2001). Prevalence of E. cervi L1 increased with age of deer (Vicente et al. 2006) which is opposite to our findings with E. alces in moose; both had higher infection rates in young males than females. The E. cervi pattern corre- sponds with that in reindeer in which E. ran- giferi infects the host late in the season, remaining at the same intensity for at least 3 years (Halvorsen et al. 1985). It appears that E. cervi and E. rangiferi have more similar and longer evolutionary relationships to each other and their respect- ive hosts than E. alces. Moose have a long, independent evolutionary history from the Alceini and the Plio-Pleistocene, suggesting a peculiar adaption and habitat restriction of the species (Niedziałkowska et al. 2014), and presumably, a relatively short evolution- ary period with E. alces that could be less adapted with its host than E. cervi and E. rangiferi. It is possible that E. alces is more pathogenic to its host because both harvested and unharvested moose of below normal or emaciated body condition were infected with E. alces. In 2-year old moose, Stuve (1986) found that infected moose were light- er (carcass weight) than uninfected moose, yet conversely, Stéen et al. (1997) found that moose experimentally infected with E. alces retained normal weight when fed ad libitum. It remains unclear, however, if poor body condition is an indirect or direct effect of the parasite, that emaciation is either directly caused by an inflammatory response due to an epidural localization, or that elaphostrongylosis causes locomotor disorders making it difficult to move and feed (Stéen and Rehbinder 1986, Stéen and Roepstorff 1990, Stéen et al. 2005). In summary, different morphology (Stéen et al. 1989, Stéen and Johansson 1990, Gibbons et al. 1991, Lankester et al. 1998), genetics (Gajadhar et al. 2000, Chilton et al. 2005, Huby-Chilton et al. 2006), location (Stéen et al. 1997, 1998) (epidural for E. alces, subdural/subarachnoid for E. rangi- feri), and life span and host age relationships with infection (Lankester 2001) suggest dif- ferent, and perhaps, ongoing evolutionary adaption in Elaphostrongylus species with their hosts. Of further consequence is that ris- ing temperatures, and a warmer and wetter climate are predicted to increase habitat, dis- tribution, and abundance of mollusk hosts (Halvorsen and Skorping 1982, Halvorsen et al. 1985), which in turn could lead to higher infection rates in cervids (Handeland and Slettback 1994, Halvorsen 2012). Although moose are not necessarily in poor condition when infected with E. alces, condition and parasite abundance were correlated. We therefore suggest continued surveillance of this disease and its specific consideration in management of moose in Sweden. ACKNOWLEDGEMENTS We thank W. E. Faber, Department of Natural Resources, Central Lakes College, Brainerd, Minnesota, USA for his feedback on the manuscript. We are grateful for all ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 25 technical help and assistance from employ- ees, and earlier employees H. Mann, S. Persson, and I. Forssell at the Department of Parasitology, National Veterinary Institute and Swedish University of Agricultural Sciences, Uppsala, Sweden. Last, but not least, we thank Swedish hunters for their help and cooperation in data collection. Financial support for this study was provided by the Swedish Environmental Protection Agency, Stockholm, Sweden. REFERENCES ANDERSON, R. C. 1964. Neurological disease in moose experimentally infected with Pneumostrongylus tenius from white-tailed deer. Veterinary Pathology 1: 289–322. doi: 10.1177/030098586400100402. BAERMANN, G. 1917. Eine einfache Metode zur Auffindung von Ancylostoma- (Nematoden-) Larven aus Erdproben. Mededeel uithet Geneesk Lab. te Wel- tevreden, Feestbundel, Batavia, pp. 41–47 (in German). BROBERG, M. 2004. Reproduction in Moose: Consequences and Conflicts in Timing of Birth. Doctoral Thesis, Gothenburg University, Gothenburg, Sweden. CHILTON, N. B., F. HUBY-CHILTON, M. W. LANKESTER, and A. A. GAJADHAR. 2005. A method for extracting genomic DNA from individual elaphostrongyline (Nematoda: Protostrongylidae) larvae and differentiation of Elaphostrongylus spp. from Parelaphostrongylus spp. by PCR assay. Journal of Veterinary Diag- nostic Investigation 17: 585–588. doi: 10.1177/104063870501700612. ENGELSEN ETTERLIN, P., A. NEIMANIS, D. GAVIER-WIDEN, and C. HÅRD AF SEGERSTAD. 2009. Postmortal Hullbe- dömning av Hull Hos Tamdjur och Vilda Djur (Postmoral Examination of Body Condition in Pet Animals and Wildlife). National Veterinary Institute, Uppsala, Sweden (in Swedish). GAJADHAR, A., T. STEEVES-GURNSEY, J. KENDALL, M. LANKESTER, and M. STÉEN. 2000. Differentation of proto- strongylid dorasal-spined larvae by PCR Amplication of ITS-2 Rdna. Jour- nal of Wildlife Diseases 36: 713–722. doi: 10.7589/0090-3558-36.4.713. GASAWAY, W. C., D. B. HARKNESS, and R. A. RAUSCH. 1978. Accuracy of moose age determinations from incisor cementum layers. Journal of Wildlife Management 42: 558–563. doi: 10.2307/3800818. GIBBONS, L. M.,O. HALVORSEN, and G. STUVE. 1991. Revision of the genus Elaphostrongy- lus Cameron (Nematoda, Metastrongyloi- dea) with particular reference to species of the genus occurring in Norwegian cer- vids. Zoologica Scripta 20(1): 15–26. doi: 10.1111/j.1463-6409.1991.tb00272.x. HALVORSEN, O. 1986a. On the relationship between social status of host and risk parasitic infection. Oikos 47: 71–74. doi: 10.2307/3565921. –––. 1986b. Epidemiology of reindeer para- sites. Parasitology Today 12: 334–339. doi: 10.1016/0169-4758(86)90053-0. –––. 2012. Reindeer parasites, weather and warming of the Arctic. Polar Biology 35: 1209. doi: 10.1007/s00300-012-1209-0. –––, and A. SKORPING. 1982. The Influence of temperature on growth and develop- ment of the nematode Elaphostrongylus rangiferi in the gastropods Arianta abus- trum and Euconulus fulvus. Oikos 38: 285–290. doi: 10.2307/3544666. –––, A. SKORPING, and K. HANSEN. 1985. Seasonal cycles in the output of first stage larvae of the nematode Elaphostrongylus rangiferi from reindeer, Rangifer taran- dus tarandus. Polar Biology 5: 49–54. doi: 10.1007/BF00446045. HANDELAND, K., and L. M. GIBBONS. 2001. Aspects of the life cycle and pathogenesis of Elaphostrongylus alces in moose (Alces alces). Journal of Parasitology 87: 1054–1057. doi: 10.1645/0022-3395 (2001)087[1054:AOTLCA]2.0.CO;2. –––, and T. SLETTBAKK. 1994. Outbreaks of clinical cerebrospinal Elaphostrongylosis in reindeer (Rangifer tarandus tarandus) 26 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 in Finnmark, Norway, and their relation to climatic conditions. Journal of Veterin- ary Medicine B 41(1–10): 407–410. doi: 10.1111/j.1439-0450.1994.tb00244.x. HEMMINGSEN, W., O. HALVORSEN, and A. SKORPING. 1993. Migration of Adult Ela- phostrongylus rangiferi (Nematoda: Pro- tostrongylidae) from the spinal subdural space to the muscles of reindeer (Rangifer tarandus). Journal of Parasitology 79: 728–732. doi: 10.2307/3283612. HUBY-CHILTON, F., N. B. CHILTON, M. W. LANKESTER, and A. A. GAJADHAR. 2006. Single-strand conformation poly- morphism (SSCP) analysis as a new diag- nostic tool to distinguish dorsal-spined larvae of the Elaphostrongylinae (Nema- toda: Protostrongylidae) from cervids. Veterinary Parasitology 135: 153–162. doi: 10.1016/j.vetpar.2005.08.001. HÖRNBERG, S. 2001. Changes in population density of moose (Alces alces) and dam- age to forests in Sweden. Forest Ecology and Management 149: 141–151. doi: 10.1016/S0378-1127(00)00551-X. LANKESTER, M. W. 2001. Extrapulmonary lungworms of cervids. Pages 228–278 in W. M. Samuel, A. A. Kocan, and M. Pybus, editors. Parasitic Diseases of Wild Mammals, 2nd edition. Iowa State University Press, Ames, IA. –––. 2010. Understanding the impact of men- ingeal worm, Parelaphostrongylus tenuis, on moose population. Alces 46: 53–70. –––, I.-M. C. OLSSON, M. STÉEN, and A.A. GAJADHAR. 1998. Extra-mammalian larval stages of Elaphostrongylus alces (Nematoda: Protostrongylidae), a para- site of moose (Alces alces) in Fenno- scandia. Canadian Journal of Zoology 76: 33–38. doi: 10.1139/z97-168. MALMSTEN, J. 2014. Reproduction and health of moose in southern Sweden. Doctoral Thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden. NIEDZIAŁKOWSKA, M., K. J. HUNDERTMARK, B.JEZDRZEJEWSKA, K.NIEDZIAŁKOWSKI, V. E. SIDOROVICH, M. GORNY, R. VEEROJA, E. J. SOLBERG, S. LAAKSONEN, H.SAND, V. A.SOLOVYEV, M. SHKVYRIA, J.TIAINEN,I.M.OKHLOPKOV,R.JUSKAITIS, G. DONE, V. A. BORODULIN, E. A. TULANDIN, and W. JEZDRZEJEWSKI. 2014. Spatial structure in European moose (Alces alces): genetic data reveal a complex population history. Journal of Biogeography 41: 2173–2184. doi: 10.1111/jbi.12362. NORDISK METODIK-KOMMITTÉ för LIVSMEDEL (NORDIC COMMITTEE on FOOD ANALYSIS). 1989. SBR (Schmid-Bondzynski-Ratslaff), No. 131. Esbo, Finland. http://www.evira.fi (accessed July 2015). OLSSON, I.-M. 2001. Elaphostrongylus alces – transmission, larval morphology and tissue migration. Veterinary Medicine Licentiate Thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden. –––, M. W. LANKESTER, A. A. GAJADHAR, and M. STEÉN. 1998. Tissue migration of Elaphostrongylus spp. in guinea pigs (Cavia porcellus). Journal of Parasit- ology 84: 968–975. doi: 10.2307/ 3284629. SAETHER, B. E., and M. HEIM. 1993. Eco- logical correlates of individual variation in age at maturity in female moose (Alces alces); the effects of environmental vari- ability. Journal of Animal Ecology 62: 482–489. doi: 10.2307/5197. SAND, H., N. JONZÉN, H. ANDRÉN, and J. MÅNSSON. 2011. Strategier för beskatt- ning av älg. (Strategies for Moose Man- agement). Forest Facts, Report from the Swedish University of Agricultural Sciences, No. 2 4, p 4 (In Swedish). SAS 2014. Version 9.4 TS Level 1M0, X64_7PRO platform, SAS Institute Inc., Cary, North Carolina, USA. STÉEN, M., C. G. M. BLACKMORE, and A. SKORPING. 1997. Cross-infection of moose (Alces alces) and reindeer (Rangi- fer tarandus) with Elaphostrongylus alces and Elaphostrongylus rangiferi (Nema- toda, Protostrongylidae): effects on para- site morphology and prepatent period. ALCES VOL. 52, 2016 STÉEN ET AL. – EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES 27 http://www.evira.fi Veterinary Parasitology 71: 27–38. doi: 10.1016/S0304-4017(97)00013-7. –––, A. G. CHABAUD, and C. REHBINDER. 1989. Species of the genus Elaphostron- gylus, parasite of Swedish Cervidae. A description of E. alces n. sp. Annales de Parasitologie, Humanie et Comparee 64: 134–142. –––, AND C. JOHANSSON. 1990. Elaphostron- gylus spp. from Scandinavian cervidae–a scanning electron study (SEM). Rangifer 10: 39–46. doi: 10.7557/2.10.1.789. –––, I.-M. OLSSON, AND E. BROMAN. 2005. Diseases in a moose population subjected to low predation. Alces 41: 37–48. –––, S. PERSSON, AND L. HAJDU. 1993. Pro- tostrongylidae in Cervidae and Ovibus moscatus; a clustering based on isoelec- tric focusing on nematode proteins. Rangifer 13: 221–223. doi: 10.7557/ 2.13.4.1121. –––, AND C. REHBINDER. 1986. Nervous tis- sue lesions caused by elaphostrongylosis in wild Swedish moose. Acta Veterinaria Scandinavia 27: 326–342. –––, and L. ROEPSTORFF. 1990. Neurological disorder in two moose calves (Alces alces L.) naturally infected with Elaphostron- gylus alces. Rangifer, Special Issue 3: 399–406. doi: 10.7557/2.10.3.887. –––, I. Y. WARSAME, and A. SKORPING. 1998. Experimental infection of reindeer, sheep and goats with Elaphostrongylus spp. (Nematoda, Protostrongylidae) from moose and reindeer. Rangifer 18: 73–80. doi: 10.7557/2.18.2.1448. STUVE, G. 1986. The prevalence of Elaphos- trongylus cervi infection in moose (Alces alces) in southern Norway. Acta Veteri- naria Scandia 27: 397–409. SYROECHKOVSKY, E. E., E. V. ROGACHEVA, and L. A. RENECKER. 1989. Moose hus- bandry. Pages 369–386 in R. J. Hudson, K. R. Drew, and L. M. Baskin, editors. Wildlife Production Systems. Economic Utilization of Wild Ungulates. Cambridge University Press, Cambridge, England. SWEDISH ASSOCIATION FOR HUNTING and WILDLIFE MANAGEMENT. 2015. http:// jagareforbundet.se (accessed July 2015). SWEDISH METEOROLOGICAL and HYDRO- LOGICAL INSTITUTE (SMHI). 2015. http:// www.smhi.se (accessed September 2015). SYSTAT. 2002. Table Curve 2D, version 5.01. SYSTAT Software Inc., San Jose, CA. –––. 2008. SigmaPlot for Windows version 11.0, Build 11.0.0.75. SYSTAT Software Inc., San Jose, California, USA. TOFT, C. A. 1991. An ecological perspective: the population and community conse- quences of parasitism. Pages 319–343 in C. A. Toft, A. Aeschlimann, and L. Bolis, editors. Parasite-Host: Association Coex- istence or Conflict? Oxford University Press, Oxford, England. VICENTE, J., I. G. FERNÁNDEZ DE MERA, AND C. GORTAZAR. 2006. Epidemiology and risk factors analysis of elaphostron- gylosis in red deer (Cervus elaphus) from Spain. Parasitology Research 98: 77–85. doi: 10.1007/s00436-005-0001-2. –––, and C. GORTÁZAR. 2001. High preva- lence of large spiny-tailed protostrongy- lid larvae in Iberian red deer. Veterinary Parasitology 96: 165–170. doi: 10.1016/ S0304-4017(00)00425-8. 28 EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES – STÉEN ET AL. ALCES VOL. 52, 2016 http://jagareforbundet.se http://jagareforbundet.se http://www.smhi.se http://www.smhi.se EPIZOOTIOLOGY OF ELAPHOSTRONGYLUS ALCES IN SWEDISH MOOSE STUDY AREA METHODS Data management and statistics RESULTS Infection, age and sex Body condition Time of infection DISCUSSION ACKNOWLEDGEMENTS REFERENCES