Transmission and Predictors of Burden of Lungworms of the Striped Dolphin (Stenella coeruleoalba) in the Western Mediterranean

The Pseudaliidae is an evolutionarily distinct family of metastrongyloid lungworms whose species almost exclusively infect cetaceans. Species have been reported in the lungs and cranial sinuses of 29 species of odontocetes worldwide (Fraija-Fernández et al. 2016). Infections can result in severe consequences, including lesions in the cranial sinuses or verminous pneumonia (Measures 2001). Little is known about the life history of marine pseudaliids and most studies have dealt with species infecting coastal or inshore populations of cetaceans (Measures 2001). Two nonexclusive routes of transmission, vertical and horizontal, have been proposed (Measures 2001). Evidence for vertical transmission exists, particularly for species of Halocercus, which have been found in fetuses and suckling calves (Fauquier et al. 2009; Reckendorf et al. 2018). In favor of horizontal transmission, Lehnert et al. (2010) detected larvae of Pseudalius inflexus in wild dab (Limanda limanda), a typical prey of harbor porpoise (Phocoena phocoena) in German waters.

Most epidemiologic surveys of pseudaliids have been focused on the potential relationship between infection levels and host age or size, with conflicting results. The prevalence or intensity of some lungworm species has been found to be negatively (Tomo et al. 2010) or positively (Geraci et al. 1978) correlated with age or length, whereas in other species, no significant relationship has been detected (Faulkner et al. 1998).

We examined a previously unstudied western Mediterranean population of the striped dolphin (Stenella coeruleoalba), an oceanic cetacean. We investigated the possibility of vertical and horizontal transmission of pseudaliids and assessed the influence of three key predictors (size, sex, and season) on lungworm infections.

A total of 87 striped dolphins stranded along the Mediterranean coast of Spain (Valencian Community, between 40°31′00″N, 0°31′00″E and 37°50′00″N, 0°45′42″W) during 1987–2018 was analyzed for lungworms (Table 1). We excluded animals that were killed by a morbillivirus in the epizootic events of 1990–91 and 2007–08 (Raga et al. 2008) to make results comparable with those from previous surveys. Only carcasses in preservation states 1–3 (Geraci and Lounsbury 2005) were selected for analysis. During postmortem examination, all individuals were sexed, total length was measured to the nearest 1 cm (Table 1), then lungs were frozen at –20 C.

Lungs were thawed for 24 h. Only one lung per individual, randomly selected, was used in this study; the other was saved for a future study on microhabitat selection. The airways were cut from the trachea to the caudal apex of the lung and examined macroscopically, and nematodes were collected. The lung was periodically rinsed over a 0.2-mm sieve to wash away any blood obscuring worm detection. All fragments and whole worms were stored in 70% ethanol. Lungworms were examined under a light microscope and identified following Anderson et al. (2009). The total worm number per lung was estimated as the number of complete worms, or incomplete worms with their caudal end intact.

Infection parameters are defined as in Rózsa et al. (2000); mean intensity is the mean number of individuals of a parasite species per infected host, mean abundance is the mean number of individuals of a parasite species per host examined, and prevalence is the number of infected hosts as a proportion of the hosts examined. To calculate the 95% confidence interval, the Sterne exact method was used for prevalence, and the bias-corrected and accelerated bootstrap method with 20,000 replications was used for mean intensity and abundance. A Kolmogorov-Smirnov test was used to investigate whether the distribution of lungworm species among hosts followed a negative binomial distribution.

We investigated the relative importance of vertical and horizontal transmission by comparing lungworm burdens between three age classes: neonate (feeding on milk only), suckling (feeding on a mix of milk and prey), and weaned (feeding on prey only). Dolphins shorter than 93 cm long were considered to be neonates (Aguilar 1991). In Mediterranean striped dolphins, weaning starts at 3 mo and stops at 18 mo, when animals average 165 cm long (Aguilar 1991). Therefore, individuals between 94 and 165 cm were classified as suckling, and individuals more than 165 cm were designated weaned. The validity of these feeding categories was confirmed by dietary data.

Lungworm prevalences and abundances between age classes were compared with Fisher's tests and Kruskal-Wallis tests, respectively. Lungworm abundance was compared between neonates and sexually mature females (>187 cm; Calzada et al. 1996) by a Mann-Whitney U-test.

A generalized linear mixed model with a negative binomial distribution and a log-link was performed to examine the effects of host body length, season, and gender on lungworm abundance. Year was included as a random effect with intercepts only. Models were fitted using restricted maximum likelihood, and F and P values were calculated by the Satterthwaite approximation for small sample sizes (Harrison et al. 2018). The full model included two-way interactions between fixed predictors. Models with different combinations of fixed and random parameters were ranked on the basis of Akaike information criteria (AIC) values corrected for small sample size (Burnham and Anderson 2002). The model with minimum AIC was considered the best model; those with values of ΔAIC≤2 were considered to have substantial empirical support, and those with ΔAIC>4 to have much less support. Residual analysis was performed to check the suitability of the final model. We used the software Quantitative Parasitology version 3.0 (Reiczigel and Rózsa 2005) to calculate parasite burden parameters. All other statistical analyses were performed with SPSS version 24 (IBM Corp., Armonk, New York, USA).

Fifty-one of 87 dolphins (59%) were infected with Skrjabinalius guevarai (total n=2,512) and Stenurus ovatus (total n=4; Table 2). All infections of Stenurus ovatus were composed of larvae only and co-occurred with those from Skrjabinalius guevarai. One of five neonates (89 cm long) was infected with 80 individuals of Skrjabinalius guevarai (Table 2), strongly suggesting vertical transmission. This calf could have acquired lungworms through transplacental infection, consumption of infected milk, or contact with contaminated spray during breathing or water during feeding, respectively (Caldwell et al. 1968; Ortiz 2001; Fauquier et al. 2009). However, the environmental pathway (through breathing or ingesting contaminated water) alone could hardly account for the high intensity of infection found, because mothers and calves are not always in close proximity and the released larvae could be easily dispersed. In Halocercus lagenorhynchi, conclusive evidence for transplacental, lactogenic, or both kinds of transmission comes from the detection of adult lungworms in the lungs of a fetus and in neonates of the common bottlenose dolphin (Tursiops truncatus; Fauquier et al. 2009). As for Skrjabinalius guevarai, future examination of fetuses and mammary glands of reproductively active females of striped dolphins will be required to shed light on the relative roles of transplacental and lactogenic transmission.

Neither prevalence (P=0.203) nor abundance (χ²=2.142, df=2, P=0.343) of Skrjabinalius guevarai differed significantly among age classes (Table 2). Also, the abundance in neonates (mean [95% confidence interval]: 16 [0–32]) and sexually mature females (40 [18–81]) did not differ significantly (U=32, P=0.114). It is unlikely that larvae directly reinfect the same individual hosts; thus, the occurrence of lungworms in adult dolphins begs horizontal transmission through contact with infected spray or water or, more likely, with infected prey (Measures 2001). The available data, however, would not suggest a prominent role of horizontal over vertical transmission in Skrjabinalius guevarai.

The generalized linear mixed model with the lowest AICc included only the three predictors: season, gender, and body length; any other model with interactions or without any of these predictors exhibited ΔAICc>16. The overall R² (percent) of the main effects model was just 17.6%, and individual contributions of predictors were all <10% (Table 3). Body length had a positive, statistically significant effect on parasite burden (Table 3 and Fig. 1), suggesting that larger (older) dolphins were exposed to infected prey for longer or consumed more infected prey because of higher metabolic requirements, as reported for Stenurus globicephalae in the Atlantic white-sided dolphin (Lagenorhynchus acutus; Geraci et al. 1978).

The effect of gender and season were weak at best, according to our analysis (Table 3). Lungworm infection was somewhat higher in males and highest in spring, followed by winter, summer, and autumn (Table 3 and Fig. 1). For a parasite using both vertical and horizontal trophic transmission, seasonal changes could reflect the host's breeding period, a seasonal variation in host diet, or both. Because the calving season of striped dolphin in the Mediterranean occurs primarily in autumn (Aguilar 1991), when the abundance of Skrjabinalius guevarai is lowest, vertical transmission could hardly account for the observed seasonal variation in lungworm abundance. This conclusion was further supported by the lack of a significant interaction between body length and season and suggested the seasonal pattern was primarily related to changes in diet. To shed further light on potential gender and season effects on the abundance of Skrjabinalius guevarai, it will be necessary to identify the prey by which this nematode infects striped dolphins.

We greatly appreciate the assistance with stranded dolphins of our colleagues from the Marine Zoology Unit, University of Valencia, Spain. We also thank two anonymous reviewers who helped to improve the article significantly. Our study was possible through support from the Servicio de Vida Silvestre, Conselleria de Agricultura, Medio Ambiente, Cambio Climático y Desarrollo Rural, and projects CGL/2012/39545 (Ministry of Economy and Competitiveness, Spain), and PROMETEO II/2015/018 (Generalitat Valenciana, Spain). R.P. benefitted from a scholarship by the Programa Santiago Grisolia de la Conselleria d'Educació, Investigació, Cultura I Esport (GrisoliaP-2017-093), CPI-17-214 (Generalitat Valenciana).