Pharodes tortugensis

Forrester, Graham E. & Finley, Rachel J., 2022, Figs. 23 - 34 in The Newly Recorded Fern-spore Feeding Moths in the Genus, Meyrick 1889 (Lepidoptera: Stathmopodidae) from Taiwan, with Notes on Life History of Three Species., Zoological Studies 61 (32), pp. 1-19 : 5-7

Effects of P. tortugensis on host body condition

To assess the effect of P. tortugensis on host body condition and energy allocation, a mix of C. glaucofraenum and C. venezuelae were collected from Muskmelon Bay and White Bay near Guana Island in from 2001–2003 ( Fig. 2 ). The gobies were fixed in 70% alcohol, measured (SL) and sexed by examining the genital papilla ( Cole and Shapiro 1990 1992). Individuals spanning the size range at which most infections occurred were used in this analysis (n = 163, 12.2–33.7 mm SL). These gobies change sex from female to male (protogynous hermaphroditism), and individuals of this size range comprise mainly juveniles and females ( Cole and Shapiro 1990 1992). Excluding larger individuals because they were rarely infected by P. tortugensis thus also excluded most males from the sample.

The otoliths (lapilli) were removed from each goby and, after clearing in immersion oil, the postsettlement age (days) was determined by counting the daily growth rings formed after the mark on the otolith that indicates settlement to the reef ( Steele and Forrester 2002). The copepods were removed from parasitized gobies, dried at 60°C until a constant mass was achieved (at least 1 h) then weighed in mg to 0.001 mg. The liver and gonads were removed from each goby, and these organs plus the body tissue (minus the alimentary tract) were each dried and weighed. Total goby body mass was calculated as the sum of liver, gonad, and body tissue mass.

To assess the effect of P. tortugensis on body condition, we measured total body mass as a function of body length (SL) ( Ogle 2016). To assess the effect of P. tortugensis on investment in energy reserves and reproduction allocation, we measured liver (LM) and gonad (GM) mass respectively as a function of total body mass (BM). To explore whether changes in condition might affect growth we examined body mass as a function of post-settlement age (A).

Fish mass-length (BM vs. SL) relationships are non-linear and are typically modeled as a power function of the form

BMi = aSLi b e ci where a and b are constants and ci is the multiplicative error term for the i th fish ( Ogle 2016). This relationship was linearized as follows log(BMi) = log(a) + blog (SLi) + ci and analysis of covariance (ANCOVA) was used to test the effect of parasitic infection (IN), a categorical v a r i a b l e (i n f e c t e d o r n o t), o n t h e t r a n s f o r m e d relationship as follows log(BMi) = log(a) + blog(SLi) + d (IN) + eIN × log(SLi) + ci

Inspection of the data suggested that the other relationships of interest (LM vs. BM, GM vs. BM, and BM vs. A) also resembled power functions rather than linear functions. ANCOVA models of the same form as the BM vs. SL model were thus used to test for effects of parasitism on these other relationships.

The effects of some macroparasites are related to the number of parasites per host, so rather than just parasite presence, we tested whether infection abundance (sensu Bush et al. 1997) was a better predictor of host impact than parasite presence. Copepods varied greatly in size, so we used the combined mass of copepods on a host, rather than the number of copepods, as our index of infection abundance. For each of the relationships just described (BM vs. SL, LM vs. BM, GM vs. BM, and BM vs. A), we substituted infection intensity (II) for parasite presence (IN) in the above linear models. Infection abundance (II) is a continuous variable and infection presence (IN) is categorical, so the model became a multiple regression rather than ANCOVA. For each pair of models (II vs. IN), we used Akaike’s Information Criterion (AICc) as a measure of relative model fit, and models differing in AIC by <6 were judged to have similar support in the data ( Richards 2005).

Spatial and temporal patterns in the distribution of hosts infected with P. tortugensis

To document the spatial distribution of P. tortugensis infections on common Coryphopterus hosts in the British and US Virgin Islands, we performed visual surveys on SCUBA at 52 sites in the area ( Table 1; Fig. 2 ). Individuals were classified as parasitized or not based on visual symptoms of infection. These counts represent the combined abundance of C. glaucofraenum , C. venezuelae and C. tortugae . We made quantitative surveys at 35 sites, in which all gobies within transects (0.5 × 4 m) were inspected and counted. At 17 other sites, we made less rigorous qualitative surveys in which gobies encountered were visually inspected and P. tortugensis was noted as either present or absent at the site.

To document changes over time in the prevalence of infection, repeated estimates were made from 1993– 2019 at a BVI site near Guana Island (Harris Ghut, Table 1, and Fig. 2 ). Coryphopterus venezuelae from this site were classified as parasitized or not based on visual symptoms of infection ( Table 2). Some estimates were based on inspection of individuals during underwater surveys, whereas other individuals were captured using hand nets for other experiments and inspected underwater while in the net prior to release back into the field.

To document the broader distribution of P. tortugensis infections on three common Coryphopterus hosts ( C. glaucofraenum , C. venezuelae and C. tortugae ), visual surveys were performed at 16 other sites throughout the Caribbean ( Table 3). Hosts were screened visually for symptoms on SCUBA as encountered. Parasitized hosts were counted individually, and the total number of hosts screened per location was recorded to the nearest 10 (n ≈ 4900; Table 3).

In addition to the four species of Coryphopterus we studied, P. tortugensis has been reported from 11 species of fish host at other locations in the western Atlantic ( NMNH 2020; WORMS 2020). Most of these other hosts are gobies ( Gobiidae Cuvier ) or blennies ( Blenniidae Rafinesque ). We therefore sought to identify potential additional hosts of P. tortugensis from these families in the British Virgin Islands ( Table 4). Some potential hosts were collected and their external body surface, gill arches, branchial chamber and underside of the operculum were carefully searched for P. tortugensis . Other hosts were inspected visually on SCUBA for the distended operculum symptomatic of infection.

Figures

Fig. 2. Locations in the United States and British Virgin Islands where Coryphopterus were found infected with Pharodes tortugensis. Numbers correspond to sites listed in table 1.