A Comparison of Injectable Fluorescent Marks in Two Genera of Darters: Effects on Survival and Retention Rates

Abstract

Visible implant elastomer (VIE) and injectable photonic dye (IPD), two types of injectable fluorescent marks, have shown promise in previous applications in a variety of fishes but have not been extensively tested on darters. We marked a species from each of two genera of darters, Percina and Etheostoma, in a laboratory experiment to determine the influence of VIE and IPD marks on survival and the influences of mark type, location, and color on mark retention. Short‐term (≤80‐d) survival was similar between marked and control specimens for both marks in both species. Over the long term (200–240 d), however, the survival rate for IPD‐marked Roanoke darters P. roanoka was significantly lower than that for controls (50% versus 80%), whereas VIE‐marked Roanoke darters had a survival rate (88%) similar to that of controls. Long‐term survival of riverweed darters E. podostemone did not differ among groups. In Roanoke darters, the mark retention rate for IPD was significantly lower than that for VIE by day 80 of the experiment (80% versus 94%), and ventral IPD marks were retained with greater frequency than were dorsal IPD marks. In riverweed darters, retention was similar for VIE and IPD (79% versus 83%) in all body locations through day 240. In both species, yellow IPD marks exhibited higher retention rates than did green IPD marks, whereas the reverse was true for yellow and green VIE marks. Overall, VIE was a superior mark in the Percina representative but performed similarly to IPD in Etheostoma. Because of interspecific and intraspecific variability in mark performance, we recommend a pilot study before initiating field use of injectable marks in untested species.

Despite the high conservation priority given to small‐bodied stream fishes in North America (Angermeier 1995; Warren et al. 1997; Warren et al. 2000), development of methodologies for tracking individuals of these species has lagged behind the need for movement and population‐size data. The practice of injecting visible liquid marks (e.g., paint, dye, and tattoo ink), a relatively simple and inexpensive method for identifying small‐bodied fishes, has been in existence since the 1920s (McFarlane et al. 1990; Guy et al. 1996). However, recent research has shown that such marks may fade over time at varying rates, depending on mark type, color, location, and fish taxon (Hill and Grossman 1987b; Goforth and Foltz 1998; Albanese 2001). Injected marks are typically applied in a batch‐specific fashion, although individual marks can be generated using different colors and body locations (Hill and Grossman 1987a; Johnston 2000; Skalski and Gilliam 2000). However, this methodology requires that mark retention be similar for different colors and body locations, an assumption that is seldom tested.

In the early 1990s, visible implant elastomer (VIE; Northwest Marine Technology, Shaw Island, Washington) emerged as an alternative to liquid injected marks. The VIE compound is composed of a colored polymer and a curing agent that, after being mixed and injected into subcutaneous tissue, solidifies into a pliable fluorescent sliver. Although VIE exhibits a short lifespan once mixed (i.e., 5 d in refrigeration; personal observation), it offers the benefits of reportedly high retention rates and low mortality effects in multiple fish taxa (Dewey and Zigler 1996; Haines and Modde 1996; Hale and Gray 1998).

Most adult darters Ammocrypta, Etheostoma, and Percina spp. are smaller than 100 mm total length (TL), thus rendering the group a likely candidate for subcutaneous marking. However, darters exhibit several features, such as a benthic habit, variable coloration, and small scales, that could affect mark retention and readability. Recent investigators have used VIE successfully in darters (Labbe and Fausch 2000; J.R. Shute, Conservation Fisheries, Inc., unpublished data). Liquid injectable marks have also been used in darters (Freeman 1995; Warren and Pardew 1998), but mark performance has yet to be rigorously evaluated. Of particular interest is injectable photonic dye (IPD; New West Technology, Arcata, California), a relatively new liquid marking material that requires no mixing, and is specifically designed for marking aquatic organisms. The literature reports low fish mortality rates from all types of subcutaneous marking (Hill and Grossman 1987b; Bonneau et al. 1995; Johnston 2000), although data on darters is lacking (but see Labbe and Fausch 2000).

Because of the diversity of marking options currently available to researchers, we sought to rigorously evaluate two of the most promising marking technologies, VIE and IPD, on a group that has received little treatment in the tagging literature, the darters. The objectives of this study were to determine the influence of VIE and IPD marks on fish survival and to determine the influences of mark type, location, and color on mark retention. We applied marks to Roanoke darters Percina roanoka and riverweed darters Etheostoma podostemone, which are regionally abundant and represent the two most speciose genera of darters.

Methods

Fish were captured from several locations in the North and South forks of the Roanoke River (Montgomery County, Virginia), between 11 and 26 June 2001, with a Smith‐Root DC backpack electrofisher and a 5‐mm‐mesh seine. During collections, we retained all Roanoke and riverweed darters 40 mm TL and longer, and these were transported to the laboratory on the Virginia Tech campus, Blacksburg. Fish were housed in two 636‐L Living Streams (Frigid Units, Inc.), each containing a 1/3‐hp chiller and recirculator. Chiller temperatures were varied seasonally to correspond to the temperature of the Roanoke River (range 10– 18°C). Additionally, we controlled the photoperiod and ambient temperature of the room containing the tanks, which we varied according to season. Tank bottoms were covered with pea‐size gravel and haphazardly distributed cobbles and hollow blocks to simulate instream cover conditions. Fish were fed frozen bloodworms every 1–2 d. We allowed specimens to acclimate to the tanks for a week before marking, and specimens were monitored daily for mortality or abnormal behaviors.

We randomly assigned individuals to the following mark type groups: tank 1, Roanoke darters (32 VIE, 32 IPD, 30 control); tank 2, riverweed darters (28 VIE, 23 IPD, 25 control). Half of all treatment specimens were randomly assigned to fluorescent yellow and the other half were assigned to fluorescent green. Furthermore, each treatment Roanoke darter was randomly assigned to one of four marking locations: midventral (MV), lower caudal peduncle (LCP), upper caudal peduncle (UCP), or middorsal (MD). In riverweed darters we chose to maximize the replication of marks by applying marks to all four body locations on each individual (thus, N = 112 VIE and 92 IPD marks in riverweed darters versus 32 marks each of VIE and IPD in Roanoke darters). During marking, specimens were anesthetized in tricaine methanesulfonate (MS‐222; Finquel). We injected 2–3‐mm‐long VIE marks subcutaneously using a 1‐cc syringe equipped with a 29‐gauge needle. For IPD marks, 0.1 mL of dye was injected subcutaneously using the Pen‐Ject tag applicator (New West Technologies, Inc.) equipped with a 30‐gauge needle. To keep handling stress consistent across treatments, control individuals were handled for the same amount of time as marked individuals, and were pricked in one (Roanoke darters) or four (riverweed darters) body locations with an empty 29‐gauge needle. To differentiate control individuals from treatment individuals that had lost marks, we applied fin clips to the caudal fin of marked individuals (upper caudal for VIE versus lower caudal for IPD). Location and color of lost marks were determined by process of elimination.

Marks were checked at 30, 80, 200, and 240 d postmarking. During each mark check, we anesthetized specimens in MS‐222 and then recorded total length, the presence of marks, the presence of a fin clip, and the presence of any anomalies associated with marks. Fish were inspected for fluorescent marks with both a blue flashlight (VIE; Northwest Marine Technologies, Inc.) and a UV light (IPD; New West Technology, Inc.). Tanks were inspected daily for mortality, and we recorded all of the above information for any dead individuals.

To test whether marks affected the probability of individual survival, we used logistic analysis (PROC GENMOD, SAS 2000) with survival (yes or no) as the response and mark type as the effect. Because comparisons across species were confounded with tank effects and because results from the four check dates were not independent, we performed models on each species at each check date separately. We evaluated model significance using likelihood ratio (LR) chi‐square tests at α = 0.05 (Stokes et al. 2000). Additionally, we designed a priori LR contrasts of VIE versus control and IPD versus control, and reported these tests when mark type was found significant by the overall LR test. We evaluated contrast significance using a Bonferroni‐adjusted α = 0.025 (for k = 2 comparisons per model).

To test whether the probability of mark retention differed among treatments, we used logistic analysis with the retention of each mark (yes or no) as the response and mark type, mark location, and mark color as the effects. To provide interpretable statistical comparisons of retention among body locations, we converted each of the four body locations to one of two binary responses: ventral (MV, LCP) or dorsal (UCP, MD). This classification is biologically meaningful because of the benthic habit of darters, which probably increases the abrasion and regeneration exhibited in ventral tissues. Individuals that died before a given check day were not included in the mark retention analysis for that day. In all mark retention models, we used LR chi‐square tests at α = 0.05 (Stokes et al. 2000) to first evaluate overall model significance and then significance of individual effects.

Results

Individual Survival

Of the 94 Roanoke darters included in the experiment, 68 (72%) survived to day 240. Survival rate exhibited a gradual decline over time in control and VIE specimens, whereas IPD specimens exhibited much lower survival rates that declined more sharply over time (Figure 1). Logistic analysis indicated that mark type did not affect the probability of Roanoke darter survival at day 80 (χ² = 5.83, P > 0.05) but did significantly affect survival at days 30 (χ² = 8.12, P < 0.05), 200 (χ² = 12.87, P < 0.01), and 240 (χ² = 12.37, P < 0.01; Table 1). Contrasts revealed that IPD‐marked individuals had significantly higher mortality than control individuals at days 200 (χ² = 6.05, P < 0.025) and 240 (χ² = 6.26, P < 0.025); no other contrasts were significant.

Riverweed darters exhibited higher survival rates than did Roanoke darters, 72 of the 76 original individuals (95%) surviving to day 240 (Figure 1). Logistic models for days 30 and 80 failed to converge, due to quasicomplete separation of the data, and could not be evaluated for significance. This condition occurs when the probability of a response (e.g., survival = yes) is very close to 1, which was a consequence of low mortalities on those check days. Mark type did not significantly affect the probability of individual survival on days 200 (χ² = 2.64, P > 0.2) or 240 (χ² = 3.34, P > 0.1; Table 1).

Mark Retention

Overall mark retention was 90% for VIE versus 44% for IPD in Roanoke darters at day 240. The model failed to converge at day 30, and mark retention did not differ among treatments at day 80 (χ² = 10.04, P = 0.12; Table 2). The VIE marks had a significantly higher probability of retention than did the IPD marks at days 200 (χ² = 15.06, P = 0.0001) and 240 (χ² = 19.36, P < 0.0001). Retention of VIE marks was similar among body locations, whereas retention of IPD marks was lower in dorsal body locations (UCP and MD; 25%) than in ventral locations (LCP and MV; 62%; Figure 2). The mark location factor was nonsignificant in all models; however, the interaction between mark type and mark location was significant at days 200 (χ² = 9.11, P < 0.01) and 240 (χ² = 9.27, P < 0.01). Yellow IPD marks exhibited higher retention rates than did green IPD, whereas green VIE had a higher retention rate than yellow VIE (Figure 3). The mark type × mark color interaction was significant at day 240 (χ² = 4.81, P < 0.05), but not at day 200. The mark color factor itself was nonsignificant in all models.

On day 240, 79% of VIE marks and 83% of IPD marks remained in riverweed darters. No individuals lost multiple marks in the short term (≤80 d), but over the long term (200–240 d), 37% of fish that lost marks lost multiple marks. Models failed to converge at days 30 and 80. Mark retention did not differ among treatments at days 200 (χ² = 1.54, P > 0.2) or 240 (χ² = 0.59, P > 0.4; Table 2). Retention of marks was similar among body locations for both VIE and IPD (Figure 2). As was the case for Roanoke darters, mark retention varied in subtle ways between mark types and colors. Although no riverweed darter models were significant, an interaction between mark type and mark color emerged; green VIE marks were retained with greater probability than were yellow VIE marks, whereas the reverse was true for IPD marks (Figure 3).

Discussion

Effects of Marking on Survival

Roanoke and riverweed darters marked with VIE suffered no short‐term or long‐term reduction in survival compared with unmarked individuals. Survival rates declined gradually over time for both species. High survival rates (>99%) following VIE application were observed in previous studies for up to a year over a range of fish taxa (Dewey and Zigler 1996; Haines and Modde 1996; Labbe and Fausch 2000; Close and Jones 2002).

In the only published study we are aware of that contrasted VIE with IPD, Catalano et al. (2001) found similar survival rates for age‐0 largemouth bass Micropterus salmoides marked in opercular tissues with VIE (77%) and IPD (78%). In our experiment, similarity or dissimilarity of survival rates between mark types depended upon the species marked. There was no short‐term (≤80 d) marking effect for either species. Over the long term (200–240 d), however, the probability of survival for IPD‐marked Roanoke darters was significantly less than that of unmarked individuals, whereas survival differences for riverweed darters were not significant.

Although there is little literature available on prior IPD use, other injectable liquid marks have shown mixed effects on mortality in previous applications. Hill and Grossman (1987b) found that acrylic paints applied with syringes to a variety of coolwater species had no measurable effects on growth or mortality. Haines and Modde (1996) found that Colorado pikeminnows Ptychocheilus lucius tagged with tattoo ink experienced a higher mortality rate (10%) than did individuals tagged with VIE (<1%) or fin clips (0%). The investigators hypothesized that the tattoo needle used to apply the ink was too invasive, causing damage to the spinal column and internal organs of marked specimens. Goforth and Foltz (1998) marked yellowfin shiners Notropis lutipinnis with latex paint in both a laboratory and a field setting and found that mortality was elevated only when marked individuals were less than 45 mm standard length or when multiple marks per individual were applied. We found that multiple marks did not cause elevated mortality, as evidenced by the high survival rates of riverweed darters. We also saw no size‐specific effect of marking on mortality rates because the mean length of the fish that died in our experiment did not differ among control and marked groups (analysis of variance: F = 0.26, P = 0.7765). Furthermore, we suspect that our technique for applying IPD marks was not as invasive as the tattoo needle used by Haines and Modde (1996) because mortality was not immediate but, rather, increased gradually over time.

Because both marks used in this study are reportedly biocompatible, both were applied using similarly‐sized needles, and both treatment groups were assigned without known bias and were subjected to the same anesthetic and handling procedures, we can offer three possible explanations for the differences in survival among marks and species that we found. First, whereas VIE cured and thus “sealed” the entry wound created by the needle, IPD did not, and may have increased the potential for secondary infection (although gross visual analysis never confirmed any such infection associated with wounds). Goforth and Foltz (1998) recognized tagging scars months after latex paint injection, whereas we have not encountered this phenomenon in any published application of or personal experience with VIE. Second, differences in swimming behavior between species may have affected mortality rates. Marking‐related injuries to swimming muscles incurred during IPD application may have been irritated by the near‐constant swimming behavior of Roanoke darters in aquaria. Such injuries would have been less aggravated by the benthic, sedentary behavior of riverweed darters in aquaria. The swimming activity levels of Roanoke darters in aquaria did not reflect their behaviors instream, where the species almost exclusively maintains contact with the substrate (J. Roberts, personal observation). This artifact of the experimental environment may have exacerbated the effect of marking on fish survival. Finally, fin clip site (upper versus lower caudal fin) was confounded with mark type and may have caused differential mortality rates between treatments, particularly for more active Roanoke darters. Previous studies, however, have shown little effect of fin clipping on fish behavior and fitness (Coble 1967; Haines and Modde 1996; Conover and Sheehan 1999; Pratt and Fox 2002).

Effects of Mark Type, Location, and Color on Retention

For marking Roanoke darters, VIE emerged as a superior technology because it was retained with high frequency in the short term (≤80 d) and with significantly greater probability than IPD over the long term (200–240 d). However, retention rates for VIE (90%) and IPD (44%) fell within the range of other studies of injectable fluorescent marks (Dewey and Zigler 1996; Haines and Modde 1996; Bailey et al. 1998; Albanese 2001; Close and Jones 2002). For marking riverweed darters, either marking technology appears appropriate because both were retained at rates exceeding 96% in the short term and 78% over the long term. Catalano et al. (2001) found similar retention rates for VIE (84%) and IPD (67%) marks in age‐0 largemouth bass. When classified by mark quality, however, the investigators found a significantly higher proportion of easily visible VIE marks than they did of easily visible IPD marks. Haines and Modde (1996) found that VIE and tattoo marks in age‐0 Colorado pikeminnows were retained with similar frequency over a 21‐d period (>97% for both) but with very different frequencies at the end of 142 d (85% and 26%, respectively).

Mark location appeared important to retention in some cases but not in others. The VIE marks were retained with similar frequencies across body locations in both species; IPD exhibited a similar pattern in riverweed darters. In Roanoke darters, however, IPD marks were lost to differing degrees depending on where they were applied. Ventral marks were retained at a higher frequency than were dorsal marks, which is counter‐intuitive for a benthic fish that would probably experience more abrasion and skin regeneration in ventral tissues. However, as previously noted, Roanoke darters did not exhibit benthic behavior in aquaria. Hill and Grossman (1987b) and Goforth and Foltz (1998) found differential mark retention rates depending on application site, better retention generally occurring in large, fleshy target areas and lower retention in less fleshy areas and near major swimming muscles. Catalano et al. (2001) found variable retention of VIE and IPD across different fin locations, although anal fin marks performed best. Additionally, a given body location may be appropriate for marking in one species, but not in another. Albanese (2001) found that chin IPD marks performed well in torrent suckers Thoburnia rhothoeca (92% retention at 83 d) but poorly in fallfish Semotilus corporalis (10% retention at 65 d).

For both species, an interaction between mark type and mark color emerged. The VIE green was retained more frequently than was VIE yellow, whereas IPD yellow was retained more frequently than IPD green. Specimens were examined thoroughly with blue and UV light, and because of the somewhat translucent musculature of the study species, it is unlikely that marks went unseen. Because retention rates were not consistent within colors, differences in retention by color probably cannot be attributed to differences in mark readability, as was described in other studies (Hill and Grossman 1987b; Goforth and Foltz 1998; Close and Jones 2002). Rather, because differences in retention by color were consistent within types and across species and because the same individual (the senior author) applied all marks using the same equipment, differences are more likely attributable to variability in the chemical constituencies of the marks. Both marking materials are prepared by their respective manufacturers in batches, where the possibility for variance in viscosity, fluorescence, and biocompatibility emerges and may confound the results of experiments in which multiple batches are used, as in this study. Only a rigorous examination of the quality of marks among versus within batches would determine whether batch‐effects are significant.

Overall, VIE appears to be the safest, most persistent mark for darters. Survival and retention rates were high in this and in previous tests of VIE. Although IPD exhibited biases that may limit its long‐term effectiveness in some species, it should still be considered as a convenient short‐term mark. Future research should compare these marking technologies in other taxa, for which IPD may be preferable (Albanese 2001). Regardless of the marking material chosen, the possibility of variability in retention among locations and colors must be addressed. Furthermore, our findings highlight the difficulties of applying the results of marking studies across taxa. We therefore encourage the use of a pilot marking study to test all marking types, colors, and locations of interest before beginning field use of subcutaneous marks in a new fish species.

Acknowledgments

Financial support for this work was provided by the U.S. Army Corps of Engineers. We thank G. H. Galbreath, J. Griffin, J. M. Kilpatrick, and B. Blood for assistance in field collections and in the laboratory, as well as K. P. Kim and S. Miller of the Virginia Tech Statistical Consulting Center for aid in analyses. Earlier drafts of this manuscript benefited from the feedback of B. W. Albanese, C. A. Dolloff, and T. J. Newcomb, and later versions were improved by the comments of T. Labbe and J.R. Shute. Mention of trade names or commercial products does not constitute endorsement or recommendation for use by the U.S. Geological Survey.

References