Introduction

Although 55% of species in the catostomid family are imperiled (Harris et al. 2014), not much is known about Carpiodes life history traits. Updated life history information of the carpsuckers (members of the genus Carpiodes) is needed as new harvest demands emerge. The Quillback Carpiodes cyprinus is a widely distributed sucker (Catostomidae) in North America (Parker and Franzin 1991; Braaten and Guy 2002). Many native fishes that were not historically harvested in large numbers by recreational anglers, including the Carpiodes species, are now being pursued in significant numbers via the growing sport of bowfishing (Scarnecchia and Schooley 2020). Bowfisheries have expanded considerably in the past decade (Scarnecchia and Schooley 2020); however, resource management’s understanding of the fishery and the species that are targeted has not kept pace. Many states use a broad, non-species-specific classification (“rough fish”) to include many native fishes (often unrelated by phylogeny, life history, or habitat), including the Carpiodes, for fisheries management (Rypel et al. 2021). For example, Carpiodes of Minnesota have no harvest limits (MNDNR 2022). Although this approach may have been adequate when these species were not subject to significant recreational harvest, it may no longer be such during the bowfishing era (Scarnecchia and Schooley 2020). Thus, an updated understanding of the biology of these species is essential as future management is adapted to meet the evolving fishery and conserve valued native resources.

Age demographics have not been validated for the carpsuckers, as age assessments have been based largely on analysis of scales or fin rays for Quillback, and exclusively on these structures for river carpsucker Carpiodes carpio and highfin carpsucker Carpiodes velifer (Table 1). Resulting age data have indicated maximum longevity of these species to be approximately 10–12 years (Vanicek 1961; Walburg and Nelson 1966; Quist and Spiegel 2012; Fish Base 2022; AnAge Database 2022), despite contrary evidence from the only two studies of Carpiodes aged using otoliths (from Quillback) (Parker 1987; Grabowski et al. 2012). However, non-otolith structures produced underestimates of fish age in a number of other species (Beamish and McFarlane 1983; Casselman 1990; Campana 2001), including catostomids (Sylvester and Berry 2006; Terwilliger et al. 2010; Radford et al. 2021). Thus, allometric analysis of the various structures (otoliths vs. scales, etc.) in Carpiodes could reveal underlying differences between their growth rates, and a deeper understanding why non-otolith structures underestimate age in catostomids.

Table 1 Review of age and growth studies on the Carpiodes compared to the present study. Note that for multiple Carpiodes spp. in a single study, the data in each column are for all Carpiodes combined; *demographic information in columns for Parker (1987) and Grabowski et al. (2012) refers only to the specimens from which otoliths were used. RC, river carpsucker; Q, Quillback; HC, highfin carpsucker; MN, Minnesota; WI, Wisconsin; TL, total length; NS, not specified

In this study, we investigate the utility of Quillback otoliths for age estimation and characterize allometry of their otolith growth relative to other body structures. We quantify age demographics based on presumed annulus counts in thin-sectioned otoliths. Using this information, we model growth and sexual maturity for a population of Quillback in Minnesota and analyze their recruitment. We also document external black pigmentation on epidermal tissue that accrues with age in Quillback. These markings have never been documented or described for any Carpiodes spp.

Methods

Sample collection

We collected a total of 81 Quillback in this study. This sample was collected during April 2018 (n = 5), December 2019 (n = 2), January–February 2020 (n = 11), November 2020 (n = 13), and March 2021 (n = 50) from angler discards downstream of Orwell Dam (46° 12′ 51.5″N 96° 11′ 03.9″W) near Fergus Falls, Minnesota.

Body dissections

We measured size of individual fish, photographed each specimen, and then dissected them to extract otoliths and determine sex. We quantified size by wet body mass (± 0.01 kg) and total length (± 1 mm). We measured exposed scale length (± 0.1 mm) for the scale immediately above (i.e., dorsal to) the lateral line at the 6th posterior scale along the lateral line as well as the midpoint radial length (± 0.1 mm) of the operculum from each fish image (see Fig. 1) using ImageJ analysis software (Rasband 1997; Abrámoff et al. 2004). Immediately after removal by dissection, we placed the otoliths in microvials pre-filled with distilled water. For 68 of 81 Quillback collected from Orwell Dam, we also removed the gonads and obtained gonad mass (± 0.005 kg) (we did not obtain gonad mass from the December 2019 and January–February 2020 collected specimens because time was limited) and calculated the gonadosomatic index (GSI) as gonad mass/body mass.

Fig. 1
figure 1

Thin-sectioned lapillus otoliths of Quillback Carpiodes cyprinus, and morphometric transects. Examples range from 3 to 44 years for Quillback otoliths. Otolith inset shows close-up of 25-year-old’s recent 16 years of growth. Dots mark each presumed annulus; triangles mark each decade. Scale bar = 600 µm (does not apply to otolith inset). Note the well-defined presumed annuli. Body-structure measurement transects: white lines indicate exposed (a) operculum length and (b) scale length

Otolith analysis

In the lab, we processed extracted otoliths to obtain photographs of their whole structure and measurements of mass. We removed residual cranial tissue and other non-otolith material under a dissecting microscope. We photographed whole otoliths in water under a dissecting microscope at 50 × , using transmitted light in light-field mode. Inspection and photography under a dissecting microscope helped determine the core and primary growth axis of the otoliths (in preparation for potential sectioning). We then air-dried otoliths and calculated lapillus otolith mass using a microbalance (± 0.1 mg) following the protocol of A. Lackmann (pers. obs. 2021). We used lapillus otoliths for calculation of otolith mass because they are the largest otolith pair from Quillback, and because they were the otolith type used for sectioning. For each of the first five Quillback, only one lapillus otolith was measured for mass because the other otolith had inadvertently already been sectioned.

We thin sectioned lapillus otoliths for age analysis. We mounted lapillus otoliths in either ACE® quick-setting or Buehler epoxy, and then sectioned them using twin diamond-embedded blades on a Buehler IsoMet™ 1000 low-speed saw to produce 300-µm sections. We sectioned otoliths through the core and along the primary growth axis. We mounted sections on a glass slide, immersed them in mineral oil, and photographed them at 75 × on a compound microscope. If a thin section was not deemed readable by the primary reader (e.g., the section fractured), then another otolith from that specimen was thin sectioned until a readable section was obtained. We produced 83 thin sections to estimate the age for each of 81 fish.

We analyzed images of thin sections to quantify individual age and otolith length. Presumed annuli were digitally marked on images by independent readers following the age-reading protocol in Lackmann et al. (2019). Thus, if a section was determined unreadable or discrepant, then another otolith from that individual was sectioned until it was resolved. We assigned year classes to fish based on collection date and ages derived from the total presumed annuli marked on the thin-sectioned otolith images. For each image, we determined whether the marginal increment on the outer edge of the otolith thin section should be counted as a presumed annulus or not by considering the date of capture relative to probable hatch date, and then estimating age to the nearest year. We assumed a probable hatch date during May because Quillback spawn during May near this latitude (Harlan et al. 1951; Behmer 1964; Parker and Franzin 1991). We measured lapillus otolith length (± 1 µm) along a single transect for each otolith using ImageJ analysis software (Rasband 1997; Abrámoff et al. 2004). Measurement transects started from the otolith core and proceeded along the primary growth axis to the edge of most recent growth, following the general direction of the transects of Terwilliger et al. (2010).

Statistical analysis

We used a likelihood ratio test to determine if the observed sex ratio deviated significantly from 1:1. To characterize (1) body length, (2) exposed scale, (3) operculum, and (4–5) otolith (length and mass) allometry, we used analysis of covariance (ANCOVA) to quantify the effects of (natural log-transformed, henceforth log-transformed) body mass, sex, and the interaction of body mass and sex on the five (log-transformed) dependent variables. We conducted this same analysis once more, except with age as the independent variable instead of body mass. If effect terms were not significant in the ANCOVA, we constructed post hoc models including only significant effect terms to develop a final allometric model.

We used the von Bertalanffy growth function (von Bertalanffy 1938) to model size at age such that \(\mathrm{TL}={L}_{\infty }\bullet \left(1-{e}^{-k\bullet \left(\mathrm{age}-{t}_{0}\right)}\right)\), where TL is the total length (cm), age is in years, L is asymptotic TL (cm), parameter k is the instantaneous rate of increase (cm/cm/d) (Schnute and Fournier 1980), and parameter t0 is age (years) at length 0. We developed four models for total length based on combinations of parameters L and k that varied by sex, and t0 was fixed (at 0). We excluded models for which t0 was unconstrained for relative lack of individuals 5 years or younger (models would be unrealistic with highly negative t0 values). We used information-theoretic methods (Burnham and Anderson 2002) to determine the highest-ranked models based on the relative Akaike’s information criterion corrected for small sample sizes (ΔAICc) when comparing multiple models in a suite (Akaike 1973).

For analysis of Quillback recruitment, years 2013–2021 were excluded based on the distribution of year classes in the Orwell Dam sample, i.e., they made up the ascending limb of the catch curve distribution (Smith et al. 2012). Thus, we concluded that recruitment from 2013 to 2021 year classes could not be quantified using catch curve analysis. We used a catch curve from the combined 2018–2021 sampling, and again with only 2021 collected fish, to obtain estimates of the annual mortality rate (Maceina 2004), the recruitment coefficient of determination (RCD; Isermann et al. 2002), and the recruitment variability index (RVI; Guy and Willis 1995) for recruitment analyses. RCD values range from 0 to 1 because it is the R2 of a catch curve, with higher values indicating more consistent recruitment (Isermann et al. 2002). The RVI = [SN/(NM + NP)] – (NM/NP), where SN is the summation of the cumulative relative frequency distribution of all possible year classes starting with the most recent year class, with year classes not yet recruited to the fishery and those prior to the oldest fish excluded. NM is the number of missing year classes, and NP is the number of year classes present (NP must be greater than NM to calculate RVI). RVI values range from − 1 to 1 with values closer to 1 indicating more stable recruitment. Note that recruitment indices in the present study require caution in interpretation because sample size (n = 81) is on the smaller end, e.g., Isermann et al. (2002) set a minimum sample size threshold for catch curve analysis at n = 50 individuals. Otolith-derived recruitment information for Quillback has never previously been reported.

We classified individuals collected between November and April as mature/immature based on threshold GSI values determined from the GSI distributions for females and males, following the method of Trippel and Harvey (1991). We used logistic regression to estimate the age at onset of sexual maturity (i.e., the age at which probability of maturity is 0.5), and to model epidermal black spot presence versus age for which we also used a chi-squared test to evaluate the logistic regression. We used JMP 16 Pro Statistical Discovery™ (JMP Statistical Discovery LLC) for statistical analysis and graphical output.

Results

We estimated the ages for 81 Quillback from presumed annulus counts of thin-sectioned lapillus otoliths. Quillback ranged from 3 to 44 years old (e.g., Fig. 1). The first annulus was readily noted as the first concentric opaque ring around the nucleus (e.g., Fig. 1), as is the case for other age-validated catostomids (e.g., Terwilliger et al. 2010; Lackmann et al. 2019; A. Lackmann pers. obs. 2021). The overall between-reader aging precision had a mean coefficient of variation (CV) of 3.8%, a median of 3.5%, range of 0–20.2%, and a standard error of 0.5%. The average percent error (APE) was 2.7%, and the correlation of age estimates between the secondary versus primary reader was explained by an R2 = 0.98. Size of the Quillback ranged from 32.2 to 49.3 cm total length (TL) and 450 to 1430 g in mass. The Quillback were comprised of 63 females and 18 males. Thus, females comprised ~ 78% of the sample and the sex ratio was significantly different from 1:1 (\({X}^{2}\) = 25.5, df = 1, p < 0.001).

We found that lapillus otolith size (mass or length) increased with body mass or age at the highest rates compared to the growth rates of the other body structures measured (total body length, scale length, and operculum length). Specifically, the allometric slope coefficients for total body, scale, and operculum length (in relation to body mass) ranged from 0.26 to 0.32 (Fig. 2a–c), whereas the same slope coefficients for the otolith metrics were much higher and ranged from 0.83 to 1.04 (Fig. 2d–e); see Table 2 for all model and parameter details. Similarly, the allometric slope coefficients for total body, scale, and operculum length (in relation to age) ranged from 0.09 to 0.11 (Fig. 2f–h), whereas the same slope coefficients for the otolith metrics were much higher and ranged from 0.48 to 0.53 (Fig. 2i–j); see Table 3 for all model and parameter details. We also found that age explains more of the variation in the otolith metrics than does body mass (83–94% of the variation explained vs. 53–58% of the variation explained; see Tables 2 and 3 for specific details).

Fig. 2
figure 2

Allometry of Quillback (n = 81) total length, scale length, operculum length, otolith mass, and otolith length from Orwell Dam as a function of body mass (a-e) and age (f-j) of the post hoc models (see the “Results” section), which indicates that otolith size (i.e., mass and length) increases with both mass (a-e) and age (f-j) at proportionately greater rates than do total length, scale length, or operculum length in Quillback. Coefficient m = slope; circles and solid red line = females; triangles and dotted line = males; solid black line indicates no significant sex effect. See Tables 2 and 3 for statistics

Table 2 ANCOVA results for the log-transformed total length (TL), scale length (ScL), operculum length (OpL), otolith mass (OM), and otolith length (OL) as a function of log-transformed body mass (BM), sex, and the interaction of sex and body mass, and corresponding results for all models. Sample size is given by n, along with estimates (with SE in parentheses) for intercept (I), body mass (BM), female sex effect (S) and interaction (BM × S), F statistic and corresponding p value, and the coefficient of determination (R2). Factors with significant effects are indicated by *. p.h., post hoc
Table 3 ANCOVA results for the log-transformed total length (TL), scale length (ScL), operculum length (OpL), otolith mass (OM), and otolith length (OL) as a function of log-transformed age (A), sex, and the interaction of sex and age, and corresponding results for all models. Sample size is given by n, along with estimates (with SE in parentheses) for intercept (I), age (A), female sex effect (S) and interaction (A × S), F statistic and corresponding p value, and the coefficient of determination (R2). Factors with significant effects are indicated by *. p.h., post hoc

Quillback exhibit sexual dimorphism both in asymptotic length and in their rate of somatic growth. In the highest-ranked von Bertalanffy growth model (Fig. 3, Table 4), the estimated L of females was 44.1 cm, whereas L of males was 41.1 cm. Females reach their asymptotic length more slowly (k = 0.258) than males (k = 0.448). According to this model, females reach 90% of their asymptotic length by age nine, whereas males have attained 98% of their asymptotic length at the same age (Fig. 3). Thus, Quillback can potentially live decades with little to no growth in their total length.

Fig. 3
figure 3

Observed total length versus age for Quillback Carpiodes cyprinus collected from Orwell Dam (open circles for females, open triangles for males; n = 81). The highest-ranked (F4,76 = 28.8, df = 4, p < 0.001, R2 = 0.53) von Bertalanffy growth model is shown with lines (solid red line = females, dotted blue line = males), with different parameters for asymptotic length (L) between females (L = 44.1, 95% CI [43.3, 44.9]) and males (L = 41.1 [39.8, 42.6]), and growth rate (k) between females (k = 0.258 [0.230, 0.295]) and males (k = 0.448 [0.322, 0.648]). The age at 0 length parameter [t0] is fixed at 0 due to the absence of 0- to 2-year-old fish in the sample

Table 4 Model selection statistics for all von Bertalanffy growth functions in the TL vs. age model suite, (n = 81). NSS, not sex-specific; SSE, sum of squares error; k, number of model parameters; AICc, Akaike’s information criterion corrected for small sample sizes; ΔAICc, delta AICc; F, F statistic; p, p value; R2, coefficient of determination

We calculated year classes for every fish in this study, with recruited classes ranging from 1972 to 2018 overall. Quillback at Orwell Dam (fish collected in 2018–2021) were from year classes 1976–2018, with 91% of the fish collected in 2020–2021 (Fig. 4). Across the span of Quillback year classes (1976–2018), we found that the 1987 year class was most abundant (Fig. 4). Excluding the 2013–2021 year classes from this population (fish age < 9 years), 37 year classes are possible over this timeframe but only 19 are represented. After the dominant 1987 year class, there is a 7-year gap until the next Quillback year class (1994) is observed (Fig. 4).

Fig. 4
figure 4

Year class distribution of Quillback Carpiodes cyprinus at Orwell Dam (n = 81) by collection year. The 1987 year class was most frequently collected across the 4 years of the study

Recruitment indices for the population of Quillback at Orwell Dam indicated variable recruitment, with an estimated annual mortality rate less than 5%. We calculated an RCD of 0.41 when data from fish collected in all years (2018–2021) were combined and age in 2021 was used for all individuals, and an RCD of 0.45 when data from fish collected only in 2021 was used (Fig. 5). The corresponding RVI was − 0.21 for data from fish collected in all years combined, but an RVI could not be calculated for fish collected only in 2021 because NM = NP. Catch curve regression indicated the annual mortality rate to be 4.8–4.9% (Fig. 5).

Fig. 5
figure 5

Catch curves showing log-transformed year class count regressed by age of Quillback Carpiodes cyprinus at Orwell Dam. a All fish (n = 81) pooled across collection years (F1,17 = 12, df = 1, p = 0.003, R2 = 0.41) revealed an annual mortality rate of A = 4.9%. b Only 2021 collected fish (n = 50) plotted (F1,13 = 11, df = 1, p = 0.006, R.2 = 0.45) revealed A = 4.8%. Note these models exclude the ascending limb of the catch curve as is routine (Smith et al. 2012; open circles represent data that is excluded from each linear regression)

We estimated that for the population of Quillback at Orwell Dam, both sexes reach onset of sexual maturity at ages older than previously estimated. We binned GSI values at 1% intervals (females: n = 51, males: n = 17) and classified maturity status based on distinct gaps in the distributions. A total of 15 females had GSIs ≤ 3%, and the rest (n = 36) had GSIs between 6 and 21%. Thus, we defined females as mature if their GSI was above 5%. Using logistic regression, we estimate the age at which a female is 50% likely to have reached sexual maturity to be ~ 8.68 years (\({X}^{2}\) = 8.6, p = 0.003, R2 = 0.14), slightly older than the age of maturity estimated for female Quillback in Canada (6–8 years; Parker 1987). For males, GSI values ranged from 0 to 3%, with no obvious “break” in the distribution. However, there was a pronounced mode in the distribution for individuals with GSIs between 1 and 2%. Indeed, 14 of the 17 males had GSIs greater than 1%. Using a GSI of 1% as the threshold value for males, we found they reach onset of sexual maturity at ~ 7.98 years (\({X}^{2}\) = 6.8, p = 0.009, R2 = 0.43), which is older than an estimate for male Quillback in Canada (4–6 years; Parker 1987).

Black external, epidermal markings may accrue with age in Quillback (Fig. S1a). Results of the logistic regression indicated the presence of black spots increased significantly with age (\({X}^{2}\) = 13.9, df = 1, p < 0.001); however, neither the intercept estimate (− 27.5 ± 15.3; \({X}^{2}\) = 3.3, df = 1, p = 0.071) nor the age parameter estimate (− 0.7 ± 0.4; \({X}^{2}\) = 3.0, df = 1, p = 0.084) were significant. This result should be interpreted with caution because of the limited number of individuals with black spotting. Of the three Quillback older than 35 years in this study, two had black spotting, and black spots were not observed on any of the other individuals.

Discussion

Our age estimates derived from the lapillus otolith of Quillback are consistent with long-lived otolith-derived estimates in other age-validated catostomids. For example, bigmouth buffalo Ictiobus cyprinellus, a closely related species for which multiple otolith types have been age-validated, has a maximum age of 112 years (Lackmann et al. 2019; A. Lackmann pers. obs. 2021), Lost River sucker Deltistes luxatus has maximum age of 57 years, shortnose sucker Chasmistes brevirostris has maximum age of 33 years (Scoppettone 1988; Hoff and Logan 1997; Terwilliger et al. 2010), notchlip redhorse Moxostoma collapsum has maximum age of 19 years (Bettinger and Crane 2011), white sucker Catostomus commersonii has maximum age of 18 years (Thompson and Beckman 1995), brassy jumprock Moxostoma sp. has maximum age of 17 years (Bettinger and Crane 2011), and golden redhorse M. erythrurum has maximum age of 12 years (Beckman and Howlett 2013). The lapillus otolith has been validated for annulus formation in all of these species. Review of these studies indicated annulus deposition is a consistent and preserved trait, appearing as regularly spaced opaque bands along the primary growth axis in thin section, across the lapillus otolith of catostomids, including our observations for Quillback in the present study. Interestingly, Quillback are also a member of the Ictiobinae, a taxon of eight species for which extremely long-lived life history characteristics are becoming known. Bigmouth buffalo are the longest-lived freshwater teleost with longevity in excess of a century, and ages up to 56 years have been reported for black buffalo Ictiobus niger (Lackmann et al. 2019). In addition, Smallmouth buffalo Ictiobus bubalus have recently been shown to live up to 62 years of age (Snow et al. 2020). Nonetheless, age validation has yet to be demonstrated for Quillback and should be pursued following this study. This would also test whether the validated age-reading methodology of Lackmann et al. (2019; A. Lackmann pers. obs. 2021) has been accurately extended to another catostomid species (e.g., Quillback in the present study), expanding upon the recommendations described by Buckmeier (2002) for testing the generality of a validated age-reading methodology.

Use of otoliths (compared to other hard parts) provides the most accurate fish-age analysis. Otoliths are acellular, metabolically inert stones (not bone) that grow continuously throughout a fish’s life unlike other hard parts of fishes that are cellular, are prone to resorption, and are asymptotic in growth (Campana et al. 1995; Campana 2001, 2005). We found proportionately greater allometric scaling with age in otolith growth (both mass and length) compared to growth in scale length, operculum length, and total body length reveals why otoliths are especially useful for age analysis: they grow steadily throughout the life of the fish at rates that are higher than rates associated with other hard parts. We found that as Quillback age, otolith length has a comparable allometric scaling exponent (0.48) to otolith mass (0.54), whereas for scale length, operculum length, and total body length, the scaling exponents are below 0.12. Therefore, much of the growth in otolith mass across the life of Quillback is retained in the length measure of the otolith’s primary growth axis. This explains why presumed annuli can be so readily distinguished along the primary growth axis of the thin-sectioned otolith—there is adequate spacing (growth) along this plane to reveal presumed annuli throughout the lifespan of the fish.

Indeed, otolith growth continues to increase whereas growth in total body length is asymptotic. This relationship has also been found in several other fishes (Pino et al. 2004; Matić-Skoko et al. 2011; Andrews et al. 2011; Hanson and Stafford 2017; Radford et al. 2021; Pacheco et al. 2021). Age estimates derived from structures other than otoliths (e.g., scales, fin rays, cleithra) in otolith-bearing fishes have been shown to be inaccurate because of the asymptotic growth of the axial skeleton in a number of cases (Beamish and McFarlane 1983, 1987; Casselman 1990; Campana 2001). These results suggest linear growth essentially ceases in scales, opercula, and fin rays at the age an individual fish reaches maturity. Thus, using non-otolith structures in age analysis of Carpiodes could lead to systemic under-aging bias, and this appears to have been the case (Table 1). More specifically, Parker’s (1987) Quillback structure-comparison analysis found pectoral fin ray age estimates are only consistent with otolith age estimates until age 8, and Grabowski et al.’s (2012) Quillback structure-comparison analysis found that scale age estimates deviate from otolith age estimates by age 10. Over the past 68 years, more than 3000 river carpsucker have been aged (from either scales or fin rays) across their full range of sizes and throughout their range, but the oldest individual ever reported was an estimated 12 years old (Table 1). We have preliminary data for three river carpsucker aged using thin-sectioned otoliths, and all are older than the previously reported maximum age with one specimen estimated at more than 45 years old. In addition, our findings for Quillback are consistent with the findings by Terwilliger et al. (2010) for shortnose and Lost River suckers. A 65- to 70-cm fork length Lost River sucker could be between 14 and 57 years old (Terwilliger et al 2010). We found that a 41- to 44-cm total length Quillback could be between 8 and 44 years old.

A long-lived life history of up to 44 years for Quillback is supported by additional evidence. Parker (1987) found ages of up to 52 years in an otolith-aged subsample of Quillback from Manitoba, and Grabowski et al. (2012) found ages of up to 30 years in an otolith-aged subsample of Quillback from Florida (Table 1). We note however that in both of these studies, otoliths were ancillary to the other structures used for aging such as fin rays or scales, and thus age estimates from non-otolith structures biased the analyses. Nonetheless, evidence of longevity of more than 30 years in Quillback is not unprecedented.

We also found evidence indicating that Quillback recruitment is more variable, their mortality rate is lower, and age of onset of sexual maturity is older than previously reported. Based on our sample of 81 fish from a single population, it was revealed that 51% of possible year classes in a 37-year span were represented, with low estimates for RCD (0.41 or 0.45) and RVI (− 0.21), and an annual mortality rate of 4.8 or 4.9%. In Iowa populations, Quist and Spiegel (2012) estimated that Quillback lived 10 years maximum from analysis of fin rays. These populations had 75–88% of possible year classes represented, RCDs ranging from 0.59 to 0.79, RVIs ranging from 0.69 to 0.78, and annual mortality rates of 25–36% depending on site (Quist and Spiegel 2012). In the Orwell Dam population, we found a dominant year class hatched more than 30 years ago (from 2018 to 2021), observed a 7-year gap in evident recruitment between 1987 and 1994, and estimated sexual maturity at 8–9 years old. Further study of sexual maturation for Quillback at Orwell Dam is warranted given the lack of smaller, younger fish in our sample.

We found black spots accrue with age in Quillback, a phenomenon not previously documented for this species. We hypothesize that these black pigment spots are homologous to the black spots described on bigmouth buffalo (Lackmann et al. 2019). The spots are similar in appearance to spots on bigmouth buffalo, and our findings indicate the spots appear at approximately the same age (35–45 years), though larger sample sizes for Quillback are needed to confirm the ages at which spotting arises. We hypothesize that extended sun exposure is the mechanism underlying the increase in black pigmentation (melanosis) over time.

The long-lived life history characteristics of the Carpiodes are significant in the context of modern fisheries management in the USA. Night bowfishing’s sophisticated technologies and popularity (Fig. S1b) make it an effective form of harvesting native fishes that has virtually no regulation or monitoring (Scarnecchia and Schooley 2020; Scarnecchia et al. 2021). For instance, over the 12 years since night bowfishing was legalized in Minnesota, The Izaak Walton League reports that resource management agencies have tracked harvest and participation of two bowfishing tournaments (both in 2019; Winter and Johnson 2021). During these two 2019 tournaments, 75% of the species taken were native fishes, with 1 fish captured every 2–3 min per bowfishing team across each 10-h bowfishing tournament (Winter and Johnson 2021). Several bowfishing tournaments have occurred across Minnesota each year since at least 2015, with more than a dozen tournaments in 2021 (A. Lackmann pers. obs. 2015–2021). Because bowfishing has no mandatory effort or harvest-reporting requirements and fish can be taken with no limit, little is known about participation, harvest rates, or ecological impact—other than that it is a highly effective and growing way of targeting fish. The pursuit of Quillback by bowfishing anglers (e.g., Fig. S1b) represents a departure from the limited pursuit of this species by hook and line anglers (Becker 1983). For example, some bowfishers have expressed that Quillback are a favorite target to night shoot from big rivers near Minneapolis, Minnesota (A. Lackmann pers. obs. 2021). The impact of bowfishing on Carpiodes spp. populations is unknown because neither harvest nor effort is consistently quantified by any agency (Scarnecchia and Schooley 2020; Scarnecchia et al. 2021).

The overall management of the Carpiodes fishery can be improved with a few basic steps. For example, currently, the ecological impact of the commercial harvest of these species is unknown because it is not actively managed. Harvest data is self-reported by commercial fishers in many states including Minnesota (Scarnecchia et al. 2021). Distinct species in the Carpiodes genus are vaguely reported collectively as “carpsucker” or “quillback” in commercial harvest reports. In addition, Carpiodes species generally have no limits established on their harvest nationwide (Rypel et al. 2021), even as members of a family (Catostomidae) in which the majority of species are imperiled (Harris et al. 2014). Accurate life history information for Quillback and other Carpiodes spp. is essential for updating management for the fishery and conservation because their longevity and population parameters have likely been systemically underestimated historically. The possible long-lived population parameters of these species have evolved in response to freshwater survival pressures we have yet to fully comprehend. These species likely require extended time to bridge unfavorable recruitment conditions like has already been demonstrated for bigmouth buffalo (A. Lackmann pers. obs. 2021). Longevity overfishing is a well-known problem in marine fisheries (Beamish et al. 2006). There is no reason to doubt that this extends to long-lived fishes in freshwater ecosystems, especially in the face of exponentially increasing, yet unregulated and unlimited, bowfishing (Scarnecchia and Schooley 2020; Scarnecchia et al. 2021). The necessary next steps include (1) investigation of population demographics (via otolith-based aging) for more Quillback and Carpiodes species across their range to more broadly understand their life history traits; (2) implementation of these new life history parameters into fisheries management, because otolith-derived life history parameters are more accurate; (3) state agencies begin consistent collection of participation and harvest data from bowfishing tournaments and commercial fishing to determine the impact on each Carpiodes species; and (4) validation of life history parameters derived from otolith-based ages for all Carpiodes species.

Freshwater fishes are rapidly declining worldwide largely due to habitat degradation and human overexploitation (He et al. 2019; WWF 2021), yet funding and study remain inadequate for many of the species that require our attention (Guy et al. 2021). Species like Quillback continue to be classified by many state agencies in the broad, non-scientifically distinguished category “rough fish,” which perpetuates attitudes inconsistent with resource conservation (Rypel et al. 2021) even as these fishes have become sportfish for many anglers (Scarnecchia and Schooley 2020; Scarnecchia et al. 2021). This is doubly concerning when we are just beginning to understand the life history complexity of these native fishes, let alone the nuanced roles they play in freshwater ecosystems.