Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2022 Sep 8;101(5):1301–1311. doi: 10.1111/jfb.15201

Otoliths suggest lifespans more than 30 years for free‐living bowfin Amia calva: Implications for fisheries management in the bowfishing era

Alec R Lackmann 1,✉, Ewelina S Bielak‐Lackmann 2, Malcolm G Butler 2, Mark E Clark 1
PMCID: PMC9826520  PMID: 36053840

Abstract

The bowfin Amia calva is an amiid (Amiiformes) relict native to North America. It is the last surviving member of the Halecomorphi, a group of fishes that evolved more than 250 million years ago. Despite the phylogenetic significance of the amiids in vertebrate evolution, little has been published about their age and growth. Recreational bowfin harvest is currently unregulated throughout most of the USA, yet new recreational fisheries are emerging. As such, bowfin are increasingly harvested by sport bowfishing without limit, in addition to their growing commercial harvest for caviar. From 2017 to 2021 we studied a total of 81 bowfin from 11 populations across the east–west gradient of Minnesota within a narrow latitudinal margin (<50 km) of the 46th parallel north. We compared the allometry and translucence of bowfin asteriscus, lapillus and sagittal otoliths and found the lapillus otoliths provide consistent readability for age estimation despite being the smallest of the set. Size‐at‐age data derived from otoliths indicated that bowfin are sexually dimorphic in asymptotic length and may live up to 33 years, which is 15 years longer than previously estimated in wild populations, but comparable to what has been reported in captivity. Overall, 28% of the otolith‐aged fish were estimated as older than the previously reported maximum age for wild bowfin populations. Our findings suggest that the bowfin life history may exhibit slower growth, greater longevity, and more variable recruitment than previously recognized, which sets the stage for more otolith‐derived population demographics across their range and age validation study. Our results have direct implications for conservation of bowfin, especially amidst the increasing rates of exploitation during the bowfishing era.

1.

STATEMENT OF SIGNIFICANCE.

Bowfin otoliths have been abandoned in previous age studies. We present the first thorough analysis of all Bowfin otoliths and find that their lapilli and asterisci are especially useful for age estimation despite being smaller than the sagittae. Using otolith‐derived data we find that Bowfin may live 33 years, which is ~2–3 times longer than previously estimated for wild populations, but comparable to known‐age Bowfin lifespans in captivity. Our findings are especially relevant because Bowfin are becoming sportfish via bowfishing while also subject to a growing caviar fishery.

2. INTRODUCTION

Bowfin Amia calva is a fish species native to North America that is the only extant member of the order Amiiformes (Burr & Bennett, 2014). The bowfin is a top predator that tends to inhabit clear waters in vegetated bays (Carlander, 1969) that likely plays a crucial role in maintaining balance of its prey (Scarnecchia, 1992). The bowfin is capable of enduring variation in the quality of its habitat in part by obtaining oxygen from air ingested into the swim bladder at water temperatures above 10°C, with reports of individuals surviving weeks in saturated sediment lacking standing water (Becker, 1983). Bowfin provide parental care to their offspring, with males building a nest and guarding their brood for up to 2 months (Becker, 1983). Understanding the natural history of bowfin has considerable scientific value because their unique lineage holds clues to the evolution of vertebrates (Thompson et al., 2021). The bowfin is still known from all the regions where it was originally documented (Burr & Bennett, 2014).

Human exploitation of bowfin has increased in the 21st century. Over the past two decades there has been increased demand for bowfin roe (Davis, 2006; Koch et al., 2009b; Porter et al., 2014; Sinopoli & Stewart, 2021), increased participation in recreational angling for bowfin (Koch et al., 2009b; Porter et al., 2014), and increased harvest of bowfin through modern bowfishing (Lackmann et al., 2022, see Figure S1b; Scarnecchia & Schooley, 2020; Scarnecchia et al., 2021). Although the rate of exploitation appears to be increasing, much of the ecology of bowfin remains unknown. Indeed, Scarnecchia (1992) argued that the ecological function of bowfin has to be reassessed because this fish has long been subject to systemic neglect stemming from the ‘rough fish’ label.

Few bowfin populations have been investigated to quantify population demographics (Porter et al., 2014). Estimates of age at sexual maturity, mortality rate, lifespan, recruitment and growth require accurate age data. Age validation is always a goal, and in its absence age estimates should be reported with caution (Beamish & McFarlane, 1983). Although bowfin have been studied for age and growth determination, no age estimates by any method have been validated for this species. Despite several attempts (Cartier & Magnin, 1967; Davis, 2006; Koch et al., 2009a), otoliths have also not been used successfully for age estimation in this species. Thus, previous characterizations of bowfin demography have relied on age estimation using scales, vertebrae, opercula, gular plates or fin rays (Cartier & Magnin, 1967; Cooper & Schafer, 1954; Daniels, 1993; Davis, 2006; Hausmann, 1998; Holland, 1964; Koch et al., 2009a; Mundahl et al., 1998; Porter et al., 2014; Sanderson‐Kilchenstein, 2015; Schiavone Jr., 1982; Table 1). Based on these studies, it is presumed that bowfin live approximately a decade under natural conditions. However, hard body parts other than otoliths have been shown to underestimate ages in other species (Beamish & McFarlane, 1983; Campana, 2001; Casselman, 1990; Lackmann, Kettelhut, et al., 2021; Radford et al., 2021). Interestingly, there are at least three cases of captive bowfin living longer than any bowfin ever recorded in the wild: a 20‐year‐old (Flower, 1925), a 24‐year‐old (Flower, 1935) and a specimen in the New York Aquarium that lived for 30 years (Breder, 1936; reviewed by Randall & Delbeek, 2009). These captive bowfin of known age contradict the life expectancy estimated for wild populations using non‐otolith structures across their range.

TABLE 1.

Review of age and growth studies on bowfin Amia calva compared to the present study

Reference Structures attempted Structures used Age range (years) TL range (cm) n Region
Cooper and Schafer (1954) Scales Scales 3–7 ~47–65 39 Michigan
Holland (1964) Scales, vertebrae, opercula, branchiostegal rays, fin rays, gular plates Gular plates 1–10 NA 178 Missouri
Cartier and Magnin (1967) Scales, gular plates, otoliths Scales 1–9 22.1–71.8 80 Quebec
Schiavone (1982) Scales Scales 1–8 ~33–73 20 NY
Daniels (1993) Gular plates Gular plates 1–3 26.2–49.5 37 WV
Mundahl et al. (1998) Scales Scales 7–18 51.5–77 48 MN
Hausmann (1998) Gular plates Gular plates 1–6 51.6–73 38 Illinois
Davis (2006) Gular plates, otoliths Gular plates 1–10 31.6–73.6 288 LA
Koch et al. (2009) Fin rays, gular plates, scales, otoliths, branchiostegal rays Fin rays 1–13 39.2–80.7 255 Iowa
Porter et al. (2014) Fin rays Fin rays 0–5 23.3–68.3 82 Georgia
Sanderson‐Kilchenstein (2015) Fin rays Fin rays 1–8 28.9–79.1 51 NY
This study Otoliths Otoliths 2–33 40.1–72.7 81 MN
Open in a new tab

Note: The information from Holland (1964) was gleaned from the review by Sanderson‐Kilchenstein (2015) and Koch et al. (2009). WV, West Virgina; NY, New York; MN, Minnesota; LA, Louisiana.

In this study we investigated the utility of all three pairs of bowfin otoliths for age estimation, characterized bowfin body growth and analysed recruitment. We quantified age demographics based on presumed annuli in thin‐sectioned otoliths and differences in otolith allometry among the three otolith types to inform otolith type usage for age estimation. Based on our age estimates, we modelled the first otolith‐derived growth characteristics of bowfin and examined variation in recruitment using catch curve analysis, offering a fresh perspective on their biology.

3. MATERIALS AND METHODS

3.1. Sample collection

From August 2017 to October 2021 we collected 81 bowfin from night bowfishing discards (Figure 1) in Minnesota. For the bowfishing groups with which we collaborated, bigmouth buffalo (Ictiobus cyprinellus), common carp (Cyprinus carpio) and white sucker (Catostomus commersonii) were the preferred targets, with bowfin being opportunistic targets for the groups. Nonetheless, opportunistic bowfishing was effective at harvesting bowfin (Figure 1).

FIGURE 1.

FIGURE 1

Open in a new tab

Representative harvest of bowfin (top and bottom panel) collected via bowfishers from two nights of bowfishing in spring 2021 in Minnesota. These fish (n = 36) ranged from 40.1 to 70.3 cm total length, and 558–2667 g in mass. Fish board in each panel = 1 m

We collected bowfin from 11 sites across east–west Minnesota, but within a narrow latitudinal gradient (<50 km). From north‐western MN, lakes sampled included North Lida in 2017 (46°35′01.3″N, 95°58′32.2″W; n = 1), West Lost in 2018 (46°23′00.1″N, 95°52′25.0″W; n = 1) and Little Pelican (46°42′38.6″N, 95°57′25.4″W; n = 5) and Rush (46°36′41.1″N, 95°59′25.3″W; n = 2) in 2020–2021. From the northcentral region of the state, sites included the Mississippi River near Riverton (46°26′49.9″N, 94°05′50.5″W; n = 11) in 2018–2021, Lower Cullen Lake (46°32′19.6″N, 94°17′30.1”W; n = 10) in 2019–2020 and Clark Lake (46°30′36.5″N, 94°15′49.8”W; n = 1) in 2021. Finally, from north‐eastern MN, lakes sampled included Lower Island (46°39′57.5″N, 92°52′31.8”W; n = 3), Tamarack (46°40′01.3″N, 92°58′55.4″W; n = 16), Big Sandy (46°46′44.5″N, 93°19′10.1″W; n = 30) and Cedar (46°31′05.9″N, 93°47′08.9”W; n = 1) in 2021.

3.2. Body dissections

We measured the size of individual fish, and dissected to extract otoliths and determine sex via gonadal examination. We quantified fish size by wet mass (±1 g) and total length (±1 mm) either immediately after the fish was landed or within a few hours after the fish had expired. For those specimens in which wet mass was quantified a few hours of harvest, fish were again immersed in water for approximately a minute prior to obtaining mass. After extraction we placed otoliths immediately in microvials pre‐filled with distilled water. Because we were initially focused on only extracting the largest otoliths from bowfin, for eight individuals we extracted only sagittal otoliths, but for the remaining 73 fish we extracted sagittal, asteriscus and lapillus otoliths.

3.3. Otolith analysis

In the laboratory 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 (Figure 2). Inspection and photography under a dissecting microscope helped determine the location of the core and primary growth axis of each otolith type prior to sectioning. We air‐dried otoliths and measured sagittal, asteriscus and lapillus otolith mass on a microbalance (±0.1 mg). A few otoliths were sectioned without obtaining an otolith mass because we had yet to adopt otolith weighing as standard protocol during pioneering stages of the study. In individuals for which both otoliths of a pair were measured for mass, we used the greatest mass of the two in subsequent analyses. We used this protocol because we assume minor amounts of otolith material could sometimes be lost during extraction and handling.

FIGURE 2.

FIGURE 2

Open in a new tab

An otolith set from an individual bowfin Amia calva. Although lapillus otoliths are the smallest pair in this species, they were most useful in estimating age. Scale bar = 3 mm

We thin sectioned processed otoliths for age analysis. We embedded otoliths in either ACE® quick‐setting or Buehler epoxy, 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. Across all specimens (n = 81), we produced and aggregate total of 101 thin sections.

We analysed images of thin sections to estimate individual age. Presumed annuli were digitally marked on images by two independent readers following the age‐reading protocol in Lackmann et al. (2019). Each reader was experienced, having processed and age scored thousands of otolith thin sections across several dozen North American freshwater fish species, with considerable experience at age validation (Lackmann et al., 2019, 2021b). High‐resolution composite images of the entire thin section were age scored (each presumed annulus marked) along the primary growth axis of each thin‐sectioned otolith. If a thin section was deemed unreadable by the primary reader, then another otolith from that specimen was thin sectioned until a readable section was obtained. 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, that is, rounding the estimated age to the nearest year while presuming a probable hatch date of mid‐May because bowfin spawn once annually during spring at this latitude (Carlander, 1969; Cartier & Magnin, 1967).

3.4. Statistical analysis

We used the von Bertalanffy growth function (von Bertalanffy, 1938) to model length at age such that TL=L∞∙1−e−k∙age−t0, where TL is total length (mm), age is in years, L ∞ is asymptotic total length (mm), parameter k is the instantaneous rate of increase (mm/mm/d) (Schnute & Fournier, 1980) and parameter t 0 is age (years) at length 0. We also estimated ω=L∞∙k, where ω is the early‐life growth rate, also known as the growth rate near t 0 (Gallucci & Quinn, 1979). We developed four models for total length based on combinations of parameters L ∞ and k that varied by sex, while t 0 was fixed (at 0). We did not estimate t 0 because the lack of individuals younger than 2 years of age would result in unrealistic (highly negative) t 0 values. We used information‐theoretic methods (Burnham & Anderson, 2002) to determine the highest ranked models for size at age based on the relative Akaike's Information Criterion corrected for small sample sizes (ΔAICc) when comparing multiple models in the von Bertalanffy model suite (Akaike, 1973). To characterize otolith allometry, we used Analysis of Covariance (ANCOVA) to quantify the effects of (natural log‐transformed) body mass (kg), sex and their interaction on (natural log‐transformed) otolith mass (mg). Similarly, we used ANCOVA to quantify the effects of (natural log‐transformed) age (years), sex and their interaction on (natural log‐transformed) otolith mass (mg). We used JMP 16 Pro Statistical Discovery™ for statistical analysis and graphical output.

For recruitment analysis we used catch curves (Maceina, 2004) and the recruitment coefficient of determination (RCD; Isermann et al., 2002). We analysed only 2021 collected fish to standardize the age distribution, and we excluded year classes 2017–2021 or 2013–2021, depending on the analysis. We used two different exclusionary ranges based on alternative assumptions of the age at which bowfin recruit to the bowfishery (age of 5 years versus age of 9 years). RCD values range from 0 to 1 because this metric is the R 2 of a catch curve, with higher values indicating more stable recruitment (Isermann et al., 2002). We pooled bowfin across sites for both recruitment and growth analyses because this allowed for a large enough sample to be analysed for any general trends. We concluded this was warranted since this is the first such otolith‐derived data for bowfin, and because all bowfin were collected within a narrow latitudinal gradient (<50 km). Furthermore, 93% of the 2021 collected fish were from hydrologically connected waters along the Mississippi River, whereas the other 7% came from adjacent waters within the Pelican River watershed (Hudson Bay drainage). Nonetheless, the results of these growth and recruitment analyses should be interpreted with caution because they are potentially limited in scope.

4. RESULTS

We estimated the ages of 81 bowfin in this study from presumed annulus counts of the otolith thin sections, with ages ranging from 2 to 33 years old (e.g., Figures 3, 4, 5). The size of these individuals ranged from 40.1 to 72.7 cm TL and 558–3329 g in mass, with 56 females and 25 males. Across all three otolith types, between‐reader aging precision had a coefficient of variation (CV; Campana et al., 1995) of 6.3%. We sectioned 20 sagittae for which eight sections were unreadable. The CV for the readable 12 sagittae was 7.0% and these fish ranged from 5 to 23 years in age (e.g., Figure 3). Because 40% of the sectioned sagittae were unreadable, we examined the lapilli for age determination (e.g., Figure 4a). We sectioned 71 lapilli for which 100% were readable ranging from 2 to 33 years old with a CV of 6.2%. We also examined asterisci for age determination by sectioning a subsample (n = 10) and found 90% were readable ranging from 9 to 33 years old with CV of 6.1% (e.g., Figure 5). In cases where multiple otolith types produced readable sections from a given individual, the age scores were consistent across otolith types (e.g., Figure 3, 4, 5). For the lapillus otolith, the otolith type that was most used because it most consistently produced readable sections, an average percent error (APE) of 4.4% was calculated between reader age scores, while the overall agreement of age scores between the secondary and primary reader was a Pearson correlation coefficient = 0.98 (Figure 4a). More specifically, the same age score (exact agreement) was assigned to 30% of all lapillus otolith sections, 76% were in agreement to within 1 year (or fewer), while 90% were in agreement to within 2 years (or fewer), while the remaining 10% differed by 3 (n = 6) or 4 (n = 1) in absolute age score difference. The APE for the lesser‐used sagittal and asteriscus otoliths was 4.9% and 4.3%, respectively.

FIGURE 3.

FIGURE 3

Open in a new tab

A comparison of lapillus versus sagittal otolith thin sections from bowfin Amia calva. (a–d) Images of thin‐sectioned otoliths from four individuals, three of which (a–c) had both otolith types sectioned. (a) Images from both sections are comparable in quality and yielded identical age estimates. (b) Presumed annuli in the image of the lapillus thin section are clearly defined, unlike presumed annuli in the image of the sagittal otolith section for the most recent years of growth, and the images yielded different age estimates. (c) The sagittal otolith thin section image was unreadable for this individual whereas the lapillus thin section image revealed consistent presumed annuli for an age estimate. (d) Although 40% (eight of the 20 sagittal otoliths sectioned) of sectioned sagittae were unreadable, some yielded clearly defined presumed annuli like this estimated 23‐year‐old specimen and are therefore potentially useful for age estimation if the lapillus otolith is damaged or lost during extraction. White dots denote presumed annuli; triangles mark decades. Scale bar = 600 μm for all

FIGURE 4.

FIGURE 4

Open in a new tab

(a) Thin‐sectioned lapillus otoliths of Bowfin Amia calva. Examples range from 3 to 30 years. Dots mark each presumed annulus; triangles mark each decade. Note the inset for the estimated 30‐year‐old individual revealing the last 20 years of this fish's ontogeny. Scale bar = 600 μm (does not apply to inset or to the age bias plot). Note the well‐defined presumed annuli. (b) Age bias plot of the secondary vs. the primary reader for the lapillus otolith (n = 71) with the 1:1 line also shown for reference (dotted grey line). The Pearson correlation coefficient of Reader 2 vs. Reader 1 was an r = 0.98

FIGURE 5.

FIGURE 5

Open in a new tab

Thin‐sectioned asteriscus otoliths of Bowfin Amia calva. Examples of an estimated 30 and 33‐year‐old individual. The 30‐year‐old from Figure 4 (lapillus) is also the 30‐year‐old specimen in this figure (asteriscus). Dots mark each presumed annulus; triangles mark each decade. Scale bar = 600 μm. Note the well‐defined presumed annuli

Analysis of covariance (ANCOVA) of otolith allometry indicated that the lapillus otolith grows proportionately more than the other otolith types (sagittal or asteriscus) in bowfin. We found that mass of the lapillus otolith increases at a greater rate than mass of the sagittal or asteriscus as bowfin increase in age or as they increase in total body size (Figure 6 and Table 2). We also found that estimated age explained more variation in otolith mass (83%–92% depending on otolith type) than did total body size (60%–62% depending on otolith type) and that it was mass of the lapillus otolith that explained the most variation in age (92%) compared to the other otolith types (Table 2).

FIGURE 6.

FIGURE 6

Open in a new tab

Allometry of bowfin Amia calva otoliths as a function of age (top panels) and body mass (bottom panels) of the final models (see results). Analysis of covariance of natural log‐transformed otolith mass (mg) versus natural log‐transformed age (years) and natural log‐transformed mass (kg) reveals otolith growth is proportionately greatest (as indicated by values of the slope) in the lapillus. Coefficient m = slope; circles and solid red line = females; triangles and dotted line = males; solid black line indicates no significant sex effect. See Table 2 for statistics

TABLE 2.

ANCOVA statistics for the natural log‐transformed otolith mass vs. natural log‐transformed age and body mass (BM) analysis (in conjunction with Figure 6)

Model n I m SexF Sex*AgeF F P R 2
Sagittal vs. age 80 3.4 (0.06) 0.426 (0.025) 0.14 (0.02) NS 125 <0.001 0.83
Sagittal vs. age p.h. 80 3.4 (0.06) 0.425 (0.025) 0.14 (0.02) 188 <0.001 0.83
Lapillus vs. age 61 1.4 (0.06) 0.594 (0.024) 0.10 (0.02) −0.06 (0.02) 218 <0.001 0.92
Aster. vs. age 73 2.3 (0.06) 0.448 (0.022) 0.13 (0.02) NS 162 <0.001 0.88
Aster. vs. age p.h. 73 2.3 (0.06) 0.448 (0.022) 0.13 (0.02) 247 <0.001 0.88
Sagittal vs. BM 80 4.1 (0.05) 0.707 (0.077) NS NS 40 <0.001 0.61
Sagittal vs. BM p.h. 80 4.0 (0.05) 0.736 (0.068) 116 <0.001 0.60
Lapillus vs. BM 61 2.4 (0.06) 0.978 (0.100) NS NS 52 <0.001 0.64
Lap. vs. BM p.h. 61 2.4 (0.06) 0.908 (0.092) 98 <0.001 0.62
Aster. vs. BM 73 3.1 (0.05) 0.739 (0.077) NS NS 42 <0.001 0.64
Aster. vs. BM p.h. 73 3.0 (0.05) 0.741 (0.068) 118 <0.001 0.62
Open in a new tab

Note: Aster, asteriscus; I, y intercept; NS, not significant; m, slope; p.h., post‐hoc; SexF, shift in y intercept of females compared to males; Sex*AgeF, change in m of females relative to males. Number in parenthesis beside estimate is the standard error.

Bowfin exhibit sexual dimorphism in asymptotic body length, but not in their rate of somatic growth. In the highest‐ranked von Bertalanffy growth model (Figure 7 and Table 3), the estimated L ∞ was 64.5 cm for females compared to 58.0 cm for males, but there was no difference in the instantaneous rate of increase (k = 0.454). According to this model, Bowfin reach ~98% of asymptotic length by an estimated age of 9 years and may live decades at asymptotic size (Figure 7). We found that the early‐life growth rate (ω) was 29.3 cm/year for females vs. 26.3 cm/year for males.

FIGURE 7.

FIGURE 7

Open in a new tab

Total length versus age for female (circles) and male (triangles) bowfin Amia calva in this study for which a length and age was determined (n = 80; one of our 81 specimens was not measured for TL). Shown is the highest‐ranked von Bertalanffy growth model (F 3,76 = 73.6, d.f. = 4, P < 0.001, R 2 = 0.66) for females (solid red line) and males (dotted blue line) in which parameters for asymptotic length (L ∞ ) differed by sex [females: L ∞  = 64.5, 95% CI (63.4, 65.7); males: L ∞  = 58.0 (56.1, 59.9)] but instantaneous growth rate [k = 0.454 (0.386, 0.549)] did not; age at 0 length was fixed (t 0 = 0)

TABLE 3.

Model selection statistics for all von Bertalanffy growth functions in the TL vs. Age model suite, (n = 80)

Model SSE k AICc ΔAICc F P R 2
k NSS, t 0  = 0 1126 3 220.1 0.0 74 <0.001 0.656
Global with t 0  = 0 1116 4 221.7 1.6 49 <0.001 0.659
L ∞ NSS, t 0 = 0 1646 3 250.5 30.4 38 <0.001 0.498
L ∞, k NSS, t 0 = 0 1796 2 255.2 35.1 64 <0.001 0.452
Open in a new tab

Note: Global has sex‐specific L ∞ and k. NSS, not sex‐specific.

Catch curve analysis indicated variable bowfin recruitment, with annual mortality rate of 5%–7% for recruits. Estimated year classes in this study ranged from 1988 to 2019 for fish collected from 2017 to 2021 (with 74% collected in 2021) (Figure 8). The 1997 and 2011 year classes comprised 25% of the sample, and were the only year classes represented at four or more sites (Figure 8b). Although 44% of the bowfin collected by bowfishers from Tamarack Lake were estimated to be 2–6 years old, 89% of the bowfin collected by bowfishing from other sites were estimated to be 9 years or older (Figure 8b). The RCD for bowfin was 0.34 if individuals recruit by age 5 and 0.49 if individuals recruit by age 9 (Figure 9), values which Isermann et al. (2002) classify as moderately variable recruitment. We estimated an annual mortality of 5.3% if individuals recruit to the bowfishery by an estimated age of 5 years and 7.1% if individuals recruit to the bowfishery by an estimated age of 9 years (Figure 9).

FIGURE 8.

FIGURE 8

Open in a new tab

Year class distribution of otolith‐aged Bowfin Amia calva (n = 81) collected by bowfishers from 2017 to 2021 in Minnesota from 11 sites across a narrow latitudinal gradient (<50 km) by drainage (a) and waterbody (b). Inline graphic Mississippi. Inline graphic Red. Inline graphic Big Sandy. Inline graphic Cedar. Inline graphic Clark. Inline graphic Lower Island. Inline graphic Lower Cullen. Inline graphic Little Pelican. Inline graphic North Lida. Inline graphic Riverton. Inline graphic Rush. Inline graphic Tamarack. Inline graphic West Lost

FIGURE 9.

FIGURE 9

Open in a new tab

Catch curves of natural log transformed year class count versus age of bowfin collected in 2021 assuming (a) bowfin recruit to the fishery by an age of 5 years (F 1,18 = 9.3, d.f. = 1, P = 0.007, R 2 = 0.34 based on a sample of n = 57 fish) with estimated annual mortality rate of A = 5.3%, and (B) bowfin recruit to the fishery by an age of 9 years (F 1,15 = 14.4, d.f. = 1, P = 0.002, R 2 = 0.49) suggesting a higher annual mortality rate of A = 7.1%

5. DISCUSSION

To our knowledge, this is the first study in which otoliths have been successfully used to estimate the age of bowfin. Evidence from presumed annuli in thin‐sectioned otoliths indicates that previously reported longevities for bowfin from wild populations are perhaps underestimates, but this remains to be conclusively determined. We found evidence that suggests bowfin may live 33 years, which is 15 years older than the maximum age previously estimated for wild populations (Mundahl et al., 1998). Interestingly, several captive bowfin for which actual ages are known (20–30 years; Breder, 1936; Flower, 1925, 1935) are comparable in lifespan to what we estimate in this study, which supports the credibility of our otolith‐derived age estimates. In this study we estimate that more than 25% of the bowfin collected were at least 20 years old, with some individuals exceeding 30 years of age. We conclude that evidence indicates bowfin may live multiple decades in the wild, contrary to all reports from other wild populations of this species, all of which used non‐otolith structures to estimate ages.

Previous studies on bowfin population demographics are characterized by younger age distributions, with age estimations based on analysis of structures other than otoliths. Of a combined total of 1116 bowfin (TL ranging from 22.1 to 80.7 cm) aged using non‐otolith structures (Cartier & Magnin, 1967; Cooper & Schafer, 1954; Daniels, 1993; Davis, 2006; Hausmann, 1998; Holland, 1964; Koch et al., 2009a, 2009b; Mundahl et al., 1998; Porter et al., 2014; Sanderson‐Kilchenstein, 2015; Schiavone Jr., 1982), at least 1068 (≥96%) were estimated to be 13 years or younger, and none was more than 18 years old (Table 1). Age estimates derived from hard parts other than otoliths have been shown repeatedly to underestimate otolith‐determined ages in a number of species (Beamish & McFarlane, 1983, 1987; Campana, 2001; Casselman, 1990; Lackmann et al., 2021a; Radford et al., 2021; Sylvester & Berry Jr., 2006), including holosteans (Buckmeier et al., 2018). Our findings suggest this may be the case for bowfin as well, but a future study that tests the validity of age estimates in bowfin is necessary to prove if this is has likely been the case.

Prior to our findings, attempts to estimate age from otoliths in bowfin have been unsuccessful. Several previous attempts (Cartier & Magnin, 1967; Davis, 2006; Koch et al., 2009a) abandoned use of otoliths due to unspecified difficulties with these structures. Cartier and Magnin (1967) do not specify which otolith type (sagitta, asteriscus or lapillus) was used, but regardless did not proceed with whole otoliths and instead used bowfin scales. We assume Cartier and Magnin (1967) used the sagittae because these are the largest otoliths in bowfin. Davis (2006) extracted sagittae and viewed them both whole and sectioned under a dissecting microscope, but was unable to estimate age from either approach. Koch et al. (2009a) reference the extraction of all otolith types, viewing thin sections under a dissecting microscope, but were also unable to use bowfin otoliths to estimate age. Neither Davis (2006) nor Koch et al. (2009a) report how many fish were examined, and Koch et al. (2009a) do not report which otolith type(s) they used before abandoning efforts to estimate bowfin age using the otolith.

Contrary to these earlier studies, we used all otolith types for age estimation of bowfin. We conclude that lapillus otoliths are the most useful structure because 100% of thin sections were readable, followed by asterisci in which 90% were readable and then sagittae in which 60% were readable. This despite the lapillus being the smallest otolith in bowfin. Typically, the largest otolith type is used in age analysis because it is the easiest to extract and process. However, we found the lowest utility for bowfin sagittae according to percentage of otolith sections that were readable.

We evaluated otolith allometry, which helped inform why lapillus and asteriscus otoliths from bowfin are more useful in age estimation compared to bowfin sagittal otoliths. Allometric patterns of otolith growth in bowfin are consistent with our findings about the relative reliability of different otolith types for age estimation in this species. As a bowfin grows older and larger, mass of the lapilli increases at a higher rate than mass of either the asterisci or sagittae. Thus, proportionately more otolith growth happens in the bowfin lapillus as the fish ages and grows, which results in relatively more space to distinguish presumed annuli along the primary growth axis of a thin section. Based on these findings, we recommend analysing whole otolith mass as a function of body mass (and estimated age, if known) among otolith types to guide selection of which otolith type to use for age estimation rather than using absolute otolith size alone. Allometric analysis may aid otolith type assessments in other species with similar otolith characteristics. In addition, high opacity in bowfin sagittae was more likely to obscure presumed annuli compared to the other otolith types, presumably due to higher concentrations of the aragonitic polymorph of calcium carbonate noted in the sagittal otoliths of other species (Campana, 1999, 2005). On the other hand, the asteriscus is known to be primarily composed of vateritic calcium carbonate, which is more translucent (Campana, 1999), and this may explain why presumed annuli were more discernible in bowfin asterisci, compared to their generally more opaque sagittae.

We found that bowfin grow to 40–50 cm total length in their estimated first 2–3 years of life, and reach asymptotic lengths by an estimated age of 8–10 years. We presume that non‐otolith structures such as scales, gular plates and fin rays plateau in growth similar to the linear axis of the skeleton and discernible annuli diminish around this same age threshold. Interestingly, the maximum age of other bowfin studies was typically 10 years (Table 1) using non‐otolith structures. Previous work on bowfin documented higher maximum ages for female bowfin than for males (Davis, 2006; Koch et al., 2009a; Porter et al., 2014; Sanderson‐Kilchenstein, 2015). In this study we found that the three oldest bowfin (30–33 years) were males. Bowfin size is sexually size‐dimorphic (with males smaller than females), consistent with previous findings (Cartier & Magnin, 1967; Cooper & Schafer, 1954; Daniels, 1993; Davis, 2006; Hausmann, 1998; Holland, 1964; Koch et al., 2009a, 2009b; Mundahl et al., 1998; Porter et al., 2014; Sanderson‐Kilchenstein, 2015; Schiavone Jr., 1982). However, bowfin growth rates (k = 0.45) from the von Bertalanffy models reported here correspond to smaller total length at age, and therefore a slower growth rate, than reported for other populations (Cartier & Magnin, 1967; Cooper & Schafer, 1954; Daniels, 1993; Davis, 2006; Hausmann, 1998; Holland, 1964; Koch et al., 2009a, 2009b; Porter et al., 2014; Sanderson‐Kilchenstein, 2015; Schiavone Jr., 1982). Age estimation using non‐otolith structures may be more likely to underestimate ages of male bowfin, in light of otolith allometry and males' smaller body size. This should be tested. Given that bowfin may be entering the bowfishing fishery as early as 4–5 years of age, new information on gonad size at age using otolith‐estimated ages is needed to estimate rates of sexual maturation in bowfin relative to their vulnerability to harvest.

The year class distribution and recruitment analysis from otolith‐aged bowfin suggests recruitment of this species is more complex and variable than patterns reported in prior studies, where ages were derived from hard parts other than otoliths (e.g., Koch et al., 2009b). Again however, these recruitment results are contingent on the assumption of unbiased harvest beyond a certain threshold (Isermann et al., 2002). Since such assumptions have yet to be tested for bowfishing, the recruitment results we document should be interpreted with caution. More otolith‐based age data are needed to understand bowfin recruitment patterns, both within and among sites, and across the species' range. We note that this study's sample was from a pooled bowfishing sample, which presents a non‐negligible risk of being a size‐selective sample that is limited in its recruitment inferences (e.g., the estimates are conditional on the assumption that the sample can be generalized as a single population). Thus, the recruitment results that we document require caution in interpretation. Nonetheless, bowfishing is a new and growing sportfishery in need of management (Bennett et al., 2015; Lackmann et al., 2019, 2022; Lackmann et al., 2021b; Quinn, 2010; Scarnecchia et al., 2021; Scarnecchia & Schooley, 2020, 2022; York et al., 2022), so the bowfishing‐specific results that we document are especially relevant. Bowfin are increasingly being targeted by bowfisheries nationwide, with most targeted fish being discarded (Scarnecchia & Schooley, 2020), while also a growing target of a caviar industry (Sinopoli & Stewart, 2021). We recommend that efforts be made to obtain otolith data from discarded bowfin. Since otoliths are widely documented as the most accurate structure for estimating age across fishes (Campana, 2001), and bowfin otoliths were not thoroughly analysed before, the data that we present could possibly be the most credible age‐demographic information on this species to date. Nonetheless, future study is needed to definitively document if this is the case (e.g., age validation; Beamish & McFarlane, 1983).

Further study of bowfin population demographics is warranted throughout their range, especially in the context of modern harvest pressures. The ability to sustainably manage fish stocks requires an accurate understanding of the life history and the exploitation rates of the species involved, neither of which is well understood for bowfin. Crucial next steps for reassessing the ecology of bowfin include (1) validating the accuracy of otolith‐estimated bowfin ages, (2) investigating bowfin age structure using otoliths from other populations, (3) obtaining information on otolith‐estimated population age structure and age at maturity, and (4) quantifying bowfishing harvest rates while capitalizing on data available from discarded fish (e.g., study bowfin shot at tournaments). We demonstrate a process to estimate bowfin ages from their otoliths, particularly using thin‐sectioned otoliths, which sets the stage for future age validation work.

Bowfin management is overdue (Scarnecchia, 1992), particularly as this species has become a targeted sportfish with the rise of modern bowfishing (Scarnecchia et al., 2021; Scarnecchia & Schooley, 2020). The approach we outline here, in which bowfishing discards are used to evaluate population structure using age estimates from otoliths, and using otolith allometry to inform otolith type assessments, also provides a blueprint for evaluations in other species now vulnerable to this rapidly growing sport fishery (e.g., Bennett et al., 2015; Lackmann et al., 2019, 2022; Quinn, 2010). Species like bowfin generally have no limits on their take and continue to be derogatorily classified as “rough fish” (Rypel et al., 2021; Sinopoli & Stewart, 2021), despite being important to the native environment of North America (Scarnecchia, 1992; Burr & Bennett, 2014). Unlimited take of wild freshwater fish with no active management amidst increasingly popular and efficient forms of exploitation warrants attention. We are just beginning to understand the complexities of these native species and the roles they play in dynamic ecosystems.

ETHICS STATEMENT

We have treated all animals in accordance with North Dakota State University and University of Minnesota Duluth Institutional Animal Care and Use Committee (IACUC) guidelines on animal care as specified in each protocol (IACUC protocols A17007 and 2007‐38272A). The collection of specimens was conducted in accordance with all applicable laws, guidelines and regulations. Specimen collection permits for this study were granted by the Minnesota Department of Natural Resources (fisheries research licence numbers: 23775, 28375, 29788, 32489).

ACKNOWLEDGMENTS

We would like to thank the bowfishers who donated fish for study, and Becky Smith for assistance obtaining otolith masses. There is no funding to declare. There is no conflict of interest declared in this article. We would like to thank the University of Minnesota Duluth Department of Biology.

Lackmann, A. R. , Bielak‐Lackmann, E. S. , Butler, M. G. , & Clark, M. E. (2022). Otoliths suggest lifespans more than 30 years for free‐living bowfin Amia calva: Implications for fisheries management in the bowfishing era. Journal of Fish Biology, 101(5), 1301–1311. 10.1111/jfb.15201

Funding information University of Minnesota Duluth

REFERENCES

  1. Akaike, H. (1973). Information theory as an extension of the maximum likelihood principle. Proceedings of the International Symposium on Information Theory, 2, 267–281. [Google Scholar]
  2. Beamish, R. , & McFarlane, G. A. (1983). The forgotten requirement for age validation in fisheries biology. Transactions of the American Fisheries Society, 112, 735–743. [Google Scholar]
  3. Beamish, R. J. , & McFarlane, G. A. (1987). Current trends in age determination methodology. In Summerfelt R. C. & Hall G. H. (Eds.), Age and growth of fish (pp. 15–42). Ames: Iowa State University Press. [Google Scholar]
  4. Becker, G. C. (1983). Fishes of Wisconsin. Madison, WI: University of Wisconsin Press. [Google Scholar]
  5. Bennett, D. L. , Ott, R. A. , & Bonds, C. C. (2015). Surveys of Texas bow anglers, with implications for managing alligator gar. Journal of the Southeastern Association of Fish and Wildlife Agencies, 2, 8–14. [Google Scholar]
  6. Breder, C. M. (1936). Long‐live fishes in the aquarium. Bulletin of the Zoological Society, 39, 116–117. [Google Scholar]
  7. Buckmeier, D. L. , Snow, R. , Smith, N. G. , & Porter, C. (2018). Are age estimates for longnose gar and spotted gar accurate? An evaluation of sagittal otoliths, pectoral fin rays, and branchiostegal rays. Transactions of the American Fisheries Society, 147(4), 639–648. [Google Scholar]
  8. Burnham, K. P. , & Anderson, D. R. (2002). Model selection and multimodel inference (2nd ed., pp. 60–66). New York, NY: Springer‐Verlag. [Google Scholar]
  9. Burr, B. M. , & Bennett, M. G. (2014). Amiidae: Bowfins. In M. L. Warren, Jr. & Burr B. M. (Eds.), Freshwater fishes of North America volume 1: Petromyzontidae to Catastomidae (pp. 279–298). Baltimore, MD: Johns Hopkins University Press. [Google Scholar]
  10. Campana, S. E. (1999). Chemistry and composition of fish otoliths: Pathways, mechanisms and applications. Marine Ecology Progress Series, 188, 263–297. [Google Scholar]
  11. Campana, S. E. (2001). Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology, 59, 197–242. [Google Scholar]
  12. Campana, S. E. (2005). Otolith science entering the 21st century. Marine and Freshwater Research, 56(5), 485–495. [Google Scholar]
  13. Campana, S. E. , Annand, M. C. , & McMillan, J. I. (1995). Graphical and statistical methods for determining the consistency of age determinations. Transactions of the American Fisheries Society, 124(1), 131–138. [Google Scholar]
  14. Carlander, K. D. (1969). Handbook of freshwater fishery biology, life history data on the freshwater fishes of the United States and Canada. Wiley‐Blackwell. [Google Scholar]
  15. Cartier, D. , & Magnin, É. (1967). La croissance en longueur et en poids des Amia calva L. de la région de Montréal. Canadian Journal of Zoology, 45(5), 797–804. [Google Scholar]
  16. Casselman, J. M. (1990). Growth and relative size of calcified structures of fish. Transactions of the American Fisheries Society, 119, 673–688. [Google Scholar]
  17. Cooper, G. P. , & Schafer, R. N. (1954). Studies on the population of legal‐sized fish in Whitmore Lake, Washtenaw and Livingston counties, Michigan. Transactions of the North American Wildlife Natural Resources Conference, 19, 239–259. [Google Scholar]
  18. Daniels, K. L. M. (1993). Reproductive biology of the Bowfin, Amia Calva Linnaeus, from the Green Bottom Wildlife Management Area, Cabell County, West Virginia. Master's thesis, Marshall University, Huntington, WV, USA.
  19. Davis, J. G. (2006). Reproductive biology, life history and population structure of a Bowfin Amia calva population in Southeastern Louisiana (Master's thesis, Nicholls State University, Thibodaux, LA, USA).
  20. Flower, S. S. (1925). Contributions to our knowledge of the duration of life in vertebrate animals.—I. Fishes. In Proceedings of the Zoological Society of London, 247–268.
  21. Flower, S. S. (1935). Further notes on duration of life in animals. 1. Fishes as determined by otolith. And scale readings and direct observation on live individuals. Proceedings of the Zoological Society of London, 2, 265–304. [Google Scholar]
  22. Gallucci, V. F. , & Quinn, T. J. (1979). Reparameterizing, fitting, and testing a simple growth model. Transactions of the American Fisheries Society, 108(1), 14–25. [Google Scholar]
  23. Hausmann, B. J. (1998). The role of bowfin, Amia calva, in multiple species management plans (Master's thesis, Eastern Illinois University, Charleston, IL, USA).
  24. Holland, H. T. (1964). Ecology of the bowfin (Amia calva Linnaeus) in Southeastern Missouri. Doctoral dissertation, University of Missouri, Columbia, Missouri.
  25. Isermann, D. A. , McKibbin, W. L. , & Willis, D. W. (2002). An analysis of methods for quantifying crappie recruitment variability. North American Journal of Fisheries Management, 22(4), 1124–1135. [Google Scholar]
  26. Koch, J. D. , Quist, M. C. , & Hansen, K. A. (2009). Precision of hard structures used to estimate age of bowfin in the upper Mississippi River. North American Journal of Fisheries Management, 29(2), 506–511. [Google Scholar]
  27. Koch, J. D. , Quist, M. C. , Hansen, K. A. , & Jones, G. A. (2009). Population dynamics and potential management of bowfin (Amia calva) in the upper Mississippi River. Journal of Applied Ichthyology, 25(5), 545–550. [Google Scholar]
  28. Lackmann, A. R. , Andrews, A. H. , Butler, M. G. , Bielak‐Lackmann, E. S. , & Clark, M. E. (2019). Bigmouth Buffalo Ictiobus cyprinellus sets freshwater teleost record as improved age analysis reveals centenarian longevity. Communications Biology, 2, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lackmann, A. R. , Bielak‐Lackmann, E. S. , Jacobson, R. I. , Butler, M. G. , & Clark, M. E. (2022). Otolith allometry informs age and growth of long‐lived quillback Carpiodes cyprinus. Environmental Biology of Fishes, 105, 1051–1064. 10.1007/s10641-022-01315-8. [DOI] [Google Scholar]
  30. Lackmann, A. R. , Kettelhut, S. , Paulson, B. L. , Anderson, C. M. , Goodchild, S. C. , Guadalupe, K. D. , & Stockwell, C. A. (2021). Thin‐sectioned otoliths reveal sexual dimorphism and a 10‐year lifespan in the endangered Pahrump Poolfish. North American Journal of Fisheries Management, 41(6), 1631–1639. [Google Scholar]
  31. Lackmann, A. R. , Kratz, B. J. , Bielak‐Lackmann, E. S. , Jacobson, R. I. , Sauer, D. J. , Andrews, A. H. , … Clark, M. E. (2021). Long‐lived population demographics in a declining, vulnerable fishery – bigmouth buffalo (Ictiobus cyprinellus) of Jamestown reservoir, North Dakota. Canadian Journal of Fisheries and Aquatic Sciences, 78(10), 1486–1496. [Google Scholar]
  32. Maceina, M. J. (2004). Verifying residuals from catch curves to detect recruitment variation in largemouth bass and crappies. North American Journal of Fisheries Management, 24(1), 231–236. [Google Scholar]
  33. Mundahl, N. D. , Melnytschuk, C. , Spielman, D. K. , Harkins, J. P. , Funk, K. , & Bilicki, A. M. (1998). Effectiveness of bowfin as a predator on bluegill in a vegetated lake. North American Journal of Fisheries Management, 18(2), 286–294. [Google Scholar]
  34. Porter, N. J. , Bonvechio, T. F. , McCormick, J. L. , & Quist, M. C. (2014). Population dynamics of bowfin in a South Georgia reservoir: Latitudinal comparisons of population structure, growth, and mortality. Journal of the Southeastern Association of Fish and Wildlife Agencies, 1, 103–109. [Google Scholar]
  35. Quinn, J. W. (2010). A survey of bowfishing tournaments in Arkansas. North American Journal of Fisheries Management, 30(6), 1376–1384. [Google Scholar]
  36. Radford, D. S. , Lackmann, A. R. , Moody‐Carpenter, C. J. , & Colombo, R. E. (2021). Comparison of four hard structures including otoliths for estimating age in blue suckers Cycleptus elongatus . Transactions of the American Fisheries Society, 150(4), 514–527. [Google Scholar]
  37. Randall, J. E. , & Delbeek, J. C. (2009). Comments on the extremes in longevity in fishes, with special reference to the Gobiidae. Proceedings of the California Academy of Sciences, 60(10), 447. [Google Scholar]
  38. Rypel, A. L. , Saffarinia, P. , Vaughn, C. C. , Nesper, L. , O'Reilly, K. , Parisek, C. A. , … Ayers, D. (2021). Goodbye to “rough fish”: Paradigm shift in the conservation of native fishes. Fisheries, 46(12), 605–616. [Google Scholar]
  39. Sanderson‐Kilchenstein, D. (2015). Aspects of bowfin and northern sunfish biology and ecology master's thesis, New York College at Brockport.
  40. Scarnecchia, D. L. (1992). A reappraisal of gars and bowfins in fishery management. Fisheries, 17(5), 6–12. [Google Scholar]
  41. Scarnecchia, D. L. , & Schooley, J. D. (2020). Bowfishing in the United States: History, status, ecological impact, and a need for management. Transactions of the Kansas Academy of Science, 123, 285–338. [Google Scholar]
  42. Scarnecchia, D. L. , & Schooley, J. D. (2022). Trophies, technology, and timescape in fisheries management, as exemplified through Oklahoma's world record paddlefish Polyodon spathula. Fisheries. [Google Scholar]
  43. Scarnecchia, D. L. , Schooley, J. D. , Lackmann, A. R. , Rider, S. J. , Riecke, D. K. , McMullen, J. , … Shattuck, Z. R. (2021). The sport fish restoration program as a funding source to manage and monitor bowfishing and monitor inland commercial fisheries. Fisheries, 46(12), 595–604. [Google Scholar]
  44. Schiavone, A., Jr. (1982). Age and growth of bowfin in Butterfield Lake, New York (Amia calva, fish). New York Fish and Game Journal, 29, 107. [Google Scholar]
  45. Schnute, J. , & Fournier, D. (1980). A new approach to length‐frequency analysis: Growth structure. Canadian Journal of Fisheries and Aquatic Sciences, 37, 1337–1351. [Google Scholar]
  46. Sinopoli, D. A. , & Stewart, D. J. (2021). A synthesis of management regulations for bowfin, and conservation implications of a developing caviar fishery. Fisheries, 46(1), 40–43. [Google Scholar]
  47. Sylvester, R. M. , & Berry, C. R., Jr. (2006). Comparison of white sucker age estimates from scales, pectoral fin rays, and otoliths. North American Journal of Fisheries Management, 26, 24–31. [Google Scholar]
  48. Thompson, A. W. , Hawkins, M. B. , Parey, E. , Wcisel, D. J. , Ota, T. , Kawasaki, K. , … Feron, R. (2021). The bowfin genome illuminates the developmental evolution of ray‐finned fishes. Nature Genetics, 53(9), 1373–1384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. von Bertalanffy, L. A. (1938). Quantitative theory of organic growth (inquiries on growth laws II). Human Biology, 10, 181–213. [Google Scholar]
  50. York, B. , Schooley, J. D. , & Scarnecchia, D. L. (2022). Oklahoma bowfishing values and perspectives toward nongame fishes and their management. North American Journal of Fisheries Management, 42, 1020–1040. [Google Scholar]

Articles from Journal of Fish Biology are provided here courtesy of Wiley

RESOURCES