Fisheries management is vital to ensure the sustainability of fisheries in the face of continued climate change and overfishing (Harley et al., 2006; Allison et al., 2009; Ekstrom et al., 2015), particularly for vulnerable coastal communities that are reliant on marine resources (Dolan and Walker, 2006; Cinner et al., 2012; Whitney and Ban, 2019). However, many exploited marine species remain understudied due to limited resources and information (Pilling et al., 2008; Wiedenmann et al., 2013; Baum and Fuller, 2016). Fossil and historical data are alternative sources that can be used to determine the extent of change due to human activities, as well as to establish pre-anthropogenic baselines (Dietl and Flessa, 2011; Mcclenachan et al., 2012; Kidwell and Tomasovych, 2013; Rick and Lockwood, 2013; Dietl et al., 2015; Kidwell, 2015; Barnosky et al., 2017; Fordham et al., 2020), but their use as a tool for fisheries management is still largely unimplemented (Jackson and Sala, 2001; Lotze et al., 2011; Toniello et al., 2019). In particular, predation scars on prey (known as repair scars) are an unused resource that can be used to infer predator abundance in both modern and fossil settings (Schindler et al., 1994; Molinaro et al., 2014; Stafford et al., 2015b), with large datasets that can be collected quickly and easily. By comparing predation traces on a common prey item from similar fossil and modern environments, we demonstrate the utility of fossil data and predation traces as alternative, cost-effective methods to determine the current status of exploited predator species.

Along the west coast of North America, cancrid crab species, such as Dungeness (Metacarcinus magister) and rock crabs (Cancer productus, Romaleon antennarium, Metacarcinus anthonyi), are important fisheries, valued at over $230 million USD in the United States in 2019 ($54 million in California) (NOAA Fisheries Landings, 2019). First-hand accounts suggest that overfishing is already affecting multispecies rock crab populations and body sizes in southern California (Fitzgerald et al., 2018, 2019). However, management of these fisheries is hampered by a lack of data other than gross landings (Culver et al., 2010; NOAA Fisheries Landings, 2019). In California, permit numbers are restricted, and a catch size limit is in place, but the effectiveness of these restrictions is unknown (Culver et al., 2010). California’s crab species also face additional pressure from rapid development/urbanization of southern California’s coast (Woodring et al., 1946; Whitaker et al., 2010), ocean acidification (Bednaršek et al., 2020), harmful algal blooms (HABs) (Moore et al., 2019), and human exploitation for at least the past 12 – 13,000 years (Rick et al., 2014; Erlandson et al., 2015). Cancrids from the Pleistocene of southern California are known mostly from occasional chela (claws), which preserve more readily than the rest of the body (Menzies, 1951; Nations, 1975). Pleistocene species and their distributions are comparable to the present, consisting of the same major cancrid taxa, including Cancer productus, Metacarcinus anthonyi, M. gracilis, M. magister, Romaleon antennarium, and R. branneri (Menzies, 1951; Nations, 1975). The only notable difference is the presence of Glebocarcinus oregonensis (the pygmy rock crab) in the Pleistocene of southern California, which does not extend south of Santa Barbara today (Nations, 1975). Given that the shell-crushing crab taxa are similar between the Pleistocene and modern, and that it is unknown if repair scars allow differentiation between species of cancrid crabs, it is assumed that repair scars represent an overall cancrid crab signal that should appear similar between the Pleistocene and modern of southern California. Faced with a lack of current or historical data, limited fisheries management resources, and a generally poor body fossil record of crabs, alternative evidence is needed to guide the management of current crab stocks.

For both ecological and paleontological studies, predation traces on prey are often the only source of information to indicate the activity or presence of predators. Crab predation scars (repair scars) on mollusks have been used to assess patterns as broad as the development of antipredatory adaptations and mollusk diversity through time (Schindel et al., 1982; Vermeij, 1982c; Alexander and Dietl, 2003; Dietl et al., 2010; Mondal et al., 2014; Mondal and Harries, 2015), to variation in predation in modern ecosystems (Vermeij, 1982a,b; Schmidt, 1989; Schindler et al., 1994; Cadée et al., 1997; Alexander and Dietl, 2001; Dietl and Alexander, 2009; Molinaro et al., 2014; Stafford et al., 2015b). Predation traces also provide useful information on how crabs and their ecosystems are affected by environmental disturbances (Tyler et al., 2019). Most importantly for stock assessments, repair frequencies accurately track crab abundances in modern coastal studies (Schindler et al., 1994; Molinaro et al., 2014; Stafford et al., 2015b), especially when limited to repairs found on a single, common prey species.

However, comparing repair frequencies over evolutionary time scales in which selection for improved defenses or weaponry may have occurred, becomes more challenging. Because repairs are caused by failed attacks, a decline in repair frequency over time may be the result of either a decrease in predator success (more failures = more repairs) or an increase in predator attacks (more attacks at the same success rate = more repairs) (Vermeij, 1987); the reverse is also true. In studies of modern crab-mollusk predator-prey systems along crab population gradients from British Columbia, Canada (Molinaro et al., 2014; Stafford et al., 2015b) and Georgia, United States (Schindler et al., 1994), repair frequency on single prey species increased with crab abundance. Although not indicated in their study, data in Cadée et al. (1997) from the northern Gulf of California, Mexico, also supports this pattern. While the intent of these studies was to test whether repairs were a function of attack frequency, and predator abundance is just one of several mechanisms for changing attack frequency, the evidence demonstrates an association of repairs with crab abundance. As such, if a change in success can be excluded as an explanation for a change in repair frequency, then it is likely that the change in repairs is caused by a change in crab abundance.

To discern potential changes in predator success, Leighton (2002) suggested that prey size, particularly the size of the largest individuals attacked, could provide a means of distinguishing between success and attack frequency as causal mechanisms for changes in repair frequency; this approach was subsequently used in Richards and Leighton (2012) and Pruden et al. (2018). For example, if a prey species has not strengthened its defenses (e.g., increased shell thickness or ornament, etc.) over time, then larger, stronger predators would not only have an increased likelihood of success, but would also be able to take larger prey. The same pattern can and has also been considered from the perspective of the prey. For example, if prey improve their defenses, predator success and the size of prey predators are able to attack and/or the severity of the damage predators can inflict is likely to decrease over time (Mondal et al., 2014). Comparing prey size and repair size has also been used to determine potential prey size refugia from sublethal shell breakage (Alexander and Dietl, 2001).

For durophagous crabs, strength refers to the force delivered by the chelae. In all animal systems, force is proportional to the cross-sectional area of the muscle bundle, perpendicular to the line of action of the muscle. The chelae closer muscle in cancrids increases in size through ontogeny; thus, larger crabs are stronger crabs. In some crab-prey systems (e.g., the stone-crab, Menippe and its oyster prey), the crab is so strong relative to the prey that its success rate approaches 100%; in such systems, there is no size refuge for the prey, and no size selection on the part of the predator. However, for many specialized durophagous crab-prey systems, the success rate is much lower than 100% (as evidenced by the existence of repairs), and in such systems, crab size, strength, success, and the ability to handle larger prey are strongly associated (Hughes and Elner, 1979; Whetstone and Eversole, 1981; Lake et al., 1987; Behrens Yamada and Boulding, 1998; Whitenack and Herbert, 2015). Note that the ability to take larger prey does not necessarily imply a preferred size range of prey.

This relationship between crab size, success, and maximum size of the prey taken also holds true for the cancrid-Tegula system (Mendonca, 2020). Not only are larger cancrids stronger and therefore more likely to be successful, but larger crabs are able to handle larger Tegula. Cancrid crabs typically attack gastropod prey by either attempting to crush the shell directly, or by peeling back the aperture with a series of chips (Stafford et al., 2015a). The latter method is time consuming (Barclay et al., 2020b) (increasing the likelihood of the crab being interrupted by its own predators and competitors) and the gastropod retracts farther into its shell with each peel, decreasing the crab’s chances of success (Stafford et al., 2015a; Mendonca, 2020). In contrast, crushing is quick, and when the shell fractures, the gastropod has no escape. The choice of crushing versus peeling is in large part a function of the relative sizes of the crab and the gastropod (Mendonca, 2020). Larger crabs have large enough chelae to grasp a large Tegula across its width and attempt a crush, whereas smaller crabs are forced to peel. Thus, aside from their greater strength, larger crabs also have the option of utilizing a more successful attack strategy more frequently. However, they also can and do take larger Tegula by peeling. Strength, size, success, and largest prey are therefore expected to correlate, and the Mendonca (2020) thesis is consistent with this prediction. Assuming that these results hold across time, maximum size at attack (SAA) is an indicator of both predator strength (relative to the prey) and success.

A change in repair frequency due to a change in success can also show a change in SAA. However, it is also possible to use SAA to distinguish changes in crab strength/success as well as evaluate whether these changes have impacted the observed repair frequency (i.e., the success of the predator compared to the prey available to them) when comparing samples. For example, consider a hypothetical scenario where SAA increases, but so does the size of prey. If the distribution of attacks (SAA) remains relative to the size distribution of prey available, the relative rate of predator success against the prey available to them has not changed. In this scenario, while we can observe that crab strength and success has increased, because prey size also increased, the relative success rate of the predator is the same and therefore changes in crab strength/success are not likely to have impacted the observed repair frequency. Therefore, by considering the size at which repairs occur, it is not only possible to distinguish changes in predator success, but the SAA can also be used to evaluate whether these changes in success have impacted the observed repair frequency. Here, we analyze repairs on Late Pleistocene and living individuals of a common and widespread crab prey item, the black turban snail (Tegula funebralis) (Geller, 1982, 1983; Molinaro et al., 2014; Stafford et al., 2015b), to assess patterns of modern crab predation relative to a pre-Industrial baseline in the Pleistocene. This approach also demonstrates the value of repair scars and paleontological data as a cost-effective, rapid assessment tool to fill the critical gap in our knowledge of the current crab fisheries in southern California.

Modern T. funebralis (1) have fewer scars (Wilcoxon rank sum test, p < 0.001), (2) are larger than their Pleistocene counterparts (p < 0.001), and (3) are repaired at larger sizes (p < 0.001) (Figure 4 and Table 1). When including prey size in mixed model analyses, the same patterns of fewer repairs and larger gastropods in the modern compared to the Pleistocene held (Table 2). Time (geologic age) still had a significant effect on whether a gastropod had at least one scar in the mixed models (Table 2). When backing out the effects of time, the presence and number of scars showed a weak, but significant correlation with size (partial R = 0.134 and 0.141, respectively, p < 0.0001). Region (Palos Verdes vs. San Diego) did not have a significant effect on scars or model performance (Table 2 and Supplementary Table 1). Random effects (lot/locality) also did not have any significant impact on model performance (Table 2). As the fossil lots from the Palos Verdes Hills area are likely attributed to the older 120,000 ya highstand (Figure 6 in Muhs et al., 2006) compared to more mixing of the 80,000, 100,000, and 120,000 ya highstands from the San Diego area (Kern, 1977; Kennedy et al., 1982; Muhs et al., 1994, 2012), this indicates that differences in time averaging between fossil localities and regions are not substantial enough to impact the results. Therefore, changes observed between the Pleistocene and modern are not the result of differences in time averaging of samples.

Changes in the frequency of repairs between the Pleistocene and modern is indicative of a change in the number of attacks, rather than a change in crab success. The number of repairs expected should be similar between the two samples because even though the average SAA is larger in the modern (p < 0.0001) (Figure 4 and Table 1), as modern gastropod body size also increases, the distribution of repair size (SAA) relative to body size are not significantly different between the two samples (Table 3 and Figure 4). Increases in modern SAA should therefore be influenced by the greater number of larger gastropods available to attack. Hypothetically, even if gastropod size did not overlap between samples, if the observed repairs show the same size distribution relative to the gastropods, as they do for our dataset (Figure 4), the relative success of crabs against those two gastropod populations would be the same, and so would the repair frequencies (e.g., Figures 3B vs. 3D). Thus, observed changes to repair frequencies should be the result of a change in attack rates.

If anything, Pleistocene crabs were relatively more successful against the gastropods available to them, as there were fewer repairs at relatively small sizes in the Pleistocene compared to today (4.3 vs. 12.7% of repairs occurred below the minimum gastropod size, [2] in Figure 3A, in Pleistocene vs. modern assemblages) (Figure 4), and the distance between the minimum SAA and the minimum gastropod size ([1] in Figure 3A) was almost twice as large in the modern compared to the Pleistocene (Table 3 and Figure 4). In addition, the maximum size of repairs is similar between the two time intervals (Table 3), indicating that absolute crab strength has not changed substantially since the Pleistocene (Figure 4). Therefore, it is unlikely that the decrease in repairs observed in the modern is a function of a change in success. Furthermore, while size had a significant effect on whether a gastropod had one or more scars, this was true regardless of time (Table 1), indicating that crabs are simply more likely to fail against larger gastropods in each assemblage/location. Fewer repairs observed in the modern therefore conservatively indicates fewer attacks, rather than a change in success between Pleistocene and modern crabs (Figure 4).

Our results demonstrate a decrease in repair frequencies in modern compared to Pleistocene assemblages, conservatively indicating fewer crab attacks today compared to the Pleistocene. Given that differences in time averaging between Pleistocene assemblage lots and modern samples did not impact the results, the signal of reduced predation in the modern can be considered a robust biological signal, as is often an advantage of time averaged assemblages (Kowalewski et al., 1998; Kidwell, 2013; Kidwell and Tomasovych, 2013). There are therefore several alternative explanations to explain the possible reductions in modern crab attacks, all of which suggest that crab fisheries in southern California require increased attention and protections to ensure sustainability of these economically important species and their ecosystems. The results of our study also exemplify the value of including paleontological data and techniques in assessments of important modern coastal resources.

The simplest, and most troubling reason for the decline in crab predation on modern T. funebralis in southern California would be a reduction in the abundance of crabs. Worldwide, crustacean fisheries in general have become increasingly socioeconomically valuable and important in recent decades (Boenish et al., 2021). Severe declines in Dungeness crab populations have been documented locally in recent decades by Indigenous fishers in British Columbia, Canada who attribute these declines to commercial and recreational fishing (Ban et al., 2017). In southern California, recent data modeling studies indicate early warning signs of crab overfishing, along with concerns from fishermen and researchers about observed changes to crab body size and abundance in southern California (Fitzgerald et al., 2018, 2019). Declines in Dungeness crabs populations have been potentially associated with increases in sea otter populations in California (only at a small, localized level) (Grimes et al., 2020), but these are reported declines in crab populations, nonetheless. Our results support a decrease in crab abundances in southern California, suggesting immediate action (e.g., closure of fisheries) is needed to protect commercially exploited crabs in southern California.

Besides overexploitation, reductions in crab predation might also be due to widespread ecosystem changes caused by human activity and climate change. For example, ocean acidification affects crab larval development along the west coast of North America (Hodgson et al., 2018; Bednaršek et al., 2020). All life stages of Dungeness crab are also predicted to be vulnerable to year-round ocean conditions (low pH, low oxygen, and increased temperatures) by the end of the century (Berger et al., 2021). Mollusks, including T. funebralis, are also expected to experience direct consequences of ocean acidification, including impaired behavior, and reduced shell growth and strength (Jellison et al., 2016; Barclay et al., 2019, 2020a). While it does appear that absolute shell strength of T. funebralis may be reduced in the modern [a shift in the minimum absolute SAA to larger sizes indicates that shells may have been easier to crush at those same sizes in the modern (Figure 4)], it does not appear to have impacted the success rate of crabs as of yet. Modern crabs may have also become more restricted vertically, either due to stress or habitat loss, and may have fewer opportunities to feed on T. funebralis in its given tidal range. Rockweeds, kelp forests, and other habitat forming macroalgae have declined in southern California recently (Whitaker et al., 2010; Jurgens and Gaylord, 2018; Beas-Luna et al., 2020). At most of our sites, rockweeds and other large macroalgae were lacking, with T. funebralis found most commonly on bare rock covered in biofilm. The loss of habitat forming macroalgae may restrict crabs to foraging mostly at lower tidal heights where there is still abundant macroalgae, as habitat forming species are critical buffers against temperature stress, moreso than latitude or tidal elevation (Jurgens and Gaylord, 2018), and provide crabs refuge from their own predators, particularly when they are younger (Robles et al., 1989; Behrens Yamada and Groth, 2016).

Crabs may have also switched to other prey items, such as mussels, and attack T. funebralis less often. Indeed, mussels are a more common prey item of crabs (Robles et al., 1989). However, we did not observe any large mussel beds at any of the modern sites, and there have also been notable declines in mussel populations in southern California in recent decades (Smith et al., 2006a,b). Reductions in mussel populations in present-day southern California might increase crab predation on T. funebralis, yet we observe the opposite. While there are other prey items that might be preferred, such as owl limpets and other species of Tegula (e.g., T. brunnea and T. eiseni), these species were also present during the Late Pleistocene. Given that mussel and macroalgal beds are important habitat formers, and there are reported declines in these and associated benthic intertidal taxa in southern California (Smith et al., 2006a,2008; Whitaker et al., 2010; Beas-Luna et al., 2020), if anything, these reductions in biodiversity should reduce the amount and/or number of alternative prey, suggesting that the probability of T. funebralis being included in crab diets, or even becoming a more preferred prey item, should increase. That we still observe a decline in predation on modern T. funebralis is therefore likely conservative with respect to decreased crab abundances.

A difference in temperature, particularly since the cooler 80,000 ya highstand, could have affected predation rates, and/or caused a change in both the size of T. funebralis and their crab predators. However, as the Pleistocene samples were a mix of both the 80,000 and 120,000 ya highstands, the latter of which was similar in temperature to today, temperature alone is an insufficient explanation for changes in body size and repairs. Furthermore, as increases in temperature also can increase predation rates (Bélair and Miron, 2009; Leighton and Tyler, 2021), we would have also expected more repair scars at modern sites, yet the opposite is true. This conservatively implies that decreases in crab attacks are driven by reduced crab abundances in the modern compared to the Pleistocene. A release from predation pressure due to a reduction in crabs might have also driven the increases in body size for T. funebralis, as T. funebralis body size is known to decrease in response to human predation intensity (Erlandson et al., 2015). Additionally, sea stars are another major predator of T. funebralis (Paine, 1969; Fawcett, 1984; Gravem and Morgan, 2017), and their populations have been severely affected by an outbreak of sea star wasting disease in 2013 (Hewson et al., 2014; Menge et al., 2016). Both crabs and sea stars are known to have disproportionately large effects on their ecosystems (Paine, 1966, 1969; Hull and Bourdeau, 2017; Leighton and Tyler, 2021). Reduced populations of both crabs and sea stars might also influence coastal ecosystems due to changes in the body size of their prey.

Regardless of the potential cause, reduced crab predation rates in southern California warrant further attention both from a fisheries/resource management and ecosystem health perspective. To avoid further depletion of current crab stocks, increased management practices may need to be enacted. Past studies show that Dungeness crab populations are quick to recover after spatial closures (Taggart et al., 2004; Frid et al., 2016). However, if the cause of declines in crab populations is also related to ecosystem changes and degradation on southern California’s coast due to anthropogenic impacts, such as changes to macroalgal and seagrass communities (Waycott et al., 2009; Whitaker et al., 2010; Beas-Luna et al., 2020) that are often nurseries for young crabs (Robles et al., 1989; Beck et al., 2003; Holsman et al., 2006; Behrens Yamada and Groth, 2016), or to seawater conditions (Bednaršek et al., 2020; Berger et al., 2021), changes to crab fishery management policies will be insufficient to sustain these resources long-term.

To improve the management and long-term sustainability of crab fisheries in southern California and beyond, we recommend: (1) the enactment of increased management practices in southern California, such as partial or spatial closures of crab fisheries that would allow for potentially rapid recovery of crab stocks, and (2) the use of repairs scar surveys in both modern and fossil assemblages as a fast, cost-effective, and easily implemented tool for initial assessments of crab population health. More broadly, there are several other areas for improvement and continued investigation. First, addressing the ever-increasing urgency of climate change and anthropogenic impacts on coastal ecosystems and resources (Beas-Luna et al., 2020; Bednaršek et al., 2020; Berger et al., 2021) is paramount to ensuring long-term sustainability of far more than just crab fisheries. For example, as macroalgal and seagrass communities are important habitat for young crabs (Robles et al., 1989; Beck et al., 2003; Holsman et al., 2006; Behrens Yamada and Groth, 2016), any restoration of these communities will also greatly benefit crabs (Anderson, 1989). Addressing anthropogenic carbon dioxide emissions is also key to addressing future ocean conditions, including warming, hypoxia, and ocean acidification that will impact crabs (Bednaršek et al., 2020; Berger et al., 2021). Second, conservation paleobiology and historical ecology offer unique perspectives, tools, and datasets that provide much needed and often missing context to ensure the sustainability and protection of ecosystems and their services (Hadlock Seeley, 1986; Erlandson and Rick, 2010; Dietl and Flessa, 2011; Mcclenachan et al., 2012; Rick and Lockwood, 2013; Dietl et al., 2015; Kidwell, 2015; Barnosky et al., 2017; Fordham et al., 2020). Related to crabs, other studies have used historical records of prey (gastropod) morphology as additional evidence of the spread of the invasive European green crab, Carcinus maenas, on the northeast coast of the United States (Hadlock Seeley, 1986). Specifically for fisheries, the regular inclusion of alternative sources of information, including repair scars and paleontological data, as well as historical and traditional knowledge/data can provide important insights to understand the status of data-limited fisheries (Lotze et al., 2011; Ban et al., 2017; Toniello et al., 2019) and should be implemented more widely in fisheries management. For example, Indigenous knowledge and first-hand accounts from fishers have proven valuable to crab fisheries modeling and stock assessments (Ban et al., 2017; Fitzgerald et al., 2018, 2019). Indigenous and first-hand knowledge provides additional context that could be used in conjunction with repair scars to assess the severity and timing of stock declines, as well as to identify areas of greatest concern where repair scar surveys could be implemented.

Our results demonstrate that crab predation traces are a novel, inexpensive initial assessment tool to determine the current health of important coastal species, particularly if current or historical records are sparse or lacking. We also demonstrate that fossil data are valuable for determining pre-human baselines that provide context for the status of current fisheries and long-term responses of both crabs and their prey amidst human-induced environmental changes. Repair scars are therefore a robust indicator that can be used to determine whether further action is needed to provide greater protections for vulnerable crab populations worldwide. We suggest that reduced crab predation in modern gastropods is most likely caused by human-induced changes (e.g., overexploitation of predators or climate-induced ecosystem disturbances), but that regardless of the cause(s), there is immediate need for a detailed investigation of current crab fishery stocks in southern California to ensure the sustainability of these critical coastal socioeconomic resources. An important avenue for continued future research would be to expand upon the current data framework to include historical and archaeological collections (Rick and Lockwood, 2013; Toniello et al., 2019) to understand the timing of crab population declines and improve sampling efforts, and to apply the methods used in this study to other locations of concern for shell-crushing crab resources. We conclude that incorporating paleontological techniques and data (i.e., conservation paleobiology) to resource management provides critical new insights that can aid in the management and sustainability of exploited biological resources.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

KB and LL conceived the study, provided funding for the study, revised, and edited the original drafts. KB collected and analyzed all the data and wrote the original drafts. LL provided guidance with analyses and acted in a supervisory role. Both authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.