Shifting headlines? Size trends of newsworthy fishes

Introduction

‘The Sea-monsters, mighty of limb and huge, the wonders of the sea, heavy with strength invincible, a terror for the eyes to behold and ever armed with deadly rage—many of these there be that roam the spacious seas…’

Oppian, Halieutica

Exploitation continues to have a pervasive impact on fish populations. Globally, fisheries catches peaked in the mid-1990s, and have been relatively static or declining since then (Pauly & Zeller, 2016). Remaining abundances of large species are now well below pre-exploitation levels (Christensen et al., 2003; Juan-Jordá et al., 2011; Myers et al., 2007; Pons et al., 2017). More locally, declining trends of abundance and/or body sizes of marine fish species are commonly reported from docks and markets (Richardson et al., 2006; McClenachan, 2009; To & Sadovy De Mitcheson, 2009). Smaller fishes in catches and in markets can arise in part from replacement of large, high-trophic level species by smaller ones (Pauly et al., 1998) or the targeting of additional small species (Essington, Beaudreau & Wiedenmann, 2006), to supplement increasingly less-productive fisheries (Sethi, Branch & Watson, 2010). Smaller fish sizes can also result from the targeted removal of the largest individuals within populations through selective fishing (Richardson et al., 2006; McClenachan, 2009; Sutter et al., 2012). Additionally, as fish age and grow they experience more lifetime fishing pressure, resulting in a disproportionate impact of fishing on old, large fish (Barnett et al., 2017).

The increasing prevalence of small-bodied fishes in markets or fishing tournaments can influence the public perception of the ‘normal’ size of fish (To & Sadovy De Mitcheson, 2009). This social phenomenon is one type of ‘shifting baseline’ syndrome. The ‘shifting baseline’ syndrome describes the general, gradual lowering of human cognitive baselines, as each generation accepts a lower standard of resource abundance as the new norm (Pauly, 1995). This has been demonstrated with fish abundance (Pauly, 1995); however, there is growing evidence of shifting baselines occurring for other measures of marine resource quality. For example, older fishers often recall both greater fish abundances and larger fish sizes than younger fishers (Sáenz-Arroyo et al., 2005; Ainsworth, Pitcher & Rotinsulu, 2008; Lozano-Montes, Pitcher & Haggan, 2008). They also report more species and fishing grounds as being depleted (Bunce et al., 2008; Katikiro, 2014). When data are available, these recollections are often confirmed empirically by declining trends in abundance or size (Richardson et al., 2006; Lozano-Montes, Pitcher & Haggan, 2008).

In this study, we ask whether the widespread trends—real and perceived—in shrinking sizes of marine fish are detectable in popular media, or whether news writers, like many marine stakeholders, have experienced a shifting baseline. Because popular media—television, newspapers, radio, and more recently, the internet—provide sources of information that shape public awareness of environmental issues (Boykoff, 2009), a shifting baseline in media reporting of the capture of large fish could bias public perception of the state of marine fish populations. The use of similar superlatives or hyperbole (e.g. ‘giant, ‘monster’, ‘enormous’) to describe a very large fish caught some years ago and a relatively smaller fish caught today might be underpinned by a constant journalistic quest for headlines. However, the result among readers might be the perception that very large fish still abound in the ocean.

To address this issue, we surveyed printed English-language news articles about large fish caught. We asked two questions. First, are captured fish described by various size-related superlatives in the English-language printed press truly among the largest specimens reported for each species? Second, is there evidence that the relative size of fish proclaimed to be superlatively large in these news articles declined over time? Such a decline would reflect both the recent rarity of truly large individuals and evidence of ‘generational amnesia’, that is, the extinction of knowledge over time (Papworth et al., 2009), among English-language newsmakers.

Methods

Results

We collected 267 news articles that met our criteria and reported either total length and/or weight of a caught fish. These articles spanned more than 100 years, from 1869 to 2015, with over half of the articles (n = 144) published between 2000 and 2015. Articles referred to 75 different species of fish (Table S2). Seven of these species had 10 or more news articles; 43 of the species in our dataset were represented by a single news article.

Articles were retrieved from a total of 28 countries, but most came from the USA (45%, n = 121), followed by Australia (11%, n = 29). While all continents were represented, Africa and South America had the lowest number of articles (five and three articles, respectively). Fourteen of the FAO major fishing areas (MFAs) were represented, ranging from as many as 53 articles (20%) for the Western-Central Atlantic (MFA 31) to as few as a single article for both the Mediterranean and Black Sea (MFA 37) and Southwest Atlantic (MFA 41).

Trends in relative fish size over time

News articles most commonly reported the total weight of a fish (n = 238) rather than total length (n = 151). Relative length (i.e. reported length divided by maximum length) of all newsworthy fishes combined decreased over time for the highest relative size quantiles (80th, p = 0.01; 90th, p = 0.02; 95th, p = 0.06), but not for lower quantiles (Fig. 1A; Table S3). In contrast, relative weight did not change significantly over time across all species combined (Fig. 1B; Table S3). There were more relatively small fish reported by weight than by length, with some articles reporting on fish that were less than 10% of the maximum weight for their species (Fig. 1).

When species were divided into subsets of similar species, we found no significant change in relative length for pelagic gamefish (all quantiles, p > 0.05; Fig. 2A; Table S4), but a significant increase in the lower quantiles of relative weight across pelagic gamefish species over time (quantiles 10–25%, p < 0.05; Fig. S2A; Table S5). For oceanic sharks, there was no change in any of the quantiles in either relative length (Fig. 2B; Table S4) or relative weight (Fig. S2B; Table S5) over time. However, the relative length of charismatic megafish declined significantly over time across six of the nine quantiles examined (i.e. all quantiles except 50th, 75th, and 80th; Table S4), although only the 25th quantile met our sample size ratio threshold (Fig. 2C). The relative weight of charismatic megafishes declined over time only for the 20th quantile (p = 0.04; Fig. S2C; Table S5).

There was no clear temporal trend in the relative length (Figs. 3A and 3B; Table S6) or weight (Figs. S3A and S3B) of newsworthy fish at unknown or low risk according to their IUCN threat status. However, the relative lengths of fishes categorised as being at high risk of extinction declined significantly over time at the higher quantiles (75th–90th quantiles, p < 0.05; Fig. 3B; Table S6). There was no detectable change in relative weight of these high-risk species over time (Fig. S3B; Table S7).

Discussion

The capture of ginormous fish is often newsworthy. Our results show that fish described as being superlatively large in the printed press are rarely the largest specimens for any species. Fish judged notable by the media were, on average, 72% of the maximum length ever recorded for their species. Moreover, we found some evidence of a shifting baseline syndrome in news media over the last 140 years, as the relative size of the largest fish worthy of a headline has declined over time. This pattern held for large, charismatic fish species, which are now reported in the media at smaller relative sizes than they were near the turn of the 20th century, and for the largest species under high risk of extinction. In contrast, there was no similar trend for pelagic gamefish and oceanic sharks, or for species under lower risk of extinction.

Is a fish that is 72% of the maximum size reported for its species really very large? The answer depends on the combined effects of life history, fishing pressure, and sampling effort on size-frequency distributions. If we only examine life histories, species that grow quickly, mature early, and die young, will tend to have body size distributions with small tails at large body sizes (Jennings, Kaiser & Reynolds, 2009). A sizeable proportion of the population is therefore likely to reach 72% of L_inf. In contrast, species with slower life histories—those that grow slowly, mature late and die old—will have long-tailed body-size distributions. In such species, the likelihood of reaching 72% of L_inf is much smaller, making the individuals that do reach this size remarkably large for their species. However, these size distributions will be heavily influenced by two human-driven factors, fishing pressure and sampling effort, which have opposing effects on the average size of the largest fish, relative to L_inf, in a population. Older fish experience fishing pressure for more years than younger fish, resulting in disproportionately fewer old fish in populations (Barnett et al., 2017), and fishing pressure often also targets large individuals (Law, 2000), so one would expect the average largest size of fish to decrease over time. However, higher fishing pressure translates into higher sampling effort, which increases the likelihood of observing a large fish, even in populations where there are few large fish remaining. When we manipulated these two factors in tandem in a simulated fish population, the average size of the largest fish caught was variable across scenarios of fishing pressure and sampling effort (Fig. S4). These results suggest that our observed average relative length of 72% only represents truly large fish under specific conditions of fishing pressures and observational effort. In a population where fishing pressure is high but there are few reported observations, the average relative size of fish will be lower than 72%. Thus, if the fishes in our dataset come from such populations, most fish that make the news are indeed newsworthy. Conversely, if fishes come from populations that are well observed and/or under low fishing pressure, the fishes that make the news at only 72% of maximum size are not particularly large.

We found some evidence of a shifting baseline syndrome among English language news reporters (Fig. 1). Two conditions must be met in order for shifting baseline syndrome to occur. First, biological changes must occur within a system. Second, people’s perceptions of a system must be consistent with current biological data, but people must also be unaware that the system was different in the past (Papworth et al., 2009). Our study meets the first condition. There is ample evidence of the widespread decline in the abundance and size of large, predatory fishes over the last 50–100 years (Christensen et al., 2003; McClenachan, 2009; Myers & Worm, 2003; Pauly et al., 1998; Sáenz-Arroyo et al., 2005; To & Sadovy De Mitcheson, 2009; Watson et al., 2012; Williams, 2003). Indeed, this declining trend is shared by species targeted primarily by the commercial fishery (Myers & Worm, 2003; To & Sadovy De Mitcheson, 2009) as well as by the recreational fishery (McClenachan, 2009; Williams, 2003), both of which are represented in our dataset. Our study also partly met the second condition. The relative size of fish proclaimed to be superlatively large in news articles declined over time, but this trend was evident only for the largest newsworthy fishes (80th–90th quantiles for length). This weak declining trend indicates that media perception is at least sometimes consistent with current biological conditions but most of the time, it is not.

There are at least three possible explanations for why we did not find strong evidence for a shifting baseline syndrome in media perception of what constitutes as a newsworthy large fish. First, greater fishing effort (e.g. more vessels) combined with new technologies that allow fishers to travel greater distances (e.g. diesel engines), engage in longer trips (e.g. refrigeration) and find fish more effectively (e.g. echo-locating equipment and helicopters) increases the incidence of fishers encountering truly large fish. Media perception of what constitutes a very large fish might therefore remained unchanged over time because fishers are still able to catch the last few remaining superlatively large individuals by either fishing more intensely (Anticamara et al., 2011; Watson et al., 2012), by fishing in previously unexploited areas (Swartz et al., 2010), by fishing near refuges (e.g. marine protected areas) out of which large fish can stray (Ahrens, Walters & Christensen, 2012), or by occasionally catching fish with traits that make them difficult to catch (e.g. shy behaviours; Biro & Post, 2008). Second, although the size and abundance of large, predatory fishes have declined over time (Christensen et al., 2003; McClenachan, 2009; Myers & Worm, 2003; Pauly et al., 1998; Sáenz-Arroyo et al., 2005; To & Sadovy De Mitcheson, 2009; Watson et al., 2012; Williams, 2003), media perception of what constitutes as a superlatively large fish has remained largely unchanged because media perceptions are based primarily on written information sources (e.g. historical records) rather than on the personal experiences of the news writers themselves or on the experiences of fishers, both of which are susceptible to generational amnesia (Papworth et al., 2009). Finally, reporting rates of large catches may have vastly increased in recent years with the rise of social media and 24-h online news cycles. The ubiquity of cell phone cameras has increased the public’s ability to visually document catches and post them online, so that catches that would not have previously been noticed by journalists are now readily available to be reported on.

Although we cannot fully rule out the first explanation, it is likely that the second had a greater influence on our results. Despite increased fishing effort and technological advances, fish sizes have declined in major commercial markets (To & Sadovy De Mitcheson, 2009) as well as in stocks targeted by the recreational fishery (Williams, 2003). These observations suggest that changes in fishing effort and technology alone are insufficient to completely mask the recent rarity of truly large fish. The defining principle of shifting baseline syndrome is that humans reset their cognitive baseline of what is considered ‘normal’ due to a lack of experience about past conditions (Pauly, 1995). This lack of past experience may occur because older generations fail to pass on their experience to younger generations (i.e. ‘generational amnesia’) or because individuals revise their own perceptions of ‘normality’ based on current conditions while forgetting that they experienced different conditions in the past (i.e. ‘personal amnesia’; Papworth et al., 2009). The latter explanation for our overall finding is more likely because news writers rarely rely exclusively on personal experiences or on the experiences of a few individuals to inform decisions about an article. Instead, the practice of ‘fact checking’ an article against multiple, existing information sources—many of which are in written form such as historical records or databases like FishBase—is common among media professionals (Elizabeth et al., 2015). This fact-checking practice ultimately promotes stability in media perception of what is considered a very large specimen for any species, regardless of current conditions. The third explanation may also have influenced our results, but is confounded here by our limited access to pre-digital media.

We surmised that we might detect stronger trends consistent with a shifting baseline syndrome for subgroupings of marine fishes in which the heterogeneity is less pronounced than within our larger, more diverse dataset. However, this was not usually the case. Media opinion of the relative size of fish worthy of a headline remained constant, regardless of all subgroupings used in this study (i.e. pelagic gamefish, oceanic sharks, and species under low or unknown risk of extinction), except for ultra-charismatic fishes and fishes under high risk of extinction. On average, news articles on charismatic species were about individuals only slightly more than half (56%) the maximum possible size for their species. Support for a shifting baseline syndrome in media reporting of this subgroup might be explained, in part, by the fact that many of the species that we classified as being ultra-charismatic are among the largest species of marine fishes (e.g. whale sharks, Rhincodon typus; basking sharks, Cetorhinus maximus; and giant manta, Mobula birostris; Table S2). Landing any individual of these species can be newsworthy simply because of its large size relative to all fishes, rather than relative to its species. These species may also draw additional journalistic interest because of their significance for both conservation and tourism; four of the six megafish in our dataset are threatened with extinction (Table S2), which drove the trend of declining relative length for high-risk species. For journalists, stories about these massive fish may attract a broader audience than the more niche interests of sportfishers. At the same time, declines in average size and abundance have been observed for at least some species within our ultra-charismatic subgroup (Bradshaw, Mollet & Meekan, 2007; Bradshaw et al., 2008; Holmberg, Norman & Arzoumanian, 2009; Dulvy et al., 2008). The declining relative size of ultra-charismatic fishes in media reporting therefore reflects current conditions in marine ecosystems. News reporters, however, are less likely to recognize that current ultra-charismatic fishes are smaller than those in the past because for these species, reporters are making size comparisons among all fish species instead of within-species.

Conclusion

Real biological changes have occurred in the world’s oceans: the maximum size of many marine fish species has fallen and is expected to continue to decrease (McClenachan, 2009; Cheung et al., 2013), and large fishes are now generally rarer than in previous decades and centuries (Christensen et al., 2003; Myers & Worm, 2003). However, the continued use of superlatives to describe fish that are a fraction of the maximum size they could reach, or a fraction of the size they used to be, does not reflect these biological changes, except perhaps for ultra-charismatic fishes. For this group, declines in the reported size reflect real shifts in body size. Given that media outlets are a powerful tool for shaping public perception and awareness of environmental issues (Anderson, 2013), there is a real concern that such stories might be interpreted as meaning that superlatively large fish still abound. This false ocean optimism might in turn diminish support for stricter fisheries management and affect support for marine conservation efforts.

Supplemental Information