Introduction

The extinction rate of species is increasing (Ceballos et al. 2015). Human activity is currently threatening tens of thousands of species with extinction, but for many of these their fate is not sealed, and preserving much of Earth’s remaining biodiversity is still a possibility (Tilman et al. 2017). This is a huge, ongoing challenge that will require the effort and ingenuity of a large number of people for many generations to come. Given this, it is imperative to the future of our world that children grow up learning about, caring about, and wanting to protect wildlife. An important part of developing environmental stewardship is bonding (Rule and Zhbanova 2012) and, as Wilson (1997) points out, we tend to bond with what we know and understand well.

Positive influence on the attitudes of students and their motivation towards learning about the natural world have been demonstrated by encouraging children to engage more with nature in outdoor areas of nurseries and schools, and on field trips (Nadelson and Jordan 2012). However, such direct exposure to nature has financial limitations and indoor learning, including exposure to both fictional and non-fictional literature with natural themes, is increasingly felt to be part of the solution (Sackes et al. 2009; Wells and Zeece 2007). As op de Beeck (2018) eloquently points out, for the past 25 years we have seen a top-down approach exhorting parents and teachers to take children outdoors, with limited attention to what the children might be reading or thinking about before or after they go there. However, picture books are recognised as an important tool in engaging students in the classroom in a variety of subjects (e.g., Mathematics Cooper et al. 2018, Evolutionary biology Browning and Hohenstein 2015). In addition, any single child’s available and easy to reach ‘outdoors’ will be unlikely to cover a wide range of ecosystems, strengthening the requirement for sources other than first-hand experience to form a vital part of developing global environmental awareness throughout childhood.

Previous studies have highlighted erroneous representations of various aspects of the natural world in the text and pictures of children’s books. For example, Trundle et al. (2008) found that 20% of children’s texts depicted the moon incorrectly, Schussler (2008) noted shortcomings in the representation of plant reproduction with 38% of books describing the production of seed or fruit having no mention of a pollinator or of pollination or fertilisation, 64% of children’s books featuring whales and dolphins were found to have basic biological errors in the text and/or the pictures (Beaumont et al. 2017), and Muthukrishnan and Kelley (2017) lament after their review of 35 children’s books focusing on sustainability, that the texts do not impart a strong understanding of eco-friendly behaviour and that a valuable opportunity to inform children is being missed.

In this study, the representation of tortoises, terrapins and turtles (Testudines) in children’s literature is examined and change in the scientific accuracy of their representation over time is assessed. Testudines were selected as the focus of this study as preliminary searches identified that animals in this order are frequently featured in children’s literature. Data on the IUCN Red List of Threatened Species (2018) certainly support focus on this order; of the 242 Testudines species listed as extant in the wild, 186 (76.9%) are currently listed as near threatened, vulnerable, endangered or critically endangered, with all of the main threats (biological resource use, competition with other species humans have introduced, agriculture and aquaculture, pollution, transportation, and residential and commercial development) being due to human actions, strongly suggesting that developing an understanding of these organisms among the next generation is vital.

Manual searches of home and local library shelves have previously been employed in numerous studies to identify samples of children’s books for examination of various aspects of representation within them (e.g., characters with disabilities Price et al. 2016, grandparents Crawford and Bhattacharya 2014, nature Marriott 2002 and cetaceans Beaumont et al. 2017). This method gives a geographical definition to the study as the books included in the sample are among those locally available to children living in the study area. Our study took place in the UK where the only native Testudines species is the Leatherback turtle (Dermochelys coriacea) (Beebee and Griffiths 2000) which does not nest locally, meaning that indirect exposure to Testudines, e.g., through books and television programmes, is likely to be the primary source of information children have about these species.

Methods

Twenty children’s picture books from the authors’ own home libraries, featuring at least one image of a turtle, terrapin, or tortoise were initially identified to review common categories of biological errors in the representation of these species. Table 1 shows the categories identified with a brief description of the kinds of errors considered to belong to each category. As the focus of this study was to assess the biological representation of these animals, Testudines communicating directly in English with members of their own species or with other species, was not necessarily regarded as an error. If, however, this communication contained factually incorrect information (e.g. ‘I remove my shell before going to sleep’), this was recorded as an error in the Text category.

Table 1 Description of the error categories used in this study
Full size table

A 6-month search was then conducted (June–November, 2017) for English language picture books suitable for children aged between 0 and 8 years that included a minimum of one image of a turtle, terrapin or tortoise. Manual shelf searches for potential texts were carried out in Derbyshire County Council libraries, Kedleston Road Library (University of Derby), and the home libraries of the authors. All encountered books meeting the criteria were included in the study.

Books were read and assessed to be fictional or non-fictional in character. Text in the first person did not exclude a book from being placed in the non-fiction category (i.e. wording such as ‘I live in the sea, but I crawl up the beach to lay my eggs’ could be found in a book placed in the non-fiction category). The date of publication of each book was recorded; where multiple issues of a book existed, the date of publication of the issue being examined was that recorded. The species depicted were recorded as terrestrial, freshwater or marine, and where possible (from the text, photographs and/or detailed drawings) the species depicted was recorded. In books containing multiple species, all species present were recorded. In addition to the errors in the categories in Table 1, where books included photographs, the use of flash photography to produce images of nesting or hatching activity at night was recorded. Texts were reviewed independently by the same two researchers and the total number of errors identified in each of the categories shown in Table 1 was determined. In the event of differences between researchers in the number of errors recorded for a text, the text was reviewed again with two researchers present to reach agreement.

χ2 Analysis was used to compare the number of categories errors were identified in between fiction and non-fiction books. Correlation between publication year and the number of error categories identified for each book was examined using Spearman’s rank correlation analysis. All statistical analyses were conducted using R version 3.0.1 (R Core Team 2018). A p value < 0.05 was considered significant for all statistical tests.

Results

A total of 204 English language picture books including at least one Testudines image was located during the study period. Of the total, 110 were placed in the fiction category, 93 in the non-fiction category, and 1, with parallel pages containing both a fictional story and non-fiction factual information, did not fit clearly in either of the categories. Data from the latter book were excluded from subsequent analysis comparing the fiction and non-fiction categories.

Terrestrial species appeared in 104 of the books, freshwater species in 28 books, and marine species in 105 of the books. Of the marine species, Green turtles (Chelonia mydas) occurred in books most frequently (40 books) and Flatback turtles (Natator depressus) and Olive Ridley turtles (Lepidochelys olivacea) least frequently (4 books each) (Fig. 1).

Fig. 1
figure 1

Number of books representing each of the seven species of marine turtle: Green turtle (Chelonia mydas), Leatherback (Dermochelys coriacea), Loggerhead (Caretta caretta), Hawksbill (Eretmochelys imbricata), Kemp’s Ridley (Lepidochelys kempii), Flatback (Natator depressus) and Olive Ridley (Lepidochelys olivacea). Total number of books including marine turtle species = 105

Full size image

In total, 799 errors were identified across all 204 books (mean = 3.9 errors/book, SD = 9.1) (Fig. 2). Errors of manoeuvrability, where Testudines were depicted or described as moving their bodies in anatomically impossible ways or making impossible movements (e.g., jumping), were most frequent (209 errors). The second most frequent category was depicting these reptiles with hair (200 errors).

Fig. 2
figure 2

Total number of errors identified within each of the 13 error categories. Total number of texts examined was 204. Error categories are described in Table 1

Full size image

Books examined were found to have errors in 0–6 of the categories shown in Table 1. Only 83 (40.7%) of the examined books were found to be error-free in the biological representation of Testudines (Table 2). Significantly more fictional books were found to have errors in one or more categories than non-fictional books [χ2(1, N = 120) = 5.79, p < 0.05].

Table 2 The proportion of books examined with errors in 0–6 of the categories outlined in Table 1
Full size table

No correlation was found between publication year and the number of categories that errors were observed in (rs = − 0.072, p = 0.307). Examining fiction (rs = 0.003, p = 0.977) and non-fiction (rs = 0.033, p = 0.751) books independently confirmed the lack of correlation between the number of categories in which errors were found and publication year.

Photographs appeared in 52 of the books examined (25.5%), either with or without additional drawings, and 44 of these books (21.6% of the total) included photographs of sea turtles. A total of 104 flash photographs of nesting sea turtles or hatchlings on nesting beaches was identified in these books (a mean of 2.4, SD = 6.4, flash photographs taken on a nesting beach/book with photographs of sea turtles).

Discussion

Of the 204 books examined in this study, almost identical numbers included marine (105) and terrestrial (104) Testudines species, while fewer (28) included fresh water species. As this study took place in the UK, where the only true member of the British fauna among Testudines is the Leatherback turtle, the balance between terrestrial and marine species is surprising, particularly as the Leatherback turtle was not even the most frequently featured marine species (Fig. 1). This supports previous work on the underrepresentation of native species in literature available locally to children (Beaumont et al. 2017) and is worrying because as Genovart et al. (2013) point out, knowledge of local fauna is crucial to promote conservation. In addition, there is evidence that familiarity can even lead to a preference for invasive exotic species over native ones (Shapiro et al. 2017).

The large number of errors observed in this study, 799 in 204 books, is a concern, particularly given that children in the geographical location where this study took place are likely to have indirect experience, including books, of these species as a primary source of information. The findings support previous studies that have identified errors in various scientific aspects of children’s literature (e.g. Beaumont et al. 2017; Schussler 2008; Trundle et al. 2008). While the errors found in this study in some categories are worrying in terms of development of biological understanding of the readers generally (e.g. reptiles with hair), other errors have obvious direct links to actions that humans are currently taking that are threatening to the survival of Testudines species (e.g. stating that sea turtles do not need to breathe air, an error recorded three times in the Text category, may lead to a reduced level of concern regarding the threats caused by entanglement in marine debris of turtles below the surface). The most frequently recorded error category was Manoeuvrability, including errors such as Testudines species moving their limbs into impossible positions or being able to jump. Hosier et al. (1981) found that Loggerhead sea turtle hatchlings took significantly longer to reach the surf after emerging from the nest on beaches with higher levels of ruts due to vehicle and pedestrian traffic, and more time on the beach equates to more exposure to terrestrial predators and a higher risk of desiccation. A belief that Testudines species can jump could reduce concern that heavy human use of nesting beaches can have an impact on hatchling survival.

This study found significantly fewer errors in non-fictional books, supporting earlier work on cetaceans (Beaumont et al. 2017), but the errors in non-fiction, while fewer, are more worrying as this information is likely to be endorsed by teachers and other adults engaging with text along with the child as fact: reading non-fictional children’s literature may form a primary source of information about the natural world for many parents and teachers too (Eamer 2018).

In this study, no errors have been recorded where potential misconceptions were resolved by the end of a book (for example, a hatchling exploring different ecosystems, where it would not be found naturally, before finding where it belongs was not recorded as a Location error). The errors recorded in this study are limited to those of basic biology that could leave a child with misconceptions about Testudines after reading the book. It could be argued that among the most important goals of children’s literature are for engagement with a book to be entertaining, and for it to be part of developing a love of reading. The importance of these goals is not disputed. However, it is noteworthy that the authors’ experiences using literature in outreach activities highlight that books found to contain no scientific errors in this study, include examples of the most entertaining and engaging texts for classroom use, for example the Davies (2008) text One Tiny Turtle, illustrated by Jane Chapman. Other important goals of children’s literature might include imparting moral lessons and promoting prosocial behaviours, roles that anthropomorphic representations of animals are often employed to fulfil. This study has avoided categorising many common anthropomorphic styles as errors, including characters speaking in English or wearing clothes (unless the clothing depicted would have required the removal of the animal’s shell to put on), as wild animals wearing clothes and speaking to each other in English are not widely held beliefs in the adult population that have the potential to impact wildlife conservation. Having said that, conservation projects often try to limit the use of anthropomorphic language and imagery as it is thought this can have negative effects on the perception of a threatened species (McCabe and Nekaris 2018) and there is some evidence that anthropomorphised animal characters might be less effective than realistic human characters when it comes to promoting prosocial behaviour through children’s literature (Larsen et al. 2017).

Reading helps children broaden their imaginative skills and abilities (Kohm et al. 2016), but artistic licence should not be used as an excuse for laziness on the part of authors and illustrators. Ganea et al. (2011) demonstrated that children as young as 4 years old are capable of transferring biological information from picture books to real world situations, and this transfer is presumably irrespective of the accuracy of the biological information a picture book contains. There is limited value in imagining, for example, sea turtles in predator/prey relationships with other animals that they would never encounter in the wild when there are so many interesting examples of inter-species interactions that do happen in nature which could provide a perfect springboard to engaging fictional works.

This study has found no evidence of an improvement in the scientific accuracy of representations of Testudines over time, and, given the 1974–2017 timeframe of publications assessed in this study, a period of time that encompasses the invention of the World Wide Web (Berners-Lee et al. 1992), this is both shocking and worrying. Given that of the total of 799 errors identified in this study, 724 (90.6%) were in the illustrations (Fig. 2), it is suggested that it is illustrators in particular who are being slow to make use of the wealth of accurate information available to them. Incidentally, contrary to a number of storylines encountered in this study, ‘slow’ is not a term that is appropriate when referring to all Testudines. Leatherback turtles, for example, have been recorded traveling in excess of 30 km/h, though they do normally swim at or below 10 km/h (Ecker 2002). Given that the human Olympic 50 m freestyle world record equates to a speed a shade under 7 km/h, it is neither fair nor accurate for human authors and illustrators to depict all Testudines as slow.

The high frequency of flash photographs of sea turtles on nesting beaches identified in this study is a concern, though these were not noted as errors. Light pollution is something that can affect sea turtle nesting and numerous conservation organisations advise against the use of flash photography on nesting beaches to reduce disturbance to sea turtles (e.g. Alabama Sea Turtle Conservation Program 2018; Sea Turtle Conservancy 2018; SEE Turtles 2018). The 104 flash photographs of nesting or hatchling sea turtles taken on nesting beaches that were recorded represent only the photographs that were used, and it is likely that a far higher number of photographs will have been taken to obtain 104 useable images. The concern is that the high frequency of such images in children’s literature normalises the behaviour of light pollution on nesting beaches, and at a time when the trend to photo-document every activity of travel is increasing. There is a risk that readers may, in the future, copy the behaviour demonstrated, and the opportunity to educate about more appropriate human behaviour on nesting beaches is being missed.

A potential criticism of this study is that it has not focussed specifically on award-winning books as some other work examining the representation of the natural world in children’s books has, for example Babb et al. (2018). While it is clear that library purchases may be influenced by book awards, the methodology employed in this study reflects the reality that children do not use award criteria when making selections and will select books from those available to them. As data collection for this study included occasions where authors were multi-tasking with parental duties, book searches were both child- and adult-led, mimicking the activities Becker (2012) reported in an observational study which took place in the children’s section of a public library. Internet searches using Amazon have also been used in previous studies to locate relevant books for analysis (e.g. Cooper et al. 2018; Muthukrishnan and Kelley 2017) but this search method was not selected in this study as the goal was for the sample to reflect books available locally.

Considering ways that children’s literature can become more of a force for good in developing environmental awareness, several approaches might be considered. More training for teachers and librarians around the consideration of biological misrepresentations when selecting books could be effective. Works with inaccuracies can be used in the classroom to link stories to early childhood science education in a way that highlights the biological misrepresentations in fictional work. Challenging children to identify parts of the plot that could not occur in the real world (‘Could the Foolish Tortoise really take off his shell?’) is a way of extending the enjoyment of the text, and personal experience of the authors suggests that this is something that can help develop a child’s critical thinking skills more generally. More emphasis on collaborative working in the publishing industry needs to be encouraged as our sample included examples of disagreement within a book between the biological accuracy of the text and illustrations probably due to the fact that, as Trundle et al. (2008) point out, author and illustrator are frequently working independently. In addition, given that the academic community are used to the review system for academic manuscripts, and that they have a vested interest in improving the understanding of the organisms they work with among the general public, the publishing industry should make more use of academics in Zoology and Environmental Science departments to offer comments prior to publication of both fictional and non-fictional work to improve the representation of the natural world in children’s literature moving forwards.

Conclusion

As Harju and Rouse (2017) point out, there is now a greater sense of urgency in the appeal of recognising our interconnectedness with nature. No one is born knowing how to read and write (Silcock 2008) nor are they born with biological understanding of the world around them. We need children to appreciate nature locally first hand in addition to having access to accurate materials that indirectly allow them to appreciate the natural world more widely. Both fiction and non-fiction have a place here, but some of the best fiction can be developed from the amazing ways organisms actually do interact with each other. Wonderful examples of symbiosis, mutualism, and commensalism can be exploited for fictional works.

In terms of attributes like race, gender and disability, it is widely accepted that children’s literature has room for improvement in ensuring that all children have the opportunity to see themselves reflected in the books they read; seeing self is critical in developing feelings of belonging, and not seeing self can lead to feelings of marginalisation (Koss 2015). It is argued that children also need to have the opportunity to see their natural environment reflected accurately in literature in order to develop feelings of belonging to, rather than marginalisation from, the natural world.

The plea to improve the scientific accuracy of the representations of organisms in children’s literature is not just about educating future biologists and policy makers (though DeWitt et al. 2014 report worrying trends in the science aspirations of primary and secondary pupils which need to be addressed). More generally, this is about developing future adults with appreciation of how their everyday actions (for example, litter disposal, shopping choices, and transportation choices) can have a profound effect on species they have grown up knowing about, caring about, and have developed an affinity strong enough to want to act to protect.