Online Noise as Illustrated by Pitfalls and Biogeography Associated With Common Names for Puma concolor

Introduction

Increasingly, conservation scientists have embraced popular news and social media channels to 1) engage the general public, 2) amplify the findings and implications of their research, 3) improve buy-in for conservation agendas and, inevitably, and 4) offset the spread of misinformation accelerated by digital networks (Papworth et al., 2015; McClain, 2017; Nanni et al., 2020). Twitter reached one billion tweets only three years after the first ever tweet (Twitter Inc., 2011) and, since 2014, people have been tweeting over 500 million times per day. Conservation practitioners now recognize internet traffic as big data that provide insights into diverse questions about how humans interact with and relate to nature (Toivonen et al., 2019), and they are leveraging online tools to engage and communicate with a vast and diverse audience (Bik et al., 2015). Culturomics, for example, is the emerging field in which researchers analyze quantitative data reflecting cultural trends in language and communication, and the sentiment associated with communication (Michel et al., 2011; Toivonen et al., 2019).

Given the current speed with which information moves across networked devices, and the sheer volume of information being shared today, word choice is more imperative than ever in effective communication. Noise is defined as the non-target search results that people encounter when searching for a particular topic of interest; noise is also the cloud of distracting data that can obscure or deflect one's conservation communication. Vernacular names for species, for example, create noise for conservation practitioners communicating with colleagues and the general public. When a species has more than one common name, it can fragment conservation attention (Ladle et al., 2016), or in cases when there is confusion over what name matches with what organism, vernaculars may hinder conservation support (Jarić et al., 2016). Further, names themselves carry negative and positive connotations (Karaffa et al., 2012; Ehmke et al., 2018). These issues are exacerbated when species names are used as product names (called theronyms), or when cultures create new meanings for animal names to represent concepts that are completely different (called homonyms) (Jarić et al., 2016; Ladle et al., 2016).

Online noise associated with large carnivores is particularly dense because their defining characteristics, such as beauty, power, stealth and dominance, have made them salient and, consequently, part of our everyday vernacular. This is important to recognize given that effective conservation communication is especially imperative for people working on the conservation of large carnivores and related ecosystem health. In general, real and perceived risks associated with predators drive human perceptions about these species (Bruskotter et al., 2017; Bombieri et al., 2018). Nevertheless, online resources about large carnivores frequently include misinformation that can undermine conservation strategies for these species, and that facilitate the rapid spread of negative sentiment that can reduce social tolerance for large carnivores more broadly (Bruskotter and Wilson, 2014); (Young et al., 2015).

Mountain lions (Puma concolor) exemplify the dilemma surrounding the use of multiple vernaculars for a species in the crosshairs of conservation conundrums. Barnes (1960) reported 84 common names for the species in English, Spanish, Portuguese and numerous indigenous languages across North and South America (Table 1). P. concolor inhabits the largest range of any native terrestrial mammal in the Americas and holds the Guinness World Record for the animal with the most common names (Guinness World Records, 2019). P. concolor is also a charismatic apex carnivore that disproportionately supports biodiversity and ecological resilience (Elbroch et al., 2017b; Barry et al., 2019), as well as a prominent character in diverse spiritual beliefs and historical narratives of the various cultures with which it overlaps (Herrmann et al., 2013), suggesting it also plays a cultural keystone role (Garibaldi and Turner, 2004). Yet despite its value to people and the ecosystems upon which humans depend, P. concolor conservation management remains controversial because the species competes with humans for space, ungulates, and other resources (Elbroch et al., 2017a), and because it poses both real and perceived risks to people, pets and livestock (Herrmann et al., 2013; Wolfe et al., 2015). Media reports also sensationalize encounters with P. concolor, contributing to negative sentiment about the animal (Wolch et al., 1997; Bombieri et al., 2018).

Table 1. Common names for Puma concolor reported in Barnes (1960) in alphabetical order, plus “Florida panther.”

Among P. concolor's most frequently used vernacular names are “mountain lion,” “cougar,” “puma” and “Florida panther.” Multiple theronyms exist for these vernaculars, including Mountain Lion, an Apple Macintosh operating system released in 2012, the Mercury Cougar, a U.S. model of vehicle manufactured from 1967–1997 and 1999–2002, and the Keystone Cougar, a popular line of recreational vehicles. “Cougar” is also slang for an older woman who dates younger men. Black Panther is the name of a Marvel comic character, as well as synonymous with black jaguars and leopards. Perhaps the most well-known theronym is PUMA, a global brand of clothing and athletic equipment that sponsors star athletes and global sporting events. “Puma” is also the name of a line of recreational vehicles competing with the Keystone Cougar. In addition, all of these vernaculars are widely used as names of professional and college sports teams, mascots, retail stores and more.

Our goal was to gather diverse usage data from the internet, social media, scientific literature and news articles to assess the cultural salience, or frequency of use, for the four most common P. concolor vernaculars, and their associated noise. This information is relevant to refining our understanding of best strategies when communicating about P. concolor to a variety of audiences across online and in print platforms, as well as across film, radio, and print communication. More broadly, we offer a suite of easy-to-navigate methods for conservation practitioners to explore, understand and penetrate the noise around topics of interest to ensure conservation communication reaches its intended audience.

Materials and Methods

Cultural Salience and Biological Relevance

We measured cultural salience in terms of total search results for “mountain lion,” “cougar,” “puma” and “Florida panther” on Google Search and Instagram. We determined the biological relevance of search results, as defined by the proportion of the top 100, or total available results if <100, for each name on Google Images, YouTube, Twitter, and Facebook, and the proportion of the first page of Google Search results, that were pertinent to P. concolor (sensu Çakir et al., 2008).

Bioregional Analysis With Google Trends

We used Google Trends to assess the use of each of the four names over a 5-year period spanning August 2, 2015 until August 2, 2020. Google Trends data illustrate “interest” in a subject for 1-week periods over time, scored on a scale ranging from 1 to 100. We did this for U.S. and Mexican states, Canadian provinces and other countries as a whole for which there were sufficient trend data. We first sampled for each name broadly, and then compared results when searching within the subcategory “Pets and Animals.”

Use in Peer-Reviewed Science

Our goal with regards to science literature was to assess the proportion of papers that use each of the vernaculars and to determine whether there has been any change in the use of vernaculars over time. We surveyed articles about P. concolor populations in the United States and Canada, separately from those from Latin America, where populations are almost always referred to as pumas.

To gather a suitable sample of published research on P. concolor, we conducted searches of ISI Web of Science and Google Scholar for empirical papers presenting new data that were published between 1950 and 2020. This search excluded book chapters, conference proceedings, state reports, reviews and student theses and dissertations. We selected search terms that reflected broad interests in ecology, but our searches were not exhaustive. We searched for 23 topics (food web, ecosystem, landscape of fear, keystone, ecosystem engineer, apex predator, trophic cascade, regulation, carrion, biodiversity, disease, risk effect, prey, social, dispersal, home range, territory, fragmentation, urban, suburban, survival, mortality, behavior) in combination with each vernacular (cougar, puma, mountain lion, panther).

We lumped our results into 5-year bins for analyses. We plotted the proportional use of each vernacular over time and tested whether the slope of any line differed from zero, as evidence of proportional change in their representation in the literature. We did this separately for papers about P. concolor populations in the US and Canada, and for populations in Latin America.

Sentiment Analysis

We measured sentiment associated with the four names for P. concolor in two ways. First, we employed a black box social listening sentiment analysis tool via Mediatoolkit (2020) to quantify and analyze social “impressions,” defined as the frequency with which content is seen, and “mentions” (i.e. use of the names in social posts) as positive, neutral, negative or unclassified. The benefits of using online social listening tools is that they simultaneously search a huge amount of internet traffic, building large sample data, and calculate numerous metrics useful to understanding the salience and sentiment associated with words and phrases. Mediatoolkit, to our knowledge, offers the largest amount of historic search data during a free trial.

We conducted searches and gathered analytical data for each P. concolor vernacular name and its plural variations to capture the breadth of relevant mentions from Sept 1—December 1, 2020; most importantly, this method included online noise (i.e., search results unrelated to P. concolor) associated with each vernacular. We combined search variations into our main categorical vernacular names as follows: “puma” represented searches for puma and pumas, “mountain lion” for mountain lion + mountainlion + mountain lions + mountainlions, “panther” for panther + panthers, and “cougar” for cougar + cougars. Then we applied a k-proportions test to determine whether one or more of these vernaculars were associated with more positive or negative sentiment than the others. When Chi-square statistics revealed that there were differences between categories, we applied a post-hoc Marascuilo procedure (i.e., contrasts; Marascuilo, 1971) to determine which proportions differed from others.

Second, we analyzed news media content from 2000–2015 archived on Newspapers.com and Google News. We searched for the four vernacular names, and then selected only those articles about P. concolor, and for which only one vernacular was used. In other words, we screened news media for biological relevance and removed potential noise before conducting any analyses. Then we determined whether articles were positive, neutral or negative based upon the lexicon employed, and sentence structure and punctuation that relays tone in the article (Table 2). We employed the same statistical procedures as described above for assessing whether there were differences in the proportion of articles with positive or negative sentiment associated with the four vernacular names.

Table 2. Example news article excerpts and associated sentiment, as determined by lexicon, sentence structure and punctuation.

Results

Cultural Salience and Biological Relevance

In terms of raw content identified by querying each of the four vernacular names, “puma” was by far the most often found on Google and Instagram, followed by “panther,” “cougar” and “mountain lion” (Table 3). Nevertheless, when viewed in terms of biological relevance, their performance was almost an exact mirror image, with “mountain lion” yielding the most relevant material across Google searches and social media, followed by “cougar.” “Panther” and “puma” were the noisiest, as the majority of content inclusive of these terms referred to theronyms (e.g., the brand PUMA) and homonyms (e.g., Black Panther, the Marvel character).

Table 3. Cultural salience and biological relevance for P. concolor vernaculars in web and social media searches.

Google Trends

A comparison of Google Trends analyses for each name with and without the Pets and Animals subcategory illustrated the noise associated with “puma” and “cougar” as compared to “mountain lion,” and the influence of mountain lion-human conflict on interest for the species (Figure 1). When we sampled only data within the “Pets and Animals” category to increase biological relevance, we found clear biogeographical patterns to the use of the four names (Figure 2). In short, Canada predominantly uses “cougar,” the United States predominantly uses “mountain lion” and all Latin countries predominantly use “puma” (Supplementary Table 1). In other parts of the world, “puma” is predominant among countries for which there are Trends data (Supplementary Table 1).

Figure 1. Screen captures of Google Trends outputs representing “interest” in cougar, puma and mountain lion over a 5-year period spanning August 2, 2015 until August 2, 2020: (A) is the results of general searches for puma, cougar and mountain lion, and (B) is the same search, but selecting the “Pets and Animals” category, to mitigate associated noise associated with these vernacular names. Mountain lion, for example, exhibits the least interest in (A) but the most interest in (B). Also note the three spikes in search attention in (B) corresponding to negative interactions between people and P. concolor: 1. The fatal attack of a mountain lion in Washington (Lacitis, 2018). 2. The non-fatal attack of a jogger by an orphaned kitten in Colorado (Schwartz, 2019). 3. The prolonged defensive behaviors of a female with small kittens and a jogger in Utah, the video of which went viral (Capron, 2020).

Figure 2. Biogeography of dominant vernacular use for P. concolor across the Americas. Proportional use for each vernacular in each region is found in Supplementary Table 2.

Use in Peer-Reviewed Science

Our search resulted in 190 peer-reviewed articles (Supplementary Appendix 1) published in scientific journals about P. concolor in the United States and Canada, and an additional 65 papers about populations in Latin America. The number of articles produced for all vernaculars in the US increased over time (Figure 3), indicating increasing scientific attention for the species overall. Nevertheless, only “puma” showed a positive slope different than zero (F = 11.02, p = 0.01, R² = 0.58) indicating an increase in use over time as compared to the proportion of papers using other names. Both “cougar” (F = 1.15, p = 0.31, R² = 0.13) and “mountain lion” (F = 0.30, p = 0.60, R² = 0.04) did not change their proportional representation in published, peer-reviewed literature over time.

Figure 3. Number of peer-reviewed science articles published using each of four P. concolor vernaculars from 1970–2020.

The first paper on a Latin American population identified by our search criteria was published in 1987, and the first to use a vernacular other than puma was published in 2002 (mountain lion in Bank et al., 2002). Puma was used in 91% of science papers covering P. concolor populations in Latin America, cougar in 6%, and mountain lion in 3%. Based upon a cut off of 0.05, the proportional predominant usage of puma did not change over time (F = 4.11, p = 0.10, R² = 0.46), despite the imbalance of usage due to the late use of alternate vernaculars other than puma.

Sentiment

Our search for the vernacular names on Mediatoolkit yielded 563,367,467 impressions and 65,910 mentions with different proportional sentiment (Table 4). Impressions and mentions yielded different patterns; for example, “mountain lion” exhibited the second highest number of impressions but the least mentions. Sentiment varied with vernacular (Figure 4). Cougar and panther mentions exhibited significantly more positive sentiment than puma and mountain lion mentions (X² = 1, 797.34, df = 3, p < 0.001) (Supplementary Figure 1). Each vernacular name exhibited statistically distinct proportions of negative mentions (X² = 1, 460.60, df = 3, p < 0.001). From least to largest, “panther” exhibited the smallest proportion of negative mentions, followed by “cougar,” “puma,” and then “mountain lion” (Supplementary Figure 2).

Table 4. Social impressions, mentions and associated sentiment for mentions for each P. concolor vernacular and associated variations, determined by Mediatoolkit (2020).

Figure 4. Above: Proportional representation of categorical sentiment for four P. concolor vernacular names from September 1—December 1, 2020, as determined by Mediatoolkit (2020)'s social listening tools and including noise. Below: Proportional representation of categorical sentiment for four P. concolor vernacular names in news media from 2000–2015, screened for noise. Blue is negative, red is neutral, gray is positive, and yellow represents the proportion of articles that were not classified.

We identified and quantified sentiment for 2221 biologically-relevant newspaper articles that used one vernacular (n = 1122 cougar, 992 mountain lion, 85 puma, 22 Florida panther) (Supplementary Materials, Appendix 2). In news media, “mountain lion” yielded disproportionately fewer articles with positive sentiment (X² = 45.92, df = 3, p < 0.001) (Figure 4), whereas the other vernacular names were statistically equivalent. We did not detect any difference in the percentage of negative articles for any of the vernaculars (X² = 4.31, df = 3, p = 0.23), which may have been influenced by the comparatively small sample of articles for “Florida panther.” In other words, when noise was removed, sentiment was nearly identical across P. concolor vernaculars.

Discussion

As conservation practitioners with specialized knowledge of specific conservation issues, we must recognize that public outreach is part of our scientific agenda (Bik et al., 2015) and be conscious of crafting communication that reaches and resonates with our intended audiences. In our case study, P. concolor suffers from what Ladle et al. (2016) called the “-onym challenge” (p. 273), meaning that the animal's many vernacular names create substantial misleading “noise.” Across internet searches, social media channels and news outlets, “puma” and “panther” exhibited higher cultural salience than “cougar” or “mountain lion.” In terms of biological relevance, however, “mountain lion” yielded the best search results. Nevertheless, geographic preference for different vernaculars complicates making just one recommendation for conservation practitioners wishing to communicate about P. concolor to as wide an audience as possible across multiple platforms. Further, our analyses of sentiment associated with P. concolor highlighted the need for increased public education about the positive roles this animal plays in natural ecosystems and human well-being, and for potential proactive outreach strategies to offset negative media coverage that spikes following conflicts between P. concolor and humans, pets and livestock (e.g. Wolch et al., 1997; Bombieri et al., 2018). Ultimately, an awareness of online noise can make conservation communication, ranging from education to fundraising campaigns, more intentional and effective.

As a vernacular polyonymous species (Correia et al., 2017), it may be that a single communication strategy cannot be applied for P. concolor across audiences, communication platforms and geographies. The most culturally salient English vernaculars for P. concolor were the noisiest. “Puma” and “panther,” for example, were used most often but yielded very little biological relevance across web searches and social media (Table 2). “Cougar” exhibited excellent biological relevance on image searches but only 2% accuracy on Twitter. The high volume of noise attributed to the slang use of “cougar,” particularly on Twitter, may warrant limiting its use on social media to local situations. “Mountain lion” outperformed every other vernacular, in terms of biological relevance, and is the most commonly used vernacular referring to the animal in the United States. “Mountain lion” also best captured spikes in internet traffic about P. concolor (Figure 1). On the other hand, mountain lion was a vernacular derived by settlers, and puma is the only one of these four vernaculars that was derived by American indigenous people (Peru). Perhaps promoting the use of puma could aid in mitigating ongoing injustices stemming from settler colonialism (e.g. Hendlin, 2014; Eichler and Baumeister, 2018). Further, puma is the predominant vernacular used for the species around the world.

Clear geographic patterns of vernacular use (Figure 2) may ultimately decide word choice for conservation practitioners communicating to targeted audiences. Likely, people in specific regions will exhibit inertia in terms of adopting new vernaculars if one is already widely used. Florida media outlets, for example, would be wise to follow the lead of their local conservation organizations and wildlife authorities and use “panther” in their coverage about P. concolor populations in their state. Similarly, Washington State media outlets should use “cougar” to communicate about P. concolor to their audiences. However, because most local media today can find a larger audience online, they should consider also including the most biologically relevant vernacular, “mountain lion,” and the most common vernacular at the global scale, “puma,” in their reporting and promotion to broaden the reach of their coverage and knowledge about the species.

Some highly regarded sources of information likely influence the cultural salience and biological relevance of search results. Google's default name for P. concolor is “cougar,” and is linked to the Wikipedia page for P. concolor; Google's decision may influence both Google Trends and web data. For example, searches for “cougar” were more biologically relevant on Google than unrelated social media channels. The genus for the animal was changed from Felis to Puma in 1996 (Nowell and Jackson, 1996), and the prominent International Union for Conservation of Nature's (IUCN) Red List of Threatened Species uses “puma” (Nielsen et al., 2015). The only professional conference dedicated to the species, which is held in the United States every three years, is called the Mountain Lion Workshop. In published science, “puma” appears to be gaining favor over other vernacular names, and perhaps with time, this trend will influence media and other public discourse as well.

Online negative sentiment among vernaculars appeared positively correlated with biological relevance, meaning that more biologically relevant results yielded higher proportions of negative sentiment and lower proportions of positive sentiment. In part, noisy vernaculars should generate more positive sentiment because theronyms are associated products that are meant to represent the positive attributes of the animal (Ladle et al., 2016). In our study, negative sentiment likely reflected the reality of print and social media patterns that follow conflicts between people and P. concolor. Puma-human conflict, including pet and livestock losses, result in real costs and trauma for the people involved, and factual reporting reflects these hardships. Occasionally, however, reporting on these conflicts include misinformation, or are framed in such a way as to propagate negative sentiment (Wolch et al., 1997; Bombieri et al., 2018). For these reasons, we would emphasize the need for conservation practitioners to engage in proactive media and education campaigns to increase positive sentiment associated with P. concolor. In India, for example, conservation biologists provided the media with training on leopard ecology, after which the portrayal of leopards in the media was more accurate and positive (Hathaway et al., 2017). We did not study the spikes in our Google Trends data that corresponded with major news about P. concolor encounters, but a future study could evaluate the attention around those particular events to quantify the broader effect of negative media on sentiment as well.

The proliferation of noise across digital platforms is uniquely evident in the case of mountain lions, but it has considerable implications for the broader conservation community. We encourage readers to utilize available resources that provide guidance on how to get involved and how to build rapport with potential audiences (Parsons et al., 2014; Wilson et al., 2016; Cooke et al., 2017). Ultimately, an important next step is to determine whether noise is impacting conservation outcomes through either misdirecting conservation communication so that audiences never receive it or obscuring conservation communication so that misinformation is difficult to disentangle from fact. Inconsistencies in shared vernaculars for species within the conservation community, for example, may hinder conservation success (Jarić et al., 2016). On the other hand, such insights might be leveraged into campaigns to reclaim noisy names or increase the frequency of use for a less-popular one. Further, if conservation practitioners could capitalize upon online noise, such as the popularity of theronyms, they may be able to initiate a positive feedback loop wherein a subject of conservation receives greater attention than it would have otherwise, leading to a more educated audience and increased positive sentiment about the topic. We believe an awareness of noise will not only help conservation practitioners become more effective communicators, but also create opportunities to ride the wave of popularity and positive sentiment attributed to the idea of wildlife and natural processes, to improve public understanding and people's perceptions of the real ones. Ultimately, when we as conservation practitioners are informed about and intentional with word choice, the messages and the organisms we want to promote can receive unprecedented visibility.

Data Availability Statement

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

Author Contributions

LE, EW, and AA conceived the project. AA, EW, and LE designed the methods. EW, CD, WO'M, and AS gathered the data. EW and LE conducted the analyses. LE, EW, and KW wrote the manuscript. All authors provided feedback and approved the final draft.

Conflict of Interest

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.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcosc.2021.692607/full#supplementary-material

References