Weaver 1 Sabertooth b lennies ( Plagiotremus azaleus ) benefit from aggressive mimicry and schooling with c onspecifics Olivia A. Weaver E cology and Evolutionary Biology University of California Santa Cruz UCEAP Tropical Biology and Conservation Spring 2019 7 June 2019 Abstract: Mimicry serves as a marker of a longstanding and evolutionary complex relationship, though studies on mimicry are notoriously difficult to execute, as there are many factors to consider. In interspecific relationships that are reported to exhibit both freq uency dependent and frequency independent mimicries, evaluating the benefits of association are especially complex. The relationship between the Sabertooth Blenny Plagiotremus azaleus and the Cortez Rainbow Wrasse Thalassoma lucasanum is proposed to be one of both aggressive and social mimicry, with the former species acting as the mimic and the latter as the model. However, previous observations as well as my own have observed large conspecific schools of P. azaleus , a phenomenon that ignores the tradition al assumptions of frequency dependent mimetic relationships. In this study, I sought to review the benefits that P. azaleus incur from their likeness and association with T. lucasanum , and to investigate what benefits P. azaleus might derive from associating with themselves. The results of this study support the established claims that P. azaleus enjoy increased predation success when associating with their model, which in turn supports the presumed aggressive mimicry between the blenny and T. lucasanum . Although sample sizes were too small to determine significant benefits they might derive from associating with each other, low P Values hint at hypotheses that conspecific schools of P. azaleus are chased by fish less often than when they are either solitary or associating with their models. I propose that the complexities in P. azaleus behavior may be the result of the interaction between aggressive and social mimicry it is reported to exhibit, or that, like P. rhinorhynchos , P. azaleus could possess facultative mimicry and so choose when to perform mimicry. Borracho Veloz ( Plagiotremus azaleus ) se benefician del mimetismo agresivo y de formar escuelas con sus conespecÂ’ficos Resumen: El mimetismo se conoce como una relaciÂ—n compleja que ha sido formada a travÂŽs de muchos aÂ–os evolutivamente hablando. Los estudios de este tema son difÂ’ciles de ejecutar, ya que hay muchos factores a considerar. En las relaciones interespecÂ’ficas donde se exhibe mimetismo dependiente e independi ente de la frecuencia de individuos, la evaluaciÂ—n de los beneficios de la asociaciÂ—n es especialmente compleja. Se ha propuesto que la relaciÂ—n entre los peces Borracho veloz Plagiotremus azaleus y el Arco Iris de CortÂŽs Thalassoma lucasanum es de mimetismo
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 2 agresivo y mimetismo social, con la primera especie actuando como imitadora y la Âœltima como modelo. Sin embargo, observaciones previas han encontrado grandes escuelas de P. azaleus , un fenÂ—meno que ignora los supuestos tradicionales de las relac iones mimÂŽticas dependientes de la frecuencia. En este estudio, investiguÂŽ los beneficios que logran P. azaleus debido a su semejanza y asociaciÂ—n con T. lucasanum, y tambiÂŽn estudiÂŽ quÂŽ beneficios podrÂ’an tener al asociarse con ellos. Los resultados de es te estudio apoyan que P. azaleus disfruta de un mayor ÂŽxito de depredaciÂ—n al asociarse con su modelo, lo que a su vez apoya el supuesto mimetismo agresivo entre P. azaleus y T. lucasanum . Aunque los tamaÂ–os de muestra fueron demasiado pequeÂ–os para determ inar los beneficios significativos que podrÂ’an derivar de asociarse entre sÂ’, los bajos valores de P apuntan a apoyar la hipÂ—tesis de que las escuelas conespecÂ’ficas de P. azaleus son perseguidas menos frecuentemente que cuando estÂ‡n solos o se asocian con sus modelos. Propongo que el complejo comportamiento de P. azaleus puede ser el resultado de la interacciÂ—n entre el mimetismo agresivo y social, o que, al igual que P. rhinorhynchos , P. azaleus podrÂ’a poseer mimetismo facultativo y asÂ’ elegir cuÂ‡ndo real izar el mimetismo a su beneficio. Introduction: Mimicry describes a phenomenon where one species has evolved to loo k like another species to incur some benefit such as reduced predation. In the case of Batesian mimicry, a harmless species has evolved to l ook like a harmful species to avoid predation. The success of this of mimicry is frequency dependent: there need to be more models than mimics so that predators learn to associate the mimicked trait with harmfulness (Turner 1987). The success of mimic ry is not frequency dependent in Mullerian mimicry, when two harmful species resemble each other, because predators can learn to associate the trait as harmful from either mimic (Turner 1987). Aggressive mimicry illustrates the "wolf in sheep's clothing" p henomenon, where a harmful species resemb les a harmless species to get closer to prey that would be otherwise deterred (Wickler 1965). Like Batesian mimicry, aggressive mimicry is also frequency dependent, and mimics must associate with a greater number of models than mimics to be successful in their predatory endeavors (Wickler 1968) Aggressive mimicry can also be related to social mimicry, which wor ks similarly to Mullerian mimicry in that it refers to similarly colored individuals that associate in order to evade predation (Moynihan 1968). Mimicry is documented to be globally common in coral reef fish, and aggressive mimicry accounts for approximat ely 48% of all reported mimetic relationships, the most prevalent form of mimicry in coral reef fish by far (Molland et al. 2005). Some families fulfill the roles of mimics and models more commonly than others, for example, 28% of reported Batesian mimics, 25% of aggressive mimics, and 38% o f social mimics in coral reefs are members of the Blenniidae family (Patzner 2009). Though they may also serve as model species, Blenniids commonly use members of the Labridae family as their models (Molland, Even et al. 2005). The members of the Labridae family are known as wrasses, a diverse group of colorful marine fish that commonly act as parasite and debris cleaners for other fish, or feed on algae on rocks (DeLoach and Humann 2004, pg 327 343). The Blenniidae fami ly is just as diverse,
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 3 though some species reputation among the reef community is not as positive as the wrasses, as they are "parasitic" predators, and feed by nipping bits of flesh and mucus from passing fish (DeLoach and Humann 2004, pg 156 175). These differences in diet create an opportunity for the evolution of aggressive mimicry, as reef fish either ignore wrasses or approach them, allowing aggressive mimics to strike. An example of such a relationship can be found between the Bluestriped Fangblenny Plagiotremus rhinorhynchos and Bluestreak cleaner wrasse Labroides dimidiatus .The cleaner wrasse, true to its name, remove ectoparasites and debris from larger fish, which actively approach them to receive these cleaning services (Patzner 2009). The mimic P. rhinorhynchos has several colour variations and does not always share coloration and association with its model, but when it does, it is able to increase its access to prey (Molland and Jones 2004). A 2004 study designed to investigate the validity of t he presumed aggressive mimicry between P. rhinorhynchos and L. dimidiatus confirmed the factors necessary to declare such a relationship. The mimic occurred less frequently than the model and was generally nearby it (so that predators would be more familia r with the model than the mimic), and suffered reduced foraging success at the removal of said model (Moland and Jones 2004). The color variations in P. rhinorhynchos that are non mimetic of L. dimidiatus are proposed to be evidence of facultative mimicry and possible social mimicry, as these alternative forms were found associating with other species of schooling fish (Patzner 2009). The Sabertooth blenny Plagiotremus azaleus does not possess multiple color variations, but it does superficially resemble th e juvenile Cortez Rainbow W r asse Thalassoma lucasanum , and this resemblance is reported to be evidence of both aggressive and social mimicry (Humann and Deloach 2004, pg 186 187; Patzner 2009). Like many wrasses, T. lucasanum occurs in three morphs as a fu nction of its age, gender and status: juvenile, initial phase (females and sneaker males) and terminal males [Fig 2, Fig 1]. Fig 1: Terminal phase and adu l t T. lucasanum Fig 2: Juvenile phase T. lucasum (left) and aggressive mimic P. azaleus (right)
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 4 It is not overly difficult for the human eye to distinguish free swimming P. azaleus from their models; they have been described well as " more slende r and utiliz[ing] anguilliform rather than labriform swimming " (Galland et al. 2010). Essentially, they sw im mostly with their tail, and are easily recognized by this unusual movement as well as their characteristic length and rounded, blunt face. Like many others i n the b lenniid family, P. azaleus are described to reside primarily in abandoned mollusk tubes o r other cylindrical holes, which they use as both protection fro m harassment and cover to ambush prey (Hobson 1968). Also , like the rest of their famil y , P. azaleus is described as territorial, and rarely appears in conspecific aggregations (Hobson 1968). Recent reports seem to contradict this latter trait, and one account from the team o f G.R Galland in Cabo Pulmo National Park in the Gulf of California reported schools of ten, twenty, and over a hundred P. azaleus that had completely abandoned associating with T. lucasanum (Galland et al. 2010). I also observed conspecific schools of P. azaleus throughout my study in the Bay of Cuajiniquil in Costa Rica , and because this phenomenon indicates a behavior that totally opposes an aggressive mimetic relationship, obvious questions are raised about the motivations for either mimetic or conspecific schooling. In my study, I sought to evaluate the effect of mimetic and conspecific associations on the predatory success and a voidance of persecution i n P. azaleus . Methods I collected the data included in this study in three bays in the gulf of Santa Elena (near the bay of Cuajiniquil), which is located on the northern Pacific coast of Costa Rica. My direct observations of P. azaleus totalled 169.5 minutes and my general observations reached 12 hours, producing a data set comprised of 86 unique schools and 138 cells of data. Here, a cell refers to the manner in which I compiled my data in Excel, so that an observation of two mutually exclusive behaviors (i.e. two attacks versus an attack and chase) in an individu a l P. azaleus pertain to two separate cells. I accumulated this data over two weeks of snorkeling in the middle part of May, a time that marks the beginning of the ra iny season for the area and the middle of nesting season for the potential prey found in vulnerable Acapulco and Giant damselfish par e nts ( Stegastes acapulcoensis and Microspathodon dorsalis , respectively). The nesting behaviors of these fish are of specia l interest to this study because their defensive behaviors required them to be both static and reactive to other fish, which provided both a reliable prey source and ample predatory opportunities f o r P. azaleus . When I found an individual P. azaleus , I began chronicling the schooling and behavioral data. The behaviors of interest to me were "chase ," where a blenny or individual o f T. lucasanum was chased by another fish, "dart," in which a blenny moved with quick, intentional speed towards potential prey , and "attack", in which a blenny made observable contact with said prey. I counted the time in seconds from when the period of observation began to when a behavior was performed, with the overall observations for a school tota ling no more than five minute s and usually two or three minutes in an attempt to ameliorate the effec t s of pseudoreplication. The direct observations were recorded in the format described in Table 1.
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 5 Table 1: Sample data collected for an individual blenny at Baja Rojo over 30 secon ds Blenny T.M Adult Juvenile Notes: 1 0 0 7 25 sec dart Acapulco 26 sec Acapulco chase This format counts the number of mimics and models as well as adult and terminal ma l e T. lucasanum . Although these other phases do not have the physical characteristics shared by the juvenile phase and P. azaleus , they are still members of the model species and so might influence the behavior of other fish. The victims (of dart or attack behavior) and aggressors (of chase behavior) were also accounted for. Observations were specific to only o n e P. azaleus individual (except in the case where a non study species chased T. lucasanum ), and I took additional notes on the surroundings of the observed blenny, such as "scho o l of barberfish above damselfish nest site". Results I constructed contingency tables and ran subsequent Chi Square tests in JMP to evaluate the significance of correlations between the likelihood of a behavior and the likelihood of a particular school type. I first created six contingency tables to calculate significance between the three studied behaviors and either model presence of presence of conspecific school. The significant results and results of interest from these tests are s een in the first three tables. The data used to generate these first three tests included all 138 cells of data, so only one factor was evaluated (presence of model or presence of conspecific school). To better evaluate the behavioral effects of model pres ence and conspecific schooling o n P. azaleus , contingency tables were constructed using data that included either solitary blennies with wrasse or multiple blennies without wrasse. These tests are the last three tables included in this results section, and allow for a direct comparison between opposing social environs. Table 2: Chi Square table for the correlation of "Attack" behavior and model presence. Likelihood R atio Prob> ChiSq =0.0024. Fisher's Exact Test P Value 0 . 0022 Prob (Attack=Yes) is greater for Wrasse=Yes than No.
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 6 The Chi Square test in Table 2 indicates that there is a significant correlation between the likelihood of there bei n g T. lucasanum and the likelihood of P. azaleus performing the "attack" behavior. This m eans th a t P. azaleus attack more when associating with their model. Table 3: Chi Square table for the correlation of "Dart" behavior and model presence. Likelihood R atio Prob>ChiSq=0.0708. Fisher's Exact Test P Value 0 . 0527 Prob (Dart=Yes) is greater for W rasse=Yes than No. The Chi Square test in Table 3 indicates that there is nearly significant correlation between the likelihood of there bei n g T. lucasanum and the likelihood of P. azaleus performing the "dart" behavior. This suggests that P. azaleus po ssibly dart more when associating with their model. Table 4: Chi Square table for the correlation of "Chase" behavior and the presence of other blennies. Likelihood R atio Prob>ChiSq=0.0744. Fisher's Exact Test P Value 0 . 0568 Prob (Chase=Yes) is greater for Multiple Blennies=No than Yes. The Chi Square test in Table 4 indicates that there is nearly significant correlation between the likelihood of there being other blennies and the likelihood o f P. azaleus experiencing the "Chase" behavior. This suggests th a t P. azaleus are possibly chased more when they are not associating with others of their species.
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 7 Table 5: Chi Square table for the correlation of "Attack" behavior and opposing social environs. Likelihood R atio Prob>ChiSq=0.02 23. Fisher's Exact Test P Value 0.0245 Prob(Attack=Yes) is greater for School Type=Solitary Blennies with Wrasse than Multiple Blennies without Wrasse. The Chi Square test in Table 5 indicates that there is significant correlation between the likelihood of there being a school of a solitary blenny with wrasse and the likelihood o f P. azaleus performing the "Attack" behavior. This suggests that P. azaleus attack more when they are associating with their models, more so than when they are schooling conspeci fically. Table 6: Chi Square table for the correlation of "Dart" behavior and opposing s o cial environs.Likelihood Ratio Prob> ChiSq=0.0176. Fisher's Exact Test P Value 0.0192 Prob(Dart=Yes) is greater for School Type=Solitary Blennies with Wrasse than Mult iple Blennies without Wrasse. The Chi Square test in Table 6 indicates that there is significant correlation between the likelihood of there being a school of a solitary blenny with wrasse and the likelihood o f P. azaleus performing the "Dart" behavior. T his suggests th a t P. azaleus attempt to attack more when they are associating with their models, more so than when they are schooling conspecifically.
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 8 Table 7: Chi Square table for the correlation of "Chase" behavior and opposing social environs. Likelihood R atio Prob>ChiSq=0.1313. Fisher's Exact Test P Value 0 . 1347 Prob(Chase=Yes) is greater for School Type=Solitary Blennies with Wrasse than Multiple Blennies wi thout Wrasse. The Chi Square test in Table 7 indicates that there is no significant correlation between the likelihood of there being a school of a solitary blenny with wrasse and the likelihood o f P. azaleus experiencing the "Chase" behavior. However, t he Fisher's Exact Test P Value to test the hypot h esis Prob(Chase=Yes) is greater for School Type=Solitary Blennies with Wrasse is greater than Prob(Chase=Yes) is greater for School Type= Multiple Blennies without Wrasse. This could imply that P. azaleus ex perience the "Chase" behavior less when they are in conspecific schools than when they are solitary and associating with their model. Considering that the sample size of n=15 for schools with multiple blennies not associating with their models, an increase d sample size could produce more significant results. Discussion In 1969, Edmund Hobson published a philosophical account as to the possible advantag e s P. azaleus might incur from associating with T. lucasanum , which was based on his considerable hours o f direct observations on both species. This account attests that, though the blenny only superficially resembl e s T. lucasanum , aggregation with this model is potentially deceptive enough to both potential predators and potential prey that P. azaleus enjoys increased success in attacks and decreased harassment through its mimetic association (Hobson 1969). My statistics on the correlation between the presence of T. lucasanum and "attack" behaviors are deemed significant by Table 2 and Table 5, confirming tha t P. azaleus enjoy significantly increased hunting success when associating with their model. The near significant and significant results for the "Dart" behavior reported in Table 3 and Table 6 also suggest that P. azaleus either attempts or is presented with more opportunities for predation when associating wi t h T. lucasanum than when it is associ a ting conspecificall y . These findings support some of the speculations made by Hobson, and provide more evidence for motivations behind the aggressive mimetic relationship between P. azaleus and T. lucasanum . However, Hobson's expected results of lowered persecution in P. az aleus when associating with its model are not supported by the findings in this study. Instead, my findings
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 9 suggest that lower rates of harassment are possibly correlated with conspecific schooling, not model association. There is a nearly significant P Va lue in Table 4 that could support the hypothesis th a t P. azaleus is chased less when schooling conspecifically than when it is not, which could potentially become significant with an increased sample size. The number of multi blenny conspecific schools tha t were included in this data only totalled 15, so statistical analyses in Table 7 could also become more significant with an increased sample size. This is of interest because Table 7 reported lower P Values in support of the hypothesis that blennies in ho mospecific schools are not chased as often as solitary blennies associating with models. Hobson did report observing small conspecific schools of blennies in the absence of T. lucasanum, but declared this phenomenon rare and stated that large conspecific schools would be unable to form because of relati v e P. azaleus scarcities (Hobson 1968, 1969). This latter assumption has been refuted by previous observations as well as my own. The previously mentioned team in Cabo Pulmo National Park in the Gulf of Cal ifornia found conspecific schools of well over twenty or even a hundr e d P. azaleus , which totally abandon the rule of frequency dependency in mimetic relationships (Galland et al. 2010). The team proposed that this irregular schooling behavior and disregar d for a mimetic relationship could be due to the area's unusually high prey densities resulting from fifteen years of conservation. This hypothesis is supported by the fact that the large schools -and overall greater number -o f P. azaleus encountered in m y study were found at Baja Rojo, a reef area with an impressive concentration of fish, especially compared to the two other sites surveyed. Hobson's previous observations gave accounts of a congregations o f P. azaleus numbering as few as six that had enou gh influence to deter large schools of potential prey fish (Hobson 1968). This could lead t o P. azaleus suffering less opportunities for predation but enjoying less instances of being chased when schooling conspecifically. This is supported by my findings of significant correlation between "attack" and model presence and the potential significance that could be found between lowered "chase" behavior and conspecific schooling. The former of my findings support aggressive mimicry i n P. azaleus , but my latter observations and statistics complicate the exact definition of this mimicry. The observations Hobson made in 1969 described a relationship that could be considered evidence of both aggressive and social mimicry, so subse q uent publishings have attributed th ese two types t o P. azaleus (Hobson 1969, Patzner 2009). The intricacies in the results I have reported in this study also support the complexity described by these sources, and the element of social mimicry between P. azaleus and T. lucasanum (or even oth er fish species) may justify the abandonment of aggressive mimicry exhibited by the large conspecific schools. Though the previously ment i oned bluestriped fangblenny Plagiotremus rhinorhynchos is described mainly as an aggressive mimic of the bluestreak c leaner wrasse Labroides dimidiatus , it too abandons its mimetic role. A 2005 experiment that studied the correlations between different color variations in P. rhinorhynchos and associations with the model L. dimidiatus reported that the blenny possessed th e ability to change its coloration based on the presence or absence of the its model (Cot e and Cheyney, 2005). This ability to "choose" when to mimic and when to not mimic is perhaps not so easily detectable in P. azaleus , as the species has yet to be desc ribed as having different color variations. However, the behavioral observations made by both myself and previous studies describing adherence to and abandonment of mimicry could be considered evidence supporting the existence of facultative mimicry i n P. azaleus . If P. azaleus does indeed possess facultative mimicry, it could explain the alternation between exhibiting aggressive mimicry and associating with T. lucasanum and aggregating in conspecific schools.
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 10 Thou g h P. azaleus are reported to be found in and to hunt primarily from old mollusk tubes (Hobson 1968), I only encountered P. azaleus using such holes in one location, Bahia Tomas. Within those observations, I only recorded one act of predation made by a P. azaleus in a hole. This low rate of tube behavior could be affected by many of the factors described so far, such as increased fish density supporting larger populations o f P. azaleus, which could increase the blenny's free swimming behavior (Galland et al. 2009). Also , i f P. azaleus possesses the same facultative mimicry ability as P. rhinorhynchos , it is possible that P. azaleus also has the capability to employ different hunting strategies based on to different environmental opportunities. I observed that P. azaleus appear to enjoy higher rates of predation when situated near damselfish nest sites: when the damselfish chased away another fish, the blenny took advantage of the moment of chaos and attacked the distracted fish. Many of the successful attacks I observed v ictimized Giant and Acapulco damselfish defending their nests. The two large conspecific schools I encountered (over 30 individu a l P. azaleus ) were "waiting" next to a damselfish nesting ground with a large school of barberfish swimming above it. When the damselfish chased the intr u ding barberfish, they resulted in pushing the barberfish towards the school of P. azaleus . The blennies then darted out and struck in a nearly sequential pattern, as if taking turns, which increased the confusion among the barber fish school. The significant positive correlations between predatory behavior in P. azaleus and association with T. lucasanum reported in this study confirm previous assertions of aggressive mimicry between the two species. However, additional observation s showed th a t P. azaleus does break this mimetic relationship and exhibit conspecific schooling. Though I was unable to test my hypothesis that P. azaleus incurs some benefit from conspecific schooling, low P values hint at possible benefits from reduced h arassment that future research might confirm. The uncertainty that remains about the motivation and benefits in the behaviors described in this study serve as a testament to how little we know -and how much we stand to learn -about the subtle yet rich int erspecific interactions that mimicry provides. Acknowledgements: This project would not have been possible without the immense efforts of my dear colleague Frank Joyce, and the knowledge and labor of the Lara fami l y of Cuajiniquil. This paper benefited greatly from the editing and logistical prowess of Federico Chinchilla and Emilia Triana. I would also like to extend my gratitude to the peers that accompanied me on my two week study: Ca m ille Ornburn, Samantha Abarca, Richard Urbina, Dhiraj Ramireddy, Ha ley Hudson, and Michelle Temby. Works Ci t ed: CÂ™tÂŽ, Isabelle M., and Karen L. Cheney. "Choosing When To Be A Cleaner Fish Mimic". Na t ure , vol 433, no. 7023, 2005, pp. 211 212. Springer Nature , doi:10.1038/433211a. Eagle, Janelle V., and Geoffrey P. Jones. "Mimicry In Coral Reef Fishes: Ecologica l And Behavioural Responses Of A Mimic To Its Model". Journal Of Zoo l ogy , vol 264, no. 1, 2004, pp. 33 43. Wiley , doi:10.1017/s0952836904005 4 73. Galland, G. R. et al. "Irreg ular Schooling Behavior And Abandonment Of Mimicry B y The Sabertooth Blenny (Blenniidae) In Cabo Pulmo National Park, Gulf Of California,
A ggressive mimicry and conspecific s chooling in sabertooth blennies Weaver 11 Mexico". Coral R e efs , vol 30, no. 1, 2010, pp. 215 215. Springer Nature , doi:10.1007/s00338 010 0702 6. Hobson, Edmun d S. Predatory Behavior Of Some Shore Fishes In The Gulf Of California . [Bureau Of Sport Fisheries And Wildlife; For Sale By The Supt. Of Docs., U.S. Govt. Print. Off.], 1968, pp. 42 43. Hobson, Edmund S. "Possible Advantages To The B l enny Runula Azalea I n Aggregating With The Wrasse Thalassoma Lucasanum In The Tropical Eastern Pacific". Copeia , vol 1969, no. 1, 1969, p. 191. JSTOR , doi:10.2307/1441 7 14. Humann, Paul, and Ned DeLoach. Reef Fish Identification: Baja To Panama . 1st ed., New World Publicat i ons, Inc, 2004, pp. 146 147, 186 187. Molland, Even, and Geoffrey P. Jones. "Experimental Confirmation Of Aggressive Mimicry By A Coral Reef Fish". Oecologia , vol 140, no. 4, 2004, pp. 676 683. Springer Nature , doi:10.1007/s0 0442 004 163 7 9. Molland, Even et al. "Ecology And Evolution Of Mimicry In Coral Reef Fishes". Oceanography And Marine Biology , R. N Gibson et al., Taylor & Francis, 2005. Moynihan, M. "Social Mimicry; Character Convergence Versus Character Displacement". Evolu t ion , vol 22, no. 2, 1968, p. 315. JSTOR , doi:10.2307/2406 5 31. Patzner, Robert A. The Biology Of Blennies . Science Publishers, 2009, pp. 380 3 8 8. " Plagiotremus Azale us Summary Page". Fishbase , https://www.fishbase.de/summary/Plagiotremus azaleus.html. Accessed 29 Apr 2019. " Thalassoma Lucasanum Summary Page". Fishbase , https://www.fishbase.se/summary/Thalassoma lucasanum.html. Accessed 29 Apr 2019. Turner, John R. G . "The Evolutionary Dynami c s Of Batesian And Muellerian Mimicry: Similarities And Differences". Ecological Entomo l ogy , vol 12, no. 1, 1987, pp. 81 95. Wiley , doi:10.1111/j.1365 2311.1987.tb00987 . x. Wickler, W. "Mimicry And The Evolution Of Animal Communi cation". Na t ure , vol 208, no. 5010, 1965, pp. 519 521. Springer Nature , doi:10.1038/20851 9 a0. Wickler, W. Mimicry in Plants and Animals . New York: McGraw Hill 1968.