How animals visually perceive the environment is key to understanding important ecological behaviors, such as predation, foraging, and mating. This study focuses on the visual system prop- erties and visual perception of color in the largemouth bass Micropterus salmoides. This study (1) documents the number and spectral sensitivity of photoreceptors, (2) uses these parameters to model visual perception, and (3) tests the model of color perception using a behavioral assay. Bass possess single cone cells maximally sensitive at 535 nm, twin cone cells maximally sensitive at 614 nm, and rod cells maximally sensitive at 528 nm. A simple model of visual perception predicted that bass should not be able to discern between chartreuse yellow and white nor between green and blue. In contrast, bass should be able to discern red from all achromatic (i.e., gray scale) stim- uli. These predictions were partially upheld in behavioral trials. In behavioral trials, bass were ﬁrst trained to recognize a target color to receive a food reward, and then tested on their ability to differ- entiate between their target color and a color similar in brightness. Bass trained to red and green could easily discern their training color from all other colors for target colors that were similar in brightness (white and black, respectively). This study shows that bass possess dichromatic vision and do use chromatic (i.e., color) cues in making visual-based decisions. Key words: bass, cones, photoreceptors, rods, vision. Many behaviors rely on visual cues, including predation, mating, In addition, the spectral sensitivity of the photopigment found in the and foraging (Loew and Lythgoe 1978; Endler 1992; Kemp et al. cones can also be altered by changes in chromophore usage. 2015; Rosenthal 2017). Thus, understanding visual capabilities is es- Photopigment (which absorbs light) is created by combining an opsin sential for understanding visual-based behavior. However, the diffi- protein with a chromophore derived from a vitamin A molecule (either culty is that animal taxa often vary in the visual system properties retinal: A1 or 3-dehydroretinal: A2). Shifting from A1 to A2 increases underlying visual perception. The way one animal perceives a scene the wavelengths to which the photopigment is maximally sensitive is different than another species, and this is particularly so for fish. (Bridges 1972; Munz and McFarland 1973; Loew and Dartnall 1976). Even among relatively shallow, diurnal species, fish vary in the num- Fish also vary in the arrangements of photoreceptors within their ret- ber of photoreceptors that they use ranging from as few as 2 to as inas (Ali and Anctil 1976). In sum, fish vision is highly variable. many as 5 or more (Partridge and Cummings 1999; Fuller et al. Variation in the perception of visual cues is also complicated by 2003; Land and Nilsson 2012; Cronin et al. 2014). These photo- the fact lighting environments vary dramatically in aquatic habitats. receptors vary in the wavelengths of light to which they are most sensi- Lighting environments can vary due to the effects of water depth, tive with some species being sensitive well into the UV range and algae, turbidity, dissolved organic matter, and time of day (Lythgoe others lacking sensitivity in the UV and violet range (Losey et al. 1999). 1968; Sondergaard and Thomas 2004; Johnsen and Mobley 2012; V C The Author(s) (2018). Published by Oxford University Press. 1 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact email@example.com Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 2 Current Zoology, 2018, Vol. 0, No. 0 Cronin et al. 2014). This alters (1) the visual backgrounds against using a food reward when bass approached target colors and mild which objects are viewed, (2) the irradiance spectrum that illumin- electric shocks when bass approached the non-target colors. He then ates objects and determines the inherent radiance reflected from an determined whether the bass could correctly discern between differ- object, and (3) the transmission of the reflected radiance between an ent colors. He found that bass could readily discern both red and object and the viewer (Endler 1990). Hence, variation in visual sys- green from all other colors, but often had problems discerning yel- tem properties and variation in lighting environments make it diffi- low from white and blue from black colors. This study lacked mod- cult to predict how different fish species perceive visual cues in their ern statistics/replication and was unable to use spectrophotometers habitats. to parameterize visual models with measures of reflectance and light Visual detection models have been developed to estimate visual environment. Regardless, Brown (1937) clearly indicated that bass perception in non-human animals (reviewed in: Kelber and Osorio can be trained to visual stimuli and that such assays can inform on 2010). Specifically, visual detection models can provide species’-spe- bass visual capabilities. cific predictions about the ability to detect and discriminate between This study had 2 goals. The first was to characterize the bass vis- different colors in different lighting environments. Predicting visual ual system and determine whether it differed among populations/ perception in non-human animals requires—at minimum—know- subspecies. Specifically, we sought to (1) characterize the number of ledge of the number and spectral sensitivities of the various photo- photoreceptors in the bass visual system and their spectral sensitiv- receptor classes, the lighting environment, and the reflectance ities and (2) determine whether the photoreceptor sensitivities varied spectra of objects in a visual scene. The model predictions can then between 2 subspecies of bass: Micropterus salmoides salmoides be tested by directly measuring the visual abilities of other species (from IL) and Micropterus salmoides floridanus (from FL). To do using behavioral assays. Animals are first trained to perform a par- this, we collected bass from Florida and Illinois and performed ticular task related to a color (e.g., pick a colored lever and strike microspectrophotometry (MSP) where we measured the spectral a colored pipette). Subsequent tests are then used to determine sensitivities of cones and rods for many individuals from each collec- the conditions under which animals can and cannot do the task tion site. The second goal was to determine which colors bass could (Vorobyev and Osorio 1998; Gerber et al. 2004; Hori et al. 2006; discriminate and whether this matched the predictions from a simple Champ et al. 2016). These types of behavioral assays are inform- model that was parameterized using our estimates of bass photo- ative because they allow researchers to ask questions such as the fol- receptor spectral sensitivities. Visual detection models provide predic- lowing: What are the visual capabilities of an organism? Does the tions of opponency and brightness for the bass visual system. We used ability to discern among visual stimuli match predictions from math- our model to identify target colors that look different to humans, but ematical models of visual detection? Does the organism truly use should appear similar to bass. We also used our model to identify col- color (i.e., chromatic signals due to differential stimulation of ors that have similar values for opponency, but differ in brightness, to cones)? In other words, can an animal discern a visual stimulus, test whether bass use opponency as a visual cue. To test these predic- such as red, from an alternate achromatic (i.e., gray scale) stimulus tions, we trained bass in the lab to approach and strike a specific tar- with identical brightness? get color and then asked whether they could discern their target color In this study, we modeled and behaviorally tested color vision in from other colors. We describe these studies below. Micropterus salmoides (largemouth bass), an ecologically and eco- nomically important fish species. Micropterus salmoides is a visually Materials and Methods oriented top predator in many freshwater systems and is one of the top sport fishes in the United States. (Schramm et al. 1991; Chen Microspectrophotometry et al. 2003; Cooke and Philipp 2009). Despite their importance, lit- We obtained adult bass from 2 populations, one from Florida and tle is known about the visual abilities of largemouth bass. Previous the other from Illinois. Bass from the Florida population (n ¼ 4) be- studies suggested that M. salmoides has dichromatic vision with longed to the subspecies M. s. floridanus and were collected by seine cone cells and that its color vision is highly sensitive to red net from the Everglades at 26-Mile Bend, Broward County, FL, in (Kawamura and Kishimoto 2002). These studies were based on elec- March 2013. Bass from the Illinois population (n ¼ 5) belonged to troretinogram readings, but did not directly measure the spectral the subspecies M. s. salmoides and were collected by electroshock sensitivity of the actual photoreceptors and did not verify these with from Lake Shelbyville, Moultrie County, IL, in June 2013. The behavioral assessment. Numerous studies have also examined the in- fishes were housed in stock tanks in a temperature-controlled green- fluences of coloration and water quality on bass prey/lure selection house with natural light and natural light: dark cycles at the (Carter et al. 2010; Huenemann et al. 2012; Moraga et al. 2015; University of Illinois. They were fed daily ad libitum with live feeder Shoup and Lane 2015). Such studies provide valuable insights on fish and bass pellets. preference, but are more limited in their ability to predict bass visual For MSP, the fish were dark adapted for 24 h, euthanized in 1% capabilities. buffered tricaine methanesulfonate (MS-222) solution, and decapi- Defining a baseline for color detection in largemouth bass is es- tated. The sexes are not dimorphic in coloration. We did not dissect sential due to the variability of light habitat these fish occupy. the fish to determine whether the individuals were male or female. Aquatic ecosystems are highly variable based on time of day, depth, The heads were packed in ice and immediately transported to and shade (Johnsen and Mobley 2012). Concurrently, bass are Cornell University, Ithaca, NY, in July 2013. All MSP procedures found in varying levels of water clarity (McMahan and Holanov were carried out under infrared light and follow methods previously 1995; Huenemann et al. 2012). Bass from these varying habitat detailed in Provencio et al. (1992), Loew (1994), and Loew et al. types may subsequently vary in their visual sensitivities; however, (2002). Briefly, enucleated eyes were hemisected and pieces of retina this remains untested prior to this study. were immersed in a simple Sorensen’s phosphate buffer (pH 7.2) Notably, early study of bass vision was conducted by Brown with 6% sucrose added. The retinas were carefully teased from the (1937). Brown (1937) trained bass to approach pipettes painted retinal pigment epithelium and macerated using razor blade frag- with particular colors (red, yellow, green, white, black, gray, etc.) ments and tungsten needles. A drop of the dispersed retina was Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Mitchem et al. Bass possess dichromatic vision 3 sandwiched between 2 cover slips and transferred to the stage of Micropterus. Given the good match between our model predictions the microspectrophotometer. Absorbance was recorded from the and the behavioral assays (particularly for yellow vs. white), we outer segments of single photoreceptor cells in 1 nm intervals from assume that the bass used in the behavioral assays had the visual 350 to 750 nm. system of adults. Template fitting was used to determine k (the wavelength at max maximum absorbance for a template-derived visual pigment best fit- Modeling color perception ting the experimental data). Whether the absorbance curves best fit A model of color perception was created that predicted the oppo- an A1 or an A2 template was determined via an Excel program nency and relative brightness of different target colors using the created by Juliet Parry, which solved for the value of k that min- max bass photoreceptor class sensitivities that were previously measured imizes deviations from visual pigment templates described by (see “Results” section). The model required estimates of the spec- Govardovskii et al. (2000). The normalized absorbance values of tral sensitivity of the viewer (AðÞ k Þ, the side-welling irradiance each photoreceptor cell was fitted to both A1 and A2 templates, and ðE ðÞ k Þ, and the reflectance of the object ðRðÞ k Þ. Side-welling irradi- the template (A1 or A2) with the least deviation from expected val- ance was measured with an Ocean Optics S2000 spectrophotom- ues (as measured by v ) was taken as the best fit for the given cell eter with a UV-vis 400 micron diameter fiber patch cord and a (example cells with fitted templates are shown in Results - terminal cosine corrector. The spectrophotometer was calibrated Microspectrophotometry). Analyses of variance (ANOVA) were using an Ocean Optics calibrated DT-3000 light source the same used to compare k values for each photoreceptor cell type be- max day as which measurements were taken. Spectrophotometer was tween the Florida and Illinois bass, with individuals nested within calibrated for the UV and short-wavelength range using the deuter- populations. All statistical tests were conducted in R version 3.0.3. ium lamp, and then again for the middle to longer wavelengths using the tungsten lamp. The 2 calibration curves were subse- quently combined. Measurements of E ðÞ k were taken in the stock Visual modeling tanks with clear water at 3:00 PM in Fall 2015 when the bass were Our goal here was to test a simple model of bass visual discrimin- being trained. The reflectance RðÞ k of numerous colored targets ation. To do this, we created a model that allowed us to predict (swatches of acrylic paint) was measured with a spectrophotometer which colors should appear similar to bass. We use the term connected to a reflectance probe (R200-7 probe, Ocean Optics “colors” loosely here to refer to different visual stimuli. We then Inc.) and a pulsed xenon lap (PX-200 Ocean Optics). Target reflect- trained bass to approach particular colors by feeding them through V R ance was measured from 350 to 700 nm. A Labsphere diffuse colored pipettes to ask whether bass could correctly identify the white spectral standard was used to calibrate the spectrophotom- color to which they had been trained versus an alternate color. Our eter. Some of the measurements had greater than 100% reflectance model predicted that some colors that humans can easily distinguish because they were brighter than our standard (see Figure 1). For re- should look similar to bass. We specifically chose target colors that flectance, the measurements and calibrations were done with the re- bass should easily be able to discern and target colors that our model flectance probe held at a 45 angle to the object. For both the predicted should look similar to bass. irradiance and reflectance measurements, the spectrophotometer was connected to a laptop and run using SpectraSuite Software (Ocean Optics). Husbandry For measures of relative brightness, we assumed that the red One hundred juvenile largemouth bass were obtained from a local cones were responsible for brightness (see Results - hatchery and kept in a naturally lit greenhouse, maintained at Microspectrophotometry). Previous studies by Neumeyer et al. 19 C, located at the Natural Resource Studies Annex—University (1991) have shown that goldfish rely on red photoreceptors for of Illinois—in September 2015. Hence, the bass were exposed to brightness perception under conditions of high illumination, but rely natural sunlight and experienced natural light: dark ratios. Bass on multiple photoreceptors for brightness perception under low illu- were separated into 12, 568-L cattle tanks, and fed cichlid pellet mination. In the results, we present the model predictions for the red food daily. Each tank was fastened with a UV-sterilizer and 4 cones. In the Supplementary Materials, we present the model predic- sponge filters mediated by air pumps to ensure clear and healthy tions for a similar analysis assuming that both cone cell types con- water. Two tanks were randomly selected to receive each training tribute to brightness perception (see Cummings (2004) for a similar color treatment. The fish grew rapidly between September and approach). The model predictions are qualitatively similar. November and were approximately 15 cm (6 inches) when we began Both opponency and relative brightness required estimates training. By the end of our assays, the bass were subadults and photon-catch of the photoreceptors. Photon-catch is also affected by ranged from 20 to 30 cm (8–12 inches) in standard length. many properties of the eye (e.g., diameter of the pupil), but these We note that the bass used in the MSP analysis were adults parameters affect both the numerator and denominator for calcula- whereas the bass used in the behavioral assays were juveniles matur- tions of both relative brightness and opponency. Hence, they can- ing into subadults. We assume that the findings for the MSP study celled out of the equations. The photon-catch (P) for a given are applicable for the behavioral assay study. However, we note that photoreceptor class (i) and given visual target (t) was calculated sunfish in the genus Lepomis is thought to possess UV photorecep- as follows: tors as juveniles but lack UV photoreceptors as adults (Dearry and Barlow 1987; Hawryshyn et al. 1988; Losey et al. 1999; Leech and k¼700 P ¼ AðÞ k R ðÞ k E ðÞ k (1) Johnsen 2006). The hypothesis is that these fish use UV vision to i;t i T h k¼350 view zooplankton such as Daphnia during the juvenile stage but then lose this sensitivity as they switch to other foods. The age at where A ðkÞ is the diffuse spectral sensitivity of receptor i; k is wave- which this happens in Lepomis is unknown. Whether such a scen- length; E ðÞ k is side-welling irradiance; and RðkÞ is the reflectance ario occurs in Micropterus, which is a close relative of Lepomis,is of the target. Integration was over the visible light spectrum ranging unknown. The visual systems of adult Lepomis are similar to that of from 350 to 700 nm. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 4 Current Zoology, 2018, Vol. 0, No. 0 For a given visual target (t), relative brightness was calculated as the sum of the photon-catch of the red photoreceptor divided by the brightness that would be created by a gray standard that reflects 25% of light from the white standard (B ) (see Baldwin and Johnsen (2012) for similar calculations). Hence, elative brightness (B) was calculated as the following: red;target B ¼ (2) red;gray25 Opponency was calculated as the difference in photon-catch be- tween the 2 photoreceptor types relative to relative brightness for a given visual stimulus (R) as the following: P P red;target green;target R ¼ (3) Opponency values ranged from negative to positive. Negative opponency indicates stimulation of mostly the green photoreceptor, Figure 1. Reﬂectance spectra of colored cards used for training and behav- and positive opponency indicates stimulation of mostly the red ioral assays. photoreceptor. Conversely, zero opponency occurs when there is equal stimulation of both photoreceptor types. Therefore, these color treatment (6 training targets 2 stock tanks ¼ 12 stock tanks target colors lack a chromatic signal and should be “colorless” to total). Preliminary training began by placing the training stimulus the viewer. pipette in the tank and dispensing the food pellets. After 1 week of association, bass were required to approach strike at the training Selecting color targets pipette to receive food. All bass in the tank were required to strike at To test our model of bass vision, we chose target colors that fit 1 of the target before food was dispensed; however, food was always dis- 3 criteria: (1) high negative opponency (i.e., high stimulation of the pensed 30 s after the first bass struck at ensure learning. If no bass green photoreceptor and low stimulation of the red photoreceptor), struck at the training pipette after 1 min, food was simply dispensed. (2) low opponency (i.e., stimulated both photoreceptors equally), Bass were trained once a day for 69 days (November 11, 2015– and (3) high positive opponency (i.e., high simulation of the red February 13, 2016). The collective action of the bass in each tank to- photoreceptor and low stimulation of the green photoreceptor). We wards the training pipette (no approach, approach within 1 body also chose a “white” and a “black” target as previous work by length, or strike) was recorded each day. A tank was considered Brown (1937) and our model (see Results) indicated that these bass trained when all the bass in a tank were observed striking at the may have difficulty distinguishing some colors from white and training stimulus for 7 consecutive days. Bass were continually fed black. Ultimately, 6 training targets were identified, namely, green, using this method in the period prior to the next training procedure. chartreuse yellow, red, blue, white, and black (Figure 1). Our model predicted that chartreuse yellow would be difficult to discern from Training to discern among target colors white and that blue and green might be difficult to discern from one Bass were trained to discern their training target from all other tar- another and from black. The model also indicated that red has par- gets presented simultaneously. For example, bass trained to ap- ticularly high opponency and should easily be discerned from any proach red had to discern red from blue, black, green, white, and achromatic cues. The reflectance of the chartreuse yellow stimulus chartreuse yellow. Bass were trained in their stock tanks. To accom- was greater than 100% because it reflected more light than the dif- plish this task, an array of all 6 training colors was created by at- fuse standard we used for our calibrations. This happened due to the taching all stimuli pipettes to a 90 30 cm foam board. The foam fluorescent properties of chartreuse yellow, where UV photons are board floated on top of the water in each tank. This allowed bass to absorbed and then emitted at a longer wavelength (Johnsen and have full visibility of the color stimuli in the water with minimal Mobley 2012; Mitchem and Fuller, unpublished data). However, interference from the researcher. The arrangement of training colors our white training target also had a reflectance slightly greater than on the foam board was rearranged every day. Training involved 100%. Regardless, our model indicated that these 2 visual stimuli placing the array of stimuli in 1 tank, then dispensing food from the should appear similar to the bass (Figure 2). pipette with the specific target color. Initially, food was simply dis- We created color cards by applying acrylic paint to 10 cm 10 cm pensed from the pipette to acclimate bass to the training conditions. stock paper, which were then laminated. We measured the reflectance of each card after lamination to ensure that reflectance spectra were Experiment #1: discerning training targets from alternates in the still within the same range of previously measured colored swatches. presence of a chemical cue We attached colored cards to large pipettes, which could be filled After 2 weeks, the bass were required to strike their designated train- with pellet food to dispense for bass. Pipettes and colored cards were ing stimulus to receive the food reward. The number of approaches then fastened with adhesive Velcro. within 1 body length and strikes to pipettes was recorded for each pipette. We visually observed only the first behavior of each individ- Training to a single target ual bass. Bass were highly responsive to the introduced targets at this First, bass were trained to strike a single colored, target pipette. point in the training process, so this process only lasted 30 s per Here, the bass could presumably smell the food. Two stock tanks tank. We calculated the sum of approaches and strikes at each train- containing 6–7 bass were randomly selected to receive each training ing color on each day. We then calculated the proportion of Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Mitchem et al. Bass possess dichromatic vision 5 approaches/strikes at each color for each day. For these trials, bass environment before beginning trials. Bass were not fed on the day likely had access to chemical cues as the target pipette contained the of acclimation. pellet food. Tanks were considered trained when all bass struck at the An array of test stimuli was created by attaching 8 pipettes to a pipette tip for 3 consecutive days. Training continued for a total of 183 30 cm foam board. Test stimuli included 7 achromatic shades 70 days (April 6, 2016–June 9, 2016). varying in brightness and the training color. Within the array of achromatic shades, gray 1 and gray 2 were similar in brightness to red. Gray 5 and black were similar in brightness to green. Hence, if Experiment #2: discern the training target from alternate targets in bass could not distinguish chromatic cues, red would be mistaken the absence of a chemical cue for gray 2 and green would be mistaken for gray 5 or black The goal here was to determine whether the bass could discern their (Figure 2). The arrangement of all 8 test stimuli was randomized for training target from the alternate targets in the absence of chemical each trial. Before trials began, the bass were isolated to 1 side of the cues. To do this, the pipettes with their color cards were placed in tank using a sheet of transparent plexiglass. A GoPro Heroþ was the stock tanks, and the number of approaches and strikes to each then placed inside the tank facing the array of stimuli to record the pipette was recorded. Bass were remained highly responsive to the bass behaviors. Next, the array of test stimuli was placed on top of introduced stimuli despite the lack of chemical cues, so the trials the water on the opposite side of the tank. Bass were then given for each tank only lasted 30 s. Again, we calculated the sum of 2 min to visualize the stimuli, then the plexiglass was removed, and approaches and strikes at each training color on each day. We then the bass were given 2 min to approach and strike at the stimuli. Bass calculated the proportion of approaches/strikes at each color for were tested on their ability to identify their training stimulus against each day. The bass were tested on 4 consecutive days. These trials the 7 achromatic, test stimuli. Response to a stimulus was defined as tested the ability of bass to identify their training target from the al- the number of seconds remaining within 1 body length of a stimulus. ternate target color in the absence of chemical cues from food. This GoPro Heroþ footage was reviewed twice to obtain an accurate assay was conducted across 4 days between May 30, 2016 and measurement of stimuli identification. June 7, 2017. We note that in both Experiments #1 and #2, group dynamics appeared to be important in these assays: 1 or 2 bold fish appeared Statistical analysis of bass behavior to do the choosing. These leader fishes struck the pipettes and then We used ANOVA to determine whether bass differed in how long it the other fish appeared to follow them. These competitive dynamics took to learn their training target color among a field of the other were helpful in the initial training because the fishes were motivated target colors with the presence of olfactory cues. The number of to reach the food source first. However, whether all of the bass were days taken to be considered trained was the dependent variable and trained to prefer a color is unclear. An alternative is that most bass the training target (i.e., our treatments) was the fixed, categorical in- were trained to follow a couple of leader fish who were genuinely dependent variable. To determine whether bass trained to approach trained to prefer a particular color. Regardless of which scenario occurred, the statistical inferences from the experiments are valid be- cause the analyses were performed at the level of tank means. Hence, we were either analyzing the behavior of all the color- trained bass or the behavior of the color-trained leaders and subse- quent followers. Experiment #3: discerning color cues from achromatic stimuli We next sought to determine whether trained bass could discern their training target from a range of achromatic cues. The hypoth- esis that bass use chromatic cues means that they compare the vis- ual inputs from the 2 cone classes. If bass fail to use chromatic cues, then they should be incapable of distinguishing their target color from an achromatic cue with a similar brightness value. Only bass trained to red and green were tested in these trials as they were the only groups that could successfully identify their training target from the alternate targets in the absence of chem- ical cues (see Results - Visual modeling). If bass do not use chro- matic cues, then bass trained to red (or green) should be unable to distinguish red (or green) from the achromatic cue equal in bright- ness. The achromatic cues are described in the following 2 paragraphs. Bass were randomly selected from each tank, and individually relocated to a 1325-L, 183 cm diameter, round tank for trials. Four Figure 2. Opponency compared with relative brightness in M. salmoides vis- bass from each tank were randomly selected for each trial. ual detection model for training colors (chartreuse yellow, white, red, blue, Meaning, 8 bass from each training color (2 tanks per color, 4 bass green, and black) and achromatic stimuli used in assay 3 (white, gray 1, gray from each tank) went through the trials. Two testing tanks were set 2, gray 3, gray 4, gray 5, and black). If bass use only achromatic cues, then up at the greenhouse at the Natural Resource Studies Annex— bass should be unable to distinguish red from gray 1 or gray 2, blue and University of Illinois—under identical conditions to the training green from gray 4/black, and yellow from white, as these stimuli have similar tanks. Bass were given 1 day to acclimate to their testing brightness. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 6 Current Zoology, 2018, Vol. 0, No. 0 different colors differed in their likelihood of approaching/striking (MWS)photopigmentwith k at 535.06 0.6 nm, and twin cones max the different colors, we performed ANOVA on the tank means on contained a long-wavelength sensitive (LWS) photopigment with k max the proportion of approaches/strikes at each color as a function of at 614.56 0.5 nm. There was no evidence for short-wavelength sensi- training color. For each tank, we calculated the average proportion tive (SWS) cone cells. Thus, the visual system of largemouth bass is di- of bass that approached/struck at each color (i.e., white, black, blue, chromatic with green-sensitive single cones and red-sensitive twin cones green, chartreuse yellow, and red) across the days when bass were (Figure 3). The Florida and Illinois populations did not differ in k for max trained. This created a data set with 12 observations (2 replicate any of the 3 photoreceptor types (Table 1, P> 0.17 in all tests). tanks per training color). We compared means among different treatments using Tukey’s post hoc tests. The analysis was repeated Visual modeling for the assays where visual discrimination was tested in the absence The bass were easily trainable. In the preliminary training to a single of chemical cues. training pipette, training colors did not differ in their learning time Finally, we used pairwise t-tests to determine whether bass that (F ¼ 1.16, P ¼ 0.43). On average, bass required 476 3.2 days to 5,6 had been trained to a specific target color could correctly identify become trained. The model of visual perception predicted that bass their color in the absence of olfactory cues in a new testing environ- should easily be able to discern red, green, and chartreuse yellow ment (target color presented with alternative achromatic stimuli). from one another. However, the model indicated that blue and green Here, the data for the red-trained and green-trained bass were ana- may be similar to one another and to black and that chartreuse yel- lyzed independently. For each set of trained bass, pairwise t-tests low may appear similar to white (Figure 2). These predictions were compared the proportion of time spent near training targets to each largely upheld in the bass behavioral trials. gray-scale target. Any trial where bass did not approach test stimuli were considered insufficient and removed from statistical analysis. Experiment #1: discerning training targets from alternates in the Statistical analysis was conducted on 39 behavioral observations re- presence of a chemical cue corded during testing trials. A total of 9 behavioral observations We tested the ability of bass to choose their training target pipette were not included in statistical analysis because of insufficient data. over alternative targets. With chemical cues present, bass correctly Statistical tests are considered significant at P < 0.05. P-values identified their training pipette resulting in significant differences in between 0.10 and 0.05 are considered as marginal trends. All data the proportion of approaches/attacks at each color as a function of have been deposited in Dryad (number to be entered on acceptance). training (Figure 4—Experiment 1, proportion red—F ¼ 353; pro- 5,6 portion green—F ¼ 463.8; proportion blue—F ¼ 27.9; propor- 5,6 5,6 tion yellow—F ¼ 384.5, proportion black—F ¼ 415; proportion 5,6 5,6 Results white—F ¼ 42.8; all F-values are significant at P < 0.0005; see 5,6 Microspectrophotometry Supplementary Figure 2 for an alternative display of the data). Bass Absorbance was measured from 246 photoreceptor cells in 9 fish trained to approach red did a particularly good job at identifying their (4 Florida bass and 5 Illinois bass), representing 41 rod cells, 76 sin- training pipettes (81.5% of approaches/strikes by bass trained to ap- gle cone cells, and 129 twin cone cells. Template fitting for photo- proach red). Likewise, bass trained to other colors rarely approached receptors was generally better with an A1 template than an A2 or struck the red pipettes. Similar results were found for green where template (rods: 29 A1, 12 A2; single cones: 57 A1, 19A2; twin bass trained to approach green correctly identified their target pipette cones: 111 A1, 18 A2). However, the difference in k as a function (72.8%) and fish trained to other colors rarely approached or struck max of using an A1 versus an A2 template was marginal (k absolute at green. max difference6 SE: rods 3.06 0.3 nm; single cones 2.36 0.1 nm; red Bass trained to approach blue, black, yellow, and white also cor- cones 1.56 0.1 nm). From here on, values of k are reported using rectly identified their training pipettes when chemical cues were pre- max the best-fit template for each photoreceptor cell. sent, but noticeable mistakes were made (Figure 4). Bass trained to Table 1 shows the average k values for each individual for the approach blue approached and struck at the blue pipette (65%) max rods, green cones, and red cones. Rods were maximally sensitive at most often. However, they also approached and struck at the black 527.96 1.00 nm, single cones contained a medium-wavelength sensitive pipette (10.8%), and they did this more often than bass trained to Table 1. Individual kmax for rods, single cones, and twin cones Rods Single cones—MWS photopigment Twin cones—LWS photopigment pop ind Mean k SE N Mean k SE N Mean k SE N max max max FL 1 532.5 1.69 5 528.5 2.53 4 NA NA 0 FL 2 532.2 2.66 4 532.7 4.33 3 613.8 2.13 10 FL 3 528.2 NA 1 535.9 1.85 5 614.4 2.18 13 FL 4 521.0 2.71 6 530.2 NA 1 614.2 1.13 3 IL 5 521.7 2.01 7 534.2 1.12 18 613.4 0.65 22 IL 6 530.9 1.69 8 535.1 1.60 9 616.0 0.75 31 IL 7 531.3 0.46 8 538.2 0.61 19 615.9 0.73 28 IL 8 525.5 3.63 2 534.4 1.48 16 612.1 1.33 21 IL 9 NA NA 0 528.7 NA 1 612.8 NA 1 ANOVA F ¼ 2.31, P ¼ 0.179 F ¼ 2.1, P ¼ 0.187 F ¼0.0, P ¼ 0.868 1,6 1,7 1,6 Notes: Sample sizes (N) and standard errors (SE) are listed for each photoreceptor cell type for each individual. “pop” refers to population. “ind” refers to individ- ual. F-tests for population differences are listed. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Mitchem et al. Bass possess dichromatic vision 7 1.2 Rod 1.0 0.8 0.6 0.4 0.2 450 550 650 1.2 Green Single Cone 1.0 0.8 0.6 0.4 0.2 350 450 650 1.2 Figure 4. The average proportion of approaches/strikes at each color as a Red Twin Cone function of training color (A–F). Means6 SE are shown. n ¼ 2 for each bar. 1.0 A–F indicate training color. Experiment 1—The data show the results of trials where olfactory cues were present. Experiment 2—The data show the results 0.8 of trials when olfactory cues were absent. The x-axis indicates the pipette color. 0.6 0.4 but similar patterns emerge with the mean number of bass within a tank that approached/struck each target (Supplementary Table 1). 0.2 Experiment #2: discern the training target from alternate targets in 450 550 650 the absence of a chemical cue The critical question was whether they could discern among target Figure 3. Examples of ﬁtted relative absorbance curves for (A) a rod, (B) a green single cone, and (C) a red twin cone. The speciﬁc rod shown and red colors in the absence of the chemical cues. Even when chemical cues twin cone cells were measured in Illinois bass. The speciﬁc green single cone were absent, bass correctly identified their training target resulting in shown was measured in Florida bass. significant differences in the proportion of approaches/attacks at each color as a function of training, with the exception of bass trained to approach red (P ¼ 0.0413) and yellow (P ¼ 0.0416), and tended to white (Figure 4—Experiment 2, proportion red—F ¼ 22.35, 5,6 do this more often than bass trained to approach green P ¼ 0.00082; proportion green—F ¼ 6.32, P ¼ 0.022; proportion 5,6 (P ¼ 0.0793). Likewise, bass that were trained to approach black blue—F ¼ 43.15, P ¼ 0.00013; proportion yellow—F ¼ 5.21, 5,6 5,6 correctly identified the black pipette (66.4%), but they also ap- P ¼ 0.034; proportion black—F ¼ 9.91, P ¼ 0.0073; proportion 5,6 proached/struck at blue at an appreciable rate (10.6%). white—F ¼ 1.309, P ¼ 0.37; see Supplementary Figure 2 for an al- 5,6 A similar pattern emerged with chartreuse yellow and white. Bass ternative display of the data). Again, bass trained to red readily identi- trained to approach chartreuse yellow correctly identified the char- fied their target color (85.4%), and bass trained to other colors rarely treuse yellow pipette (61.2%), but they also approached/struck at the approached/struck at red (P < 0.005 in all post hoc tests). Bass trained white pipette (21.8%), and they did this more often than bass trained to green also identified their target color well (72.3%), and bass to green (P ¼ 0.0456), and tended to approach/strike at the white trained to other colors rarely approached/struck at green (P < 0.05 in pipette more often than bass trained to red (P ¼ 0.0787) or black all post hoc tests). (P ¼ 0.097) (Figure 4). Finally, bass trained to approach white were Bass trained to blue, black, yellow, and white performed less more likely to approach/strike at white than other colors (57.4%), well. Bass trained to blue approached/struck at blue pipettes more but they also approached and struck at chartreuse yellow (16.3%), than the others (48.3%) and they approached/struck at blue pipettes and they did this more often than bass trained to approach black at high rates than bass trained to other colors (P < 0.02 in all post (P ¼ 0.00128), blue (P ¼ 0.00147), green, (P ¼ 0.00969), or red hoc tests). However, they also approached/struck at the black pip- (P < 0.001) (Figure 4). The data presented here are on the proportions ette at a high rate (39.2%). Bass trained to black correctly identified Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 relative absorbance relative absorbance relative absorbance 8 Current Zoology, 2018, Vol. 0, No. 0 the black pipettes at a high rate (70.8%), but they also approached/ struck at the blue pipette (29.2%). Bass trained to black ap- proached/struck the black pipette at a higher rate than bass trained to red (P ¼ 0.0098), green (P ¼ 0.019), or yellow (P ¼ 0.0085) pip- ettes and tended to strike at them at higher rates than bass trained to white pipettes (P ¼ 0.052). However, there was no statistically sig- nificant difference in the rates of approaches/strikes at black be- tween bass trained to black and bass trained to blue (P ¼ 0.23). Bass trained to chartreuse yellow approached/struck at the char- treuse yellow pipette at an appreciable rate (45.8%), but they also approached/struck at the white (25%) and blue (29.2%). Bass trained to chartreuse yellow tended to approach/strike at the chartreuse yellow pipette at a greater rate than bass trained to red (P ¼ 0.077), green (P ¼ 0.066), blue (P ¼ 0.066), or black (P ¼ 0.066). There were no significant differences in the rates of approaches/strikes at yellow pipettes between bass trained to yellow and bass trained to white (P ¼ 0.75). Likewise, bass trained to white approached/struck at the white pipette (33.3%), but they also struck at yellow (29.2%), black (22.9%), and blue (14.5%). There were no statistically significant (or trending) differences between the propor- tions of approaches/strikes at white among the bass trained to differ- ent colors. The data presented here are on the proportions but similar patterns emerge with the mean number of bass within a tank that approached/struck each target (Supplementary Table 2). Figure 5. Experiment 3—Proportion of time spent near simultaneously pre- sented stimuli during test trials (mean6 SE). Bar represents target colors pre- sented to bass where TC ¼ training color, W ¼ white, G1–G5 ¼ gray targets, Experiment #3: discerning color cues from achromatic stimuli and B ¼ black. Each grid represents bass trained to A ¼ red and B ¼ green. We next asked whether bass use chromatic cues to select trained tar- gets. The critical test here is whether bass trained to red and green could identify their target among a series of achromatic alternates. sensitivity. Our study found little evidence that the k of the differ- max In these trials, bass were less likely to perform strikes, and bass were ent photoreceptor classes or the degree of A1 versus A2 template use generally less likely to approach the pipettes. Despite this, bass differed between the 2 populations. The fact that we fit both A1 and trained to red and green were able to accurately select their training A2 templates to different photorecptors within the same individual target among alternative gray targets (Figure 5). An analysis of the most likely reflects noise in the data and not within population (or time spent associated with each target indicated that bass trained to even within individual) variation in chromophore usage. These re- red more often selected their training target compared with all gray sults imply that a single model of bass vision can be used for mul- targets except gray 1 (all targets except gray 1: pairwise t-tests tiple populations. P < 0.0065, gray 1: P ¼ 0.11). Similarly, bass trained to green spent Our visual detection model of bass vision indicated that dichro- more time near their training target compared with all gray targets matic bass vision limits the perception of yellow coloration. In par- (P < 0.05) with the exception gray 2 and black (gray 2: P ¼ 0.074, ticular, chartreuse yellow should appear similar to white. This black: P ¼ 0.070) where the differences were marginal. Interestingly, happens because chartreuse yellow equally stimulates both the green unlike our prediction, bass did not select grays that were similar in and red cone cells at similar frequencies. Hence, there is no oppo- brightness to their training colors during gray trails. Instead, bass nency resulting from chartreuse yellow. Our behavioral assays sup- trained to red and green selected targets that were brighter than their ported this hypothesis. In trials with olfactory cues, bass trained to target color (gray 1 and gray 2). chartreuse yellow and white could correctly identify their target col- ors. Yet even here, they often made mistakes and frequently chose Discussion the other. This pattern was amplified when bass were tested in the Largemouth bass possess dichromatic color vision, with green sensi- absence of olfactory cues. Here, the bass were incapable of distin- tive single cones and red sensitive twin cones. This finding agrees guishing white from chartreuse yellow and vice versa. Similar results with Kawamura and Kishimoto’s (2002) prediction for a red- were found by Brown (1937) who used light electric shocks to train sensitive eye in largemouth bass. Kawamura and Kishimoto sug- bass. Taken together, these results provide strong support for the gested that the largemouth bass eye provides better color analysis at idea that chartreuse yellow appears similar to white in the bass vis- long wavelengths over shorter wavelengths. ual system. There was little evidence to suggest substantial phenotypic vari- Our visual detection model also predicted that blue, green, and ation in visual sensitivity between the Florida and Illinois popula- black would appear similar to the bass. These results partially sup- tions. Other fish species have been shown to harbor phenotypic ported this prediction. Bass trained to blue frequently struck at variation among populations (Boughman 2002; Fuller et al. 2003, black, and bass trained to black frequently struck at blue. In trials 2004; Carleton et al. 2005; Fuller and Noa 2010), but the mechan- with olfactory cues, bass trained to black also selected blue at an ap- isms underlying this variation varies among systems. Shifts in k , preciable rate and vice versa for bass trained to blue. In the absence max A1 versus A2 retinal templates, relative cone/opsin expression, and of olfactory cues, bass were incapable of distinguishing between lens transmission can contribute to phenotypic variation in visual black and blue colors. Again, Brown (1937) found a similar pattern Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 Mitchem et al. Bass possess dichromatic vision 9 where blue and black were indistinguishable to bass. Brown (1937) sunfish retina contained rods with k at 525 nm, single cones with max also found that bass could discern green from black and grays, but k at 535 nm, and twin cones with k at 621 nm. These photo- max max that they had difficulty discerning green and blue, which is in keep- receptor sensitivities are a close match to the largemouth bass sensi- ing with our visual model. The results of the behavioral assays taken tivities, with both the rods and single cones being almost identical with the results of Brown (1937) indicate that this pattern is robust. (largemouth bass rod: 527.94 nm; single cone: 534.98 nm). Green Our model of bass vision may need to include other aspects of sunfish twin cones were slightly red-shifted compared with large- visual morphology to account for the discrepancy in blue perception mouth bass (green sunfish twin cone: 621 nm; largemouth bass twin between model predictions and behavioral results. The behavioral cone: 614.48). In addition, Dearry and Barlow also did not find evi- assays and visual detection model indicate that there is a genuine dence for blue or UV sensitive cones in adult fish. Darters also have chromatic stimulus for green that bass can detect. The larger ques- a similar dichromatic visual system with rods maximally sensitive tion is why this does not occur for blue. The model indicates that from 529 to 525 nm, single cones maximally sensitive from 508 to blue should create a similar chromatic stimulus that differs from 531 nm, and twin cones maximally sensitive from 602 to 608 nm black, but the behavioral assays do not support this. One possibility (Gumm et al. 2012; Zhou et al. 2015). Further studies are needed to is that there are filtering properties of the bass eye that we did not determine whether all centrarchids possess a similar suite of photo- consider in our model (Thorpe et al. 1993; Aksnes and Utne 1997; receptor cells. Kawamura and Kishimoto 2002). Work in other centrarchids In conclusion, this study showed that bass possess dichromatic (Lepomis cyanellus and Lepomis gibbossus) indicates the presences vision with red and green cells in addition to a rod cell. A simple vis- of pigments in the lens that filter light lower than 400 nm (Thorpe ual model of this visual system indicated that there are colors such et al. 1993). Whether such filtering pigments are present in the bass as chartreuse yellow that bass should perceive as being similar to lens and cornea is currently unknown. white. Our behavioral assays provided good support for the model In the bass system, chromatic cues, and particularly red, are eas- prediction that chartreuse yellow is indistinguishable from white. ier to identify. Bass trained to red and green had high rates of The behavioral assays also indicated that blue is indistinguishable approaches/strikes at their respective targets, and bass trained to from black. Bass could readily identify red and green and could dis- other targets rarely mistakenly approached/struck at red or green. tinguish these colors from achromatic alternatives. Bass from Illinois Bass trained to red and green were also able to identify their targets and Florida populations possess similar photoreceptor sensitivities among a panel of achromatic cues. These results indicate that bass despite differences in environmental light composition. Whether can more readily associate meaning to chromatic cues of high oppo- bass from these light environments have innately different learning nency. However, bass had difficulty associating meaning to achro- abilities or preferences for colors in currently unknown. Our find- matic cues (white, black, and for the bass system, yellow). These ings have implications for the recreational fishing industry and for results are in keeping with a long literature in the field of visual natural systems where bass are often a top predator. psychology showing that chromatic cues are easier to learn for many species (Kelber et al. 2003; Hori et al. 2006; Roth et al. 2007). Acknowledgments Red coloration has long been thought to be particularly attract- ive to largemouth bass (Howick and Obrien 1983; Kawamura and We would like to thank Joel Borowics, Sean Bruyere, Shun Kobayashi, and Drew Costenbader for assistance in animal husbandry and data collection. Dr Kishimoto 2002; Ciccotto and Mendelson 2016). Red was particu- Alison Bell, Dr John Epifanio, Michelle St. John, and Rachel Moran provided larly easy to identify (for bass trained to red) and avoid (for bass comments that greatly improved this manuscript. trained to other colors). A study by Ciccotto and Mendelson (2016) found that largemouth bass had a strong preference for red color- ation over blue or black. These were presumably “innate” prefer- Funding ences whereas the behaviors shown here were learned. Whether or This project was funded by the Animal Behavior Society Student Research not innately preferred colors have high opponency and are also eas- Grant and Illinois Natural History Survey via the R. Weldon Larimore/Jordan ily learned is unknown. Creek Endowment Fund. Most large, predatory fish are dichromatic, meaning they rely on only 2 photoreceptor classes to perceive color (Loew and Lythgoe 1978; Cronin et al. 2014). Lythgoe (1968) proposed that under- Supplementary Material water predators perceive optimally with an offset, dichromatic sys- Supplementary material can be found at https://academic.oup.com/cz. tem, where 1 photoreceptor optimally perceives the background illumination spectrum, and 1 photoreceptor contrasts the back- ground spectrum. An offset dichromatic system creates high contrast References between background lighting and prey illuminated by overhead sun Aksnes D, Utne ACW, 1997. A revised model of visual range in ﬁsh. Sarsia 82: (Loew and Lythgoe 1978). In aquatic environments, long wave- 37–41. lengths (orange–red spectrum) are reflected in background lighting, Ali MA, Anctil M, 1976. Retinas of Fishes: An Atlas. Berlin: Springer-Verlag. whereas short wavelength (blue–green spectrum) contrasts that Baldwin J, Johsnen S, 2012. The male blue crab, Callinectes sapidus, uses both background (Lythgoe 1968; Johnsen and Mobley 2012). These chromatic and achromatic cues during mate choice. J Exp Biol 215: aquatic spectral properties generally align with the M. salmoides vis- 1184–1191. ual system. Boughman JW, 2002. How sensory drive can promote speciation. Trends Ecol Similar visual system properties have been found in other cen- Evol 17:571–577. trarchids (Dearry and Barlow 1987; Hawryshyn et al. 1988). Dearry Bridges CDB, 1972. Photochemistry of Vision. New York: Springer-Verlag. and Barlow (1987) previously characterized the photoreceptor sensi- Brown FA, Jr., 1937. Responses of the large-mouth black bass to colors. tivities of the green sunfish L. cyanellus. They found that the green Illinois Nat Hist Surv Bull 21:33–55. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018 10 Current Zoology, 2018, Vol. 0, No. 0 Carleton KL, Parry JWL, Bowmaker JK, Hunt DM, Seehausen O, 2005. Kelber A, Osorio D, 2010. From spectral information to animal colour vision: Colour vision and speciation in Lake Victoria cichlids of the genus experiments and concepts. Proc Biol Sci 277:1617–1625. Pundamilia. Mol Ecol 14:4341–4353. Kelber A, Vorobyev M, Osorio D. 2003. Animal colour vision–behavioural Carter MW, Shoup DE, Dettmers JM, Wahl DH, 2010. Effects of turbidity tests and physiological concepts. Biol Rev Camb Philos Soc 78:81–118. and cover on prey selectivity of adult smallmouth bass. Trans Am Fish Soc Kemp DJ, Herberstein ME, Fleishman LJ, Endler JA, Bennett ATD et al., 139:353–361. 2015. An integrative framework for the appraisal of coloration in nature. Champ CM, Vorobyev M, Marshall NJ, 2016. Colour thresholds in a coral Am Nat 185:705–724. reef ﬁsh. R Soc Open Sci 3:160399. Land MF, Nilsson DE, 2012. Animal Eyes. Oxford: Oxford University Press. Chen RJ, Hunt KM, Ditton RB, 2003. Estimating the economic impacts of a Leech DM, Johnsen S, 2006. Ultraviolet vision and foraging in juvenile bluegill trophy largemouth bass ﬁshery: issues and applications. North Am J Fish Lepomis macrochirus. Can J Fish Aquat Sci 63:2183–2190. Manag 23:835–844. Loew ER, 1994. A third, ultraviolet-sensitive, visual pigment in the Tokay Ciccotto PJ, Mendelson TC, 2016. The ecological drivers of nuptial color evo- gecko Gekko gekko. Vision Res 34:1427–1431. lution in darters (Percidae: etheostomatinae). Evolution 70:745–756. Loew ER, Dartnall HJA, 1976. Vitamin A1/A2-based visual pigment mixtures Cooke SJ, Philipp DP, 2009. Centrarchid Fishes: Diversity, Biology and in cones of the rudd. Vision Res 16:891–896. Conservation. 1st edn. West Sussex: Wiley-Blackwell. Loew ER, Lythgoe JN, 1978. The ecology of cone pigments in teleost ﬁshes. Cronin TW, Johnsen S, Marshall NJ, Warrant EJ, 2014. Visual Ecology. Vision Res 18:715–722. Woodstock, Oxfordshire: Princeton University Press. Loew ER, McFarland WN, Margulies D, 2002. Developmental changes in the Cummings ME, 2004. Modelling divergence in luminance and chromatic visual pigments of the yellowﬁn tuna Thunnus Albacares. Mar Freshw detection performance across measured divergence in surfperch Behav Physiol 35:235–246. (Embiotocidae) habitats. Vision Res 44:1127–1145. Losey GS, Cronin TW, Goldsmith TH, Hyde D, Marshall NJ et al., 1999. The Dearry A, Barlow RB, 1987. Circadian rhythms in the green sunﬁsh retina. UV visual world of ﬁshes: a review. J Fish Biol 54:921–943. J Gen Physiol 89:745–770. Lythgoe JN, 1968. Visual pigments and visual range underwater. Vision Res 8: Endler JA, 1990. On the measurement and classiﬁcation of colour in studies of 997–1012. animal colour patterns. Biol J Linnean Soc 41:315–352. McMahan TE, Holanov SH, 1995. Foraging success of largemouth bass at dif- Endler JA, 1992. Signals, signal conditions, and the direction of evolution. Am ferent light intensities: implications for time and depth of feeding J Fish Biol Nat 139:S125–S153. 46:759–766. Fuller RC, Carleton KL, Fadool JM, Spady TC, Travis J, 2004. Population Moraga AD, Wilson ADM, Cooke SJ, 2015. Does lure colour inﬂuence catch variation in opsin expression in the blueﬁn killiﬁsh Lucania goodei:a per unit effort, ﬁsh capture size and hooking injury in angled largemouth real-time PCR study. J Comp Physiol A Neuroethol Sensory, Neural, Behav bass? Fish Res 172:1–6. Physiol 190:147–154. Munz FW, McFarland WN, 1973. The signiﬁcance of spectral position in the Fuller RC, Fleishman LJ, Leal M, Travis J, Loew E, 2003. Intraspeciﬁc variation rhodopsins of tropical marine ﬁshes. Vision Res 13:1829–1874. in retinal cone distribution in the blueﬁn killiﬁsh Lucania goodei. JComp Neumeyer C, Wietsma JJ, Spekreijse H, 1991. Separate processing of color Physiol A Neuroethol Sensory, Neural, Behav Physiol 189:609–616. and brightness in goldﬁsh. Vision Res 31:537–549. Fuller RC, Noa LA, 2010. Female mating preferences, lighting environment, Partridge JC, Cummings ME, 1999. Adaptations of visual pigments to the aquatic and a test of the sensory bias hypothesis in the blueﬁn killiﬁsh. Anim Behav environment. In: Archer SN, Djamgoz MBA, Loew ER, Vallerga S, editors. 80:23–35. Adaptive Mechanisms in the Ecology of Vision. Dordrecht: Kluwer, 251–283. Gerber B, Scherer S, Neuser K, Mechels B, Hendels E et al., 2004. Visual learn- Provencio I, Loew ER, Foster RG, 1992. Vitamin A2-based visual pigments in ing in individually assayed Drosophila larvae. J Exp Biol 207:179–188. fully terrestrial vertebrates. Vision Res 32:2201–2208. Govardovskii VI, Fyhrquist N, Reuter T, Kuzmin DG, Donner K. 2000. In Roth LSV, Balkenius A, Kelber A, 2007. Colour perception in a dichromat. search of the visual pigment template. Visual Neurosci 17:509–528. J Exp Biol 210:2795–2800. Gumm JM, Loew ER, Mendelson TC, 2012. Differences in spectral sensitivity Rosenthal GG, 2017. Mate Choice: The Evolution of Sexual Decision Making within and among species of darters (genus Etheostoma). Vision Res 55:19–23. from Microbes to Humans. Princeton (NJ): Princeton University Press. Schramm HL, Armstrong ML, Funicelli NA, Green DM, Lee DP et al., 1991. Hawryshyn CW, Arnold MG, McFarland WN, Loew ER, 1988. Aspects of The status of competitive sport ﬁshing in North America. Fisheries 16:4–12. color vision in bluegill sunﬁsh (Lepomis macrochirus): ecological and evolu- Shoup DE, Lane WD, 2015. Effects of turbidity on prey selection and foraging re- tionary relevance. J Comp Physiol A 164:107–116. turn of adult largemouth bass in reservoirs. North Am J Fish Manag 35:913–924. Hori S, Takeuchi H, Arikawa K, Kinoshita M, Ichikawa N et al., 2006. Sondergaard M, Thomas D, 2004. Dissolved Organic Matter (DOM) in Associative visual learning, color discrimination, and chromatic adaptation Aquatic Ecosystems: A Study of European Catchments and Coastal Waters. in the harnessed honeybee Apis mellifera L. J Comp Physiol A Neuroethol A publication by the EU project DOMAINE (EVK3-CT-2000-00034). Sensory, Neural, Behav Physiol 192:691–700., Thorpe A, Douglas RH, Truscott RJW, 1993. Spectral transmission and Howick GL, Obrien WJ, 1983. Piscivorous feeding behavior of largemouth short-wave absorbing pigments in the ﬁsh lens-I. Phylogenetic distribution bass: an experimental analysis. Trans Am Fish Soc 112:508–516. and identity. Vision Res 33:289–300. Huenemann TW, Dibble ED, Fleming JP, 2012. Inﬂuence of turbidity on the Vorobyev M, Osorio D, 1998. Receptor noise as a determinant of colour foraging of largemouth bass. Trans Am Fish Soc 141:107–111. thresholds. Proc Biol Sci 265:351–258. Johnsen S, Mobley C, 2012. The optics of life: a biologist’s guide to light in na- Zhou MC, Loew ER, Fuller RC, 2015. Sexually asymmetric colour-based ture. Am J Phys 80:937. species discrimination in orangethroat darters. Animal Behav 106: Kawamura G, Kishimoto T, 2002. Color vision, accommodation and visual 171–179. acuity in the largemouth bass. Fish Sci 68:1041–1046. Downloaded from https://academic.oup.com/cz/advance-article-abstract/doi/10.1093/cz/zoy019/4924236 by Ed 'DeepDyve' Gillespie user on 08 June 2018
Current Zoology – Oxford University Press
Published: Mar 7, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera