Printer Friendly

Color pattern variation in a shallow-water species of opisthobranch mollusc.


Opisthobranch sea slugs (Mollusca: Opisthobranchia) are among the most strikingly colored marine animals. The bright and contrasting external colorations of many opistho-branchs have been considered to be aposematic (Edmunds, 1987; Gosliner and Behrens, 1989; Rudman, 1991) because of the wealth of chemical defenses in these organisms (Cimino and Ghiselin, 2009) and experimental data suggesting unpalatability (Gosliner, 2001; Long and Hay, 2006; Haber et al., 2010). Aposematism is a defensive adaptation in which a conspicuous coloration is used to warn potential predators that a species is chemically or otherwise defended (Cott, 1940; Cortesi and Cheney, 2010). Aposematism is widespread and well studied in terrestrial ecosystems. Several recent studies have investigated the complex evolution of aposematic signaling both empirically (Darst et al., 2006; Lindstedta et al., 2011; Nokelainen et al., 2012; Barnett et al., 2012) and theoretically (Speed and Ruxton, 2005; Marples et al., 2005; Grant, 2007, Stevens and Ruxton, 2012). In marine environments aposematic defensive strategies are not well understood and are most often restricted to shallow-water tropical ecosystems (Ritson-Williams and Paul, 2007).

Opisthobranchs are nearly blind. Opisthobranch eyes are so simple in structure (Hughes, 1970) that it is virtually certain that they are unable to perceive an image or the color pattern of another individual (Edmunds, 1987). Thus, color or color patterns may not be useful for intraspecific communication, excluding the possibility of evolution by sexual selection. Most authors recognize that color in opistho-branchs has evolved mainly for defensive purposes (Edmunds, 1987), although Crozier (1916) argued that the bright external coloration of Hypselodoris zebra is a "metabolic accident" with no defensive role. Gosliner and Behrens (1989) reviewed the biological significance of color in opisthobranchs. They estimated that about 50% of opisthobranch species are aposematic, whereas the rest use other color-related defensive strategies such as camouflage. Several species of opisthobranchs and other marine invertebrates appear to mimic other opisthobranchs (Gosliner and Behrens, 1989; Rudman, 1991; Gosliner, 2001), providing further support for the defensive ecological role of color. Moreover, experimental data show that opisthobranch species with bright external colors avoid predation more successfully than distasteful cryptic species (Gosliner, 2001), and predators learn to avoid them (Tullrot, 1994). One caveat of most studies on the role of color in marine organisms is that human visual perception, which does not necessarily reflect how other visual predators would perceive those color patterns, is generally used to identify striking color patterns (Rosenberg, 1989; Guilford and Cuthill, 1991). However, Cortesi and Cheney (2010), using a phylogenetic comparative analysis, found a significant correlation between the conspicuousness of opisthobranch species as perceived by two fish species and their estimated toxicity.

A second caveat of these hypotheses on the biological significance of color is one of the most prevalent ideas in opisthobranch systematics--that is, individual members of the same species share similar color patterns. Although Harris (1973) demonstrated that many pigments are taken up by opisthobranchs with the food and incorporated into the notum, color pattern is recognized as an important trait in taxonomic descriptions (e.g., Rudman, 1984; Dayrat, 2010; Gosliner, 2011), and field guides heavily rely on color patterns for species identifications (Gosliner et al., 2008). Color variation has been recognized in some species, but this is often viewed as the exception rather than the rule (Edmunds, 1987). However, some recent molecular studies have shown that, at least in some species, color patterns are considerably variable within members of the same species (Pola et al., 2006; Ornelas-Gatdula et al., 2011). This finding calls into question not only the usefulness of color patterns for species identifications, but also the hypothetical ecological and evolutionary roles that color patterns might play in this group of organisms.

In this paper we study a putative species of benthic opisthobranch mollusc with several intriguing characteristics that makes it suitable for testing hypotheses on the ecological role of color. Philinopsis pusa (Marcus and Marcus, 1966) is a western Atlantic aglajid sea slug (Cephala-spidea) that is consistently found crawling on white sand during the day around Stocking Island, Bahamas, actively preying upon other opisthobranchs. Individual specimens are therefore visually exposed to potential predators (fish), although--like most other aglajid species--they burrow in the sand when disturbed. This putative species has a broad range of colors (Fig. 1), varying from mostly white (with dark spots) to brown with light spots, uniformly brown, and black (with bright neon blue lines). The lighter color forms appear to be camouflaged on sandy environments, whereas the dark forms clearly stand out on the light sandy background. The dark animals, particularly those of a black color morph, have bright neon blue lines on the edge of the parapodia and the tip of the dorsal crest, making them more conspicuous. Other animals have intermediate color patterns. Because of this broad color range, new species names have recently been introduced for some of the color forms (Ortea et al., 2007). In recent years some papers have used reproductive behavior (Anthes and Michiels, 2007; Turner and Wilson, 2012) and molecular data (Anthes et al. 2008) to investigate species boundaries in aglajid species complexes. However, none of these studies looked specifically at genetic differentiation between color forms using a large sample size.

The goals of this study are to determine the molecular population structure of all the color forms of this putative species (based on mitochondrial markers) and the relationship between color pattern and the size and nonreproductive behavior of individual organisms. Because of the difficulty of conducting crossbreeding experiments with opistho-branchs, the potential genetic basis for color was not fully investigated. Although this paper is based on a single species, it attempts to contribute new data to the broader issue of the evolutionary and ecological significance of color pattern in marine organisms.

Materials and Methods

Field observations

Specimens were observed in the field over a span of 6 years by scuba diving and snorkeling in several localities around Stocking Island, Exumas, Bahamas (Table 1). Specimens were photographed in situ, and the time of day and date of observation for all specimens was recorded. Additionally, the defensive escape behavior of the animals was observed in the field and timed from the initial disturbance of the animal to the time at which the animal was completely burrowed in the sand. Because some of the individuals were partially buried after being photographed, the complete burrowing time could not be determined.


Table 1

Date and time of observation, habitat, color with
phenotypic class number), burrowing time, and locality of
specimens found in the Bahamas

Time (24h) and  Depth    Body  Color (PC)  Burrowing  Locality
dale              (m)  length               time (s)

14:50--Dec 19,    1.5      10  Brown              --  HH2,
2003                           spotted                Stocking
                               (3)                    Is.

12:01--Dec 24,      2       7  Brown             115  HH2,
2003                           spotted                Stocking
                               (3)                    Is.

11:43--Dec 31,    1.5       9  Light (2)          --  HH2,
2003                                                  Stocking

15:17--Feb 23,      2      17  Brown (4)         116  HH2,
2004                                                  Stocking

14:53-Dec 17,     2.5      12  Brown (4)          --  HH2,
2004                                                  Stocking

12:30--Jan 9,       2       9  Brow               --  SDB,
2005                           spotted                Stocking
                               (3)                    Is.

14:05--Jan 14,      2      15  Brown (4)          --  HH2,
2005                                                  Stocking

15:07--Dec 20,      2       5  Black (5)          69  GB,
2005                                                  Stocking

14:55--Jan 11,      2       5  Black (5)          --  HH2,
2006                                                  Stocking

15:06--Jan 28,      2      18  Light (2)         125  HH1,
2007                                                  Stocking

15:27--Jan 28,      2      18  Black (5)          --  GB,
2007                                                  Stocking

12:35--Feb 1,     1.5       5  White              --  HH2,
2007                           spotted                Stocking
                               (1)                    Is.

09:13--Feb 1,     1.5      14  Light (2)         126  HH2,
2007                                                  Stocking

11:27--Feb 10,      2      18  Black (5)          --  SDB,
2007                                                  Stocking

11:58--Fob 10,      2      18  Brown (4)          NB  SDB,
2007                                                  Stocking

12:14--Feb 10,      2      17  Brown              59  SDB,
2007                           spotted                Stocking
                               (3)                    Is.

12:23--Feb 10,      2      17  Brown (4)          81  SDB,
2007                                                  Stocking

12:30-Feb 10,       2      15  Brown              --  SDB,
2007                           spotted                Stocking
                               (3)                    Is.

12:50--Feb 10,      2      13  While              --  SDB,
2007                           spotted                Stocking
                               (1)                    Is.

08:02--Apr 18,      2      15  Brown              91  SDB,
2007                           spotted                Stocking
                               (3)                    Is.

08:24--Apr 18,      2      14  Black (5)          --  SDB,
2007                                                  Stocking

08:36--Apr 18,      2      15  Brown (4)          85  SDB,
2007                                                  Stocking

08:55--Apr 18,      2      14  Light (2)          78  SDB,
20O7                                                  Stocking

09:20--Apr 18,      2      12  White              --  SDB,
2007                           spotted                Stocking
                               (1)                    Is.

16:33--Dec 21,      2      21  Brown (4)          61  HH2,
2007                                                  Stocking

15:23--Jan 14,      2      15  Light (2)          --  HH2,
2008                                                  Stocking

14:17--Jan 14,      2      16  Brown (4)          --  HH2,
2008                                                  Stocking

09:30--Jan 6,       2      15  Brown (4)          NB  HH2,
2009                                                  Stocking

09:54--Jan 6,       2      12  Brown              --  HH2,
2009                           spoiled                Stocking
                               (3)                    Is.

10:45--Jan 6,       2      16  Brown (4)          NB  HH2,
2009                                                  Stocking

11:00--Jan 18,      2      10  White              --  HH2,
2009                           spotted                Stocking
                               (1)                    Is.

13:30--Feb 12,      2      10  Black (5)          63  HH2,
2009                                                  Stocking

14:52-,Dec 21,      2      15  Black (5)          --  HH2,
2009                                                  Stocking

15:13--Dec 21,    1.6       6  Brown (4)          --  HH2,
2009                                                  Stocking

09:56--Dec 28,    1.5      14  Brown (4)          --  HH2,
2009                                                  Stocking

14:38--Feb 8,       2      13  Brown (5)          --  HH1,
2010                                                  Stocking

14:54--Feb 8,       2      11  Brown              --  HH1,
2010                           spotted                Stocking
                               (3)                    Is

15:11--Feb 8,       2      10  Light (2)          --  HH1,
2010                                                  Stocking

15:28--Feb 8,       2      10  Light (2)          --  HH1,
2010                                                  Stocking

15:40--Feb 8,       2       5  Brown              --  HH1,
2010                           spotted                Stocking
                               (3)                    Is

15:40--Feb 8,       2       7  Brown              --  HH1,
2010                           spotted                Stocking
                               (3)                    Is

15:58--Feb S,       2      10  Brown (4)          --  HH1,
2010                                                  Stocking

PC = phenotypic class number; NB = observed for more than
60 seconds and did not attempt to burrow; SDB = Sand Dollar
Beach, GB = Gaviota Bay, HH1 = Hurricane Hole 1. HH2 =
Hurricane Hole 2.

Specimens examined in the laboratory

Forty-two specimens covering the entire color range of the putative species Philinopsis pusa were sequenced in this study (Table 2). Specimens were collected and narcotized using a 1 mol [1.sup.-1] solution of Mg[C1.sub.2], and preserved in 70% EtOH. Tissue samples were transferred to 99% EtOH. All the specimens examined have been deposited at the Natural History Museum of Los Angeles County (LACM) and in the research collection of the Biological Sciences Department of the California State Polytechnic University (CPIC). For outgroup comparison, sequences of the closely related species Philinopsis cyanea and Philinopsis depicta were obtained from GenBank or sequenced.

Table 2 Specimens sequenced, including locality information,
collection voucher numbers, and GenBank accession numbers for
Philinopsis species


Species    Locality           Dale  Voucher        16S       CO1

P. depict  Spain                              AM421831  AM421892
a          (Mediterranean)

P.         Isla Coiba,         May     LACM         --  KC603849
cyanea     Panama              20.   153372

P. pusa    HH2, Stocking       Dec     CPIC   KC611173  KC611215
           Is.                 16.    00734

P. pusa    HH2, Stocking       Dec     CPIC   KC611174  KC611216
           Is.                 24.    00735

P. pusa    HH2, Stocking       Feb     CPIC   KC611175  KC611217
           Is.                 23.    00736

P. puna    HH2, Stocking       Jan     LACM   KC611176  KC611218
           Is.                 14.   172291

P. pusa    GB, Stocking        Dec     LACM   KC611177  KC611219
           Is.                 20.   173220

P. pusa    GB, Stocking        Dec     LACM   KC611178  KC611220
           Is.                 20.   173221

P. pusa    GB, Stocking        Jan     CPIC   KC611179  KC611221
           Is.                 28.    00737

P. pusa    HH2, Stocking    Feb 1.     CPIC   KC611180  KC611222
           Is.                2007    00738

P. pusa    SDB, Stocking       Feb     CPIC   KC611181  KC611223
           Is.                 10,    00739

P. pusa    SDB, Stocking       Feb     CPIC   KC611182  KC611224
           Is.                 10.    00740

P. pusa    SDB, Stocking       Feb     CPIC   KC611183  KC611225
           b.                  10.    00741

P. pusa    SDB, Stocking       Feb     CPIC   KC611184        --
           Is.                 10.    00742

P. pusa    SDB, Stocking       Apr     CPIC   KC611185  KC611226
           Is.                 18.    00743

P. pusa    SDB, Stocking       Apr     CPIC   KC611186  KC611227
           Is.                 18.    00744

P. pusa    SDB, Stocking       Apr       --   KC611187  KC611228
           Is.                 18.

P. pusa    SDB, Stocking       Apr       --   KC611188  KC611229
           Is.                 18.

P. pusa    HH2, Stocking       Dec     CPIC   KC611189  KC611230
           Is.                 21.    00745

P. pusa    HH2, Stocking      Jan.     CPIC   KC611190  KC611231
           Is.                 14.    00746

P. pusa    HH2, Stocking      Jan.     CPIC   KC611191  KC611232
           Is.                 14,    00747

P.         HH2, Stocking      Jan.     CPIC   KC611192  KC611233
pusII      Is.                 14.    00748

P. pusa    HH2, Stocking       Jan     CPIC   KC611193  KC611234
           Is.                 14.    O0749

P. pusa    HH2, Stocking       Jan     CPIC   KC611194  KC611235
           Is.                 14.    00750

P. pus     HH2, Stocking       Jan     CPIC   KC611195  KC611236
a          Is.                 14.    00751

P. pusa    HH2, Stocking       Jan     CPIC   KC611196        --
           Is.                 14.    00752

P. pusa    HH2, Stocking    Jan 6.     CPIC   KC611197  KC611237
           Is.                2009    00753

P. pusa    HH2, Stocking    Jan 6.     CPIC   KC611198  KC611238
           Is.                2009    00754

P. pusa    HH2, Stocking       Jan     CPIC   KC611199  KC611239
           Is.                 18.    00755

P. pusa    HH2, Stocking       Feb     CPIC   KC611200  KC611240
           Is.                 12.    00756

P. pusa    HH2, Stocking       Feb     CPIC   KC611201  KC611241
           Is.                 12.    00757

P. pusa    HH2, Stocking       Dec     CPIC   KC611202        --
           Is.                 21.    00758

P. pusa    HH2, Stocking       Dec     CPIC   KC611203        --
           Is.                 21.    00759

P. pusa    HH2, Stocking       Dec       --   KC611204        --
           Is.                 28.

P. pusa    HH2, Stocking       Jan     CPIC   KC611205        --
           Is.                 28.    00760

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611206        --
           Is.                2010    00761

P. pusa    HH2, Stocking    Feb 8.       --   KC611207        --
           Is.                2010

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611208        --
           Is.                2010    00762

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611209        --
           Is.                2010    00763

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611210        --
           Is.                2010    00764

P. pusa    HH2, Stocking    Feb 8.       --   KC611211        --
           Is.                2010

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611212        --
           Is.                2010    00765

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611213        --
           Is.                2010    00766

P. pusa    HH2, Stocking    Feb 8.     CPIC   KC611214        --
           Is.                2010    00767

SDB = Sand Dollar Beach. GB = Gaviota Bay, = Hurricane Hole
1, HH2 = Hurricane Hole 2.

Statistical analyses of color pattern, body length, and burrowing time

In order to estimate the relationship between genetic structure, body length, burrowing behavior, and color pattern, specimens were assorted into five phenotypic classes depending on their color pattern ranging from lighter to darker: (1) white spotted, body color mostly white with dark spots; (2) light, body color mostly light brown with dark and lighter spots; (3) brown spotted, body color brown with lighter spots; (4) brown, body color uniformly brown with no spots; (5) black, body color uniformly black with no spots. Analysis of variance (ANOVA) tests were conducted in SAS 9.3 (SAS Institute, 1989) to determine possible associations between phenotypic class (color pattern) and body length, and between phenotypic class and burrowing time. In the case of body length, the ANOVA analysis was conducted for all phenotypic classes; but for burrowing time, class 1 (white spotted) was excluded due to lack of data. Additionally a multivariate analysis of variance (MANOVA) test was run in SAS to determine possible associations between the three variables: phenotypic class (color pattern), body length, and burrowing time. The frequency at which specimens of each phenotypic class were found in the field was tested for randomness using a chisquared goodness-of-fit against a uniform distribution implemented in SAS.

DNA extraction

DNA was extracted using either a hot Chelex protocol or the DNeasy Blood & Tissue Kit (Qiagen). About 1-3 mg of the foot was cut into fine pieces for extraction for both protocols. For the Chelex extraction, the foot tissue was rinsed and rehydrated using 1.0 ml of TE buffer (10 mmol [1.sub.-1] Tris, 1 mmol [1.sub.-1] EDTA. pH 8.0) for 20 min. A 10% (w/v) Chelex 100 (100-200 mesh, sodium form, Bio-Rad) solution was prepared using TE buffer. After rehydration, the tissue mixture was then centrifuged, 975.00 [micro]l of the supernatant was removed, and 175.00 [micro]l of the Chelex solution was added. Samples were then heated in a 56 [degrees]C water bath for 20 min and heated in a 100 [degrees]C heating block for 8 min: the supernatant was used for PCR. The DNeasy protocol supplied by the manufacturer was followed, with some modifications. The elution step was modified such that the first elution was collected using 100.00 [micro]l of Buffer AE and was allowed to incubate at room temperature for 5 min before centrifugation. In a new test tube, a second elution step was conducted using 200.00 [micro]l of Buffer AE and was also allowed to incubate at room temperature for 5 min before centrifugation. The first elution was used for PCR.

PCR amplification and sequencing

Palumbi's universal 16S rRNA primers (16S ar-L 5'-CGCCTGTTTATCAAAAACAT-3', 16S br-H 5'-CCG-GTCTGAACTCAGATCACGT-3' developed by Palumbi, 1996), and Folmer's universal CO1 primers (LC01490 5'-GGTCAACAAATCATAAAGATATTGG-3 ' HCO2198, 51-TAAAC1TCAGGGTGACCAAAAAATCA-3' developed by Fol mer et al., 1994) were used to amplify the regions of interest for all specimens.

The master mix was prepared using 34.75 [micro]l of [H.sub.2]O, 5.00 [micro]l of Buffer B (ExACTGene, Fisher Scientific), 5.00 I.A1 of 25 mmol [1.sup.-1] Mg[Cl.sub.2], 1.00 [micro]1 of 40 mmol [1.sup.-1] dNTPs, 1.00 [micro]l of 10 mmol [1.sup.-1] primer 1, 1.00 [micro]l of primer 2, 0.25 [micro]l of 5 mg/ml Tag, and 2.00 [micro],1 of extracted DNA. Reaction conditions for 16S were as follows: an initial denaturation for 2 min at 94 [degrees]C, 30 cycles of (1) denaturation for 30 s at 94 [degrees]C, (2) annealing for 30 s at 50 [degrees]C, and (3) elongation for 1 min at 72 [degrees]C, and a final elongation for 7 min at 72 [degrees]C. Reaction conditions for CO1 were an initial denaturation for 3 min at 95 [degrees]C, 35 cycles of (1) denaturation for 45 s at 94 [degrees]C. (2) annealing for 45 s at 45 [degrees]C, and (3) elongation for 2 min at 72[degrees] C, and a final elongation for 7 min at 72 [degrees]C.

PCR products yielding bands of appropriate size (approximately 475 bp for 16S and 700 bp for C01) were purified using the Montage PCR Cleanup Kit (Millipore). Cleaned PCR samples were quantified using a NanoDrop 1000 spectrophotometer (Thermo Scientific). Each primer was diluted to 2.0 pmol/[micro]l to send out for sequencing with the PCR products. PCR products were diluted to 7.5 and 11.5 ng/[micro]l for 16S, and CO1, respectively. Samples were sequenced at the City of Hope DNA Sequencing Laboratory (Duarte, CA) using chemistry types BigDye V1.1 for fragments less than 500 bp and BigDye V3.1 for fragments larger than 500 bp.

DNA data analysis

Sequences for each gene were assembled and edited using Geneious Pro 4.7.4 (Drummond et al., 2010). Ge-neious was also used to extract the consensus sequence between the primer regions, to construct the alignment for each gene using the default parameters, to concatenate the alignments, and to calculate uncorrected pairwise distances. The sequences were not trimmed after alignment. A total of 412 bp for 16S and 658 bp for CO1 were used for the phylogenetic analyses. Uncorrected pairwise identity and proportion of invariant sites for each gene were also calculated in Geneious.

The Bayesian phylogenetic analysis was conducted for both genes concatenated. To assess whether 16S and COI have significantly conflicting signals, the incongruence length difference (ILD) test (Mickevich and Farris, 1981; Farris et al., 1994), implemented in PAUP*4.0 as the partition homogeneity test (Swofford, 2002), was calculated for all genes combined with 2000 replicates. The levels of saturation for each gene and for the first and second versus third codon positions of COI were investigated using the substitution saturation test developed by Xia et al. (2003) and Xia and Lemey (2009) implemented in the program DAMBE (Xia and Xie, 2001).

The Akaike information criterion (Akaike, 1974) was executed in MrModeltest (Nylander, 2004) to determine the best-fit model of evolution (Table 3). The Bayesian analysis was executed in MrBayes 3.2.1 (Huelsenbeck and Ronquist, 2001), partitioned by gene (unlinked). The Markov chain Monte Carlo analysis was run with two runs of six chains for 10 million generations, with sampling every 100 generations. The default 25% burn-in was applied before constructing majority-rule consensus trees.

Table 3 Results from the ANOVA and MANOVA tests

                      DF      MS       E  F Value  P Value

Phenotypic class vs.   3  794.27  581.51     1.37   0.3210
body length

Phenotypic class vs.   4   31.94   15.31     2.09   0.1024
burrowing time

MANOVA (three          6      --      --     2.77   0.0544

DF, degrees of freedom; MS, model mean square; E. error mean square.

For the analysis of the population genetic structure in relation to color pattern, haplotype networks for CO1 and 16S sequences were constructed using TCS 1.21 (Clement et al., 2000) with a 95% connection limit. The color phenotypic classes mentioned above were coded in the haplotype groups identified with TCS.

Based on color pattern (phenotypic classes), five population groups were tested for genetic structure using the analysis of molecular variance (AMOVA) implemented in Arlequin (Excaffier and Lischer, 2010). Significance of the AMOVA was tested using 16,000 permutations of individuals between groups. Although AMOVA analysis is typically used to determine molecular variance between population groups, here it was used to estimate genetic differentiation between groups of individuals with different color patterns. The intention is to provide statistical evidence of possible genetic differentiation between organisms with different color patters. A similar approach was successfully implemented by Lin et al. (2008). Additionally, the [] value as a measure of pairwise differences between the two haplotype groups recovered in the TCS analysis was calculated using Arlequin using the two genes concatenated. The significance of the pairwise [] value was estimated by performing 10,000 permutations.

Results and Discussion

The sequence data obtained in this study were informative for the Bayesian analysis. The saturation analyses showed insignificant levels of saturation for both genes, CO1: Iss (0.0373) < Iss.c (0.7304), P = 0.000; 16S: Iss (0.0126) < Iss.c (0.6933), P = 0.000, even when the third codon positions of COI were analyzed independently: Iss (0.1027) < Iss.c (0.6395), P = 0.000. Thus, the entire sequences were determined to be useful for phylogenetic reconstruction. Additionally, the ILD test showed no significant conflicting signals between the two genes combined (P = 0.68); therefore both genes were concatenated and analyzed together.

The analysis of the molecular data reveals limited genetic variation among specimens of Philinopsis pusa collected in the Bahamas in the two mitochondrial genes examined. The uncorrected pairwise distances for the specimens here examined (CO1 = 99.6% and 16S = 99.8%, proportion of identical sites for CO1 = 98.2% and 16S = 98.3%) are consistent with intraspecific variation in other species of cephalaspidean opisthobranchs from the region. For example, Omelas-Gatdula et al. (2011, 2012) found uncorrected pairwise distances ranging from 95.5% to 100% for 16S and 94.6% to 99.5% for CO1 in several species of Navanax and Chelidonura from the Atlantic and Eastern Pacific.

The Bayesian consensus tree (Fig. 2A) confirmed the monophyly of the Bahamian specimens assigned to P. pusa regardless of their color pattern. Moreover, it recovered a poorly supported clade (posterior probability 0.93) including specimens representing all the phenotypic classes. The rest of the specimens (also covering the entire range of color patterns) form a large basal polytomy (Fig. 2A). The CO1 haplotype network (Fig. 2B) shows results similar to the phylogenetic analysis, but in this case two distinct haplotype groups were recovered. These two haplotype groups differ in four substitutions and correspond to the poorly supported clade in the Bayesian tree and the basal polytomy. Both haplotype groups include specimens representing all phenotypic classes. There is no correlation between the two haplotype groups and depth or time of the year in which the specimens were collected. The 16S haplotype network (Fig. 2C) is also similar, as it recovered two large haplotype groups which correspond to the haplotype groups found in the CO1 network, but in this case they differ in only one substitution. The result of the pairwise [] analysis between these two haplotype groups was significant ([] = 0.4535; P = 0.000), confirming that the two haplotype groups are genetically distinct. Finally, the AMOVA test (Table 4) showed no significant genetic differentiation between the phenotypic classes (color patterns).

Table 4 Matrix of Fcr values based on AMOVAs between groups for
both genes concatenated (lower triangular)

            White     Light     Brown    Brown (4)       Black (5)
          spotted       (2)   spotted
              (1)                 (3)

While          --       P =       P =  P = 0.58143  P = 0.24872 [+
spotted             0.59699   0.53156     [+ or -]   or -] 0.00356
(1)                [+ or -]  [+ or -]      0.00388
                    0.00414   0.00373

Light     0.05306        --       P =  P = 0.54787  P = 0.03418 [+
(2)                           0.38451     [+ or -]   or -] 0.00133
                             [+ or -]      0.00429

Brown    -0.30421  -0.08930        --  P = 0.68666  P = 0.63336 [+
spotted                                  + 0.00359   or -] 0.00394

Brown    -0.07393  -0.01804  -0.01635           --  P = 0.13842 [+
(4)                                                  or -] 0.00302

Black     0.32603   0.59061  -0.03870      0.21720

After Bonferroni correction (10 comparisons) significant values
are P < 0.005.

The haplotype network and [] results indicate that there is genetic structure in Bahamian populations of P. pusa, although it was only poorly recovered in the Bayesian analysis, but this structure is not correlated with color pattern. Moreover, there are no significant genetic differences between the different phenotypic classes identified, suggesting that color pattern is a poor predictor of genetic structure for the two neutral markers here studied. However, we did not test for the possible genetic basis of color pattern due to the difficulty of conducing crossbreeding experiments with this species.

Ortea et at. (2007) introduced species names for specimens with different color patterns assigned to P. pusa, suggesting that color (along with the shell morphology) is a useful trait to recognize and distinguish species in this group (see also Ortea et al., 2012). White spotted forms were named Philinopsis batabanoensis Ortea et al., 2007, whereas darker animals were named Philinopsis bagaensis Ortea et at., 2007. The name P. pusa was left for darker specimens with dark pigment in the center of the light spots. If this species hypothesis is correct and different color patterns represent distinct, yet closely related sympatric species, it might suggest that disruptive selection acting on color could have played a role in speciation. We could envision a scenario in which the species is chemically defended and lighter, better-camouflaged individuals avoid attacks by blending in, whereas more conspicuous darker animals with neon blue lines prevent attacks by advertising their chemical defenses. If the species is not chemically defended, lighter forms would rely on camouflage, whereas conspicuous animals should evolve alternative defensive mechanisms such as faster burrowing to escape attacks. Disruptive selection is known to cause sympatric speciation (Fournier and Giraud, 2007; Elmer et al., 2010), but it requires the formation of genetic barriers (e.g., niche partitioning, allochrony, mate choice) between the incipient species. In freely interbreeding sexual populations such as P. pusa, the distribution of phenotypes is constrained by the processes of segregation and recombination, which cause many individuals to have the maladaptive intermediate phenotypes (Dieckmann and Doebeli, 1999). However, recent studies with birds have shown that foraging preferences of predators can regularly (though not always) result in the increase to fixation of a novel color morph (Marples et al., 2005). More importantly, such fixation events occur even if both novel and familiar prey are fully palatable and despite the novel color morph being much more conspicuous than others (Marples et al., 2005). Additionally, computer modeling revealed that when crypsis imposes opportunity costs (for example, when the species moves over visually heterogeneous habitats), prey evolve secondary defenses that facilitate raised behavioral conspicuousness as they exploit opportunities within their environment (Speed and Ruxton, 2005). Regardless of whether P. pusa is chemically defended or not, it is theoretically possible that selection resulted in the evolution of different color morphs in this species. However, our data suggest that such selection has not occurred or at least has not resulted in genetic differentiation in the neutral markers examined.

With the available data, we cannot provide a viable hypothesis to explain the existence of two genetically distinct groups of P. pusa in the Bahamas, as shown by the [] analysis. However, the existence of pairs of genetically distinct groups of aglajid sea slugs in this region seems to be a recurring pattern; this was found in populations of Phili-nopsis petra and Chelidonura berolina (Ornelas-Gatdula et al., 2011; Omelas-Gatdula and Valdes, 2012). However, in these two other examples, the genetic diversification was accompanied by consistent phenotypic differences, and the two groups were considered to be distinct species. Omelas-Gatdula and Valdes (2012) argued that in the case of P. petra the genetic divergence most likely occurred allopatrically. Whatever the reason for the genetic divergence among populations of P. pusa, P. petra, and C. berolina in the Bahamas, it is apparently unrelated to color pattern. In all three species there is a considerable color pattern variation among, but not between, genetically distinct groups.

Another goal of this study was to determine possible correlations between color pattern and other biological characteristics of these animals (such as size and behavior) in an attempt to investigate the source of color variation in this species. For example, if individuals of different sizes have different color patterns, it could be an indication of ontogenetic variation. Alternatively, if the relative abundance of different color patterns varies seasonally or with depth, it would suggest the influence of environmental factors or diet. Finally, if individuals possessing certain color patterns display different defensive (escape) behaviors, it would suggest a possible ecological and evolutionary role of color pattern. It should be expected that aposematic individuals would have slower escape behaviors than cryptic ones in order to maximize the effectiveness of aposematism (Jackson et at., 1976; Cooper et at., 2009).

The ANOVA tests (Table 3) revealed that there is no significant difference in length among the five phenotypic classes (P = 0.1024) and no significant difference in burrowing time among the four phenotypic classes for which burrowing time was measured (P = 0.3210). The MANOVA test also revealed a nonsignificant association between the three variables (phenotypic class, body length, and burrowing time) (P = 0.0544). The goodness-of-fit test revealed that phenotypic classes do not occur randomly: the observed frequencies are significantly different from those expected from a 1:1:1:1:1 ratio ([x.sup.2] = 10.6, DF = 4, P = 0.03). Further investigation revealed that intermediate phenotypes (brown spotted and brown) do not occur significantly more often than the rest of the phenotypic classes (white spotted, light, and black) ([x.sup.2] = 0.15, DF = 2, P = 0.98), darker phenotypic classes (brown spotted, brown and black) occur significantly more often than the light phenotypic classes (white spotted and light) ([x.sup.2] = 16.3, DF = 1, P = 0.0001), and all the dark phenotypic classes are equally frequent ([x.sup.2] = 1.94, DF = 2, P = 0.4). Figure 3 provides a graphical representation of the data.

The absence of correlation between burrowing speed and color pattern also suggests that color does not play a significant defensive role in this species. Specimens of P. pusa in the Bahamas were found crawling on white sand during the day, readily exposed to visual predators. When disturbed, the animals burrow in the substrate, but the process is slow (89 s on average, and as long as 126 s) (Table 1) and very unlikely to prevent predation from fast-moving predators such as fish. The shell of P. pusa is a vestigial, internal plate. Thus it appears that this species must rely on chemical defenses. However, there is no evidence to support that supposition, beyond circumstantial evidence that all other members of Philinopsis for which the chemistry is known are chemically defended (Cimino and Ghiselin, 2009).

Animals of all color forms were collected during the same time of the year, in the same locality, at the same time of the day, and were subjected to similar environmental conditions; we found no significant relationship between phenotypic class and body length (a proxy for age). This appears to suggest that color is not determined by environmental factors (maturity, temperature, depth, exposure to light, etc.). However, it is difficult to determine the influence that diet may play in the determination of color pattern, as we do not have data on the historic food source of the specimens collected. We have observed P. pusa feeding on other opisthobranchs sea slugs--Gastropteron chacmol and Aglaja fellx--as well as on others of its own species. If the color pattern in P. pusa were influenced by diet, it would be plausible that specimens that preyed more often upon darker species such as A. felix would have darker colorations, whereas animals that preferably fed on lighter species such as G. chacmol would be lighter. However, this hypothesis does not explain the presence of bright neon blue lines on darker specimens. It is also difficult to explain why animals living in the same area and active during the same period of time would have distinct feeding habits. The relatively higher abundance of darker phenotypic classes could be an artifact of collecting biases, as darker animals are more visible to a human collector on a light background.

Two recent studies investigated the reproductive behavior of different color morphs in another aglajid sea slug, the Chelidonura tsurugensis-sandrana species complex (Anthes and Michiels, 2007; Turner and Wilson, 2012). Both papers described behavioral differences between color morphs, but Turner and Wilson (2012) found no assortative mating by color, suggesting that there is no support for the classification of the different color forms as separate species or subspecies. This conclusion is similar to our results for P. pusa based on molecular evidence. However, Turner and Wilson (2012) showed that color forms can have a greater tendency to mate with a color form different from themselves, more often and/or for a longer time. Those authors suggested several scenarios in which color forms could be maintained in the absence of assortative mating, and they proposed lines of research to elucidate the mechanism that maintains these color polymorphisms.

The implications of the results of the present study are twofold. First of all, our genetic data do not support the existence of molecular population structure associated with the color forms of Philinopsis pusa. This result contradicts the species hypotheses based on color pattern proposed by Ortea et at. (2007). Because the type localities (Cuba and Brazil) of the species considered valid by Ortea et at. (2007) were not sampled in this study, we cannot determine whether the species names should be synonymized. Examination of the shells of several specimens with different color patterns from the Bahamas showed no morphological differences, but these data cannot be extrapolated to other regions in the Caribbean. Second, we found no relationships between color pattern and the size (ontogeny) or the nonreproductive (escape) behavior of individuals of P. pusa. These results, along with the absence of genetic structure associated with color pattern, cast doubt upon the biological significance of external coloration in this species. Although these data cannot be extrapolated to other groups of opistho-branchs, and some of the most conspicuously colored groups (such as chromodorid nudibranchs) could have achieved rapid diversification because of aposematic coloration defense strategies (Johnson and Gosliner, 2012), the present paper points out the need for caution in making broad generali7ations about the biological and evolutionary role of color in opisthobranchs and other marine organisms.


We are extremely grateful to David Moriarty (Cal Poly Pomona) for his help with the statistical analyses and to Richard Zimmer, Paul Beardsley, and three anonymous reviewers for their constructive comments and criticisms. This paper was supported by a Cal Poly Pomona Research, Scholarship, and Creative Activity grant to A. Valdes. Additional material and travel support was provided by student grants from Conchologists of America and the Southern California Academy of Sciences to E. Ornelas-Gatdula. Finally, E. Ornelas-Gatdula was supported by a NIH MBRS Research Initiative for Scientific Enhancement (RISE) grant to Cal Poly Pomona (2 R25 GM061190-05A2). Lindsey Groves from the Natural History Museum of Los Angeles County helped with access to the collection and curation of specimens.

Literature Cited

Akaike, H. 1974. A new look at the statistical model identifications. IEEE Trans. Automat. Contr. 19: 716-723.

Anthes, N., and N. K. MichieLs. 2007. Reproductive morphology, mating behavior, and spawning ecology of cephalaspid sea slugs (Aglaji-dae and Gastropteridae). Invertehr. Biol. 126: 335-365.

Anthes, N., H. Schulenburg, and N. K. Michiels. 2008. Evolutionary links between reproductive morphology, ecology and mating behavior in opisthobranch gastropods. Evolution 62: 900-916.

Barnett, C. A. J. Skelhorn, M. Bateson, and C. Rowe. 2012. Educated predators make strategic decisions to eat defended prey according to their toxin content. Behav. Ecol. 23: 418-424.

Cimino, G., and M. T. Ghiselin. 2009. Chemical defense and the evolution of opisthobranch gastropods. Proc. Calif Acad. Sci. 60: 175-422.

Clement, M., D. Posada, and K. A. Crandall. 2000. TCS: A computer program to estimate gene genealogies. Mol. Ecol. 9: 1657-1659.

Cooper, W. E., J. P. Caldwell, and L. J. Vitt. 2009. Conspicuousness and vestigial escape behaviour by two dendrobatid frogs. Dendrobates auratus and Oophaga purnilio. Behaviour 146: 325-349.

Cortesi, F., and K. L. Cheney. 2010. Conspicuousness is correlated with toxicity in marine opisthobranchs. J. Evol. Biol. 23: 1509-1518.

Cott, H. B. 1940. Adaptive Coloration in Animals. Methuen. London.

Crozier, W. J. 1916. On the immunity coloration of some nudibranchs. Proc. Natl. Acad. Sci. 2: 672-675.

Darst, C. R., M. E. Cummings, and D. C. Cannatella. 2006. A mechanism for diversity in warning signals: conspicuousness versus toxicity in poison frogs. Proc. Natl. Acad. Scl. 103: 5852-5857.

Dayrat, B. 2010. A monographic revision of discodorid sea slugs (Gastropoda. Opisthobranchia, Nudibranchia, Doridina). Proc. Calif Acad. Sci. 61: 1-403.

Dieckmatm, U.. and M. Doebeli. 1999. On the origin of species by sympatric speciation. Nature 400: 354-357.

Drummond, A. J., B. Ashton, M. Cheung, J. Heled, M. Kearse, R. Moir, S. Stones-Havas, T. Thierer, and A. C. Wilson. 2010. Geneious v 4.8 [Online]. Biomatters. Available: [2010, March 3].

Edmunds, M. 1987. Color in opisthobranchs. Am. Malacal. Bull. 5: 185-196.

Elmer, K. R., T. K. Lehtonen, A. F. Kautt, C. Harrod, and A. Meyer. 2010. Rapid sympatric ecological differentiation of crater lake cichlid fishes within historic times. BMC Biol. 8: 60. doi:10.1186/1741-7007-8-60.

Excoffier, L., and H. E. L. Lischer. 2010. Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Mol. Ecol. Resour. 10: 564-567.

Farris, J. S., M. Kallersjo, A. G. Kluge, and C. Bait. 1994. Testing significance of incongruence. Cladistics 10: 315-319.

Folmer, 0., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar Biol. Biotech. 3: 294-299.

Fournier, E., and T. Giraud. 2007. Sympatric genetic differentiation of a generalist pathogenic fungus, Botrytis cinerea, on two different host plants, grapevine and bramble. J. Evol. Biol. 21: 122-132.

Gosliner, T. M. 2001. Aposematic coloration and mimicry in opistho-branch mollusks: new phylogenetic and experimental data. Boll. Malacol. 37: 163-170.

Gosliner, T. M. 2011. Six new species of aglajid opisthobranch mollusks from the tropical Indo-Pacific. Zootaxa 2751: 1-24.

Gosliner, T. M., and D. W. Behrens. 1989. Special resemblance, apose-matic coloration and mimicry in opisthobranch gastropods. Pp. 127138 in Symposium on the Adaptive Significance of Color in Invertebrates, M. Wicksten, ed. Texas A&M University Press. College Station, TX.

Gosliner, T. M., D. W. Behrens, and A. Valdes. 2008. Indo-Pacific Nudibranchs and Sea Slugs: A Field Guide to the World's Most Diverse Fauna. Sea Challengers, Gig Harbor, WA.

Grant, J. B. 2007. Ontogenetic colour change and the evolution of aposematism: a case study in panic moth caterpillars. J. Anim. Ecol. 76: 439-447.

Guilford, T., and I. Cuthill. 1991. The evolution of aposematism in marine gastropods. Evolution 45: 449-451.

Haber, M., S. Cerfeda, M. Carbone, G. Calado, H. Gaspar, R. Neves, V. Maharajan, G. Cimino, M. Gavagnin, M. T. Ghiselin, and E. Mollo. 2010. Coloration and defense in the nudibranch gastropod Hypselodoris .fontandraui. Biol. Bull. 218: 181-188.

Harris, L. G. 1973. Nudibranch associations. Curr. Top. Comp. Patho-biol. 2: 213-315.

Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes: Bayesian inference of phylogeny. Bioinformatics 17: 754-755.

Hughes, H. P. I. 1970. A light and electron microscope study of some opisthobranch eyes. Z. Zellforch. Microsk. Anat. Abr. Histochem. 106: 79-98.

Jackson, J. F., W. Ingram, and H. W. Campbell. 1976. The dorsal pigmentation pattern of snakes as an antipredator strategy: a multivariate approach. Am. Natur. 110: 1029-1053.

Johnson, R. F., and T. M. Gosliner. 2012. Traditional taxonomic groupings mask evolutionary history: a molecular phylogeny and new classification of the chromodorid nudibranchs. PLoS ONE 7(4), e33479.

Lin, H.-C., S.-H. Li, J. Fong, and S. M. Lin. 2008. Ventral coloration differentiation and mitochondrial sequences of the Chinese Cobra (Naja arra) in Taiwan. Conserv. Genet. 9: 1089-1097.

Lindstedt, C., H. Eager, E. Ihalainen, A. Kahilainen, M. Stevens, and J. Mappes. 2011. Direction and strength of selection by predators for the color of the aposematic wood tiger moth. Behav. Ecol. 22: 580-587.

Long, J. D., and M. E. Hay. 2006. Fishes learn aversion to a nudi-branch's chemical defense. Mar. Ecol. Prog. Ser. 307: 199-208.

Marples, N. M., D. J. Kelly, and R. J. Thomas. 2005. Perspective: the evolution of warning coloration is not paradoxical. Evolution 59: 933-940.

Mickevich, M. F., and J. S. Farris. 1981. The implications of congruence in Menidia. Syst. Zool. 30: 351-370.

Nokelainen, 0., R. H. Hegna, J. H. Rendler, C. Lindstedt, and J. Mappes. 2012. Trade-off between warning signal efficacy and mating success in the wood tiger moth. Proc. R. Soc. B 279: 257-265.

Nylander, J. A. A. 2004. MrModeltest ver. 2 [Online]. Evolutionary Biology Centre. Uppsala University, Uppsala. Available: http:// [2010. October 28].

Ornelas-Gatdula, E., and A. Valdes. 2012. Two cryptic and sympatric species of Philinopsis (Mollusca, Opisthobranchia, Cephalaspidea) in the Bahamas distinguished using molecular and anatomical data. J. Molluscan Stud. 78: 313-320.

Ornelas-Gatdula, E., A. DuPont, and A. Valdes. 2011. The tail tells the tale: taxonomy and biogeography of some Atlantic Chelidonura (Gastropoda: Cephalaspidea: Aglajidae) inferred from nuclear and mitochondrial gene data. Zool. J. Linn. Soc. 163: 1077-1095.

Ornelas-Gatdula, E., Y. Camacho-Garcia, M. Schrodl, V. Padula, Y. Hooker, T. M. Gosliner, and A. Valdes. 2012. Molecular systematics of the "Navanax aenigmaticus" species complex (Mollusca, Opisthobranchia): coming full circle. Zool. Scr. 41: 374-385.

Ortea, J., L. Moro, and J. Espinosa. 2007. Descripcion de dos nuevas especies de Philinopsis Pease, 1860 (Mollusca: Opisthobranchia: Cephalaspidea) de Cuba y Bahamas con comentarios sobre las especies atlanticas del genero. Rev. Acad. Canaria Cienc. 18: 33-52.

Ortea, J., J. Espinosa, L. Moro, M. CabsIler, and J. J. Bacallado. 2012. Notas en Opisthobranchia (Mollusca, Gastropoda) 5: Sobre el uso de la concha interna como catheter sistematico de primer orden en el inven-tario de las especies Atltinticas de la familia Aglajidae (Mollusca: Cephalaspidea). Rev. Acad. Canaria Cienc. 24: 183-195.

Palumbi, S. R. 1996. Nucleic acids 11: The polymerase chain reaction. Pp. 205-247 in Molecular Systematics. D. M. Hillis, C. Moritz, and B. K. Mable, eds. Sinauer, Sunderland, MA.

Pola, M., Y. Valles, J. L. Cervera, M. Medina, and T. M. Gosliner. 2006. Taxonomic status of Tambja abdere and Tambja fitsca based on morphological and molecular evidence, with comments on the phylogeny of the subfamily Nembrothinae (Nudibranchia: Polyceri-dae). Ann. Zool. Fenn. 43: 52-64.

Ritson-Williams, R., and V. J. Paul. 2007. Marine benthic invertebrates use multimodal cues for defense against reef fish. Mar. Ecol. Prog. Ser. 340: 29-39.

Rosenberg, G. 1989. Warning coloration evolves by individual selection: evidence from marine gastropods with pelagic larvae. Evolution 43: 1811-1813.

Rudman, W. B. 1984. The Chromodorididae (Opisthobranchia: Mollusca) of the Indo-West Pacific: a review of the genera. Zool. J. Linn. Soc. 81: 115-273.

Rudman, W. B. 1991. Purpose in pattern: the evolution of colour in chromodorid nudibranchs. J. Molluscan Stud. 57: 5-21.

SAS Institute. 1989. SAS/STAT User's Guide, ver. 6, Vol. 1, 4th ed. Cary, NC.

Speed, M. P., and G. D. Ruxton. 2005. Aposematism: what should our starting point be? Proc. R. Soc. B 272: 431-438.

Stevens, M., and G. D. Ruxton. 2012. Linking the evolution and form of warning coloration in nature. Proc. R. Soc. B 279: 417-426.

Swofford, D. L. 2002. PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods). ver. 4. Sinauer, Sunderland, MA.

Tullrot, A. 1994. The evolution of unpalatability and warning coloration in soft-bodied marine invertebrates. Evolution 483: 925-928.

Turner, L. M., and N. G. Wilson. 2012. The Chelidonura tsurugensis-sandrana (Gastropoda: Cephalaspidea) species complex: Do reproductive decisions maintain colour polymorphism? J. Molluscan Stud. 78: 166-172.

Xia, X., and P. Lemey. 2009. Assessing substitution saturation with DAMBE. Pp. 615-630 in The Phylogenetic Handbook: A Practical Approach to DNA and Protein Phylogeny, 2[nd] ed, P. Lemey, M. Salemi, and A.-M. Vandamme, eds. Cambridge University Press, Cambridge, UK.

Xia, X., and Z. Xie. 2001. DAMBE: Data analysis in molecular biology and evolution. J. Hered. 92: 371-373.

Xia, X., Z. Xie, M. Salemi. L. Chen, and Y. Wang. 2003. An index of substitution saturation and its application. Mol. Phylogenet. Evol. 26: 1-7.


(1.) Department of Biological Sciences, California State Polytechnic University, 3801 West Temple Avenue, Pomona, California 91768; and (2.) 4070 NW 7th Lane, Delray Beach, Florida 33445

Received 21 August 2012: accepted 19 February 2013.

* To whom correspondence should be addressed. E-mail:
COPYRIGHT 2013 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Valdes, Angel; Ornelas-Gatdula, Elysse; Dupont, Anne
Publication:The Biological Bulletin
Article Type:Report
Geographic Code:1USA
Date:Feb 1, 2013
Previous Article:Low functional redundancy in sponges as a result of differential picoplankton use.
Next Article:Microtransplantation of cellular membranes from squid stellate ganglion reveals ionotropic GABA receptors.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |