Phylogenetic signal and diversity of visceral pigmentation in eight anuran families

PHYLOGENETIC SIGNAL AND VARIATION OF VISCERAL PIGMENTATION IN

EIGHT ANURAN FAMILIES

Zoologica Scripta 2012 41(6): 547-556

DIOGO B. PROVETE et al.

Biological traits

¨  Models of trait evolution

¤ Plasticity vs. Conservatism => phylogenetic constraints and the role of environment

¤ Meaning of phylogenetic signal => related species have similar traits

Pigmented cells

Wallin,2002

Visceral melanocytes

Differential occurrence among organs

Goals

¨  Function of visceral pigmentation not yet understood ¤ Hypothesis relate the role of melanin as a protective

agent against toxic agents and also in thermoregulation

¨  Analyse the distribution of visceral pigmentation among species

¨  Test for a phylogenetic signal using a comparative method (Pavoine et al. 2010)

Methods

¨  Pigmentation categories on the surface of 16 visceral organs of 35 anuran species, belonging to 13 genera from 8 families nested within Meridianura (sensu Frost et al. 2006)

Methods

¨  Data collection ¤ Field: We collected calling adult anurans in several

locations in the state of São Paulo, southeastern Brazil

¤ Literature: Franco-Belussi et al. (2011, 2012)

¤ Museum specimens: DZSJRP, CCLZU, JJ-MNRJ

Visceral melanocytes quantification

Franco-Belussi et al., 2009 Anat. Rec.

Figure 3. (Legend on page 179.)



180 FRANCO-BELUSSI ET AL.

peritoneum, kidneys, renal veins, urinary bladder, testes,fatty bodies, lumbar plexus nerves (lumbar portion), parie-tal peritoneum, and intestinal mesenterium. We recordedthe pigmentation on these organs ⁄ structures based on col-oration intensity, following the protocol of Franco-Belussiet al. (2009), which is similar to those used by Grant et al.(2006). The intensity of pigmentation on organs wasdivided into four categories, ranging from absence of pig-mentation to entirely pigmented, as follows: category (0)lack of pigment cells on the surface of organs, in whichthe usual colour of the organ is evident; category (1) a fewscattered pigment cells, giving the organs a faint pigmen-tation; category (2) presence of a large amount of pigmentcells; category (3) presence of a massive amount of pig-ment cells, rendering an intense pigmentation to the struc-ture, changing its usual colour and superficialvascularization (Franco-Belussi et al. 2009). We assigned acertain category of pigmentation to an organ ⁄ structureconsidering the pigmentation of the majority of individu-als. For example, when three specimens out of five hadpigmentation category 0, and the remaining two hadcategory 1, we assigned category 0 to that organ in that

species, since the method used (Pavoine et al. 2010; seebelow also) does not incorporate intraspecific trait variabil-ity and only uses a single value.

Taxon sampling and statistical analysisSpecies selection for this study took into account availabil-ity of phylogenetic information, published data aboutvisceral pigmentation, and its availability in scientific col-lections. We included species from the following familiesin the analysis: Brachycephalidae, Craugastoridae, Cyclor-amphidae, Hylidae, Hylodidae, Leiuperidae, Leptodactyli-dae and Strabomantidae (see Appendix S2). To conductthe analysis, we assembled by hand a supertree for the spe-cies sampled using the software MESQUITE 2.74 (Maddison& Maddison 2010), considering the topological relation-ships proposed by the following studies: Ponssa (2008) forspecies of Leptodactylus; Wiens et al. (2010) for species ofScinax and Dendropsophus; Hedges et al. (2008) for speciesof Ischnocnema, Holoaden and Haddadus; Tarano & Ryan(2002) for species of Physalaemus. We followed the phylog-eny of amphibians proposed by Frost et al. (2006), as mod-ified by Grant et al. (2006) and Hedges et al. (2008), for

Fig. 1 Phylogenetic tree constructed for the analysis of trait diversity showing the higher-order groups (left), families (right) and speciesgroups within Leptodactylus.

D. B. Provete et al. d Visceral pigmentation in anurans

ª 2012 The Authors d Zoologica Scripta ª 2012 The Norwegian Academy of Science and Letters, 41, 6, November 2012, pp 547–556 549

Decomposing trait diversity along the nodes of a phylogeny

possible permutations of the nodes that directly descendfrom node j, and Ti is the set of interior nodes in thesubtree rooted at node i. Consequently, the test does notdepend on a single tree representation used to displaythe phylogenetic tree. In contrast, it uses all possible treerepresentations to order the nodes. The subtree rootedat node i is complex if the number of possible treerepresentations is high. We propose an additionalcriterion when branch lengths are available: the sum ofbranch lengths between the node and the root (Fig. 3).Note that if the phylogeny is dated, ordering the nodesfrom the oldest to the most recent is exactly equal toordering the nodes from the closest to the root to thefarthest from the root. Nodes are ordered from low tohigh values of complexity or alternatively low to highvalues of the sum of branch lengths between the nodeand the root. The third statistic is

S3 !1

N

XN

i!1

ivi "7#

where, in contrast to Ollier et al. (2006), we scale the testby N, the number of interior nodes to obtain a valuebetween zero and unity. This third test provides adescription of phylogenetic signal where species havemore different trait values if they are distantly related onthe phylogeny (trait diversity is skewed to root) and adescription of the absence of phylogenetic signal whereclosely related species have highly different trait values(e.g., due to convergent evolution and trait diversity is

then skewed to tips; Fig. 2A, B). This tips/root skewnesstest is related to the phylogenetic signal randomizationtest of Blomberg et al. (2003) based on branch-lengthtransformations and to the divergence order test (DOT)developed by Ackerly et al. (2006). In the statistic ofAckerly et al. i is replaced with the age of the node and viwith unstandardized contrasts for a single trait acrossthe nodes (according to Felsenstein 1985). In contrast toour statistic S3, statistics of Ackerly et al. and Blomberget al. can only be applied to a single quantitative trait ata time, although the values taken by the statistic ofAckerly et al. for two traits can be compared todetermine if one trait diverged earlier than the other.In S3, several traits from any statistical type (quantita-tive, nominal, binary, circular) can be combined.All tests are performed by permuting the species

across the tips of the phylogenetic tree, which corre-sponds to the permutation scheme 1p of Hardy (2008).Other permutation schemes are given by Hardy (2008).A nominal a error must be chosen: we set a ! 5%. Thestatistics are calculated after permutations, leading totheoretical distributions under the hypothesis that thefrequencies and trait states of species are interchange-able across the tips of the phylogeny.In general, the three tests allow the phylogenetic

signal underpinning trait diversity to be determined (see,for example, Fig. 2). Indeed, four contrasting situationscould, for instance, be obtained from the application ofour methods (Fig. 2). In Fig. 2A, B, the diversity in trait

FIG. 2. Four different situations that occur in the decomposition of trait diversity among the nodes of a phylogenetic tree: (A)‘‘close-to-root’’ model, (B) ‘‘close-to-tips’’ model, (C) random model, and (D) even model. For this simple example, one trait only isconsidered. The values are provided by Cleveland (1994) dot plots. Panels (A) and (B) show exact values of the traits considered(structured). In panel (C), values have been taken from a normal Gaussian distribution with mean 0 and variance 1 (representingrandomness). Values of the quadratic entropy (QE) applied on the Euclidean distances among trait states in panels (A)–(C), and onequal distances in panel (D), are given at the bottom of each panel. The relative abundances of species were considered equal in allexamples. The circles at nodes provide the contribution of nodes to trait diversity. The scale is given at the bottom left-hand cornerof each panel. No circles or small circles mean null or low contributions. The sum of all contributions is equal to the total traitdiversity as measured by the quadratic entropy (QE values). Results of the permutation tests are given at the bottom of each panel:SN, single-node skewness test; FN, few-nodes skewness test; Ro, root/tips skewness test (two sided); NS, nonsignificant. Whenequal distances are considered (panel D), no test is available because the permutations of the species across the tips of thephylogeny do not affect the phylogenetic pattern of trait diversity since species are considered as equivalent.

SANDRINE PAVOINE ET AL.490 Ecological MonographsVol. 80, No. 3

Pavoine et al. 2010 Ecol. Monog

Phylogenetic signal Plasticity Random

Results

¨  Phylogenetic signal in each organ/structure is significantly biased towards the root in the heart, testicle, lumbar parietal peritoneum, lumbar nerve plexus, as well as considering all organs together.

¨  Analysis was run to 12 out of 16 organs (4 organs did not have pigmenation)

influence melanocyte dynamics and the occurrence of pig-mentation on testes. However, experimental studies on thefunction of the pigmentation on testes are still scarce, andfurther information is needed to properly infer the under-lying process producing this phylogenetic signal.

Conversely, pigmentation on the pericardium, mesente-rium, vertebral column, rectum, lungs, cardiac blood ves-sels, kidneys, and renal veins seems to be more labile.Indeed, the pigmentation in those organs was convergentin our phylogeny. Specifically, the diversity of pigmenta-tion categories was skewed towards the tips of the phylog-eny in the rectum and lungs. This can occur whendistantly related species evolve towards similar phenotypictraits, because of similar environmental conditions (Pavo-ine et al. 2010). In fact, the pigmentation in these organsseems to have a strong phenotypic plasticity and couldchange according to the local environment. The visceral

pigmentation may vary according to physiological (e.g. age,nutritional status and diseases; Agius & Agbede 1984)and ⁄ or environmental factors (e.g. temperature). Tempera-ture is an environmental factor that varies geographicallyand may change pigmentation and metabolism of the liver(Barni et al. 2002). However, in our analysis, pigmentationseems to vary around a pattern that is species specific.

We did not find pigmentation in fatty bodies, urinarybladder, intestine and stomach. A similar pattern is alsoreported for Dendropsophus (Franco-Belussi et al. 2011) andScinax (Franco-Belussi et al. 2012). The absence of pig-mentation may be due to tissue types and embryonic ori-gins, or even cannot be visible on organ’s surface.

In the analysis that took into account all organs, thenodes corresponding to Athesphatanura and Cruciabatra-chia had the highest diversity of pigmentation categories.The pattern of visceral pigmentation in anurans found in

Heart Pericardium Cardiac blood vessels

TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09

SNtTest = 0.02FNtest = 0.01Ro test = 0.01

TQE = 0.244SN test = 0.87FN test = 0.73Ro test = 0.84


KidneyLung Renal veins




TesƟcle Rectum Mesenterium

Lumbar parietalperitoneum




Lumbarnerve plexus

All organsVertebralcolumn





A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028

Fig. 6 Decomposition of the diversity of categories of pigmentation among the nodes of the anuran phylogenetic tree. (A–L) Variation ofpigmentation categories as measured in a single specified organ. —M. Variation of pigmentation categories as measured considering thepigmentation categories of all organs among the 32 species. Circles at nodes provide the contribution of nodes to trait diversity, scale aregiven below each tree. The larger the circle, the larger the trait diversity in that node. Results of the permutation tests are given at thebottom of each tree: SN, single-node skewness test; FN, few-nodes skewness test; Ro, root ⁄ tips skewness test (two sided). Total quadraticentropy (TQE) represents the overall value of trait diversity, the higher the TQE, the higher the diversity of pigmentation categories in agiven organ. See Fig. 1 for node labels. [Correction added on 8 October 2012, after first online publication: Fig. 6 was replaced with thecorrect test values for section K].

Visceral pigmentation in anurans d D. B. Provete et al.

554 ª 2012 The Authors d Zoologica Scripta ª 2012 The Norwegian Academy of Science and Letters, 41, 6, November 2012, pp 547–556

Pigmentation in the urogenital system







TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028










TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028










TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028




Cardio-respiratory system

The species from Terrarana, Leptodactylidae, Hylodidae and Proceratophrys had no pigmentation in the pericardium, whereas Odontophrynus an Leiuperidae had a small amount of pigmentation

Cardio-respiratory system







TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028










TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028




Digestory system







TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028




Digestory system







TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028










TQE = 0.130

0.005 0.015 0.025 0.035 0.045 0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.002 0.008 0.012 0.002 0.008 0.0120.018

0.022 0.0280.01 0.03 0.05 0.07 0.09













Lumbarnerve plexus






A

D E F J K L M

B C G H I

0.002 0.008 0.012 0.005 0.015 0.025 0.035 0.005 0.015 0.0250.002 0.008 0.012

0.002 0.008 0.012

0.018 0.022

0.002 0.008 0.012 0.018 0.002 0.008 0.012 0.0180.022

0.028




Discussion

¨  Pigmentation protecting against bacterial infections in testicles (Franco-Belussi & Oliveira 2011) => increase fitness ¤ Clade Calamitophrynia

¨  Role of environment in dictating plasticity ¤ Temperature => increase pigmentation => species that

lives in areas with high temperature

¨  Physiological factors (e.g., disease, age, MSH hormone)

Conclusion

¨  Overall, the pigmentation of organs is a very conserved trait.

¨  Future studies ¤ Histological analyses ¤  Experiments manipulating factors that influence melanin

production and melanocyte migration (Franco-Belussi et al. in press)

¨  More pictures in MorphoBank http://bit.ly/phylosignal

Shortcommings

¨  Dependency of the topology ¤  Incorporate Bayesian methods to account for

phylogenetic uncertainty

¨  Qualitative methods to describe pigmentation ¤  Incorporate quantitative methods

¨  Expand taxon sampling to other speciose anuran families

New directions

¨  Incorporate intraspecific variation

¨  How pigmentation varies along space

¨  Climatic and topographic correlates (Extended RLQ analysis) ¤ Which environmental factors influence the expression and

variation of pigmentation

Funding

Webite bit.ly/DiogoBrazil

Email: [email protected]

Thank you!!

Phylogenetic signal and diversity of visceral pigmentation in eight anuran families

Education

Transcript of Phylogenetic signal and diversity of visceral pigmentation in eight anuran families