The interplay of morphological adaptations and neuromuscular control

The complex behaviors of animals and their evolution arise from an interplay of morphological specialization and neuromuscular control mechanisms. Teasing apart the respective roles of these contributing factors can often be difficult, such as in the case of avian vocal behavior. Understanding the respective roles of morphology and neuromuscular control, however, is crucial for understanding how behavior is generated. Whereas the role of the central nervous system in avian vocal behavior has been studied extensively, the functional morphology of the vocal organ is poorly understood despite the well-described interspecific diversity in its skeletal and neuromuscular anatomy. Nevertheless, it is widely accepted that both neuromuscular and morphological specializations contribute to the production of avian vocalizations and their speciesspecific acoustic features. Here we investigate the link between avian vocal diversity and morphological diversity.

THE INTERPLAY OF MORPHOLOGICAL ADAPTATIONS AND NEUROMUSCULAR CONTROL by Sarah M. Garcia A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology The University of Utah December 2018 Copyright © Sarah M. Garcia 2018 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Sarah M. Garcia has been approved by the following supervisory committee members: Franz Goller , Chair Date Approved Frederick R. Adler , Member 10/26/18 Michael D. Shapiro , Member 10/26/18 Neil J. Vickers , Member 10/26/18 Gabriel B. Mindlin , Member Richard M. Clark , Dir. of Grad. Studies 10/26/18 M. Denise Dearing , Biol. Dept. Chair 10/26/18 Date Approved Date Approved Date Approved Date Approved Date Approved Date Approved and by M. Denise Dearing the Department/College/School of and by David B. Kieda, Dean of The Graduate School. , Chair/Dean of Biology ABSTRACT The complex behaviors of animals and their evolution arise from an interplay of morphological specialization and neuromuscular control mechanisms. Teasing apart the respective roles of these contributing factors can often be difficult, such as in the case of avian vocal behavior. Understanding the respective roles of morphology and neuromuscular control, however, is crucial for understanding how behavior is generated. Whereas the role of the central nervous system in avian vocal behavior has been studied extensively, the functional morphology of the vocal organ is poorly understood despite the well-described interspecific diversity in its skeletal and neuromuscular anatomy. Nevertheless, it is widely accepted that both neuromuscular and morphological specializations contribute to the production of avian vocalizations and their speciesspecific acoustic features. Here we investigate the link between avian vocal diversity and morphological diversity. To Keith, Kai, and Finnley. “There are more things in heaven and earth, Horatio, / Than are dreamt of in your philosophy.” –Hamlet, (1.5.167-8), Hamlet to Horatio TABLE OF CONTENTS ABSTRACT....................................................................................................................... iii LIST OF FIGURES ........................................................................................................... ix LIST OF TABLES ............................................................................................................. xi ACKNOWLEDGEMENTS .............................................................................................. xii Chapters 1 INTRODUCTION ........................................................................................................... 1 1.1 References ............................................................................................................ 7 2 EVOLUTION OF VOCAL DIVERSITY THROUGH MORPHOLOGICAL ADAPTATION WITHOUT VOCAL LEARNING OR COMPLEX NEURAL CONTROL........................................................................................................................ 13 2.1 Summary ............................................................................................................ 14 2.2 Results and discussion ....................................................................................... 14 2.3 Star ★ methods .................................................................................................. 18 2.4 Supplemental information.................................................................................. 19 2.5 Author contributions .......................................................................................... 19 2.6 Acknowledgments.............................................................................................. 19 2.7 References .......................................................................................................... 19 2.8 Star ★ methods .................................................................................................. 21 2.8.1 Key resources table ..................................................................... 21 2.8.2 Contact for reagent and resource sharing.................................... 21 2.8.3 Experimental model and subject details ..................................... 21 2.8.3.1 Individuals...................................................................... 21 2.8.3.2 Netting............................................................................ 21 2.8.3.3 Husbandry ...................................................................... 21 2.8.4 Method details............................................................................. 21 2.8.4.1 Distress calls before and after manipulations ................ 21 2.8.4.2 Nerve cut ........................................................................ 21 2.8.4.3 vMT rupture ................................................................... 22 2.8.4.4 vMT gluing .................................................................... 22 2.8.4.5 Observation and membrane block with a fiberscope .... 22 2.8.4.6 Mathematical modeling of the syrinx ............................ 22 2.8.5 Quantification and statistical analysis ......................................... 23 2.8.5.1 Frequency extraction ...................................................... 23 2.8.5.2 Quantification of duty cycle........................................... 23 2.8.5.3 Signal to noise ratio quantification ................................ 23 2.8.6 Data and software availability .................................................... 23 3 NEUROMUSCULAR CONTROL OF THE AVIAN VOCAL ORGAN PLAYS DIFFERENT ROLES IN VOCALIZATIONS OF SUBOSCINES AND OSCINES ...... 24 3.1 Introduction ........................................................................................................ 25 3.1.1 Vocal complexity and the syrinx ......................................................... 25 3.1.2 Syringeal morphology and function .................................................... 26 3.1.2.1 Membranes and labia .............................................................. 26 3.1.2.2 Muscles ................................................................................... 27 3.1.3 Neural control of the oscine and suboscine syrinx .............................. 28 3.1.4 Neuromuscular control of vocalizations .............................................. 29 3.1.5 Distress calls ........................................................................................ 30 3.2 Methods.............................................................................................................. 31 3.2.1 Individuals and field sites .................................................................... 31 3.2.2 Collection, husbandry, and audio recording ........................................ 31 3.2.2.1 Short-term individuals ........................................................... 31 3.2.2.2 Long-term individuals............................................................ 32 3.2.3 Surgical transection of nerves ............................................................. 33 3.2.3.1 Short-term individuals ........................................................... 33 3.2.3.2 Long-term individuals............................................................ 33 3.2.4 Quantification in Praat and statistical analysis .................................... 34 3.2.4.1 Frequency extraction and histogram generation .................... 34 3.2.4.2 Duty cycle calculation............................................................ 34 3.2.4.3 Signal intensity (Signal to Noise Ratio [SNR]) ..................... 35 3.3 Results ................................................................................................................ 35 3.3.1 Frequency distribution ......................................................................... 36 3.3.2 Spectral quality .................................................................................... 37 3.3.3 Intensity (volume) ............................................................................... 37 3.4 Discussion .......................................................................................................... 38 3.4.1 Acoustic parameter assessment using distress calls ............................ 38 3.4.2 Effects of nerve cut on frequency range .............................................. 39 3.4.3 Tyrannid neural control as an outlier................................................... 39 3.4.4 Effects of nerve cut on intensity .......................................................... 41 3.4.5 Effects of nerve cut on duty cycle ....................................................... 42 3.4.6 Role of neuromuscular control and hypotheses................................... 43 3.4.7 Neuromuscular control and histological composition of oscillating tissue ............................................................................................................. 44 3.4.8 Degree of coupling of neuromuscular control and vocal learning behavior ........................................................................................................ 44 vii 3.5 References .......................................................................................................... 45 4 SUBOSCINE VOCAL MEMBRANES CONTAIN HISTOLOGICALLY DISTINCT LAYERS ........................................................................................................................... 58 4.1 Introduction ........................................................................................................ 59 4.1.1 Gross mechanics of the syrinx ............................................................. 60 4.1.2 Oscillatory behavior and acoustic features .......................................... 60 4.1.3 Passive influence and active control of oscillatory behavior .............. 61 4.1.4 Relevant extra-cellular matrix components ......................................... 61 4.1.5 Known histology in Passerines ............................................................ 62 4.2 Methods.............................................................................................................. 64 4.2.1 Individuals and collection at field sites ............................................... 64 4.2.2 Perfusion and tissue preparation .......................................................... 64 4.2.3 Tissue processing, staining, and imaging ............................................ 64 4.2.4 Tissue analysis ..................................................................................... 65 4.2.5 Frequency extraction ........................................................................... 65 4.2.6 Statistical analyses ............................................................................... 66 4.3 Results ................................................................................................................ 66 4.3.1 Number of layers ................................................................................. 67 4.3.2 Acoustic analysis ................................................................................. 67 4.3.3 Layers and frequency range relationship ............................................. 67 4.3.4 Qualitative layer assessment................................................................ 68 4.3.4.1 Cartilage .................................................................................. 68 4.3.4.2 General labia characteristics ................................................... 68 4.3.4.3 ECM components and layer trends ......................................... 69 4.4 Discussion .......................................................................................................... 69 4.4.1 Pipridae histological trends and possible implications ........................ 69 4.4.2 Left-syrinx/right-syrinx labia asymmetry............................................ 71 4.4.3 Tyrannid histology and syringeal neuromuscular control ................... 71 4.4.4 Histologically unique cartilages adjacent to vocal tissue .................... 72 4.4.5 Differential lateral and medial labia histological composition............ 73 4.4.6 Fundamental frequency range and layers ............................................ 73 4.5 References .......................................................................................................... 76 viii LIST OF FIGURES Figures 1.1 Components of the avian vocal system. ...................................................................... 11 1.2 Phylogeny of species used in Chapters 2-4................................................................. 12 2.1 Fiberscopic visualization of the F. rufus syrinx reveals sound sources. ..................... 15 2.2 Manipulation of the ventral MT reveals role in sound production and interactions with labial sound sources in F. rufus ........................................................................................ 16 2.3 Sound frequency shifts after manipulation of MT in tracheophones. ......................... 17 2.4 Manipulation of the ventral MT increases duty cycle and decreases sound amplitude in tracheophone species. ................................................................................................... 18 3.1 Syringeal neuromuscular control does not significantly impact acoustic parameters of suboscine species Myiarchus tyrannulus distress calls. .................................................... 53 3.2. Syringeal neuromuscular control significantly affects acoustic parameters of spontaneous vocalizations in oscine species Melospiza melodia. .................................... 54 3.3 Bilateral tracheosyringeal nerve cut significantly affects fundamental frequency distribution in passerellid oscines but has little (furnariids, piprids and thamnophilids) to no (tyrannids) effect in suboscines. .................................................................................. 55 3.4 Bilateral tracheosyringeal nerve cut significantly affects spectral quality in passerellid oscines but has little to no effect on spectral quality in suboscines. ................................. 56 3.5 Effects of bilateral tracheosyringeal nerve cut on intensity in suboscines and passerellid oscines............................................................................................................. 57 4.1 Air sac pressurization causes air to flow through the syrinx and may cause vocal tissue to oscillate, producing sound. ................................................................................. 83 4.2 Histological stains of Ceratopipra mentalis (piprid suboscine) syrinx reveals distinct layer structure.................................................................................................................... 84 4.3 Schematics of qualitative assessment of medial and lateral labia layer structure in select suboscines and passerellid oscines. ........................................................................ 85 4.4 Frequency range of vocalizations in select suboscines and passerellid oscines. ........ 86 4.S1 Syrinx histological section stains of Manacus vitellinus. ......................................... 87 4.S2 Syrinx histological section stains of Lepidothrix coronata. ..................................... 88 4.S3 Syrinx histological section stains of Cercomacra tyrannina.................................... 89 4.S4 Syrinx histological section stains of Furnarius rufus. .............................................. 90 4.S5 Syrinx histological section stains of Tyrannus verticalis. ........................................ 91 4.S6 Syrinx histological section stains of Conotopus sordidulus. .................................... 92 4.S7 Syrinx histological section stains of Spizella breweri. ............................................. 93 4.S8 Syrinx histological section stains of Melospiza melodia. ......................................... 94 4.S9 Syrinx histological section stains of Passerella iliaca. ............................................ 95 4.S10 Syrinx histological section stains of Junco hyemalis. ............................................ 96 4.S11 Syrinx histological section stains of Pipilo maculatus. .......................................... 97 4.S12 Syrinx histological section stains of Pipilo chlorurus. ........................................... 98 x LIST OF TABLES Tables 3.1 Study individuals. ....................................................................................................... 50 3.2 Significance of effects of nerve cut on duty cycle and intensity by species ............... 51 3.3 Average % change and standard error in intensity and duty cycle by family............. 52 4.1 Study individuals. ....................................................................................................... 80 4.2 Effective frequency range, vocal tissue layer count, and body mass of the species investigated. ...................................................................................................................... 81 4.3 Linear regression results of medial, lateral, and total layer counts and effective frequency range for suboscines and oscines. .................................................................... 82 ACKNOWLEDGEMENTS Many people and entities made this project possible. I would first like to thank my spouse, Keith Luscinski, for his unwavering support in all aspects of this long endeavor. I would like to thank my advisor, Franz Goller, for insight and assistance throughout this entire process. I would like to thank the rest of my committee for their feedback, participation, and support. I would like to thank my collaborators, Cecelia Kopuchian, Gabriel Mindlin, Matthew Fuxjager, and Pablo Tubaro for the ability to conduct research abroad and the difficulties entailed, and their feedback and support with publication of my second chapter. I would also like to thank Tobias Riede for his help and support. I am grateful to Natasha Verzhbitskiy for the husbandry of test subjects and general lab support. I would like to thank Jay Love, Lindsey Reader, and Amanda Hoepfner for feedback throughout the years. I would like to thank Richard Clark and Shannon Nielsen for their assistance in this process. I thank Andy Smith and Mark Holton for always inspiring, supporting, and guiding me. I also thank Allison Quirk, David Katz, and Emily DiBlasi for their support. Lastly, I would like to thank NSF and the University of Utah for general funding. I would like to thank STRI and the province of Corrientes, AR, for allowing my collaborators and me to conduct research. CHAPTER 1 INTRODUCTION The complex behaviors of animals and their evolution arise from an interplay of morphological specialization and neuromuscular control mechanisms. Teasing apart the respective roles of these contributing factors can often be difficult, such as in the case of avian vocal behavior. Understanding the respective roles of morphology and neuromuscular control, however, is crucial for understanding how behavior is generated. Whereas the role of the central nervous system in avian vocal behavior has been studied extensively (e.g., Nottebohm et al., 1976; Nottebohm, 1980; Bottjer et al., 1984; Bottjer et al., 1989; Sohrabji et al., 1990; Suthers and Goller, 1997; Shi et al., 2018), the functional morphology of the vocal organ is poorly understood despite the well-described interspecific diversity in its skeletal and neuromuscular anatomy (e.g., Müller, 1847; Ames, 1971). Nevertheless, it is widely accepted that both neuromuscular and morphological specializations contribute to the production of avian vocalizations and their species-specific acoustic features (Scharff and Nottebohm, 1991; Riede and Goller, 2010; Riede et al., 2010; Prince et al., 2011; Riede and Goller 2014). Vocalizations can vary in a number of acoustic features, such as fundamental frequency (F0), amplitude (volume), spectral composition, and temporal patterning. The range of these parameters and extent to which they are modulated determine the degree of complexity of a 2 vocalization. The unique combinations and modulations of these parameters provide species-specific vocalizations, and have allowed birds to fill a variety of acoustic niches. Vocal learning is the ability to imitate acquired sounds, and this rare behavior involving neural specializations in songbirds (oscines) is thought to have contributed, at least in part, to their successful and massive radiation as they comprise approximately half of all extant bird species (Vu et al., 1994; Vates et al., 1996; Ellers and Slabbekoorn, 2003; Mason et al., 2016). Vocal learning is different from vocal contextual learning (Janik and Slater, 2000), and species exhibiting this behavior require exposure to adult conspecific vocalizations during a critical period of ontogeny to properly develop their species-specific vocalizations (e.g., Thorpe, 1958, 1961; Nottebohm, 1969; Marler, 1970, 1991; Doupe and Kuhl, 1999; Wilbrecht and Nottebohm, 2003). In addition to songbirds, this behavior has been found in hummingbirds (Trochilidae; e.g., Baptista and Schuchmann, 1990), parrots (Psittaciformes; e.g., Durand et al., 1997), and beyond Aves, bats (e.g., Esser, 1994; Prat et al., 2015), cetaceans (Janik, 2014), pinnipeds (Reichmuth and Casey, 2014), elephants (Poole et al., 2005), and humans. In certain songbird species (e.g., zebra finch, Taeniopygia guttata; canary, Serinus canaria), the central neurological bases of this behavior have been investigated extensively (for review, see Nottebohm, 2005; Fitch and Jarvis, 2013). The behavior requires specific brain adaptations, and is initiated in specialized forebrain regions. Starting at the HVC (formerly known as the hyperstriatum ventrale, pars caudalis), the signal travels to the tracheosyringeal branch of the hypoglossal nerve by way of the robust nucleus of the arcopallium (RA). In addition to this motor pathway, additional dedicated circuitry (anterior forebrain pathway) controls vocal plasticity and includes Area X, the dorsolateral anterior thalamic nucleus 3 (DLM) and the lateral magnocellular nucleus of the nidopallium (LMAN). These forebrain and thalamic regions are dedicated to motor production and vocal learning and are absent in vocal non-learners (though see Liu et al., 2013). The signal originating from the forebrain travels ipsilaterally via the hypoglossal nerve to the vocal organ and its muscles (Fig. 1). In contrast, this process is controlled in the midbrain (dorsomedial nucleus, DM) in passerine vocal non-learners (suboscines). While suboscines arguably have less complex vocalizations, they are speciose and fill a wide variety of acoustic niches and do make up a large part of extant bird species (Raikow and Bledsoe, 2000). They accomplish this without vocal learning, posing the question of how this acoustic diversity is achieved. Suboscine taxa, unlike oscines, exhibit substantial morphological diversity of the syrinx, suggesting that these structural differences of their vocal organs underlie the observed vocal differences (Gaunt, 1983; Garcia et al., 2017). Detailed study of their vocal behavior will be required to test this hypothesis in detail. The avian vocal organ, the syrinx, sits deep within the body cavity of birds, just above the heart where the trachea bifurcates into the primary bronchi (Fig. 1.1). This stands in distinction with larynx-based sound production, where sound originates much higher in the vocal tract. During vocal production, the air sacs of the bird are pressurized and cause air to flow through the syrinx, which in turn draws the labia into the airstream. The airstream flowing past the labia causes them to passively oscillate (Nottebohm, 1980; Fee, 2002; Mindlin and Laje, 2006; Riede and Goller, 2010; Elemans et al., 2015), and these oscillations give rise to sound. The oscillatory behavior of the tissue is determined in large part by its intrinsic myoelastic properties (Riede and Goller, 2014). Previous studies suggest inter- and intraspecies differences in histological composition of the labia, 4 and this is likely a significant source of acoustic variation (Riede et al., 2010; Riede and Goller, 2010). In addition, as air flows through the syrinx and vocal tissue oscillates, muscles can exert force on various syringeal structures (e.g., cartilage adjacent to the labia), resulting in modulation of the vocal tissue oscillations. By altering the position of cartilage adjacent to the labia, the tension and length of the labia are changed, both of which determine the frequency at which the labia oscillate, and therefore the fundamental frequency of sound produced. Control of labial position in the airstream (gating) also affects acoustic parameters such as sound amplitude and thus constitutes one mechanism of generating amplitude modulation. The functional morphology of the syrinx therefore dictates the possible range of acoustic features produced by a given species. While these general sound production and modulation mechanisms are understood, the specific functional morphology for any species has not been worked out in sufficient detail. Across species, the syrinx shows significant morphological diversity (especially in suboscines; e.g., Ames, 1971), the function of which is mostly unknown. Detailed study of individual species will likely yield new insights into how morphology contributes to acoustic features. In Chapter 2 of this dissertation (Garcia et al., 2017), we therefore assess the function of tracheal membranes in tracheophone suboscines. We describe that their syrinx has three sound sources that interact to produce sound, as opposed to the two sound sources found in most birds. The interaction of these three sources enables production of specific acoustic features. Such experimental data on unique syringeal structures enhance our understanding of the relationship between morphological diversity and acoustic diversity in birds. In addition, the role of syringeal muscles is especially unclear in suboscines. In 5 Amador et al. (2008), transection of syringeal nerves yielded no appreciable change in spontaneous vocalization acoustic parameters despite activation of syringeal muscles during vocalization. In Chapter 3 of this dissertation, we investigate the role of neuromuscular control in other tyrannid suboscines, and this study yields similar results, suggesting a family-wide trait. The function of syringeal muscles in tyrannids is still unclear, though we suggest their activation plays an important role in ensuring syringeal patency. This striking family trend is absent in other suboscines, however. Lack of neuromuscular control in other suboscine species altered acoustic features of spontaneous vocalization, though to a lesser extent than in oscines. This is particularly important, as we can now begin to understand the role of syringeal muscles in suboscine passerines, and compare it to that of oscines. Finally, in Chapter 4 of this dissertation, we investigate the intrinsic myoelastic properties of the oscillating tissue within the syrinx. The presence of elastin, collagen, and hyaluronic acid are of particular importance, and these substances can form discrete layers within the vocal tissue. Such layers have been found in select oscine species and correlate with their fundamental frequency range (Riede and Goller, 2010, 2014). We therefore identify these vocal tissue components in studied species, and found that suboscines do also contain discrete layers, though on average fewer layers than oscines. Suboscines in general are understudied, though there is a general assumption that they do not exhibit vocal learning (though see Kroodsma et al., 2013). While there are few studies specifically investigating this, the suboscine/oscine divide nevertheless provides an ideal framework within which to study functional morphology of the syrinx. While the acoustic repertoires of oscines (vocal learners) must have physiological 6 constraints, vocal learning can confound interspecies comparisons of the link between vocal diversity and morphological diversity. For this reason, we investigated the link between acoustic diversity and functional morphological diversity focusing specifically on suboscines, and use select oscines for comparison. Both groups exhibit a degree of diversity in vocal parameters, and our guiding question is: Do they arrive at this interspecific vocal diversity via the same method? In the following studies, we show that 1) morphological diversity of the syrinx can give rise to vocal diversity even in species without vocal learning and highly complex neuromuscular control, 2) direct neural control of the syrinx is used by both vocal learners and vocal non-learners to achieve vocal diversity, and 3) both groups contain unique histological properties of their vocal tissue. Collectively, these data indicate that the only trait truly unique to songbirds is their specialized forebrain circuitry that enables vocal learning and highly complex neuromuscular function. All other mechanisms used by oscines to achieve vocal diversity are also exploited by suboscines (non-learners)(e.g., muscular control, differential histological composition of the oscillating tissue). Therefore, beyond vocal learning, it is not a matter of unique mechanisms, so much as degree of use and complexity (both of which have obviously allowed songbirds to achieve incredible levels of interspecies vocal diversity). This therefore suggests that, to achieve vocal diversity, suboscines may have relied more heavily on morphological specialization, while oscines relied on neural specialization. 7 1.1 References Amador, A., Goller, F. and Mindlin, G.B. (2008). Frequency modulation during song in a suboscine does not require vocal muscles. J. Neurophysiol. 99, 2383–2389. Ames, P.L. (1971). The morphology of the syrinx in passerine birds. Peabody Mus. Nat. Hist. Bull. 37, 1-194. Baptista, L.F. and Schuchmann, K.L. (1990). Song learning in the Anna hummingbird (Calypte anna). Ethology 84, 15-26. Bottjer, S.W., Miesner, E.A. and Arnold, A.P. (1984). Forebrain lesions disrupt development but not maintenance of song in passerine birds. Science 224, 901-903. Bottjer, S.W., Halsema, K.A., Brown, S.A. and Miesner, E.A. (1989). Axonal connections of a forebrain nucleus involved with vocal learning in zebra finches. J. Comp. Neurol. 279, 312-326. Doupe, A.J. and Kuhl, P.K. (1999). Birdsong and human speech: common themes and mechanisms. Annu. Rev. Neurosci. 22, 567-631. Durand, S.E., Heaton, J.T., Amateau, S.K. and Brauth, S.E. (1997). Vocal control pathways through the anterior forebrain of a parrot (Melopsittacus undulatus). J. Compar. Neurol. 377, 179-206. Elemans, C.P.H., Rasmussen, J.H., Herbst, C.T., Düring, D.N., Zollinger, S.A., Brumm, H., Srivasta, K., Svane, N., Ding, M., Larsen, O.N. et al. (2015). Universal mechanisms of sound production and control in birds and mammals. Nat. Commun. 6, 8978. Ellers, J. and Slabbekoorn, H. (2003). Song divergence and male dispersal among bird populations: a spatially explicit model testing the role of vocal learning. Animal Behav. 65, 671-681. Esser, K.H. (1994). Audio-vocal learning in a non-human mammal: the lesser spearnosed bat Phyllostomus discolor. Neuroreport 5, 1718-1720. Fee, M.S. (2002). Measurement of the linear and nonlinear mechanical properties of the oscine syrinx: implications for function. J. Comp. Phys. A 188, 829-839. Fitch, W.T. and Jarvis, E.D. (2013). Birdsong and other animals models for human speech, song, and vocal learning. In Language, music, and the brain (ed. M.A. Arbib), pp. 499–539. Cambridge: MIT Press. Garcia, S.M., Kopuchian, C., Mindlin, G.B., Fuxjager, M., Tubaro, P.L. and Goller, F. (2017). Evolution of vocal diversity through morphological adaptation without 8 vocal learning or complex neural control. Curr. Biol. 27, 2677-2683. Gaunt, A.S. (1983). An hypothesis concerning the relationship of syringeal structure to vocal abilities. Auk 100, 853-862. Janik, V.M. and Slater, P.J. (2000). The different roles of social learning in vocal communication. Animal Behav. 60, 1-11. Janik, V.M. (2014). Cetacean vocal learning and communication. Curr. Opin. Neurobiol. 28, 60-65. Kroodsma, D.E., Hamilton, D. Sánchez, J.E. Byers, B.E., Fandiño-Mariño, H., Stemple, D.W., Trainer, J.M. and Powell, G.V. (2013). Behavioral evidence for song learning in the suboscine bellbirds (Procnias). Wilson J. Ornithol. 125, 1-14. Liu, W.C., Wada, K., Jarvis, E.D. and Nottebohm, F. (2013). Rudimentary substrates for vocal learning in a suboscine. Nat. Commun. 4, 2082. Marler, P.A. (1970). Comparative approach to vocal learning: song development in white-crowned sparrows. J. Comp. Physiol. Psychol. 71, 1–25. Marler, P. (1991). Song-learning behavior: the interface with neuroethology. Trends Neurosci. 14, 199-206. Mason, N.A., Burns, K.J., Tobias, J.A., Claramunt, S., Seddon, N. and Derryberry, E.P. (2017). Song evolution, speciation, and vocal learning in passerine birds. Evolution 71, 786-796. Mindlin, G.B., and Laje, R. (2006) The Physics of Birdsong. Berlin: Springer. Müller, J. (1847). Über die bisher unbekannten typischenVerschiedenheiten der Stimmorgane der Passerinen. Berlin: Abhandlungen der Königlichen Akademie der Wissenschaften, 321–391pp. Nottebohm, F. (1969). The “critical period” for song learning. Ibis 111, 386-387. Nottebohm, F., Stokes, T.M. and Leonard, C.M. (1976). Central control of song in the canary, Serinus canarius. J. Comp. Neurol. 165, 457–486. Nottebohm, F. (1980). Brain pathways for vocal learning in birds. In Progress in psychobiology and physiological psychology (ed. J.M. Sprague, A.N Epstein), pp. 86-125. San Francisco: Academic Press. Nottebohm, F. (2005). The neural basis of birdsong. PLoS Biol. 3, e164. Poole, J.H., Tyack, P.L., Stoeger-Horwath, A.S. and Watwood, S. (2005). Animal 9 behaviour: elephants are capable of vocal learning. Nature 434, 455. Prat, Y., Taub, M. and Yovel, Y. (2015). Vocal learning in a social mammal: demonstrated by isolation and playback experiments in bats. Sci. Adv. 1, e1500019. Prince, B., Riede, T. and Goller, F. (2011). Sexual dimorphism and bilateral asymmetry of syrinx and vocal tract in the European starling (Sturnus vulgaris). J. Morphol. 272, 1527-1536. Raikow, R.J. and Bledsoe, A.H. (2000). Phylogeny and evolution of the passerine birds: independent methods of phylo-genetic analysis have produced a well-supported hypothesis of passerine phylogeny, one that has proved particularly useful in ecological and evolutionary studies. AIBS Bull. 50, 487-499. Reichmuth, C. and Casey, C. (2014). Vocal learning in seals, sea lions, and walruses. Curr. Opin. Neurobiol. 28, 66-71. Riede, T. and Goller, F. (2010). Functional morphology of the sound-generating labia in the syrinx of two songbirds species. J. Anatom. 216, 23-36. Riede, T., Fisher, J.H. and Goller, F. (2010). Sexual dimorphism of the zebra finch syrinx indicates adaptation for high fundamental frequencies in males. PLoS One 5, e11368. Riede, T. and Goller, F. (2014). Morphological basis for the evolution of acoustic diversity in oscine songbirds. Proc. Roy. Soc. B 2281, 20132306. Scharff, C. and Nottebohm, F. (1991). A comparative study of the behavioral deficits following lesions of various parts of the zebra finch song system: implications for vocal learning. J. Neurosci. 11, 2896-2913. Shi, Z., Piccus, Z., Zhang, X., Yang, H., Jarrell, H., Ding, Y., Teng, Z., Tchernichovski, O. and Li, X. (2018). miR-9 regulates basal ganglia-dependent developmental vocal learning and adult vocal performance in songbirds. eLife 7, e29087. Sohrabji, F., Nordeen, E.J. and Nordeen, K.W. (1990). Selective impairment of song learning following lesions of a forebrain nucleus in the juvenile zebra finch. Behav. Neural Biol. 53, 51-63. Suh, A., Paus, M., Kiefmann, M., Churakov, G., Franke, F.A., Brosius, J., Kriefs, J.O. and Schmitz, J. (2011). Mesozoic retroposons reveal parrots as the closest living relatives of passerine birds. Nat. Commun. 2, 443. Suthers, R.A. and Goller, F. (1997). Motor correlates of vocal diversity in songbirds. In Current Ornithology, (ed. J. Nolan), pp. 235-288). New York: Plenum Press. 10 Thorpe, W.H. (1958). The learning of song patterns by birds, with especial reference to the song of the chaffinch Fringilla coelebs. Ibis 100, 535-570. Thorpe, W.H. (1961). Bird-Song: The biology of vocal communication and expression in birds. New York: Cambridge Univ. Press. Vates, G.E., Broome, B.M., Mello, C.V. and Nottebohm, F. (1996). Auditory pathways of caudal telencephalon and their relation to the song system of adult male zebra finches (Taeniopygia guttata). J. Comp. Neurol. 366, 613–642. Vu, E.T., Mazurek, M.E. and Kuo, Y.C. (1994). Identification of a forebrain motor programming network for the learned song of zebra finches. J. Neurosci. 14, 6924– 6934. Wilbrecht, L. and Nottebohm, F. (2003). Vocal learning in birds and humans. Ment. Retard. Dev. Disabil. Res. Rev. 9, 135-148. 11 Figure 1.1 Components of the avian vocal system. Signals to produce vocalizations are initiated in the forebreain (oscines) or midbrain (suboscines), and travel ipsilaterally by the tracheosyringeal nerve of the hypoglossal branch to the syringeal muscles. These muscles attach to cartilages adjacent to the medial and lateral labia, and may rotate and otherwise alter the position of the cartilate. This in turn alters the tension of the labia. Pressurization of the airsacs by abdominal muscl activation induces airflow out of the vocal tract, pushes the labia into the airspace of the bronchi, and passively cause the labia to oscilllage. These oscillations alter the flow of air, which can be perceived as sound. As the sound exits the body caviy, it is shaped by the upper vocal tract (beak, oropharyngeal esophageal cavity [OEC], glottis, and trachea). 12 Acanthisittidae (NZ wrens) Passeriformes suboscine Thamnophilidae Furnariidae F. rufus C. cinnamomeus L. angustirostris P. ruber S. rufosuperciliata Pipridae M. vitellinus L. coronata C. mentalis Tityridae P. validus Tyrannidae oscine T. caerulescens C. tyrannina Passerellidae T. verticalis C. sordidulus E. mesoleuca E. varius H. margaritaceiventer P. sulphuratus M. rixosa P. pitangua M. tyrannulus P. murina S. breweri M. melodia P. iliaca J. hyemalis P. maculatus P. chlorurus Figure 1.2, Phylogeny of species used in Chapters 2-4. Chapters 2-4 describe results from 21 species spanning 5 suboscine families and 6 species within Passerellidae. 13 CHAPTER 2 EVOLUTION OF VOCAL DIVERSITY THROUGH MORPHOLOGICAL ADAPTATION WITHOUT VOCAL LEARNING OR COMPLEX NEURAL CONTROL Reprinted with permission from Garcia, S.M., Kopuchian, C., Mindlin, G.B., Fuxjager, M.J., Tubaro, P.L. and Goller, F. (2017). Evolution of vocal diversity through morphological adaptation without vocal learning or complex neural control. Current Biology, 27(17), 2677-2683. September 11, 2017 © 2017 Elsevier Ltd. 14 15 16 17 18 19 20 21 22 23 24 CHAPTER 3 NEUROMUSCULAR CONTROL OF THE AVIAN VOCAL ORGAN PLAYS DIFFERENT ROLES IN VOCALIZATIONS OF SUBOSCINES AND OSCINES Avian vocalizations are highly complex and exhibit a wide breadth of diversity among species. These sounds are produced by the syrinx, a vocal organ unique to birds, which is controlled by syringeal muscles innervated by the tracheosyringeal branch of the hypoglossal nerve. The role of neural control of the syrinx in sound production has been studied in some oscine species (e.g., Taeniopygia guttata, Cardinalis cardinalis, Serinus canaria, Toxostoma rufum), but its role in suboscine species is relatively understudied and limited to two species of Tyrannidae. Here we present data on the effects of eliminating neural control of the syringeal muscles on vocalizations of 20 species spanning 5 suboscine families (Tyrannidae, Furnariidae, Thamnophilidae, Pipridae, and Tityridae) and 1 oscine family (Passerellidae). Coupled with already published data in oscines, these findings indicate a striking difference in the role of neural control in generating vocal features between suboscines and oscines. Using surgical resection of the syringeal nerves, we measured acoustic parameters of vocalizations before and after treatment. Oscines had significant reduction in intensity (>55%) after treatment while suboscines did not (<27%). Consistent with other published data, Tyrannidae species 25 showed no change in frequency range, and oscine species showed significant reduction. The frequency range of some suboscines did shift (Furnariidae, Thamnophilidae, Pipridae), but these shifts were relatively small compared to the shifts seen in Passerellidae. Including the data presented in this chapter, a total of 10 species of Tyrannidae have now been investigated, and none show any appreciable change in measured acoustic parameters, strongly suggesting that syringeal muscle activity is not required for acoustic features. These data also indicate a variable role of neural control of vibratory dynamics of the labia and thus spectral quality of sound (here measured as duty cycle). Collectively, these results strongly suggest that neural control of syringeal muscles plays a much smaller role in vocal production in suboscines than oscines. 3.1 Introduction The avian vocal organ, the syrinx, generates the vast array of vocalizations produced by birds, ranging from simple to highly complex songs. Vocalizations are produced in a variety of contexts, such as mating, maintaining territory, parenting, and predatory heterospecific interactions. While the specific vocal behaviors of species vary, the majority of birds produce vocalizations within these common contexts (reviewed in Thielcke, 1970b). This study addresses to what extent these vocalizations rely on neural control of the muscles of the syrinx. 3.1.1 Vocal complexity and the syrinx The degree of complexity of a vocalization can be described in terms of temporal patterning, fundamental frequency (F0) range, F0 modulation, harmonic emphasis and 26 spectral quality, amplitude range, and amplitude modulation; and different combinations thereof produce distinct types/series of vocalizations (e.g., song types). We assume the widely implied framework that oscine vocalizations are more complex than suboscines (e.g., Raikow, 2000; Ríos-Chelén et al., 2012). While there may be outliers, it is generally clear that oscine song is more complex (e.g., large F0 range, larger song repertoire, variable temporal patterning) than suboscine song. We therefore do not assess degree of complexity between suboscines and oscines, but focus on the comparison of acoustic effects of neural control of the vocal organ in these two groups. For example, we would not compare the rates of amplitude modulation between the suboscine Furnarius rufus and the oscine Junco hyemalis. Instead we would assess whether or not amplitude modulation occurs before and after nerve transection in each species, and when applicable, to what degree does its modulation change. Here, we investigate the degree to which neural control of the syrinx in select suboscines and oscines plays a role in production of acoustic parameters. 3.1.2 Syringeal morphology and function 3.1.2.1 Membranes and labia The morphology of the avian vocal organ, the syrinx, varies between species, and it is unclear to what degree this morphological diversity relates to vocal diversity (Warner, 1972; Gaunt, 1983). Unlike the human larynx, the syrinx sits deep within the body cavity of birds, and is located at the tracheal-bronchial juncture. In many species, including Passerines, it comprises two different sets of oscillating tissue, generally called labia, in each bronchus. Some species have other oscillating tissues, such as tracheal 27 membranes. When referring to these structures collectively, we use the term oscillating tissues. 3.1.2.2 Muscles Various sets of intrinsic and extrinsic muscles control the syrinx and were identified over a century ago (e.g., Müller, 1847; Myers, 1917; Ames, 1971; Elemans et al., 2008; Christensen et al., 2017). The number of syringeal muscles, their sizes and insertion differ across species (e.g., Ames, 1971). However, oscine gross syringeal morphology displays less interspecific variation than that seen in suboscines (Warner, 1972). These muscles exert forces on syringeal structures and thereby facilitate production of different frequencies and temporal patterns (Gaunt and Gaunt, 1977; Goller and Larsen, 1997; Düring et al., 2017; Döppler et al., 2017). Vocal membranes oscillate as air flows past them, and their degree of tension, as well as intrinsic morphological properties, will determine the way in which they vibrate and therefore the frequency of sound they produce (Mindlin and Laje, 2006). Activity of syringeal muscles can change the tension of these membranes and therefore the frequencies they produce (Goller and Suthers, 1996a; Goller and Larsen, 1997; Riede et al., 2010; for review see Goller and Riede, 2013). By rapidly closing and opening (gating) the bronchial lumen with the labial valve, these muscles can facilitate the production of intensity modulation (Goller and Suthers, 1996b). These muscles also play a critical role in maintaining open airways during quiet breathing, as they are activated during the expiratory phase of each breath (e.g., Vicario, 1991; Goller and Suthers 1996a,b). 28 3.1.3 Neural control of the oscine and suboscine syrinx Passerines, which make up over half of all bird species, comprise mainly two sister groups, oscines and suboscines. Oscines are vocal learners and therefore require exposure to adult conspecific song to properly develop their own song (Thorpe, 1961; Marler, 1970), while suboscines are vocal non-learners (Kroodsma and Konishi, 1991; Gahr, 2000; Liu et al., 2013). Published studies detail the neurophysiological differences specific to vocal learners, mainly the presence of specialized forebrain nuclei (e.g., Vu et al., 1994; Vates et al., 1996). In both suboscines and oscines, the tracheosyringeal branches of the hypoglossal nerve run down the trachea and innervate the ipsilateral sets of syringeal muscles. Individual syringeal muscles map onto areas in the hypoglossal nucleus and to some degree even into specific areas of the motor cortical song control nucleus (robust arcopallial nucleus; Vicario, 1991). There are no other known nerves innervating the syrinx (Goller and Suthers, 1996a). Consistent across studied species, which have been almost exclusively oscines (e.g., chaffinches, Fringilla coelebs, Nottebohm, 1971; canary, Serinus canaria, and white-crowned sparrow, Zonotrichia leucophrys, Nottebohm and Nottebohm, 1976; canary, Hartley and Suthers, 1990; zebra finch, Vicario, 1991; zebra finch, Taeniopygia guttata, and canaries, Vicario, 1995; Suthers and Zollinger, 2004; though see Nottebohm, 1976), severing the tracheosyringeal nerves drastically reduces frequency range of vocalizations produced. The exception, however, is the great kiskadee, Pitangus sulphuratus (Amador et al., 2008), the only suboscine species investigated (though see Garcia et al., 2017), and a member of the Tyrannidae family. In this species, there was no appreciable change in vocalizations after nerve 29 transection. Here we investigate whether these results from the great kiskadee study are unique, or if other tyrannids and other suboscines share this trait. 3.1.4 Neuromuscular control of vocalizations Data currently available on the neuromuscular control of frequency, intensity, and spectral quality are from mostly oscine species, and some non-passerines (e.g., chicken, Gallus gallus; duck, Anas platyrhynchos; pigeon, Columba liva). Oscine studies generally indicate that acoustic features of song are largely controlled and modulated by the syringeal muscles, which specifically can abduct and adduct the labia and modulate their tension (Goller and Suthers 1996a,b; for a review of specific muscles and their function see Suthers et al., 1999). Gating action may facilitate induction of oscillations at lower phonation threshold pressures (e.g., Goller and Suthers, 1996b; Düring et al., 2017). Frequency control through syringeal muscles enhances the achieved frequency range of vocalizations by decoupling it from pressure modulation driven adjustments. All of these functions act on the labia, which can also have morphological specialization that gives rise to acoustic diversity (Riede and Goller, 2014). Depending on the species and the type of vocalization, activation of syringeal muscles may vary substantially (e.g., Vicario, 1991; Goller and Suthers, 1996a,b; Goller and Cooper, 2004; Mendez et al., 2010). Furthermore, different vibratory regimes of the labia (pulse tone or modal vibration) may contribute to broader ranges of acoustic parameters (e.g., Jensen et al., 2007; Sitt et al., 2008). Because acoustic features arise from a combination of all these different sources, assessing the specific role of neural control may be quite difficult in a species. 30 3.1.5 Distress calls To enable investigation of a variety of species in the field, we mainly used distress calls to assess the effects of neural control on acoustic parameters (for certain species, we were able to record long-term repertoires). Individuals produce distress calls when they are physically restrained (Perrone, 1980), such as by a predator, in a mist net, or when held by a human. It is thought that these calls either elicit release by the predator restraining the individual, or attract attention from other conspecifics and/or heterospecifics, and possibly secondary predators (Stefanski and Falls, 1972; Conover, 1994). Distress calls are similar across taxa (Jurisevic and Sanderson, 1998), and are characterized by high intensity, minimal frequency modulation, broad and complex spectra, and rich harmonics, making them harsher in sound and more easily localized, though the exact acoustic signature varies among species (Stefnaski and Falls, 1972; Jurisevic and Sanderson, 1998). Data suggest that the frequency of distress calls across taxa is negatively related to body size (Jurisevic and Sanderson, 1998; Martin et al., 2011). While current data suggest neural control of syringeal muscles plays a role in oscine vocal production, and is less important in suboscines, such a conclusion requires more extensive sampling among suboscine species. To compare the role of neural control within the suboscine/oscine framework, we investigated the effects of syringeal denervation on vocalizations in 17 suboscine species representing 5 different families (Tyrannidae, Furnariidae, Pipridae, Tityridae, and Thamnophilidae) and 3 oscine species (Passerellidae). To quantify the effects of nerve cut on vocalizations, we measured frequency, intensity, and spectral quality (here measured as duty cycle) before and after 31 treatment (Garcia et al., 2017). Oscines exhibited the greatest change post treatment for all of these measurements; they had drastically reduced frequency range and intensity (volume) of vocalizations, and largely decreased duty cycles. The non-tyrannid suboscines exhibited a smaller shift to lower frequencies and intensities and varying changes in duty cycle than found in oscines, whereas there was no clear change in frequency and intensity in tyrannid species. 3.2 Methods 3.2.1 Individuals and field sites Individuals (see Table 3.1) were captured with mist nets in November of 2014 and 2015 at the Estación Biológica de Corrientes (EBCo) in Corrientes, Argentina (27,55095o S 58,68441o W), under permits issued by the Fauna Province Direction and the Provincial Natural Reserve Area; March of 2015 in Gamboa, Panama, with permission from the Autoridad Nacional del Ambiente and the Smithsonian Tropical Research Institute (STRI); and March-September of 2013-2015 in Salt Lake City, Utah, with permission from the U.S. Fish and Wildlife Service. Experimental procedures were approved (17-11- 2014, 20-10-2015, Corrientes; SE/A-6-15, Gamboa). All experiments were conducted at the respective field sites. Individuals are listed in Table 3.1. 3.2.2 Collection, husbandry, and audio recording 3.2.2.1 Short-term individuals Field sites did not enable the year-long captivity and recording required for longterm studies. To enable neuromuscular control assessment of field-site individuals, a 32 short-term procedure was used. This allowed the study to include a diverse set of suboscine species. Individuals (Table 3.1) were processed within 3 hours of capture, during which time they were not provided food or water to prevent emesis, and were kept indoors in a climate controlled environment. Birds were taken out of the cloth bag and held during audio recording of distress calls, which were emitted spontaneously. Vocalizations were recorded with an Audiotechnica AT 8356 microphone and Marantz (PMD 660) recorder while birds were held ~30 cm away from the microphone. 3.2.2.2 Long-term individuals Long-term studies were conducted on oscine individuals (S. breweri, M. melodia, J. hyemalis) and one suboscine individual (T. verticalis) to test neuromuscular control effects on larger repertoires of an individual. Individuals were captured throughout the breeding season and the fall in Utah, using mist nets, and placed in small cloth bags to minimize stress. Individuals were returned to the lab within 3 hours. They were placed in individual cages with food and water. Food consisted of mealworms, crickets, moths, mixed seed, shredded apple, carrot, and broccoli, which was replaced daily. Individuals were placed together in either an isolated room maintained at 27 ̊C and on a light/dark cycle set to reflect outdoor conditions, or were kept in a multipurpose room with natural lighting. Individuals were maintained in captivity through the winter and into the next breeding season during which time their vocalizations were recorded. To maximize audio quality, sound-insulating foam was placed around individual cages, and directional microphones were stationed in front of cages (~20 cm away from the perch). Microphones (Audiotechnica—AT 8356) were connected to multichannel amplifiers and 33 recorded vocalizations using AviSoft. Vocalizations were recorded before and after treatment. 3.2.3 Surgical transection of nerves 3.2.3.1 Short-term individuals After recording spontaneous distress calls, we performed a bilateral tracheosyringeal nerve cut. This procedure took no more than ~3 minutes. Topical Cetacaine gel was applied to the skin in the neck region just over the trachea. After allowing a minute for the anesthetic to work, an incision was made to expose a small section of the trachea. The tracheosyringeal nerves (both left and right) were dissected of connective tissue and a 1-2 mm segment of nerve was removed, ensuring a thorough cut. We closed the small incision with tissue glue (Vetbond) and then repeated the recording of spontaneous distress calls. 3.2.3.2 Long-term individuals Food and water were removed at least 1 hour prior to surgery to prevent emesis. Individuals were anesthetized with a gaseous mixture of isoflurane and oxygen. An incision of variable size (based on bird size) was made over the trachea, which was then visualized. Left and right tracheosyringeal nerves were dissected of connective tissue and a 1-2 mm segment of the nerve was removed, ensuring a thorough cut. After closing the incision with tissue adhesive, individuals were placed back in their cage to recover from anesthesia under observation. Water and food were replaced after the individual was awake and perching. 34 3.2.4 Quantification in Praat and statistical analysis 3.2.4.1 Frequency extraction and histogram generation Fundamental frequency was extracted at ~5 ms intervals using the pitch listing function in Praat (sound analysis application, Boersma and Weenink, Mac version 6.0.16). If intensity modulation was present, both the modulation frequency and the carrier frequency were extracted. These extracted frequencies were placed in 100 Hz bins, and bins were then normalized relative to the bin with the most occurrences. We used the Kolmogorov-Smirnov test to see whether distributions of fundamental frequency measurements differed statistically before and after manipulation. 3.2.4.2 Duty cycle calculation The duty cycle was measured as the ratio of the pulse duration (time passed from the initial peak to the first trough of the waveform) to the period of the fundamental frequency of the vocalization (Garcia et al., 2017). Therefore, a purely sinusoidal waveform has a duty cycle of 0.5. The duty cycle was calculated using 20 cycles from 13 representative high-quality vocalization(s) per individual before and after nerve cut. Cycles were generally measured every 1-5 ms per call depending on call quality and duty cycle length. We used an unpaired t-test to test the statistical significance of the change in duty cycle after nerve cut. Percent change in duty was calculated as (duty cycle after treatment - duty cycle before treatment)/(duty cycle before treatment*100). Family averages were calculated. 35 3.2.4.3 Signal intensity (Signal to Noise Ratio [SNR]) Signal and noise level were extracted from the output voltage of the microphone using the intensity feature in Praat. Specifically in short-term recordings, the gain on the Marantz recorder was adjusted between individuals and individual distress calls to accommodate field conditions and optimize recording quality, but it was held constant during an individual distress call. To account for different recording levels, we determined the intensity of the vocalization as the amount of signal above the noise level. The noise levels in each recording room were relatively constant and were measured just prior to the vocalization. The signal was then extracted every 2.5 milliseconds over the duration of the vocalization. The number of data points used to calculate the overall SNR therefore depended on how many times the individual vocalized. We used an unpaired ttest to test the statistical significance of the change in intensity after nerve cut, the results of which can be found in Table 3.2. We list average percent change in intensity by family (Table 3.3). For a given species, this was calculated as (intensity after treatment intensity before treatment - 1)/(intensity before treatment - 1). We subtracted 1 from each measurement since a value of 1 represents direct background noise. 3.3 Results As expected, passerellid oscine spontaneous vocalizations show drastic changes in acoustic features, and distress calls in suboscines showed relatively minimal changes. Similar to results seen in Amador et al. (2008), tyrannid suboscines showed no measurable changes in the vocalizations investigated (Fig. 3.1). The changes seen in passerellid species are consistent with all other oscine data on the role of neuromuscular 36 control in vocalizations (e.g., Suthers and Zollinger, 2004). The data presented here include non-tyrannid suboscines, and these species do show some changes in acoustic parameters, but they are not as substantial as those changes seen in oscines (see summary in Figs. 3.2-3.5). 3.3.1 Frequency distribution There is some variation in the changes to the frequency distribution of distress calls (and spontaneous vocalizations in the long-term study of T. verticalis) post nerve cut within suboscines. The degree to which these distress calls represent the frequency range found in full repertoires of these species varies; in some species, distress calls span similar ranges of more complete repertoires (see Fig. 3 of Garcia et al., 2017). The frequency range of spontaneous vocalizations in long-term studied oscines was clearly drastically reduced (Fig. 3). K-S values for all species are less than 0.01 and therefore, the distributions before and after treatment are significantly different. Some suboscine species exhibited a shift to lower frequencies (e.g., L. coronata, C. tyrannina, Fig. 3), or a loss of the upper frequency range (e.g., F. rufus, Fig. 3). While these shifts and losses seem small compared to the frequency distribution changes in oscines, the changes in suboscines (families Furnariidae, Thamnophilidae, Pipridae, Tityridae) are nevertheless significant (Table 3.2). In both Pipridae and Thamnophilidae, one species exhibited a shift to lower frequencies while the frequency distribution did not change in the other species. In contrast, the change in frequency range of tyrannids (e.g., M. tyrannulus, Fig. 3) was strikingly smaller. 37 3.3.2 Spectral quality We have represented the spectral quality of sounds as duty cycle (see Garcia et al., 2017 for further description of duty cycle). Nerve cut had a strong effect on passerellid duty cycle of spontaneous vocalizations recorded in long-term studies. Prior to treatment, the time waveform of these individuals was nearly sinusoidal, and post treatment decreased dramatically in duty cycle, resulting in a mean decrease of 69% (Table 3.3). The degree to which distress calls of short-term studies are sinusoidal (tonal) vary. For example, E. varius has a nearly tonal distress call (similar to passerellids), and post nerve cut, this call decreases minimally in tonality. All other short-term distress calls, including more sinusoidal ones, showed little effect post nerve cut (Table 3.3, Fig. 3.4). 3.3.3 Intensity (volume) The range of mean % change of intensity in suboscine families is -7.25 to 26.59% (Table 3.3), while it is -55.21% in Passerellidae. Compared to passerellids, intensity did not change substantially in suboscines (Fig. 3.5). Interestingly, furnariids and the one tityrid species investigated showed an increase in the mean intensity change after nerve cut. Within individual species of most suboscine families, there is some variation in response to nerve cut; however, these changes (both increases and decreases) are much smaller than the changes seen in passerellid species. 38 3.4 Discussion These data strongly indicate a differential role of neuromuscular control of the syrinx in suboscines and oscines. From the various measured acoustic parameters before and after treatment, it is clear that without neuromuscular control of the syrinx, oscines are not able to generate appreciable frequency modulation or intensity (volume), and are unable to produce sinusoidal (less spectrally complex) vocalizations. These parameters do not change in distress calls of tyrannids post nerve transection, and change minimally in other suboscines. 3.4.1 Acoustic parameter assessment using distress calls It is important to note that the majority of the recordings used in suboscine species (all but T. verticalis) were from distress calls, a single call type. These distress calls may exhibit different acoustic characteristics than those seen in other call types, limiting the applicability of our data. As stated in section 3.1.5, distress calls are produced when the individual is being attacked or physically restrained, and are characteristically harsh (complex, broad spectra) (Fig. 3.1). This likely varies from other vocalizations of the repertoire of a species, which could be more sinusoidal in their time-waveform and exhibit more frequency modulation (Fig. 3.2). However, from the long-term study of T. verticalis, we can assess the effect of bilateral nerve transection on other vocalizations of the repertoire, and the results fit within data from the distress calls of other species in tyrannids (Fig. 3.3). These long-term recordings show no appreciable change in frequency range, a result consistent across tyrannids (Figs. 3.1, 3.3). Since tyrannids generally have a small overall frequency range, short-term data are representative of the 39 effect on frequency range for the rest of the repertoire. As stated in section 3.3.2, duty cycle (tonality) of distress calls varies by species, and some distress calls (E. varius) are comparably tonal to that of passerellids prior to nerve cut and still show relatively minimal changes post nerve cut. 3.4.2 Effects of nerve cut on frequency range The loss of upper frequency range and shift to lower frequencies in some suboscines investigated is consistent with published data on the effects of nerve cut in oscine species (though the reduction is much less prominent). There is, however, clearly a spectrum for the role of neural control in acoustic parameters. It is interesting that fundamental frequencies in P. ruber shifted up after the nerve cut (Fig. 3.3). One possibility is that the loss of muscle tone resulting from the bilateral nerve cut altered the way in which the tracheal membranes and bronchial labia interact in this tracheophone syrinx (Garcia et al., 2017). However, this effect did not occur in other tracheophone species (Fig. 3.3). Indeed, F. rufus loses its upper frequency component post nerve cut. These seemingly contradictory data may merely be indicative of the variation in tracheal/bronchial membrane interactions within tracheophones (Garcia et al., 2017). 3.4.3 Tyrannid neural control as an outlier While consistent with published data, it is interesting that tyrannids display no measurable requirement of syringeal muscles since they 1) have prominent syringeal muscles (extrinsic and intrinsic; Ames 1971) and 2) EMG recordings from these muscles indicate they are active during vocalization (A. Amador, F. Goller and G.B. Mindlin, 40 unpublished observations/data; J. Peltier and F. Goller, unpublished observations/data). Since Amador et al. (2008) did not present data on distress calls of P. sulphuratus, they identified them as a possible call type requiring syringeal muscles and therefore the presence of such muscles. As distress calls are harsher and may rely on neural control differently from other less spectrally complex calls, they could require muscles whereas the calls recorded by Amador et al. (2008) did not. However, in our study, such distress calls were recorded before and after nerve cut in tyrannids, including P. sulphuratus, and the results were the same: there was no apparent effect on measured acoustic parameters post nerve cut. Together, these data strongly suggest that tyrannids do not explicitly rely on their syringeal muscles to generate specific frequencies, intensities (volume), or spectral features. EMG activity of syringeal muscles during vocal behavior suggests that these muscles play some role, perhaps in stabilizing the orientation of the syrinx, but tyrannids are able to produce “normal” vocalizations after denervation of the syringeal muscles. Alternatively, muscle activity could permit sound production at lower phonation threshold pressure levels than is possible without active control. Additionally, we have evidence that these muscles function in keeping the airways open during labored breathing. In an enclosed aviary, P. sulphuratus individuals with bilateral nerve cuts wheezed while flying (S.M. Garcia, C. Kopuchian and F. Goller, unpublished observations/data). Further studies are required to test these possibilities. Although we only looked at a small number of individuals in each species, the breadth of species does strengthen the conclusion that neural control plays a much larger role in the control of intensity, spectral complexity (duty cycle), and frequency range in oscines than in suboscines. It is important to note, however, that the suboscines do show a 41 change in these parameters. This means that the initial findings in P. sulphuratus (no effect of neural control on vocalizations; Amador et al., 2008) is a phenomenon likely restricted to tyrannids. Our data also support this, as acoustic parameters do not seem to change post nerve cut in the tyrannids we investigated whereas members of the other suboscine families show modest changes (relative to oscines) in duty cycle, intensity, and frequency range after syringeal denervation. 3.4.4 Effects of nerve cut on intensity As with other parameters, the change in intensity post nerve cut was largest in passerellids. This is consistent with previously published data on syringeal muscle activity during intensity control (e.g., Suthers, Goller, and Hartley, 1994). However, the increase in intensity post nerve cut in furnariids is unique, and occurred in 4 out of 5 species (Fig. 3.3). We know that intensity is drastically reduced in tracheophones when the tracheal membranes alone are disabled (Garcia et al. 2017), but it is unclear as to why intensity would increase post nerve cut. This could be caused by an alteration in tracheal/bronchial dynamics, and perhaps a shift to sound generation predominantly from the tracheal membranes, which would also explain the observed frequency shift away from higher frequencies. The effects on intensity in tyrannids are minimal relative to other studied species, and there is no overall consistent effect on intensity in this family (the average % change is extremely small). To better clarify these effects, a larger number of individuals per species is required in future work. 42 3.4.5 Effects of nerve cut on duty cycle Nerve-cut-induced changes in duty cycle alter the spectral composition of sounds. The long-term recordings of passerellids show drastic changes in duty cycle post nerve cut. The changes in distress calls of suboscines are relatively small. Such small changes are not likely due entirely to the fact that these are distress calls (see section 3.4.1 for further discussion on the use of distress calls). For example, the distress calls of E. varius are nearly as tonal as passerellids and yet they show little to no change in duty cycle post nerve cut. Conversely, the more complete spontaneous repertoire of T. verticalis is less tonal than E. varius prior to nerve cut, and also changes very little compared to passerellids. Duty cycle is a direct indication of labial dynamics. Decreases in duty cycle toward values smaller than 0.5 result in a less sinusoidal time waveform, and therefore pulse-like quality, which is expressed as an increase in upper harmonic content and generally less harmonic sounds. Since vocalizations with higher fundamental frequency tend to be less pulse-like, the drastic shift in duty cycle seen in oscines indicates that labial dynamics are shifting with the decrease in frequency after the nerve cut. These changes are all caused by changes in the production of sound, i.e., changes in labial vibration patterns. Shorter duty cycles indicate a longer closed-phase of the labia (Mindlin and Laje, 2005). The perceptual salience of these changes to the spectral composition of sounds is largely unknown. However, denervation of the syringeal muscles may not be the best method for testing this, as intensity and frequency range are affected as well. 43 3.4.6 Role of neuromuscular control and hypotheses The way in which syringeal muscles can effect change is determined by the functional gross morphology of the syrinx itself, including where and how the muscles insert, presence of varying cartilage structures, and perhaps most importantly, morphology of the sound generating tissue itself (Riede et al., 2010; Düring et al., 2013; Riede and Goller, 2014; Düring et al., 2017). By eliminating this control, birds are left with aerodynamic means (i.e., modulating air sac pressure) and passive mechanisms (i.e., labial composition) by which they can modulate acoustic features. In tyrannids, our and previous data (Amador et al., 2008; Garcia et al., 2017) strongly suggest that these nonmuscular mechanisms are sufficient for the modulation of acoustic parameters observed in their vocalizations. In other suboscines, a small degree of active modulation seems to be present. Neural control of the syrinx may also provide a passive effect on vocalizations by providing “muscle tone”. While no data are available, cutting of the tracheosyringeal nerves may alter the overall orientation and/or location of the syrinx. This may passively impact vocalizations by preventing optimal orientation, and could have an effect on vocal production. The presence of intrinsic syringeal musculature in Passeriformes suggests that these muscles play some role during vocal behavior. Our data suggest that in suboscines, this role is limited and may have been secondarily reduced in tyrannids (see also Suh et al., 2011). This role of syringeal muscles is interesting in light of the evolution of vocal learning, which either might have been ancestral to the last common ancestor of parrots and Passeriformes, or it might have evolved independently in parrots and oscines (Prum et al., 2015; and perhaps contingids, Saranathan et al., 2007, and Kroodsma et al., 2013). 44 It appears that extensive use of syringeal musculature for vocal behavior is paired with vocal learning (e.g., Heaton et al., 1995; Riede et al., 2010), and either evolutionary scenario requires a combined view of the evolution of these two major adaptations, neural reorganization for vocal learning and evolution of intrinsic musculature on the syrinx. 3.4.7 Neuromuscular control and histological composition of oscillating tissue In oscines, the histological composition and other morphology of the labia plays a critical role in the frequency range that can be generated (Riede and Goller, 2014). However, this broad range is the result of an interaction between neural control and labial morphology. For example, the drastic reduction in frequency range after denervation suggests that recruitment of the different layers of the labia into oscillation may either not be possible or is not as effective in broadening the frequency range (e.g., Riede et al., 2010). This differential reliance on active control of the sound generating labia in suboscines versus oscines should therefore be reflected in the histological composition of this tissue. To date, however, there are no data on the histology of labia in suboscines (but see Chapter 4). 3.4.8 Degree of coupling of neuromuscular control and vocal learning behavior The degree to which vocal learning and direct neural control of the syringeal muscles are coupled is complex. Neural control of syringeal muscles may affect vocalizations made by a species regardless of whether or not they are a vocal learner (these data). However, there does seem to be a link between vocal learning and degree of use of syringeal muscles since nerve cut in studied vocal learners results in significant 45 reduction of frequency (these data; Heaton et al., 1995; Riede et al. 2010). These data therefore indicate that neural control of syringeal muscles and vocal learning are, to an extent, decoupled; neural control is required for vocal learning, but vocal learning is not required for neural control, and the degree of neural control absent vocal learning varies by species. This study ultimately expands our knowledge of the role of neural control of the syrinx and its effects on acoustic parameters. These data clearly indicate that lack of acoustic changes post nerve transection is a predominantly tyrannid phenomenon. There are measurable changes in acoustic parameters in other suboscine families, though these changes are unremarkable compared to changes seen in oscines (in this study and others). Some degree of reliance on direct neural control of the syrinx for generating acoustic parameters may have therefore been ancestral and present before the suboscine/oscine split. The increased role of neuromuscular control of the syrinx now found in oscines likely contributed to their successful radiation. 3.5 References Amador, A., Goller, F. and Mindlin, G.B. (2008). Frequency modulation during song in a suboscine does not require vocal muscles. J. Neurophysiol. 99, 2383–2389. Christensen, L.A., Allred, L.M., Goller, F. and Meyers, R.A. (2017). Is sexual dimorphism in singing behavior related to syringeal muscle composition? Auk 134, 710-720. Döppler, J.F., Busch, A., Goller, F. and Mindlin, G.B. (2017). From electromyographic activity to frequency modulation in zebra finch song. J. Comp. Physiol. A 204, 209-217. Düring, D.N., Ziegler, A., Thompson, C.K., Ziegler, A., Faber, C., Muller, J. and Elemans, C.P.H. (2013). The songbird syrinx morphome: a three-dimensional, high resolution, interactive morphological map of the zebra finch vocal organ. BMC Biol. 46 11,1. Düring, D.N., Knörlein, B.J., Elemans, C.P.H., Schmidt, M., Goller, F., Riede, T. and Rome, L. (2017). In situ vocal fold properties and pitch prediction by dynamic actuation of the songbird syrinx. Sci. Rep. 7,11296. Elemans, C.P.H., Mead, A.F., Rome, L.C. and Goller, F. (2008). Superfast vocal muscles control song production in songbirds. PLoS ONE 3, e2581. Gahr, M. (2000). Neural song control system of hummingbirds: comparison to swifts, vocal learning (songbirds) and nonlearning (suboscines) passerines, and vocal learning (budgerigars) and nonlearning (dove, owl, gull, quail, chicken) nonpasserines. J. Comp. Neurol. 426, 182–196. Garcia, S.M., Kopuchian, C., Mindlin, G.B., Fuxjager, M., Tubaro, P.L. and Goller, F. (2017). Evolution of vocal diversity through morphological adaptation without vocal learning or complex neural control. Curr. Biol. 27, 2677-2683. Gaunt, A.S. (1983). An hypothesis concerning the relationship of syringeal structure to vocal abilities. Auk 100, 853-862. Gaunt, A.S. and Gaunt, S.L.L. (1977). Mechanics of the syrinx in Gallus gallus. II. Electromyographic studies of Ad Libitum vocalizations. J. Morphol. 152, 1–19. Goller, F. and Riede, T. (2013). Integrative physiology of fundamental frequency control in birds. J. Physiol. Paris 107, 230-242. Goller, F. and Larsen, O.N. (1997). A new mechanism of sound generation in songbirds. PNAS 94, 14787-14791. Goller, F. and Cooper, B.G. (2004). Peripheral motor dynamics of song production in the zebra finch. Ann. NY Acad. Sci. 1016, 130-152. Goller, F. and Suthers, R.A. (1996b). Role of syringeal muscles in gating airflow and sound production in singing brown thrashers. J. Neurophysiol. 75, 867– 876. Goller, F. and Suthers, R.A. (1996a). Role of syringeal muscles in controlling the phonology of bird song. J. Neurophysiol. 76, 287–300. Goller, F. and Larsen, O.N. (1997). In situ biomechanics of the syrinx and sound generation in pigeons. J. Exp. Biol. 200, 2165-2176. Hartley, R.S. and Suthers, R.A. (1990). Lateralization of syringeal function during song production in the canary. Development. Neurobiol. 21, 1236-1248. Heaton, J.T., Farabaugh, S.M. and Brauth, S.E. (1995). Effect of syringeal 47 denervation in the budgerigar (Melospittacus undulatus): the role of the syrinx in call production. Neurbiol. Learn. Mem. 64, 68-82. Jensen, K. K., Cooper, B. G., Larsen, O. N. and Goller, F. (2007). Songbirds use pulse tone register in two voices to generate low-frequency sound. Proc. R. Soc. Lond. B 274, 2703-2710. Kroodsma, D.E., Hamilton, D. Sachez, J.E. Byers, B.E., Fandino-Marino, H., Stemple, D.W., Trainer, J.M. and Powell, V.N. (2013). Behavioral evidence for song learning in the suboscine bellbirds (Procnias). Wilson J. Ornithol. 125, 1-14. Kroodsma, D.E. and Konishi, M.A. (1991). Suboscine bird (eastern phoebe, Sayornis phoebe) develops normal song without auditory feedback. Anim. Behav. 42, 477– 487. Liu, W.C., Wada, K., Jarvis, E.D. and Nottebohm, F. (2013). Rudimentary substrates for vocal learning in a suboscine. Nat. Commun. 4, 2082. Marler, P.A. (1970). Comparative approach to vocal learning: song development in white-crowned sparrows. J. Comp. Physiol. Psychol. 71, 1–25. Martin, J. P., Doucet, S. M., Knox, R. C. and Mennill, D. J. (2011). Body size correlates negatively with the frequency of distress calls and songs of Neotropical birds. J. Field Ornithol. 82, 259-268. Mendez, J.M., Dall’Asen, A.G. and Goller, F. (2010). Disrupting vagal feedback affects birdsong motor control. J. Exp. Biol. 213, 4193-4204. Mindlin, G.B. and Laje, R. (2005). The Physics of Birdsong. Berlin: Springer. Müller, J. (1847). Über die bisher unbekannten typischenVerschiedenheiten der Stimmorgane der Passerinen. Berlin: Abhandlungen der Königlichen Akademie der Wissenschaften, 321–391pp. Myers, J.A. (1917). Studies on the syrinx of Gallus domesticus J. Morphol. 29, 163-215. Nottebohm, F. (1972). Neural lateralization of vocal control in a passerine bird. II. Subsong, calls, and a theory of vocal learning. J. Exp. Zool. 179, 35-49. Nottebohm, F., Stokes, T.M. and Leonard, C.M. (1976). Central control of song in the canary, Serinus canarius. J. Comp. Neurol. 165, 457–486. Prum, R.O., Berv, J.S., Dornburg, A., Field, D.J., Townsend, J.P., Lemmon, E.M. and Lemmon, A.R. (2015). A comprehensive phylogeny of birds (Aves) using targeted next-generation DNA sequencing. Nature, 526, 569. 48 Riede, T. and Goller, F. (2010). Functional morphology of the sound-generating labia in the syrinx of two songbirds species. J. Anatom. 216, 23-36. Riede, T., Fisher, J.H. and Goller, F. (2010). Sexual dimorphism of the zebra finch syrinx indicates adaptation for high fundamental frequencies in males. PLoS One 5, e11368. Riede, T. and Goller, F. (2014). Morphological basis for the evolution of acoustic diversity in oscine songbirds. Proc. Roy. Soc. B 281, 20132306. Ríos‐Chelén, A.A., Salaberria, C., Barbosa, I., Macías Garcia, C. and Gil, D. (2012). The learning advantage: bird species that learn their song show a tighter adjustment of song to noisy environments than those that do not learn. J. Evol. Biol. 25, 21712180. Saranathan, V., Haminlton, D. Powell, G.V.N., Kroodsma, D.E. and Prum, R.O. (2007). Genetic evidence supports song learning in the three-wattled bellbird Procnias triarunculata (Cotingidae). Mol. Evol. 16, 3689-3702. Sitt, J.D., Amador, A., Goller, F. and Mindlin, G.B. (2008). Dynamical origin of spectrally rich vocalizations in birdsong. Phys. Rev. E 78, 011905. Suh, A., Paus, M., Kiefmann, M., Churakov, G., Franke, F.A., Brosius, J., Kriegs, J.O. and Schmitz, J. (2011). Mesozoic retroposons reveal parrots as the closest living relatives of passerine birds. Nat. Commun. 2, 443. Suthers, R.A. and Zollinger, S.A. (2004). Producing song: the vocal apparatus. Ann. NY Acad. Sci. 1016, 109–129. Suthers, R.A., Goller, F. and Pytte, C. (1999). The neuromuscular control of birdsong. Philo. Trans. R. Soc. Long. B 354, 927-939. Suthers , R.A., Goller, F. and Hartley, R.S. (1994). Motor dynamics of song production by mimic thrushes. Development. Neurobiol. 25, 917-936. Thorpe, W. H. (1961) Bird-Song. Cambridge: Cambridge Univ. Press. Vates, G.E., Broome, B.M., Mello, C.V. and Nottebohm, F. (1996). Auditory pathways of caudal telencephalon and their relation to the song system of adult male zebra finches (Taeniopygia guttata). J. Comp. Neurol. 366, 613–642. Vicario, D.S. (1991). Contributions of syringeal muscles to respiration and vocalization in the zebra finch. J. Neurobiol. 22, 63-73. Vicario, D.S. (1991). Organization of the zebra finch song control system: II. Functional organization of outputs from nucleus Robustus archistriatalis. J. Comp. Neurol. 309, 49 486-494. Vicario, D.S. and Simpson, H.B. (1995). Electrical stimulation in forebrain nuclei elicits learned vocal patterns in songbirds. J. Neurophysiol. 73, 2602-2607. Vu, E.T., Mazurek, M.E. and Kuo, Y.C. (1994). Identification of a forebrain motor programming network for the learned song of zebra finches. J. Neurosci. 14, 6924– 6934. Warner, R.W. (1972). The anatomy of the syrinx in passerine birds. J. Zool. 168, 381393. 50 Table 3.1: Study individuals Treatment Field site Corrientes, Argentina Gamboa, Panama suboscines Shortterm Corrientes, Argentina oscines Longterm Salt Lake City, USA Family Species Certhiaxis cinnamomeus Furnarius rufus Lepidocolaptes Furnariangustirosrtis idae Phacellodomus ruber Phacellodomus striaticolis Synallaxis frontalis Manacus vitellinus Pipridae Lepidothrix coronata, Cercomacra tyrannina ThamnoThamnophilus philidae caerulescens Tityridae Pachyramphus validus Elaenia mesoleuca,nn Empidonomus varius Hemitriccus margaritaceiventer Pitangus sulphuratus TyrannMachetornis rixosa idae Megarhynchus pitangua Myiarchus tyrannulus Phaeomyias murina Tyrannus verticalis Spizella breweri PasserellMelospiza melodia idae Junco hyemalis Sex/age, n= M/M, 2 M/M, 3 M/J, F/M, 2 M/M, 1 F/M, 1 M/M, 1 M/M, 2 M/M, 1 M/M, 2 M/M, 3 M/M, 3 M/M, 1 F/M, 1 M/M, 1 F/M, M/M 4 M/M, 1 M/M, 1 M/M, 1 M/M, 1 M/M, 3 M/M, 2 M/M, 4 M/M, 4 Species used in this study are listed with their treatment type, collection location, family, scientific name, common name, and sex/age (e.g., F/M is female/mature, M/J is male/juvenile). 51 Table 3.2 Significance of effects of nerve cut on duty cycle and intensity by species. Family Furnariidae Pipridae Thamnophilidae suboscines Tityridae Tyrannidae oscines Passerellidae Species C. cinnamomeus F. rufus L. angustirosrtis P. ruber P. striaticolis S. frontalis M. vitellinus L. coronata, T. caerulescens C. tyrannina P. validus E. mesoleuca, E. varius H. margaritaceiventer P. sulphuratus M. rixosa M. pitangua M. tyrannulus P. murina T. verticalis S. breweri M. melodia J. hyemalis Unpaired t-test p-values Duty cycle Intensity 0.0404 < 0.4243 0.2805 0.9476 < 0.0002 < 0.9642 < < < 0.0108 0.0843 0.0990 < 0.7491 < 0.0025 0.0236 0.9561 0.3072 0.1053 < < < 0.0233 < 0.1378 < 0.0001 < 0.9704 < 0.9495 0.2663 0.19 0.2116 0.1154 < 0.001 < < < < < n/a < Un-paired t-test reveals that the changes in duty cycle (see Fig. 3.4) for all passerellids, and some suboscines are significant; and the changes in intensity (see Fig. 3.5) for all passerellids and most suboscines are significant. While most changes are significant, the degree of change varies greatly (see Table 3.3). t-test values less than 0.00005 are indicated by the symbol “<”. 52 Table 3.3 Average % change and standard error in intensity and duty cycle by family. Family (# of species) Furnariidae (5) Pipridae (2) suboscines Thamnophilidae (2) Tityridae (1) Tyrannidae (9) oscines Passerellidae (2) Duty cycle Intensity Average % Δ St. err. Average % Δ St. err. -1.72 0.05 26.59 0.05 -0.2 0.04 -7.25 0.04 -6.45 0.05 -3.16 0.02 -10.84 --10.11 --2.36 0.03 0.24 0.03 -68.91 0.03 -55.21 0.15 Standard error values are not available for Tityridae since only one species was investigated in this family. 53 Figure 3.1 Syringeal neuromuscular control does not significantly impact acoustic parameters of suboscine species Myiarchus tyrannulus distress calls. Example oscillograms, spectrograms, and power spectra of the M. tyrannulus distress call (short term treatment) are presented before (A) and after (B) bilateral tracheosyringeal nerve transection. Power spectra were taken at points indicated by numbered, red markers. Acoustic parameters, including frequency range and modulation, did not change appreciably after nerve cut as seen in the similarity of (A) and (B). 54 Figure 3.2. Syringeal neuromuscular control significantly affects acoustic parameters of spontaneous vocalizations in oscine species Melospiza melodia. Representative oscillograms, spectrograms, and power spectra of spontaneous vocalizations (long-term treatment) of M. melodia are presented before (A) and after (B) bilateral tracheosyringeal nerve transection. Differences between the two spectrograms of (A) and (B) are readily visible. Fundamental frequency (F0) range and modulation were drastically reduced post-treatment (B). Pre-treatment (A), F0 of vocalizations span a wide range and have fewer spectra compared to post-treatment vocalizations. Power spectra were taken at points indicated by numbered, red markers. A variety of syllable types pre-treatment can be seen, and therefore power spectra were taken at three different instances. Post-treatment vocalizations contain less variability and we therefore show power spectra at only two instances. 55 Figure 3.3 Bilateral tracheosyringeal nerve cut significantly affects fundamental frequency distribution in passerellid oscines but has little (furnariids, piprids and thamnophilids) to no (tyrannids) effect in suboscines. Histograms of vocalization fundamental frequencies before (blue) and after (red) bilateral trachealsyringeal nerve transection are presented, showing relative abundance by use of color gradient (larger bins are correspondingly darkly shaded). Species are grouped by family (Furnariidae, Pipridae, Thamnophilidae, Tyrannidae, and Passerellidae). Data from long-term studies are indicated by an asterisk (*), and all other data are from shortterm studies. Tyrannids show virtually no changes in frequency range post nerve cut. 56 Figure 3.4 Bilateral tracheosyringeal nerve cut significantly affects spectral quality in passerellid oscines but has little to no effect on spectral quality in suboscines. Spectral quality is quantified using duty cycle (the beginning of the time waveform apex to the first initial peak divided by the period). Duty cycle before (blue) and after (red) bilateral trachealsyringeal nerve transections is presented, and standard error bars are included (not visible if smaller than the square symbol used to identify the datum). Significantly different averages before and after treatment are indicated by the symbol “!”. Species are grouped by family. Data from long-term studies are indicated by an asterisk (*), and all other data are from short-term studies. 57 Figure 3.5 Effects of bilateral tracheosyringeal nerve cut on intensity in suboscines and passerellid oscines. Intensity was quantified using the signal to noise ratios of individual vocalizations before (blue) and after (red) bilateral trachealsyringeal nerve transection. Standard error bars are included (not visible if smaller than the symbol used to identify the datum). Significantly different averages before and after treatment are indicated by the symbol “!”. Species are grouped by family, and family averages can be seen in Table 3.3. Data from longterm studies are indicated by an asterisk (*), and all other data are from short-term studies. Post-treatment data are not available for L. angustirostris. 58 CHAPTER 4 SUBOSCINE VOCAL MEMBRANES CONTAIN HISTOLOGICALLY DISTINCT LAYERS The avian vocal organ, the syrinx, produces highly diverse vocalizations ranging from the simple to the highly complex. The syrinx has been studied for over a century and displays remarkable morphological diversity, though it is unclear to what degree this morphological diversity is linked to acoustic features and vocal diversity. The gross morphology of the syrinx is well described across taxa, and recently there have been studies on the functional morphology of the sound generating tissue itself. During phonation, air passes through the syrinx and passively sets vocal membranes (labia) into oscillation. These oscillations periodically alter the air column exiting the syrinx, which is perceived as sound, and the way in which they oscillate and any changes thereof will be reflected in the sound. Identifying the intrinsic oscillatory properties of the labia is therefore relevant for understanding the link between vocal diversity and morphological diversity, and may in part be able to explain the broad array of vocalizations birds produce. Here we present data on the histological composition of labia from 6 passerellid oscines and 7 species of suboscines. In each species, we identify the presence of elastin, collagen, and hyaluronan, three major components of the extracellular matrix of vocal tissue that have been identified previously as relevant to sound production and the 59 intrinsic oscillatory behavior of tissue. These data indicate that 1) suboscines, like oscines, do have discrete layers comprising their labia; 2) on average, suboscines have 55 fewer discrete layers than oscines; 3) suboscines do not exhibit the same positive correlation between layers and frequency range found in oscines (their sister group) in previously published data. 4.1 Introduction Complex animal behavior, such as avian vocal behavior, arises from an interplay of specialized morphology and neuromuscular control (Galis, 1996; Dickinson et al., 2000; Fuxjager and Schlinger, 2015). Avian vocalizations function in a variety of roles (e.g., mate attraction, courtship, territoriality, alarm, contact, etc.), each of which has influenced the evolution of a given species’ unique vocal repertoire (e.g., Searcy, 1992). This diversity has long been studied through the lens of vocal learning (most prominently in oscines) (e.g., Scharff and Nottebohm, 1991; Searcy, 1992; Riede and Goller, 2014; Mason et al., 2017), the phenomenon where birds must be exposed to adult conspecific song to properly develop their own song (Nottebohm, 1972). However, many birds are vocal non-learners (e.g., suboscines, the sister group of oscines) (Kroodsma and Konishi, 1991; Gahr, 2000; Liu et al., 2013) and still display a wide range of vocal diversity. Independent of vocal learning, acoustic features of vocalizations are determined by 1) the morphology of the vocal organ (syrinx) and its neuromuscular control, 2) respiration, and 3) the upper vocal tract (Riede and Goller, 2010a). Whereas gross morphology of the syrinx has been studied for over a century (e.g., Müller, 1847; Ames, 1971), its functional morphology has only relatively recently been given some attention (e.g., Goller and 60 Suthers, 1996a, b; Goller and Larsen, 1997; Riede and Goller, 2014). 4.1.1 Gross mechanics of the syrinx Each bronchus of the avian vocal organ, the syrinx, contains a pair of labia oriented dorso-ventrally that are supported by adjacent cartilages (though see Garcia et al., 2017) (Fig. 4.1). Substantial variation in the cartilaginous framework as well as the attachment of syringeal muscles exists between species (e.g., Ames, 1971). Irrespective of this highly variable structural voice box, the sound-generating mechanisms are strikingly similar in the avian syrinx and even the mammalian larynx (Riede and Goller, 2010a). The syrinx sits within the interclavicular air sac such that pressurization for expiration not only drives air through the syringeal lumina, but also applies this pressure externally onto the vibrating elastic tissue, thus moving the labia further luminally (Fig. 4.1). The repositioning of the labia and the increased air velocity through the syrinx initiate oscillation of the labia. 4.1.2 Oscillatory behavior and acoustic features The oscillatory behavior of the labia shapes the acoustic features of the sound produced. Labial movement during oscillations (Fig. 4.1) is complex and can be described as movement along different axes: medial-lateral movement opening and closing the syringeal lumina, cranio-caudal movement, as well as oscillations along the length of the labia (Goller and Larsen, 1997; Elemans et al., 2015). These movements create a convergent and divergent profile of the labia during oscillation, and the ratio of these two phases to the overall period of this movement influences the time waveform of 61 the sound produced. In other words, two sounds may have the exact same period (i.e., fundamental frequency), but drastically different time waveforms due to the open versus closed phase of the oscillations of the labia (see Chapter 2 of this dissertation) and therefore different spectral composition. 4.1.3 Passive influence and active control of oscillatory behavior The oscillatory behavior of the labia is heavily dependent on the degree of stress placed on the tissue. Driving pressure of the air sacs can passively change this level of stress and therefore the frequency (e.g., Cardoso and Atwell, 2011). However, fairly large changes in pressure are required for substantial changes in frequency. Muscles achieve more effective control of frequency by either acting directly on the labia or indirectly via movement of cartilages to change labial tension. This mechanism for changing labial tension achieves a wider range of frequencies (Gaunt and Gaunt, 1977; Goller and Larsen, 1997; Düring et al., 2017; Döppler et al., 2017). The syrinx of passerines has both intrinsic and extrinsic muscles (e.g., Gaunt, 1983; Düring et al., 2013), and these different muscles can affect the labial tension either directly or indirectly by positioning them in the syringeal lumen. 4.1.4 Relevant extracellular matrix components Vibrating tissue (vocal folds and labia) is composed of an extracellular matrix consisting of mainly elastin, collagen, and hyaluronan (Goller and Riede, 2013) (Fig. 4.2). Whereas a labium may be composed of a uniform arrangement of the elastic proteins, it may also consist of layers in which different ratios of the proteins or different 62 alignments of the fibers generate differential physical properties. These materials enable the labia to withstand oscillatory demands (thousands of oscillations per second depending on the fundamental frequency, F0) by allowing elongation and elastic recoil of the tissue (e.g., Gray et al., 2000) and also impart acoustic features (e.g., Fee, 2002; Riede and Goller 2010a, b; Titze, 2011). Elastin specifically functions in the recoil of tissue. As a cross- linked polymer, it is able to stretch during stress as tropoelastin monomers uncoil that then recoil when the stress is removed (Chan et al., 2007; Moore and Thibeault, 2012; Titze et al., 2016). Collagen fibrils facilitate this by preventing tissue damage (Riede and Goller 2010a, b; Moore and Thibeault, 2012; Titze et al., 2016). Hyaluronan is found in all extracellular matrices, including vocal tissue, and contributes to the viscoelastic properties and hydration of the labia, and is thought to reduce shear stress of the tissue as well as heal the tissue after damage (Fraser et al., 1997; Titze, 2001). 4.1.5 Known histology in Passerines Much of our knowledge of labial and membrane histology in birds stems from non-passerines [e.g., dove (Abd El-Rahm, 2011); duck (Frank, 2006; Yilmaz, 2012) goose (Onuk, 2010); pigeons (Yildiz, 2005); chicken (Pal, 1971)] and ten species of oscines [golden-crowned kinglet (Regulus satrapa), ruby-crowned kinglet (Regulus calendula), zebra finch (Taeniopygia guttata), white-crowned sparrow (Zonotrichia leucophrys), European starling (Sturnus vulgaris), American robin (Turdus migratorius), yellow- headed blackbird (Xanthocephalus xanthocephalus), black-billed magpie (Pica hudsonia), Riede and Goller, 2014, and others; and the meadowlark (Sturnella magna and S. neglecta), Ellis, 1973]. Oscines exhibit a correlation in the number of distinct 63 layers in the labia (up to five have been identified) and the vocal range of a species’ repertoire (Riede and Goller, 2014). Because the role of histological variation of the labia in shaping acoustic features has only been explored in oscines, which are vocal learners, the relationship is complicated by elaborate neuromuscular control. Looking at their sister group, suboscines, may expand our understanding of the functional morphology of the syrinx, as neuromuscular control is more limited in these vocal non-learning taxa (Chapters 2 and 3; Amador et al., 2008). Like oscines, they have a dual sound source syrinx with two pairs of labia in the bronchi (Ames, 1971; though see Garcia et al., 2017). If the use of complex labial composition for enhancing the range of possible acoustic features depends on neuromuscular control, we would expect that differential labial layer composition may be absent in suboscines. Alternatively, if labial layers can be recruited differentially into oscillation via aerodynamic changes, suboscine taxa should also exhibit layers. Differential labial layer recruitment by 1) neuromuscular control and 2) aerodynamic changes may both play a role in shaping acoustic features, and it is possible that suboscines rely more heavily on the latter given the reduced role of neuromuscular control in this group relative to oscines (Chapter 2 of this dissertation; Amador et al., 2008; Garcia et al., 2017). Here we identify the histological makeup of 7 suboscine species and 6 passerellid oscine species. These data, coupled with previously published data, indicate that 1) suboscine labia do contain discrete layers, though 2) the mean number of layers is greater in oscines, and 3) while our passerellid oscine data fit within the broader oscine histology data, there is no appreciable relationship of frequency range and number of layers in 64 suboscines. 4.2 Methods 4.2.1 Individuals and collection at field sites Individuals were captured with mist nets in November of 2014 and 2015 at the Estación Biológica de Corrientes (EBCo) in Corrientes, Argentina (27,55095o S 58,68441o W), under permits issued by the Fauna Province Direction and the Provincial Natural Reserve Area; March of 2015 in Gamboa, Panama, with permission from the Autoridad Nacional del Ambiente and the Smithsonian Tropical Research Institute (STRI); and March-September of 2013-2015 in Salt Lake City, Utah, with permission from the U.S. Fish and Wildlife Service. Experimental procedures were approved (17-112014, 20-10-2015, Corrientes; SE/A-6-15, Gamboa). All tissue collection was conducted at the respective field site. Individuals are listed in Table 4.1. 4.2.2 Perfusion and tissue preparation Individuals were anesthetized (ketamine/xylazine) and perfused intracardially with phosphate-buffered saline following methods of Riede and Goller (2010, 2014). The syrinx was excised and immediately placed in 10% formalin for 1 week. The tissue was then decalcified for 8 hours, and again placed in 10% formalin for 1-2 weeks. 4.2.3 Tissue processing, staining, and imaging The tissue was embedded in paraffin and cross-sectioned in 5 µm increments at Research Histology, ARUP labs, Salt Lake City, Utah. The F. rufus tissue was processed 65 and embedded in Corrientes, Argentina, by Dr. Hugo Domitrovic. To enable consistency, this sample was re-embedded in Salt Lake City before sectioning. Sections were stained with hemotoxylin and eosin (HE; general), elastica van Gieson (EVG; elastin), trichrome (TRI; collagen), alcian blue (pH 2.5) (AB; mucopolysaccharides and glycosaminoglycans), and AB after digestion with bovine testicular hyaluronidase (AB with hyaluronan removed). Micrographs were taken at ARUP Laboratories, Salt Lake City, Utah, with a digital camera (AxioCam HRc, Carl Zeiss, Germany) and an Axioplan Zeiss microscope (Axioplan, Carl Zeiss; 10x, 20x and 80x magnification) using computer software (AXIOVISION 40, v. 4.6.3.0, Carl Zeiss). 4.2.4 Tissue analysis Tissue was qualitatively described and assigned a number of layers based on information from all 5 stain types. Layer identification generally followed Riede and Goller (2010, 2014). Different layers were identified by different orientation of fibers and distinct density differences of elastin and collagen fibers and hyaluronan (Fig. 4.2). Schematics of these layers were generated (Fig. 4.3) for each species. Suboscines exhibited different layer structure, such as more cranio-caudally oriented layers, which were still counted as different layers based on histological composition. 4.2.5 Frequency extraction Representative sound files for each species were downloaded from Xenocanto.org. These files were analyzed in Praat (sound analysis application, Boersma and Weenink, Mac version 6.0.16) and fundamental frequency (F0) was extracted at ~5 ms 66 intervals using the pitch listing function. If amplitude modulation was present, both the modulation frequency and the carrier frequency were extracted. These extracted frequencies were placed in 100 Hz bins, and bins were then normalized relative to the bin with the most occurrences. The mode was identified as the largest bin. Effective frequency range was calculated as 100*(# non-zero bins), where 100 is the bin size (Hz). 4.2.6 Statistical analyses An un-paired t-test was used to test the statistical significance of the average modes of frequency range. Simple linear regression was used to calculate the relationship between layers and effective frequency range, and was analyzed with ANOVA test. The Mann-Whitney non-parametric rank test was used to test whether the distributions of the labia layer counts in suboscines and oscines differed significantly. A Pearson correlation test was used to identify potential relationships between frequency range, effective frequency range, mode, and body mass (body masses were sourced from online databases and articles). 4.3 Results The data presented here identify discrete layers within suboscine labia (both medially and laterally in most cases). While they do show discrete layers, on average, they have fewer layers than passerellid oscines, and do not display a positive relationship between the number of layers and their frequency range as has been found in oscines (Riede and Goller, 2014). Despite this, suboscines show nearly comparable F0 ranges. 67 4.3.1 Number of layers Qualitative assessments of labia layers clearly indicate distinct histological labial layers in suboscines (Fig. 4.2), which have been summarized for all species (Fig. 4.3, Table 4.2). Suboscines on average do have fewer medial labial layers than passerellid oscines (1.4 and 3.5, respectively; median 1 and 3.5, respectively). The medial layer distributions in the two groups differed significantly (Mann-Whitney U=0, n1=7, n2=6, p=0.0025) whereas the lateral layer count did not differ significantly (Mann-Whitney U=26.5, n1=7, n2=6, p=0.381). 4.3.2 Acoustic analysis Suboscines have a lower average mode for their F0 range (2236 Hz) compared to oscines (4417 Hz) (unpaired t-test, p = 0.0031). However, their overall F0 range (F0 maxF0 min) is similar to that of oscines (Fig. 4.4). When considering their effective F0 range, oscines have a slightly larger range (Table 4.2, suboscine average is 4025.86Hz, ±562.36, oscine average is 5360Hz, ±766.80). Oscines also achieve higher frequencies than suboscines, though only by ~1.5kHz (Fig. 4.4). Pearson correlation test revealed no correlation between mode and frequency range (r = -0.02896), frequency range and body mass (r = 0.27889), nor mode and body mass (r = -0.26777); and a weak correlation between effective frequency range and body mass (r = 0.420741). 4.3.3 Layers and frequency range relationship All correlations of effective frequency range and number of labial layers were weak (Table 4.3). For medial labia, there is a weak negative correlation in suboscines and 68 a weak positive correlation in oscines (Table 4.3). When oscine data are combined with previously published data (Riede and Goller, 2014), a relatively stronger positive correlation results. While all correlations are weak, lateral layers have the smallest r2 for both oscines and suboscines. 4.3.4 Qualitative layer assessment 4.3.4.1 Cartilage As shown by Ames (1971), there is more variation in syringeal morphology in suboscines than oscines, and the cartilages of the suboscine species show unique aspects. C. mentalis (Fig. 4.2) and M. vitellinus also have elastin in their first bronchial cartilage, but not their second. C. tyrannina bronchial cartilages that support the lateral labia (rings 1 and 2) are compositionally different from the other bronchial and tracheal cartilages, and more similar to oscines (Fig. 4.S7-4.S12). They contain densely packed elastin fibers, and are also oblong in shape, and are likely elastic cartilages. Consistent with other oscine syringeal histology (Riede and Goller, 2010; 2014), oscines have large, specialized cartilages underlying the lateral labia. In addition, M. vitellinus also distinctly shows fusion of the 2nd and 3rd bronchial cartilages. 4.3.4.2 General labia characteristics The location of the labia in the syringeal cartilage framework (i.e., relative to the pessulus) varies between suboscines and oscines. The labia are generally located between cartilages 1 and 2 in suboscines, whereas in oscines cartilages, 1-3 appear embedded in the labia, or are directly under the labia. While suboscines do have histologically distinct 69 layers in the medial labia, the overall relative thickness of these labia is substantially smaller than that seen in oscines. Additionally, suboscines present relatively symmetrical labia of the right and left bronchi (Fig. 4.1, Figs. 4.S1-4.S6), whereas there is some asymmetry in oscines (M. melodia, Fig. 4.S8; P. maculatus, Fig. 4.S11). All oscine species exhibit more medial layers than lateral, and this trend is not consistently seen in suboscines that may have more, fewer, or the same number of lateral layers as medial layers (Fig. 4.3). 4.3.4.3 ECM components and layer trends All species (both oscine and suboscine) contained some degree of elastin, collagen, and hyaluronan in or adjacent to the labia, though the orientation and size of these fibers varied from species to species. Consistent across the three Pipridae species, the medial labia, which have just one layer, are virtually entirely collagen (with few small elastin fibers) (Table 4.2, Fig. 4.1, Fig. 4.S1, 4.S2). This simple composition is in contrast to the relatively complex medial labia in all other species, including other suboscine families. All oscines (passerellids) had similar medial labia: a luminal layer of collagen and a more basal layer of elastin with little variation (Fig. 4.3). 4.4 Discussion These data suggest a number of interesting histological trends, and together strongly suggest that histologically complex vocal tissue in birds is not restricted to vocal learners. Suboscine species (vocal non-learners) investigated clearly have discrete layers, and these layers vary among the families investigated. While oscines species on average 70 have labia with a greater number of layers, it is interesting to note their relatively equal ranges, indicating that other factors influence frequency range. Below we address the visible histological trends and their potential functions. 4.4.1 Pipridae histological trends and possible implications The densely packed single collagen layer found in manakins seems unique in regard to the other suboscines investigated, but their frequency ranges are similar (Table 4.2, Fig. 4.4). Given the density of collagen, this tissue can likely withstand significant tension and therefore greater changes in length (enabled by the presence of diffuse/sparse elastin fibers) (Titze, 2016). This change in length would support a broad range of fundamental frequencies despite arising from a single layer, which would facilitate vocal flexibility (Titze, 2016), explaining how manakins achieve comparable F0 ranges of other suboscines with only one layer comprising their medial labia. Additionally, manakins produce very long whistles containing frequency modulation, and the dense collagen layer could facilitate this sustained production by not dissipating stress quickly. Interestingly, it is these manakin data that give suboscines a negative correlation between F0 range and number of medial layers (Table 4.3). Without these data, this negative trend line would have a slope much closer to zero, suggesting that the positive trend (increased layers increases effective frequency range) seen in oscines does not exist in suboscines. 71 4.4.2 Left-syrinx/right-syrinx labia asymmetry Some oscines have prominently asymmetric labia on the two sides of the syrinx, which allows them to specialize the F0 range of each set of labia (e.g., Allan and Suthers, 1994; Prince et al., 2011; Goller and Riede 2013). In line with these findings, our data show size asymmetry in some of the passerellids, but none in suboscines. This aspect of morphology is consistent with the decreased role of neural control (Amador et al., 2008; Garcia et al., 2017; Chapter 3 of this dissertation), lack of vocal learning (Kroodsma and Konishi, 1991; Gahr, 2000; Liu et al., 2013), and the lack of two-voice phenomena in suboscines. It is difficult to assess small-scale symmetry of a species from few samples as orientation of the tissue when embedded can affect the relative appearance of size symmetry. However, large asymmetries such as those seen in M. melodia (Fig. 4.S8) and P. maculatus (Fig. 4.S11) are not likely from misalignment of the tissue upon embedding, and it is highly likely that these species do have left/right labia asymmetry. While suboscines and oscines seem to span similar F0 ranges (Fig. 4.4), oscines do have larger effective ranges and produce slightly higher frequencies, which could be facilitated by the asymmetry of the labia in the two sound generators. 4.4.3 Tyrannid histology and syringeal neuromuscular control Syringeal muscles also play a critical role in maintaining open airways during quiet breathing, as they are activated during the expiratory phase of each breath (e.g., Vicario, 1991; Goller and Suthers, 1996a,b). The sectioning of the tyrannid syringes supports the observation from Chapter 3 that tyrannids may use syringeal muscles to open airways during labored breathing. While not visible in Fig. 4.S5 and Fig. 4.S6, more 72 dorsal syringeal sections of T. verticalis and C. sordidulus show syringeal muscle inserting more ventrally on the first bronchial ring. This shift in muscle attachment supports a role in opening airways, but may not enable fine control of the labia to facilitate changes in acoustic features. 4.4.4 Histologically unique cartilages adjacent to vocal tissue The histological composition of the cartilaginous framework in some species indicates specialized cartilages that are associated with the labia. Most strikingly, bronchial cartilages 1 and 2 in suboscine C. tyrannina (Fig. 4.S3) (but not F. rufus, another tracheophone species, Fig. 4.S4) contain densely packed elastin fibers, which are not seen in the other bronchial cartilages. These cartilages appear to be elastic cartilage, which is found in very flexible structures, such as the human ear. This higher density of elastin in bronchial cartilages associated with the labia is found in all of the oscines investigated, as well as suboscines L. coronata and C. mentalis. These elastic cartilages to which the labia attach will be more flexible than non-elastic cartilages, and this could support control of a larger range of frequencies. The frequency range of C. tyrannina is more limited than that of the F. rufus, another tracheophone species, which does not have such high elastin content in the cartilages. It is possible that within the syringeal morphological framework of C. tyrannina, these elastic cartilages do enhance the F0 range, but perhaps this does not enhance its range relative to other species’ ranges. This highlights the complexity by which morphological parameters interact to give rise to vocal ability, and one parameter, such as increased F0 range, can be achieved in a variety of ways. Additionally, this morphological difference between the C. tyrannina and the F. 73 rufus may be one possible basis for the difference in the way in which the three sources of the tracheophone syrinx interact (Garcia et al., 2017). 4.4.5 Differential lateral and medial labia histological composition This study addresses both medial and lateral layers, the latter of which have not been addressed in other histological studies of passerines (Riede and Goller, 2010, 2014; Goller and Riede, 2013). Our data suggest that the positive correlation between the number of layers and F0 range in oscines may only be present for the medial labia, and not the lateral labia. Passerellid oscine data presented here show a very weak positive correlation; however, they do fit within the broader dataset of oscine histology available (Riede and Goller, 2014). It is unclear why lateral labia present no correlation with frequency range, and no data are available to describe the implications of difference in layers for a given set of labia for F0 range. This difference is present in both oscines and suboscines (e.g., this study; Prince et al., 2011) and despite this difference, they do oscillate relatively uniformly, suggesting entrainment of the opposing labia (Mindlin and Laje, 2005). Because lateral and medial labia often also differ substantially in size, the differential layer structure might function to balance vibrating masses of the two labia. Detailed study of the vibratory behavior of labia could be used to test this speculative interpretation. 4.4.6 Fundamental frequency range and layers The positive relationship between frequency range and medial labia layer count is not visible in passerellid oscines studied here. However, these passerellid data are not far 74 from other published data on frequency range and layer count (Riede and Goller, 2014). If both of these data sets are combined, the relationship improves (y = 1560.63x+75.07; r2=0.544; ANOVA p-value = 0.004). The lack of a positive correlation between layers and F0 range in suboscines is not entirely unexpected. Suboscines do have drastically different morphology (i.e., tracheophones) relative to oscines, whose syrinx shows surprisingly similar morphology across the otherwise remarkably divergent and speciose clade (Warner, 1972). Because their gross and functional morphology is so different (Ames, 1971; Garcia et al., 2017), it is difficult to isolate the effects of one parameter (i.e., number of layers) on F0 range. In addition, the smaller span of number of layers in suboscines makes a trend difficult to resolve. To better understand this relationship in suboscines, we need to investigate more species, and perhaps focus specifically on species with extremely large or extremely small F0 ranges. This is the first study to address the functional morphology of the labia in the suboscine syrinx, and from these data, it is clear that suboscines, vocal non-learners, do have histologically distinct layers comprising their labia. While our data do not suggest a positive relationship between the number of layers and F0 range in suboscines, it is possible that upon expansion of the number of species studied such a trend will emerge, or, given the substantial variation of the morphology of the syrinx in suboscines, the trend will only exist on a smaller scale (i.e., in families, and families with more conserved syringeal morphology at that; or the trend may only be visible when considering extremes in F0 range). However, it is clear that while vocal learning and histologically complex labia go hand-in-hand, a complex labial histology is not restricted to vocal learners. If 75 there is a positive correlation yet to be seen in suboscines regarding number of labial layers and F0 range, the complex neural control seen in oscines is not required to make use of such labial morphology to achieve acoustic diversity, especially in tyrannids. At least in the parameter of F0 range, suboscines do achieve similar diversity to that of oscines. It is unclear exactly how increased labial layers give rise to increased F0 range even in oscines. Mainly, layers are thought to function in a manner consistent with the body-cover hypothesis (Story and Titze, 1995). In other words, different layers of the labia are recruited, allowing for effective differences in labial characteristics based on which layers are recruited. This situation is likely, and would explain how neuromuscular control of the oscine syrinx gives rise to diverse acoustic features, since some muscles of the oscine syrinx insert into the labia themselves (Fee, 2002; Düring et al., 2013), enabling presumably precise control of different layers of the labia. In suboscines, such differential recruitment of the labia would be more passive, especially in tyrnannids where neural control of the syrinx is not absolutely required for normal sound production (Chapter 3; Amador et al., 2008). While suboscines do have discrete labial layers, they have fewer and perhaps less complex layers than oscines, and exhibit less complex vocalizations. It therefore seems as if labial layer complexity increases with increasing acoustic complexity and that some degree of morphological complexity of the labia was ancestral to oscines and suboscines. While these and other data suggest that morphological specialization in suboscines gives rise to comparable F0 ranges seen in oscines, it may not allow single species to achieve the same range as is possible in vocal learning species. So while 76 morphological specialization can achieve specific acoustic features, as suggested by Garcia et al. (2017), it may reduce the degrees of freedom available, whereas diverse neuromuscular control is permissive of a broad range of acoustic features using one syringeal morphology. Both strategies (diverse morphology and diverse neuromuscular control) seem to have facilitated vocal diversity between species and may have played a significant role in the remarkable radiation in suboscines and oscines. 4.5 References Abd El-Rahman, F.A.A. (2011). Comparative morphological and histological studies on the syrinx of Streptopelia senegalensis (Linneaus, 1766) and Coturnix coturnix (Linneaus, 1758). Egypt. J. Zool. 174, 1-44. Allan, S.E. and Suthers, R.A. (1994). Lateralization and motor stereotype of song production in the brown‐headed cowbird. J. Neurobiol., 25, 1154-1166. Amador, A., Goller, F. and Mindlin, G.B. (2008). Frequency modulation during song in a suboscine does not require vocal muscles. J. Neurophysiol., 99, 2383–2389. Ames, P.L. (1971). The morphology of the syrinx in passerine birds. Peabody Mus. Nat. Hist. Bull. 37, 1-194. Cardoso, G.C. and Atwell, J.W. (2011). On the relation between loudness and the increased song frequency of urban birds. Animal Behav. 82, 831-836. Chan, R.W., Fu, M., Young, L. and Tirunagari, N. (2007). Relative contributions of collagen and elastin to elasticity of the vocal fold under tension. Ann. Biomedic. Engin. 35, 1471-1483. Dickinson, M.H., Farley, C.T., Full, R.J., Koehl, M.A.R., Kram, R. and Lehman, S. (2000). How animals move: an integrative view. Science 288, 100–106. Döppler, J.F., Busch, A., Goller, F. and Mindlin, G.B. (2017). From electromyographic activity to frequency modulation in zebra finch song. J. Comp. Physiol. A 204, 209-217. Düring, D.N., Ziegler, A., Thompson, C.K., Ziegler, A., Faber, C., Muller, J. and Elemans, C.P.H. (2013). The songbird syrinx morphome: a three-dimensional, high resolution, interactive morphological map of the zebra finch vocal organ. BMC Biol. 77 11, 1. Düring, D.N., Knörlein, B.J., Elemans, C.P.H., Schmidt, M., Goller, F., Riede, T. and Rome, L. (2017). In situ vocal fold properties and pitch prediction by dynamic actuation of the songbird syrinx. Sci. Rep. 7, 11296. Elemans, C.P.H., Rasmussen, J.H., Herbst, C.T., Düring, D.N., Zollinger, S.A., Brumm, H., Srivasta, K., Svane, N., Ding, M., Larsen, O.N. et al. (2015). Universal mechanisms of sound production and control in birds and mammals. Nat. Commun. 6, 8978. Ellis, C.J. (1973). Syringeal histology, II. Meadowlark (Sturnella magna, S. neglecta) Iowa State J. Res. 48, 175–191. Fee, M.S. (2002). Measurement of the linear and nonlinear mechanical properties of the oscine syrinx: implications for function. J. Comp. Physiol. A 188, 829–839. Fraser, J.R.E., Laurent, T.C. and Laurent, U.B.G. (1997). Hyaluronan: its nature, distribution, functions and turnover. J. Inter. Med. 242, 27-33. Frank, T., Walter, I., Probst, A. and König, H. E. (2006). Histological aspects of the syrinx of the male mallard (Anas platyrhynchos). Anat. Histol. Embryol. 35, 396401. Fuxjager, M.J. and Schlinger, B.A. (2015). Perspectives on the evolution of animal dancing: a case study of manakins. Curr. Opin. Behav. Sci. 6, 7–12. Gahr, M. (2000). Neural song control system of hummingbirds: comparison to swifts, vocal learning (songbirds) and nonlearning (suboscines) passerines, and vocal learning (budgerigars) and nonlearning (dove, owl, gull, quail, chicken) nonpasserines. J. Comp. Neurol. 426, 182–196. Galis, F. (1996). The application of functional morphology to evolutionary studies. Trends Ecol. Evol. 11, 124–129. Gaunt, A.S. and Gaunt, S.L.L. (1977). Mechanics of the syrinx in Gallus gallus. II. Electromyographic studies of Ad Libitum vocalizations. J. Morphol. 152, 1–19. Gaunt, A.S. (1983). An hypothesis concerning the relationship of syringeal structure to vocal abilities. Auk 100, 853-862. Goller, F. and Larsen, O.N. (1997). A new mechanism of sound generation in songbirds. PNAS 94, 14787-14791. Goller, F. and Riede, T. (2013). Integrative phyiology of fundamental frequency control in birds. J. Phys. Paris 107, 230-242. 78 Goller, F. and Suthers, R.A. (1996a). Role of syringeal muscles in controlling the phonology of bird song. J. Neurophysiol. 76, 287–300. Goller, F. and Suthers, R.A. (1996b). Role of syringeal muscles in gating airflow and sound production in singing brown thrashers. J. Neurophysiol. 75, 867– 876. Gray, S., Titze, I.R., Alipour, F. and Hammond,T. (2000). Biomechanical and histologic observations of vocal fold fibrous proteins. Ann. Otol. Rhino Laryngol. 109, 77–85. Kroodsma, D.E. and Konishi, M.A. (1991). Suboscine bird (eastern phoebe, Sayornis phoebe) develops normal song without auditory feedback. Anim. Behav. 42, 477– 487. Liu, W.C., Wada, K., Jarvis, E.D. and Nottebohm, F. (2013). Rudimentary substrates for vocal learning in a suboscine. Nat. Commun. 4, 2082. Mason, N.A., Burns, K.J., Tobias, J.A., Claramunt, S., Seddon, N. and Derryberry, E.P. (2017). Song evolution, speciation, and vocal learning in passerine birds. Evolution 71, 786–796. Mindlin, G.B. and Laje, R. (2005). The Physics of Birdsong. Berlin: Springer. Moore, J. and Thibeault, S. (2012). Insights into the role of elastin in vocal fold health and disease. J. Voice 26, 269-275. Müller, J. (1847). Über die bisher unbekannten typischenVerschiedenheiten der Stimmorgane der Passerinen. Berlin: Abhandlungen der Königlichen Akademie der Wissenschaften, 321–391pp. Nottebohm, F. (1972). The origins of vocal learning. Am. Nat. 106, 116-140. Onuk, B., Hazıroğlu, R.M. and Kabak, M. (2010). The gross anatomy of larynx, trachae and syrinx in goose (Anser anser domesticus). Kafkas Univ Vet Fak Derg 16, 443-450. Pal, G. and Bharadwaj, M. B. (1971). Histological and certain histochemical studies on the respiratory system of chicken. II. Trachea, syrinx, bronchi and lungs. Indian J. Anim. Sci. 41, 37-45. Prince, B., Riede, T. and Goller, F. (2011). Sexual dimorphism and bilateral asymmetry of syrinx and vocal tract in the European starling (Sturnus vulgaris). J. Morphol. 272, 1527–1536. Riede T. and Goller F. (2010b) Functional morphology of the sound generating labia in 79 the syrinx of two songbird species. J. Anat. 216, 23–36. Riede, T. and Goller, F. (2010a). Peripheral mechanisms for vocal production in birds– differences and similarities to human speech and singing. Brain Lang. 115, 69-80. Riede, T. and Goller, F. (2014). Morphological basis for the evolution of acoustic diversity in oscine songbirds. Proc. Roy. Soc. Lond. B 281, 1779. Scharff, C. and Nottebohm, F. (1991). A comparative study of the behavioral deficits following lesions of various parts of the zebra finch song system: implications for vocal learning. J. Neurosci. 11, 2896-2913. Searcy, W.A. (1992). Song repertoire and mate choice in birds. Integrat. Compar. Biol. 32, 71–80. Story, B. H. and Titze, I. R. (1995). Voice simulation with a body‐cover model of the vocal folds. J. Acoust. Soc. Am. 97, 1249-1260. Suthers, R.A., Goller, F. and Pytte, C. (1999). The neuromuscular control of birdsong. Phil. Trans. R. Soc. Lond. B 354, 927–939. Titze, I.R. (2011). Vocal fold mass is not a useful quantity for describing F0 in vocalization. J. Speech Lang. Hear Res. 54, 520–522. Titze, I.R., Riede, T. and Mau, T. (2016). Predicting achievable fundamental frequency ranges in vocalization across species. PLoS Comput. Biol. 12, e1004907. Vicario, D.S. (1991). Organization of the zebra finch song control system: II. Functional organization of outputs from nucleus robustus archistriatalis. J. Comp. Neurol. 4, 486-494. Yildiz, H., Yilmaz, B. and Arican, I. (2005). Morphological structure of the syrinx in the Bursa roller pigeon (Columba livia). Bull. Veterin. Instit. Pulway 49, 323. Yilmaz, B., Yilmaz, R., Arican, I. and Yildiz, H. (2012). Anatomical structure of the syrinx in the mallard (Anas platyrhynchos). Harran Univ. Vet Fak Deg 1, 111-116. Warner, R.W. (1972). The anatomy of the syrinx in passerine birds. J. Zool. 168, 381393. 80 Table 4.1 Study individuals. Field site Gamboa, Panama Suboscine Family Pipridae Thamnophilidae Corrientes, Argentina Furnariidae Tyrannidae Oscine Salt Lake City, USA Passerellidae Species Manacus vitellinus Lepidothrix coronata Ceratopipra mentalis Cercomacra tyrannina Furnarius rufus Tyrannus verticalis Conotopus sordidulus Spizella breweri Melospiza melodia Passerella iliaca Junco hyemalis Pipilo maculatus Pipilo chlorurus Common name (n=) Golden-collared manakin (3) Blue-crowned manakin (3) Red-capped manakin (2) Dusky antbird (2) Rufus ovenbird (6) Western kingbird (3) Western wood pewee (2) Brewer’s sparrow (2) Song sparrow (2) Fox sparrow (2) Dark-eyed junco (3) Spotted towhee (3) Green-tailed towhee (2) Species used in this study are listed with their collection location, family, scientific name, and common name. All tissue samples were taken from adult males. 81 Table 4.2 Effective frequency range, vocal tissue layer count, Family Oscine # ML # LL # MT layers layers layers 1 3 n/a 1 3 n/a 1 2 n/a Thamnophilidae 2 2 3 Furnariidae 1 2 1 2 1 n/a Tyrannidae 2 2 n/a 3 2 n/a 4 2 n/a 3 1 n/a Passerellidae 3 2 n/a 4 2 n/a 4 2 n/a and body mass of the species investigated. Pipridae Suboscine Species M. vitellinus L. coronata C. mentalis C. tyrannina F. rufus T. verticalis C. sordidulus S. breweri M. melodia P. iliaca J. hyemalis P. maculatus P. chlorurus Effective F0 range (Hz) 3319 5630 4285 2732 6311 3632 2272 2229 7029 5697 5384 4439 7382 Body mass (g) 18 15 16 18 50 40 13 6 19 38 19 40 28 82 Table 4.3 Linear regression results of medial, lateral, and total layer counts and effective frequency range for suboscines and oscines. Suboscine Oscine Labia type Medial Lateral Total Medial Lateral Total Regression equation y = -2007.52x+6893.65 y = 451.74x+3057.77 y = -1254.63x+8506.58 y = 1846.66x-1103.29 y = -404.23x+6101.09 y = 729.94x+1467.00 R2 0.521 0.044 0.203 0.290 0.007 0.101 ANOVA p-value 0.067 0.652 0.310 0.270 0.869 0.669 83 Figure 4.1 Air sac pressurization causes air to flow through the syrinx and may cause vocal tissue to oscillate, producing sound. The syrinx is located at the tracheal-bronchial juncture, which sits within the airsacs. When the air sacs are pressurized, labia are drawn in by Bernoulli forces, and are also pushed into the bronchial air space by relative higher pressure in the air sacs. Intrinsic properties of the labia, including length, mass, and viscoelasticity, will dictate their oscillatory behavior. Air flow in the bronchi passively set the labia into self-sustained oscillation, indicated by dashed lines. Relevant structures are labeled in the magnified schematic: tracheal cartilage (TC), pessulus (P), syringeal muscle (SM), bronchial cartilage (BC), lateral labium (LL), medial labium (ML), bronchial space (BS), medial tympanic membrane (MTM). 84 Figure 4.2 Histological stains of Ceratopipra mentalis (piprid suboscine) syrinx reveals distinct layer structure. A) Schematic and overview stain of the syrinx with relevant structures labeled (ST, sternotrachealis muscle in pink; TS, tracheal space; TC, tracheal cartilage; P, pessulus; BC, bronchial cartilage; LL, lateral labium in green; ML medial labium in thicker blue; MTM, medial tympaniform membrane in thinner blue; BS, bronchial space). The inset of the right lateral labium of the overview image is B) expanded and stained for 1) collagen (fibers indicated with arrows), 2) elastin (fibers indicated with arrows, see C) inset for close-up), 3) hyaluronan (indicated with arrows), and 4) hyaluronan digest. Note syringeal muscles inserted on first bronchial cartilages, indicating potential neuromuscular control of lateral labia. These stains indicate two lateral labia layers (inner layer near the bronchial space is predominantly elastin with some collagen and hyaluronan; deep layer away from bronchial space is diffuse collagen, elastin, and some hyaluronan), and one densely packed collagen layer in the medial labium. 85 Figure 4.3 Schematics of qualitative assessment of medial and lateral labia layer structure in select suboscines and passerellid oscines. For each A) suboscine and B) oscine species, from left to right, the schematic presents lateral (L) layer(s), bronchial space (br. sp.), and medial (M) layer(s). Hyaluronan (light blue), collagen (dark blue), and elastin (black) are presented. Fibers oriented parallel to the craniocaudal plane are elongated, while fibers oriented parallel to the dorsoventral plane are circular. 86 Figure 4.4 Frequency range of vocalizations in select suboscines and passerellid oscines. Gray bars indicate frequency ranges and black lines indicate mode. Average mode for suboscine species (blue) and emberizid oscines (red) are indicated by horizontal lines (un-paired t-test, p = 0.0031). Pearson correlation test revealed no significant relationship between mode and frequency range (r = -0.02896), frequency range and body mass (r = 0.27889) nor mode and body mass (r = -0.26777); and a weak correlation between effective frequency range and body mass (r = 0.420741). Furnariidae (Furn.), Thamnophilidae (Tham.). 87 Figure 4.S1 Syrinx histological section stains of Manacus vitellinus. A) Hematoxylin and eosin overview stain with inset of B) left labia. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. C) Close-up inset of the lateral labia with three arrows identifying collagen presence in discrete layers. Note the syringeal muscle attachment to the first bronchial cartilage and fusion of the second and third cartilages, identified by the arrow in the overview stain of B). The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 88 Figure 4.S2 Syrinx histological section stains of Lepidothrix coronata. A) Hematoxylin and eosin overview stain with inset of B) left labia. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. Note the syringeal muscle attachment to the first bronchial cartilage. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 89 Figure 4.S3 Syrinx histological section stains of Cercomacra tyrannina. A) HE overview stain with close up inset of B) the membrana trachealis (MT) and C) the right labia. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. 90 Figure 4.S4 Syrinx histological section stains of Furnarius rufus. A) HE overview stain with close up inset of B) the membrana trachealis (MT) and C) the right labia. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 91 Figure 4.S5 Syrinx histological section stains of Tyrannus verticalis. A) HE overview stain with close up inset of B) the left labia and C) the left labia stained for elastin. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black and hyaluronan appears aqua. Collagen staining is unavailable for this species. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 92 Figure 4.S6 Syrinx histological section stains of Conotopus sordidulus. A) HE overview stain with close up inset of B) the left labia and C) the left labia stained for elastin. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 93 Figure 4.S7 Syrinx histological section stains of Spizella breweri. A) HE overview stain with close up inset of B) the right labia and C) the right labia stained for elastin and collagen. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 94 Figure 4.S8 Syrinx histological section stains of Melospiza melodia. A) HE overview stain with close up inset of B) the right labia. Arrows indicate stainrelevant components of the extracellular matrix (for example, the arrows of the elastinstained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 95 Figure 4.S9 Syrinx histological section stains of Passerella iliaca. A) HE overview stain with close up inset of B) the left labia and C) the left labia stained for elastin and collagen. Arrows indicate stain-relevant components of the extracellular matrix (for example, the arrows of the elastin-stained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 96 Figure 4.S10 Syrinx histological section stains of Junco hyemalis. A) HE overview stain with close up inset of B) the right labia. Arrows indicate stainrelevant components of the extracellular matrix (for example, the arrows of the elastinstained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 97 Figure 4.S11 Syrinx histological section stains of Pipilo maculatus. A) HE overview stain with close up inset of B) the right labia. Arrows indicate stainrelevant components of the extracellular matrix (for example, the arrows of the elastinstained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining. 98 Figure 4.S12 Syrinx histological section stains of Pipilo chlorurus. A) HE overview stain with close up inset of B) the left labia. Arrows indicate stainrelevant components of the extracellular matrix (for example, the arrows of the elastinstained slide indicate the presence of elastin fibers). Elastin fibers appear dark blue/black, collagen appears blue, and hyaluronan appears aqua. The presence of hyaluronan (indicated by arrows in the slide stained for hyaluronan) is confirmed by its absence in the hyaluronan digest stain where hyaluronan was digested prior to staining.