i CHARACTERIZATION OF SEA LAMPREY PHEROMONE COMPONENTS By Cory O. Brant A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Fisheries and Wildlife Doctor of Philosophy 2015 i ABSTRACT CHARACTERIZATION OF SEA LAMPREY PHEROMONE COMPONENTS By Cory O. Brant S ea lamprey ( Petromyzon marinus L.) rely upon chemical and environmental information to modulate key aspects of their single reproductive season , including; stream temperature, odor of juveniles , and a male - released mating pheromone containing a main component 3 - keto petromyzonol sulphate ( 3kPZS). In this dissertation , I test the overall hypothesis that pheromone mediated behavior is dependent upon multiple factor s ( i.e . environmental, physiological and social cues, or pheromone components versus mixtures in specific ratios ) . In Chapter 1, field tests suggest 3kPZS functions as a migratory cue in pre - spawn sea lamprey specifically during cold (< 15 °C) stream temperatures . In Chapter 2, field tests confirm a functional shift of 3kPZS from a migratory cue to a proximal mating pheromone as spawning commences . Chapter 3 marks the next step in characterizing the behavioral function of a new mating pheromone component 3,12 - diket o - 4,6 - petromyzonene - 24 - sulfate (DkPES) that functions as a mating pheromone when mixed with 3kPZS. In Chapter 4, an additional male - released compound, identified as petromyzonamine - 24 - monosulfate (PAMS - 24) , is examined in field tests and hypothesize d to function as a mating pheromone that advertises nest boundaries . In Chapter A - 1, a novel dyhydroxylated tetrahydrofuran ( THF) diol fatty acid, compiled of four stereoisomers, is currently under investigation for behavioral activity, and is therefore include d as an a ppende d c hapter to this dissertation. Studies here advance our understanding of the multiple contexts at which pheromones modulate behavior , the origins of chemical communication , and the utility of pheromones for integrated invasive species control . iii To my grandpa George W. Anderson who took me fishing. iv AKNOWLEDGEMENTS I would like to thank my advisor Dr. Weiming Li and members of my graduate committee , Dr s . Michael Wagner, Cheryl Murphy, and Charles Krueger , for their support and insights during these studies. Dr. Ke Li conducted all theoretical and analytical chemistry and Dr. Mar Huertas conducted all electrophysiological experiments for these studies. I thank all personnel at the United States Geological Survey /Great Lakes Science Center Hammond Bay Biological Stat ion in Millersburg, Michigan, Fisheries and Oceans Canada , and the United States Fish and Wildlife Servi ce Marquette Biological Station in Marquette, Michigan, for providing sea lamprey and techn ical support during these projects. I also thank Dolly Trump and Lydia Lorenz for allowing access to the Ocqueoc River field site, and Andy Meyers and family for allowing access to the Trout River field site . Special thanks are well deserved of all field technicians for their work on these projects in day or night, rain or shine : L. Racey, E. Buchinger, J. Riedy, Z. Smillie, B. Vieder, A. S. Chun, D. Partyka, D. Strohm, J. Olds, K. Hill, B. Grieve, , and M. P omranki . Thanks to Dr. Yu - Wen Chung - Davidson for assistance with histology and Dr. Ugo Bussy for assistance with analytical chemistry. I thank the Great Lakes Fishery Commission for funding this research . v TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ..................... viii LIST OF FIGURES ................................ ................................ ................................ ..................... xi INTRODUCTIO N TO DISSERTATION ................................ ................................ .................. 1 CHAPTER 1 A PHEROMONE OUTWEIGHS TEMPERATURE IN INFLUENCING MIGRATION OF SEA LAMPREY ................................ ................................ ................................ .......................... 15 ABSTRACT ................................ ................................ ................................ .............................. 16 INTRODUCTION ................................ ................................ ................................ ..................... 17 METHODS ................................ ................................ ................................ ................................ 19 Animals and tagging ................................ ................................ ................................ .............. 19 Pheromones ................................ ................................ ................................ ........................... 19 Field behavioral t ests ................................ ................................ ................................ ............. 20 Data analysis ................................ ................................ ................................ .......................... 24 RESULTS ................................ ................................ ................................ ................................ .. 26 Movement upstream ................................ ................................ ................................ .............. 26 Movement towards treatment channels ................................ ................................ ................. 30 DISCUSSION ................................ ................................ ................................ ........................... 32 Data accessibility . ................................ ................................ ................................ .................. 32 Competing interests ................................ ................................ ................................ ............... 32 ACKNOWLEDGEMENTS ................................ ................................ ................................ ...... 33 Funding statement . ................................ ................................ ................................ ................. 33 Permission to use published material ................................ ................................ ..................... 34 CHAPTER 2 FUNCTIONAL TRANSITION OF A CHEMICAL CUE INTO A CHEMICAL SIGNAL IN SEA LAMPREY ................................ ................................ ................................ .................... 36 ABSTRACT ................................ ................................ ................................ .............................. 37 INTRODUCTION ................................ ................................ ................................ ..................... 38 METHODS ................................ ................................ ................................ ................................ 41 Field behavioral assay ................................ ................................ ................................ ........... 41 Data analyses ................................ ................................ ................................ ......................... 49 RESULTS ................................ ................................ ................................ ................................ .. 52 Preference for 3kPZS becomes increasingly targeted as females approach ovulation .......... 52 Similar compounds did not yield a behavioral response in migratory subjects .................... 56 Larval sea lamprey release 3kPZS ................................ ................................ ......................... 58 DISCUSSION ................................ ................................ ................................ ........................... 6 0 ACKNOWLEDGMENTS ................................ ................................ ................................ ......... 64 vi CHAPTER 3 MIXTURES OF TWO BILE ALCOHOL SULFATES FUNCTION AS A PROXIMITY PHEROMONE IN SEA LAMPREY ................................ ................................ ........................ 65 ABSTRACT ................................ ................................ ................................ .............................. 66 INTRODUCTION ................................ ................................ ................................ ..................... 67 RESULTS ................................ ................................ ................................ ................................ .. 70 Spectra of synthesized DkPES match those of purified DkPES ................................ ........... 70 DkPES is discriminated from 3kPZS in sea lamprey olfactory epithelia .............................. 72 Females are attracted to the 3kPZS and DkPES mixture at ratios similar to those identified in SMW extracts ................................ ................................ ................................ .................... 74 DISCUSSION ................................ ................................ ................................ ........................... 78 METHODS ................................ ................................ ................................ ................................ 82 Test Subjects ................................ ................................ ................................ .......................... 82 Purity analysis of DkPES ................................ ................................ ................................ ...... 83 Synt hesized Pheromone Components ................................ ................................ ................... 84 Electro - olfactogram (EOG) Recording ................................ ................................ ................. 84 Passive Integrated Transponder (PIT) Tagging Procedures ................................ .................. 85 Field Bioassays ................................ ................................ ................................ ...................... 86 Details of Treatments ................................ ................................ ................................ ............. 87 Swim T rack Mapping ................................ ................................ ................................ ............ 88 Statistical Analyses of Behavioral Data ................................ ................................ ................ 89 ACKNOWLEDGEMENTS ................................ ................................ ................................ ...... 91 SUPPLEMENTAL INFORMATION ................................ ................................ ....................... 92 CHAPTER 4 A TERRITORIAL PHEROMONE THAT DEFINES NEST BOUNDARY IN THE SEA LAMPREY ................................ ................................ ................................ ................................ .. 95 ABSTRACT ................................ ................................ ................................ .............................. 96 INTRODUCTION ................................ ................................ ................................ ..................... 97 RESULTS ................................ ................................ ................................ ................................ .. 98 DISCUSSION ................................ ................................ ................................ ......................... 107 METHODS ................................ ................................ ................................ .............................. 109 Sea lamprey ................................ ................................ ................................ ......................... 109 Wild nesting sea lamprey observations ................................ ................................ ............... 110 Extraction of sea lamprey - conditioned water ................................ ................................ ...... 110 Separation of PAMS - 24 ................................ ................................ ................................ ....... 111 Structural analysis ................................ ................................ ................................ ................ 111 Determin ation of PAMS - 24 concentrations in extracts and wash - water samples ............... 112 PIT tagging and marking for field behavioral studies ................................ ......................... 115 Electro - olfactogram (EOG) responsiveness to PAMS - 24 ................................ ................... 116 Synthesized odorant treatments ................................ ................................ ........................... 116 Experimental designs for preliminary behavioral tests using telemetry .............................. 117 Details of treatments for behavioral tests using telemetry ................................ .................. 119 Statistical analysis of PIT data ( shown in Supplementary Table S4 - 2) .............................. 119 Swim track and plume mapping ................................ ................................ .......................... 120 Statistical analysis of swim tracks and plume mapping ................................ ...................... 121 vii ACKNOWLEDGEMENTS ................................ ................................ ................................ .... 123 SUPPLEMENTAL INFORMATION ................................ ................................ ..................... 124 C ONCLUSION TO DISSERTATION ................................ ................................ ................... 134 CONTROL IMPLICATIONS ................................ ................................ ................................ . 137 APPEN DI X ................................ ................................ ................................ ................................ 142 REFERENCES ................................ ................................ ................................ .......................... 169 viii LIST OF TABLES Table 1 - 1. Multiple effects, their interactions, and fit statistics for each model examined for best explaining the variation in the proportion of sea lamprey that mo ved upstream across treatments. **Final model selected with corresponding AIC and logLik values. i AIC and Akaike weights ( w i ) were calculated using equations ( 1 ) and ( 2 ) above. Relative likelihoods were calculated with the following equation: exp( - i ) . Odor ant refers to 3kPZS, methanol (control), or larval extract treatments. Period refers to early (2000 - 2300h) or late (0000 - 0300h) trials. Temp refers to stream temperature (°C) .. 28 Table 1 - 2. Treatment comparisons at the lower, median, and upper quartiles of str eam temperature in Figure 1 - 2. Odor ant treatments include methanol (control), synthesized 3kPZS (5x10 - 13 M), and larval extract (LE; 5x10 - 14 M PADS). Italicized P - values indicate a significant difference (two - tailed t of the temperature range (Figure 1 - 29 Table 1 - 3. Behavior responses (percentage) of pre - spawn sea lamprey to treatments. Treatments: 3 - keto petromyzonol sulfate (3kPZS, dissolved in 50% methanol), extracted water conditioned with larvae (larval extract, dissolved in 50% methanol), and control (veh icle, 50% methanol). Response variables: percentage swimming upstream (Upstream), percentage entering the sub - channel containing each treatment (Treatment channel), and the percentage entering within proximity of the treatment source (Treatment source). Di fferent lower - case letters within 0.05) 31 Table 2 - 1: Additional preference responses (in addition to Figure 2 - 2a) of migratory female sea lamprey to conspecific - released compounds. See Figure 2 - 2 for explanation of treatments. Trials were conducted during E arly (May) and L ate (June) 2013 migratory season. Additional preference response Down shows the percentage ( n ) of subjects that moved down from release cages, and did not come back upstream during trials. Up refers to subjects that moved upstream from the release cage and continued to swim 205 m to the confluence of the two sub - channels. Treatment source refers to subjects that entered within 0.5 m of the treatment source after entering the sub - channel activa ted with each treatment. Responses that share a letter across treatments are not significantly different ( logistic regression ; = 0.05) 55 Table 2 - 2: Preference responses of migratory female sea lamprey to similar conspecific - released compound s. Treatments included methanol vehicle controls (vehicle vs. vehicle), larval extract at 5x10 - 14 molar (M) benchmark PADS vs. vehicle, a mixture of PADS (1x10 - 12 M ), PSDS (5x 10 - 13 M ) , PZS ( 5x10 - 13 M ) and 3kPZS (5x 10 - 13 M ) vs. vehicle, the same mixture minus 3kPZS vs. vehicle, and synthesized 3kPZS (3kPZS) at 5x10 - 13 molar vs. vehicle. Response variables are consistent with those described in Figure 2 - 2, Table 2 - 1. Responses that share a letter across treatments are not significantly different ( logistic regression ; = 0.05) ............... 57 ix Table 2 - 3: Release rates of larval - released compounds. Larval sea lamprey were sampled from for tributaries of Lakes Michigan and Huron, located in the northern lower peninsula of Michigan. Compounds examined included: 3 - keto petromyzonol sulfate ( 3k PZS) , petromyzonamine disulfate (PADS), petromyzosterol disul fate (PSDS), and petromyzonol sulfate (PZS) . Mean release rates (ng/g - larvae/hr ) from triplicate samples of each batch of larvae ( n ) are shown for each date of collection. Release rates within each compound that share a letter are not significantly differe nt ( ANOVA ; = 0.05) . Compounds PADS and PSDS were not detected ( ND ) in our samples during these dates ... 5 9 Table S3 - 1. High resolution mass spectrum report for synthesized DkPES ammonium salt (HR - ESI - MS) .. .. 92 Table S3 - 2. Percentage of sexually mature female sea lamprey that moved upstream 45 m (Up) to side - by - side nest antennas activat ed with pheromone treatments. Treatments included: 3kPZS (5E - 13 M) vs. 3kPZS (5E - 13 M), spermiated male washings (SMW, applied at 5E - 13 M 3kPZS benchmark) vs. river water, ratio 1:1 (5E - 13 M 3kPZS:5E - 13 M DkPES) vs. 3kPZS (5E - 13 M), ratio 10:1 (5E - 13 M 3kPZS:5E - 14 M DkPES) vs. 3kPZS (5E - 13 M), ratio 20:1 (5E - 13 M 3kPZS: 2.5E - 14 M DkPES) vs. 3kPZS (5E - 13 M), and ratio 30:1 (5E - 13 M 3kPZS:1.67E - 14 M DkPES) vs. 3kPZS (5E - 13 M) . ..93 Table S4 - 1. 1 H (900 MHz, J in Hz) and 13 C NMR (225 MHz) spectroscopic data for PAMS - 24 in DMSO - d 6 . Table S4 - 1 is curtesy of Dr. Ke Li, Michigan State University. The symbol * indicates an overlap with H 2 O in DMSO - d 6 , symbol indicates an overlap with DMSO - d 6 , and a indicates a ssignments are interchangeable ... 125 Table S4 - 2 Numbers of sexually mature female ( a. MF) and mature male ( b. MM) sea lamprey that approached conspecific odorants during field tests, as recorded by telemetry. These responses were recorded with passive integrated transponder telemetry. Treatments are described in Figure 4 - 2. Response variables were monitored with telemetry and include: Down - the percentage of test subjects that moved downstream 5 m or more from release cages, Up - the percentage of subjects that moved upstream at least 5 m or more from the release cage, Enter treatment nest - the percentage of su bjects that entered each respective treatment nest, Enter control nest - the percentage of animals that entered the adjacent control nest. Each response variable was evaluated with logistic re gression. Responses that share a letter are not significantly di fferent ( = 0.05) ... 126 Table A - 1 - 1 . Responses of migratory sea lamprey to (a) fraction pools (1 - 4) and (b) sub - pools (3.1 - 3.3) from larval odor. Trials were conducted over the 2010 and 2011 migratory seasons in the Upper Ocqueoc River, Millersburg, MI. Treatments Vehicle (50% MeOH) and Larval Extract (extracted raw larval odor applied to one sub - channel at a volume achieving 5E - 14 M PADS benchmark and vehicle applied to the adjacent sub - channel) were controls. In 2010, pools in Figure A - 1 - 1b were tested. In 2011, three sub - pools from active Pool 3 were tested, including; Sub - pool 3.1 (Fraction 5), Sub - pool 3.2 (Fraction 6), and Sub - pool 3.3 (Fractio n 7). Responses include Up (percentage moving 200 m upstream to the confluence of the two sub - channels), Treatment channel (of the subjects that moved up, percentage that enter the sub - channel containing each treatment), and Treatment nest (of the subjects that entered the treatment x channel, the percentage that swam through a 1 m 2 nest fixed to the center of the stream bed at the upstream end of the respective sub - channel). Each response across treatments was evaluated with logistic regression. Values that share a letter within each response variable are not significantly different ( = 0.05) ... 149 Table A - 1 - 2. Responses of migratory female sea lamprey to new tetrahydrofuran diol compounds 971 and 973 during Early May and Late June 2013 and 2014 migratory seasons . Trials were conducted over the 2013 and 2014 migratory season in the Upper Ocqueoc River, Millersburg, MI, USA. Treatments included vehicle controls (50% MeOH), Larval Extract controls (Larval extract applied to one sub - channel and vehicle applied to the adjacent sub - channel, 971:973 (1:1, 1E - 12 molar M total), 971 alone (5E - 13 M), and 973 alone (5E - 13 M). Responses include Down (percentage moving downstream of the release cages and not coming back up during the 2.5 hour - long trial), Up (percentage moving 200 m upstream to the confluence of the two sub - channels), Treatment channel (percentage that enter the sub - channel containing each treatment), and Treatment source (of the subjects that entered the treatment channel , the percentage that then passed through a 1 m 2 source antenna fixed to the center of the stream bed at the upstream end of the respective sub - channel). Responses were evaluated with a generalized linear model and binomial distribution. Responses that sha re a letter are not significantly different ( 152 xi LIST OF FIGURES Figure I - 1. Hypothesized role of pheromones in modulating the life history of the sea lamprey. (A) Parasitic adults detach from host fish, cease feeding, navigate the coast, and migrate at night into freshwater tributaries that are activated with the smell of juvenile conspecifics (Plume) in the early spring (Feb. - April) . (B) Nocturnal migrating adults continue to move upstream in the spring (May - June) , guided in part by the smell of conspecific larvae. Males move upstream , mature sexually, establish nests in rocky/ riffle areas of the stream, and begin to release a mating pheromone (June - July) . The main component of the male mating pheromone consists of a bile alcohol identified as 3 - keto petromyzonol sulphate (3kPZS), which functions to draw ovulated females to the male for courtship. Spent adults die after a single spawning season . (C) Eggs hatch and j uvenile larvae remain burrowed in upstream sediments for 4 17+ years where they filter - feed on algae and plant particles carried with the current, releasing the migratory odor as a metabolic by - product with their waste (comprised of bile acids and steroi d derived compounds). (D) During winter months, larval sea lamprey transform into parasitic adults, and move downstream into the lake or ocean where they feed on fish for roughly 1.5 years, thus completing the cycle 5 Figure I - 2. De tails of the bioassay guided fractionation system developed by our laboratory at Michigan State University for identification and characterization of sea lamprey pheromone components. Subjects are kept in holding tanks for collection of sea lamprey conditi oned wash - water. Wash - water is extracted with solid - phase extraction (SPE) across a bed of aliphatic adsorbent resin ( Fine et al., 2006 ) . Extracts from resin are subjected to chromatography to separate fractions or pools of fractionated compounds ( Li et al., 2012 ) . Fractions are tested for olfactory poten cy using electro - olfactogram (EOG) assays ( Li et al., 2014 ) . Active pools, fractions, or components at the olfactory level are tested in the field to evaluate behavioral activity ( Li et al., 2013 b ) . Further separation into pure compounds is conducted with chromatography, and nuclear magnetic reson ance (NMR) is used to identify chemical structure ( Li et al., 2013 b ) . Upon purification or chemical synthesis, behavioral activity is confirmed in the field ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) .. 7 Figure 1 - 1. ( a ) Schematic of the Upper Ocqueoc River, Millersburg, MI, U.S.A. field for female testing only. ( b ) Schematic of Upper Trout River, Rogers City, MI, U.S.A. field site for male testing only. Transecting right ( C R) and left (C L ) PIT antennas recorded the number of test subjects that move into the treatment sub - channel (altern ated each trial). Square - frame O R and O L PIT antennas recorded animals entering near the treatment source .. 22 Figure 1 - 2. Upstream movement of pre - spawn female ( a ) and male ( b ) sea lamprey. trials where 3kPZS was administered to the stream (dashed regression line, females R 2 = 0.59, males R 2 reg ression line, females R 2 = 0.68, males R 2 = 0.57), and = trials where larval extract was administered to the stream (dot - dashed regression line, females R 2 = 0.10, males R 2 = 0.85). q 1 , q 2 , and q 3 = the lower, median, and upper quartiles of temperatures. See Table 1 - 2 for statistical comparisons at each quartile of temperature 27 xii Figure 2 - 1. The section of the Upper Ocqueoc River, Millersburg, MI, U.S.A (T35N, R3E, Sec. 27) used to examine behaviors of sea lamprey in relation to pheromone treatments. ( a ) Full 250 m - long section and details of the naturally bifurcated sub - channels used for examining behaviors of migratory female subjects . Black - dashed lines indicate channel transec ting PIT antennas placed at the confluence of the left ( C L ) and right ( C R ) sub - channels, and upstream and downstream of the release point. ( b ) The 45 m - long section used for examining behaviors of sexually mature female subjects . Hollow boxes represent left (T L ) and right (T R ) 1 m 2 PIT antennas where, in the center of each, treatments were administered. Scale bars = 25 m . 44 Figure 2 - 2. Preference response of sea lamprey to conspecific - released compounds. Treatments included methanol vehicle controls (vehicle vs. vehicle), larval extract (LE) at 5x10 - 14 molar benchmark PADS vs. vehicle, and synthesized 3kPZS (3kPZS) at 5x10 - 13 molar vs. vehicle. ( a. ) The proportion of immature female sea lamprey entering the sub - channel containing each treatment during the Early (May, 2013) and L ate (June, 2013) migratory season. Different upper - case letters indicate a significant proportion moved into the activated treatment sub - channel (logistic regression: X 2 5 = 20.35, P = 0.001). Key histological features examined in immature oocytes included the follicular cell layer ( arrow ) that encompasses an adhesive cell layer, nucleus ( I. in Early , III. in Late ), and build - up of fluid of ovulation ( II. in Early , IV . in Late ). ( b . ) The proportion of Mature ovulated female sea lamprey entering the treatment vs. vehicle source (1 m 2 ) during the mating season in early July. Different upper - case letters indicate a significa nt proportion moved into the activated treatment source ( X 2 2 = 57.25, P < 0.001). Key features examined in Matur e oocytes included a broken (absent) follicular cell layer with exposed, lysed, adhesive cells ( arrow ), and lack of nucleus ( V. ). Scale bars = 100µm ... 54 Figure 2 - 3. Conceptual model of the evolution of pheromone 3 - keto petromyzonol sulfate (3kPZS) in P. marinus . ( a. - product by larvae residing and feeding in streams, along with a suite of other b. ritualized as a navigational cue to migrating adults. ( c. ) Males adapted to massively upregulate - adapted bias to 3kPZS in female receivers. Females adapted to f ine tune their movement towards a 3kPZS source (nest) to synchronize reproduction upon maturation of their gonads. Both parties continued to mutually benefit as 3kPZS evolved from a cue to a signal 63 Figure 3 - 1. Spectral comparisons o f synthesized and natural (purified) DkPES. (A) Comparison of 1 H NMR spectra of natural and synthesized DkPES obtained from 600 MHz NMR spectrometry (Varian Inova) in methanol - d 4 . (B) Comparison of 13 C NMR spectra of natural and synthesized DkPES obtained from 600 MHz NMR spectrometry (Varian Inova) in methanol - d 4 71 Figure 3 - 2. Olfactory detection and discrimination of DkPES and 3kPZS by adult sea lamprey . (A) Semi - logarithmic plots of electro - olfactogram (EOG) responses t o different concentrations of DkPES and 3kPZS . The response amplitude was corrected for the blank response amplitude and normalized against the response amplitude of a standard odorant, L - arginine at 10 - 5 Molar. ( B) Qualitative differences in DkPES and 3kPZS a ssessed by cross - adaptation. Percentage unadapted response is the response amplitude to an odorant (treatment) when the olfactory epithelium was pre - adapted to an odorant, expressed as a percentage of the xiii respons e amplitude to the same testing odorant when the olfactory epithelium was not pre - adapted. 3kPZS adapted and DkPES adapted indicate when the olfactory epithelium was pre - adapted to 3kPZS and DkPES, respectively. D ifferent letters indicate significant diff erences in responses amplitude with respect to the corresponding adaptation ( DkPES: t = 3.78, P = 0.004; 3kPZS: t = 2.81, P = 0.019 ). Vertical bars represent one standard error, n = 6 73 Figure 3 - 3 . Behavioral responses of mature female sea lamprey to DkPES in a natural spawning stream. (A) Schematic of t he 45 m - long section of the Upper Ocqueoc River used for field bioassays. The point at which treatment concentrations reached that of our target whol e - stream molarity was calculated based on rhodamine concentration, and is indicated at S . The plume map and estimated treatment concentrations (Molarity) are shown, and were mapped following described methods ( Johnson et al., 2009 ) . Locations of passive integrated transponder (PIT) antennas are shown. Downstream release cages are indicated with solid black boxes. ( B ) Swim tracks of individual test subjects during 3kPZS (5x10 - 13 M) vs. 3kPZS (5x10 - 13 M) control treatments and 3kPZS ( 5x10 - 13 M) vs. ratio 30:1 (5x10 - 13 M 3kPZS:1.7x10 - 14 M DkPES) treatments are shown starting at point S . ( C) Mean sinuosity (track length/shortest connecting line) of tracks for each treatment (± 1 SEM) was calculated from point S up to adjacent nests (~ 15 m) during treatments: 3kPZS (5x10 - 13 M), spermiated male washings (SMW, applied at 5x10 - 13 M 3kPZS benchmark), ratio 1:1 (5x10 - 13 M 3kPZS:5x10 - 13 M DkPES), ratio 10:1 (5x10 - 13 M 3kPZS:5x10 - 14 M DkPES), ratio 20:1 (5x10 - 13 M 3kPZS: 2.5x10 - 14 M DkPES), and ratio 30:1. Treatments that share a letter are not significantly different ( ANOVA and post - hoc HSD: F 5,74 = 3.19, P = 0.012) . The number of responding subjects ( n ) are indicated within each column ... ............................................................................................................................. ........... . 75 Figure 3 - 4 . Pair - wise comparison of male pheromone components at various ratios for their induction of p reference r esponse in mature female sea lamprey. Treatments included: 3kPZS (5x10 - 13 M, n = 16) vs. 3kPZS (5x10 - 13 M, n = 17), spermiated male washings (SMW, applied at 5x10 - 13 M 3kPZS benchmark, n = 26) vs. river water (RW, n = 1), ratio 1:1 (5x10 - 13 M 3kPZS: 5x10 - 13 M DkPES, n = 16) vs. 3kPZS (5x10 - 13 M, n = 9), ratio 10:1 (5x10 - 13 M 3kPZS:5x10 - 14 M DkPES, n = 9) vs. 3kPZS (5x10 - 13 M, n = 11), ratio 20:1 (5x10 - 13 M 3kPZS: 2.5x10 - 14 M DkPES, n = 8) vs. 3kPZS (5x10 - 13 M, n = 7), and ratio 30:1 (5x10 - 13 M 3kPZS:1.7x10 - 14 M DkPES, n = 12) vs. 3kPZS (5x10 - 13 M, n = 3). Dashed vertical grey lines separate pair - wise comparisons. (A ) Percentage of subjects that entered each treatment nest. Horizontal grey line indicates 50%. Treatments that share a letter are not significantly different (Logistic regression: X 2 5 = 25.51, P < 0.001). (B) Mean (± 1 SEM) retention (min.) of subjects inside respective treatment nests. Treatments that share a letter are not significantly different ( ANOVA and post - hoc F 11,123 = 3.55, P < 0.001 ) 77 Figure 3 - 5 . A c onceptual model for a female sea lamprey encountering the conspecific male pheromone. Nesting males release compound 3kPZS that induces upstream movement in mature (ovulated) females ( Johnson et al., 2009 ; Li et al., 2002 ) . Background 3kPZS concentrations remain very high throughout spawning grounds ( Xi et al., 2011 ) , indicated by (3kPZS) under stream flow. Females are able to detect and discriminate between 3kPZS and minute compounds such as DkPES at the olfactory level, while increasing the sinuosity of their swim path aroun d the odor plume ( Johnson et al., 2012 ) , as they approach the source . 81 xiv Figure S3 - 1. All tracks and plumes for all treatments during field trials. Treatments and ratios are described in Figure 3 - 3 ..94 Figure 4 - 1. ( a . ) Structure of PAMS - 24 ( 1 ) including key 1 H - 1 H COSY (bold lines) and HMBC (arrows) correlations of PAMS - 24 recorded in DMSO - d 6 using a Brüker NMR spectrometer ( 1 H NMR, 900 MHz; 13 C NMR, 225 MHz). ( b . ) Semi - logarithmic plot of normalized electro - olfactogram (EOG) amplitudes recorded in sea lamprey in response to different concentrations of PAMS - 24 purified from spermiating male lamprey washings and PAMS - 24 in synthesized form. Data are the means ± SEM ( n = 6), and are blank corrected and n ormalized to the amplitude of response to 1E - 5 M L - arginine 101 Figure 4 - 2: Details of field behavioral studies testing mature male sea lamprey. ( A ) The 18.5 m - long section of the Upper Trout River used for field bioassays. Downstream release cages are shown as solid black boxes Plumes were mapped and concentrations of treatments (Molarity) were estimated using rhodamine dye concentrations following Jo hnson et al. ( Johnson et al., 2009 ) . The point at which treatm ent concentrations reached that of our target whole - stream concentration is indicated ( S ). ( B ) Swim tracks of mature male subjects during trials where 3kPZS (5x10 - 13 M) was applied to both nests, mature male washings (MMW, applied at 5x10 - 13 M benchmark 3kPZS) was applied to the left nest vs. vehicle, a 100:1 3kPZS:PAMS - 24 ratio (3kPZS 5x10 - 13 M:PAMS - 24 5x10 - 15 M was applied to the left nest vs. 3kPZS at 5x10 - 13 M in adjacent nest, and when 3kPZS alone (5x10 - 13 M) was applied to one nest and Vehicle was applied to the adjacent nest. ( C ) Mean number of sharp turns ( X > 90° ± 1 SEM) of mature male subjects as they approached within 5 m of the treatment sources. Treatment 1:1 was 3kPZS:PAMS - 24 (5x10 - 13 M: 5x10 - 13 M) vs. 3kPZS (5x10 - 13 M). Treatm ents that share a letter are not significantly different (ANOVA and post - hoc F 4,74 = 3.19, P = 0.012). Sample sizes ( n ) are shown within columns . 104 Figure 4 - 3. The proportion of mature male sea lamprey that approached the source of treatments and either directly entered (Enter) the nest (0.5 meter 2 ), or sharply avoided (Avoid) the boundary of the nest. Treatments are explained in Figure 4 - 2 . Responses within each treatment were evaluated with logistic regression (GLM: X 2 4, 64 = 15.66, P = 0.0285). Responses that share a letter are not significantly different ( = 0.05) ... 10 5 Figure 4 - 4. Mean (± 1 SEM) weight (g) of mature male sea lamprey that either entered (Enter) or sharply avoided each treatment. Treatments are explained in Figure 4 - 2. Responses within each treatment was evaluated with a t - test. Responses that share a letter are not significantly different ( = 0.05) .. 106 Figure S4 - 1. In - stream behavioral field sites for o bserving behaviors of sea lamprey to pheromone components. ( A ) The 45 m - long section of the Upper Ocqueoc River, Millersburg, MI, USA, used for observing movement patterns of sexually mature female sea lamprey to male - released pheromones. ( B ) The 18.5 m - long section of the Upper Trout River, Rogers City, MI, USA, used for observing movement patterns of sexually mature male sea lamprey in relation to mature male pheromones. At the upstream end of each site, odorants were applied into the cente r of a square passive integrated transponder (PIT) antennas (hollow boxes, 1 m 2 and 1.5 m apart in xv A, 0.5 m 2 and 1 m apart in B ) . Transecting PIT antennas were placed within each site (grey rectangles) to observe the proportion of subjects moving out of re lease cages (solid boxes), upstream, and hitting on a nest PIT antenna 127 Figure S4 - 2. In - stream swim tracks of mature sea lampr ey to treatments of mature male conditioned wash - water (MMW). Treatments are maintained at consistent stream concentrations based on an in - stream concentration of 5E - 13 M benchmark 3kPZS. Mature female ( MF ) subjects were tested in Figure S1A, where MMW was applied to the right ( a. ) and left ( b. ) nests respectively. Mature male (MM) subjects were tested in Figure S1B, where MMW was also applied to the left ( a .) and right ( b. ) nests respectively. Swim tracks were mapped by manually tracking and recording the path of each subject. Transecting strings (dashed lines) were strung every 1 m downstream of the source to aid in swim track mapping. Plumes (outlined in grey) were mapped using rhodamine dye following Johnson et al. ( Johnson et al., 2009 ) . Scale bars = 1 m . 128 Figure S4 - 3. Wild nesting sea lamprey. ( A ) A single male maintains a nest in the lower Cheboygan River, below the Cheboygan Dam, Cheboygan, MI, USA. ( B ) One male accompanied by 7 females in a nest in the Lower Ocqueoc River, Millersburg, MI, USA 129 Figure S4 - 4. Observations of natural sea lamprey nests . Observations suggest th at larger males (MM) are often accompanied with greater numbers of mature females (MF) per nest ( A : Linear regression: F 1,57 = 3.67, P = 0.060), while mature males rarely join other mature m ales in a nest ( B: Linear regression: F 1,57 = 11.95, P = < 0.001) . 130 Figure S4 - 5. Comparison of carbon resonances of PAMS - 24 with squalamine ( Wehrli et al., 1993 ) and PADS ( Hoye et al., 2007a ) . 13 C NMR data of PAMS - 24 and squalamine were acquired in DMSO - d 6 and displayed in black and blue, respectively. 13 C NMR data of PADS were acquired in methanol - d 4 and displayed in green Figure S4 - 6. Concentrations of washings and release rates of PAMS - 24 compared to 3kPZS. a. M eans +/ - 1 SEM of PAMS - 24 concentrations in washings for larval sea lamprey, immature males (IMs, n = 8), mature males head region only (MM - H) and same mature m al es tail region only (MM - T, n = 7 ), and ovulated females (MF, n = 15). * t 14 = 2.14, P <0.001. b. Mean +/ - 1 SEM of PAMS - 24 concentrations in extracted (SPE) and concentrated washing from mature males head region only (MM E - H ) compared to the same m ature males tail region only (MM E - T, n > 20 ) . c . Release rates of PAMS - 24 and 3kPZS by weight (g) of 6 mature males sampled from a natural spawning stream . The natural release ratio of the two is roughly 1:0.01, 3kPZS:PAMS - 24. PAMS - 24 (Linear: R² = 0.3305) and 3kPZS (Linear: R² = 0.2731) regression lines are shown . 132 Figure S4 - 7. Qualitative differences in a) 3kPZS and PAMS and b) PAMS, PADS and PSDS assessed by cross - adaptation in immature male lamprey. Data are expressed as a percent age of the unadapted response. (SAC), self - adapted control; odorant 1 v. odorant 2, odorant 1 against an adapting solution of odorant 2. Values are means ± S.E. ( n = 6) ... 133 xvi Figure A - 1 - 1. Bioassay guided fractionation pinpointed active compounds. ( A ) Olfactory responses to larval fractions F1 - F9 measured by EOG. ( B ) Mass spectra of pools 1 to 4 of larval sea lamprey fractions, components of 1 ( m/z 329), petromyzonin ( m/z 308), and petromyroxols ( m/z 273) are occurring in pool 3 147 Figure A - 1 - 2. Schematic of the field site in a 250 m - long section of the Upper Ocqueoc River, Millersburg, MI, USA, used for behavioral testing of component 1. The downstream rele ase point is shown, along with the Up and Down passive integrated transponder (PIT) antennas used to monitor subjects moving upstream or downstream, respectively, after release. The upper 45 m of the section is naturally bifurcated by an island. Proportion s of subjects entering each sub - channel was monitored by respective Treatment channel PIT antennas. The proportion of subjects entering the treatment source was then monitored by 1 m 2 Treatment source antennas, where treatments were administered into the s tream ... 148 Figure A - 1 - - - 1b (972), (+) - 1a (973), and (+) - 1b (974 ) 150 Figure A - 1 - 4. Mean ± 1 SEM retention ( sec. ) of female sea lamprey inside artificial nests (within 0.5 m of the source) while pheromone treatments were administered. Treatments included 3kPZS at 5x10 - 13 - 1a ( 971 , 5x10 - 13 M) and (+) - 1a ( 973 , 5x10 - 13 M) at a 1:1 ratio (totaling 1x10 - 12 M), 971 alone (5x10 - 13 M), 973 alone (5x10 - 13 M), and vehicle control (methanol). Trials were conducted in the Ocqueoc River, Millersburg, MI (Figure A1 - 2). Columns with different lower - case letters are significantly different, ANOVA and post hoc Tukey HSD: ( F 4, 123 = 15.58, P = < 0.0001) . 154 1 INTRODUCTION TO DISSERTATION 2 All organisms continuously emit non - communicative compounds into the environment as by - products of life and death . If released or excreted on a reliable basis under specific contexts , some of these compounds can become ritualized ( Tinbergen, 1952 ) to a variety of evolutionary trajectories from ambient odor, to a chemical or mixture that serves as public information (cue) , to a unique chemical signature that functions between members of the same species as a signal ( Symonds and Elgar, 2008 ) . I refer to a cue throughout this dissertation as a source of publi c information that only benefits the receiving individual. C ues can range from environmental /physical ( i.e . changes in stream temperature which may deliver information to fishes that spring is approaching ) , to visual /audible ( i.e. the sight of the turkey vulture ( Cathartes aura ) flying overhead which may indicate a potential food source to other scavengers ) , to chemical ( i.e . the smell of compounds contained in the waste of juvenile s which may indicate that successful reproduction has occurred in the past to adults ) . While cues have no direct benefit to the sender ( i.e. vultures did not evolve to fly overhead to draw other animals to their food source) , a signal is referred to in this dissertation as a form of communication between two members of the same s pecies that evolved to benefit both parties ( Symonds and Elgar , 2008 ) . Pheromones are discussed throughout this dissertation as a chemical signal released from an individual tha t elicit a behavioral or physiologic al response in members of the same species ( Karlson and Lüscher, 1959 ) . Pheromone research over the past several decades has primarily focused on insects (over 80%) with Lepidoptera (butterflies and moths) as the predominant taxa ( Symonds and Elgar, 2008 ) . Studies of pheromone communication in vertebrates remain substantially less prominent ( Symonds and Elgar, 2008 ) . This creates large taxonomic gaps in our knowledge of the origins of c h emical communication , the diversity of compounds that have e volved to function as pheromone mixtures and components, and the multiple factor s 3 (environmental, physiological, and social) that influence behavioral responses to pheromones . The studies presented here focus on pheromone communication in sea lamprey ( Petromyzon marinus , L.), a jawless fish that represents one of the most primitive , extant, vertebrates known to mankind ( Oisi et al., 2013 ) . I refer to the c haracterization of pheromone components throughout this dissertation as elucidation of chemical structure of new compounds , investigation of release and detection of these compounds in conspecifics, investigation of behavioral functions in the field , and investigation of the multiple factor s ( i.e. environmental, chemical, physiological, and social) that may influence behavioral activity of these compounds. The invasive and destructive nature of sea lamprey in the Laurentian Great Lakes, coupled with their dependency upon chemical information to modulate their migration and mating season, has prompted many large scale investigations of sea lamprey pheromone c ommunication ( Li et al., 2007 ; Teeter, 1980 ) . Yet only within the past decade have we truly begun to understand how important pheromones are in modulating the single migratory and mating season of sea lamprey. Parasitic adult sea lamprey detach from their host in early springtime, cease feeding and navigate to river plumes to stage their single reproductive migration ( Figure I - 1A) . Sea lamprey do not home to their natal streams like Salmonids ( B ergstedt and Seelye, 1995 ) ; yet, a nosmic sea lamprey are unsuccessful in locating the river mouth, emphasizing their dependency upon olfactory senses at this time ( Vrieze et al., 2010 ) . A dults are hypothesized to implement a series of search tactics to navigate the coast and locate a river plume, navigate the river mouth relying partially on olfactory information provided by conspecific juveniles, and enter a tributary ( Meckley et al., 2014 ; Vrieze et al., 2011 ) . The odor of juvenile larvae is hypothesized to indicate the presence of high - quality reproductive habitat upstream ( Bjerselius et al., 2000 ; Sorensen et al., 2005 ; Wagner et al., 2009 ) . Sea lamprey 4 migrate upstream in the springtime where they sexually mature and establish spawning nests ( Figure I - 1B). Migration primarily occurs during the night ( Applegate, 1950 ; Manion and Hanson, 1980 ) , and migrants navigate the stream using larval odor ( Bjerselius et al., 2000 ; Teeter, 1980 ; Wagner et al., 2009 ) , stream temperatures ( Binder and McDonald, 2008 ) , and likely other environmental cues ( Binder et al., 2010 ) . Males establish spawning nests in rocky/riffle areas of stream ( Applegate, 1950 ; Manion and Hanson, 1980 ) , and begin to release a mating pheromone ( Li et al., 2002 ; Siefkes et al., 2005 ; Teeter, 1980 ) . The main component of the male mating pheromone has been identified as a bile alcohol 3 - keto petromyzonol sulphate, or 3kPZS ( Li et al., 2002 ; Yun et al., 2003 ) . The spawning season lasts for roughly two weeks in early summer (June) foll owed by the death of spent adults ( Applegate, 1950 ; Manion and Hanson, 1980 ) . Upon hatching, juvenile larvae burrow into stream sediments ( Figure I - 1C) where they can remain for 4 17+ years (average 4 5 years), filter - feeding on algae and other plant p articles carried by stream flow during this time while subsequently releasing a suite of bile acids and steroids as metabolic by - products of their feeding which constitute the larval odor ( Li et al., 1995 ; Sorensen et al., 2005 ) . In the winter months, larvae transform into parasitic adults and complete a migration downstream into the lake or ocean ( Figure I - 1D) . Parasitic adults attach and feed upon other fishes for roughly 1.5 years while steadily increasing their body mass ( Applegate, 1950 ) , thus comple ting the life cycle. 5 Figure I - 1. Hypothesized role of pheromones in modulating the life history of the sea lamprey. (A) Parasitic adults detach from host fish, cease feeding, navigate the coast, and migrate at night into freshwater tributaries that are activated with the smell of juvenile conspecifics (Plume) in the early spring (Feb. - April) . (B) Nocturnal migrating adults continue to move upstream in the spring (May - June) , guided in part by the s mell of conspecific larvae . M ales move upstream , mature sexually, establish nests in rocky/ riffle areas of the stream , and begin to release a mating pheromone (June - July) . The main component of the male mating pheromone consists of a bile alcohol identified as 3 - keto petromyzonol sulphate (3kPZS), which functions to draw ovulated females to the male for courtship. Spent adults die after a single spawning season . (C) Eggs h atch and juvenile larvae remain burrowed in upstream sediments for 4 17+ years where they filter - feed on algae and plant particles carried with the current, releasing the migratory odor as a metabolic by - product with their waste (comprised of bile acids and steroid derived compounds) . (D) During winter months, larval sea lamprey transform into parasitic adults, and move downstream into the lake or ocean where they feed on fish for roughly 1.5 years, thus completing the cycle. 6 Much of the research within the last decade regarding chemical communication in sea lamprey, including the studies presented in this dissertation, has been supported by the Great Lakes Fishery Commission, a binational treaty formed in between the U.S. and Canada (1954) with a mission to control the destructive sea lamprey in the Laurentian Great Lakes and protect the fishery ( Commission, 2001 ) . The fishery in the Great Lakes is worth an estimated $7 billion annually, and the sea lamprey invasion undoubtedly added to the collapse of the fishery throughout the mi d - 20 th century ( Commission, 2001 ; Smith and Tibbles, 1980 ) . Controlling sea lamprey populations in the Great Lakes is an ongoing battle. Fuell ed by a mission to identify and integrate new pheromones into the current arsenal of control techniques used for sea lamprey in the Great Lakes (see Control Implications for a description of the current control techniques) , our laboratory developed a bioassay guided fractionation system at Michigan State University that would allow us to further characterize the structure and function of known pheromone components and identify new components that elicit behavioral or olfact ory responses in sea lamprey ( Figure I - 2). We assembled an integrated team of researchers at Michigan State University to develop the bioassay guided fractionation system. Led by primary investigator Dr. Weiming Li, our team is comprised of Dr. Ke Li (anal ytical and theoretical chemist), Dr. Mar Huertas (ele ctrophysiologist), and myself (b ehavior). Several new compounds have been identified and published by our team using this system, and more are continuing to show stereochemistry across a broad range of c hemical diversity ( Appendix, Publication List ) . 7 Figure I - 2. Details of the b ioassay guided fractionation system developed by our laboratory at Michigan State University for identification and characterization of sea lamprey pheromone components. Subjects are kept in holding tanks for collection of sea lamprey conditioned wash - water. Wash - water is extracted with solid - phase extraction (SPE) across a bed of aliphatic adsorbent resin ( Fine et al., 2006 ) . Extracts from resin are subjected to chromatography to separate fractions or pools of fractionated compounds ( Li et al., 2012 ) . Fractions are tested for olfactory potency using electro - olfactogram (EOG) assays ( Li et al., 2014 ) . Active pools, fractions, or components at the olfactory level are tested in the field to evaluate behavioral activity ( Li et al., 2013 b ) . Further separation into pu re compounds is conducted with chromatography , and nuclear magnetic resonance (NMR) is used to identify chemical structure ( Li et al., 2013 b ) . Upon purification or chemical synthesis, behavioral activity is confirmed in the field ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) . 8 My overall hypothesis tested in this dissertation is that behavioral responses to pheromones are modulated by multiple environmental, physiological, chemical, and social factors . Specifically, I predict animals are faced with a hierarchy of information based on certain contexts f rom environmental factors such as temperature, physiological factors s uch as sexual maturation, chemical factors such as pheromone components or mixtures of pheromones, and social factors such as inter - and intraspecific interactions with others , which ultimately must be sorted for optimal fitness payoffs on the signaller and receiver ends . Over the past decade, t he hypothesized main mating pheromone component, 3kPZS , was extensively studied in the field for its mating function in sea lamprey , as well as for its applicability for sea lamprey control . Male sea lamp rey were found to und ergo a massive upregulation of bile salt synthesis in their livers upon sexually maturing ( Brant et al., 2013 ) , at which point 3kPZS was found to be released across specialized gill epithelia cells at a rate of 0.5 mg/hr ( Sie fkes et al., 2003 ) . In the field, 3kPZS was found to draw significant numbers of ovulated femal es to the odorant source point consisting of 1 m 2 area of the stream ( Johnson et al., 2006 ; Johnson et al., 2012 ; Johnson et al., 2009 ; Siefkes et al., 2005 ) . Therefore 3kPZS was targeted as a potential lure to increase trap efficacy ( Johnson et al., 2013 ) . Extensive field tests of 3kPZS began to suggest that the component may not solely function in mating, but may have a behavioral function in pre - spawn adults as well which interacts with water temperature (Brant, 2011 MSc The sis, Michigan State University) . E nvironmental factors modulate movement in organisms across the animal kingdom. In aquatic organisms, environmental cues such a s water temperature, river discharge, turbidity, and lunar cycle ( Forsythe et al., 2012 ; Quinn and Adams, 1996 ) have been shown to modulate behavioral aspects of fish migration. Temperature fluc tuations have been hypothesized as a 9 primary migratory cue in teleost fishes ( Moore et al., 2012 ; Skov et al., 2010 ) , as well as in lampreys including Pacific lamprey ( Entosphenus tridentatus ) ( Clemens et al., 2012 ) , river lamprey ( Lampretra fluviatilis ) ( Kemp et al. , 2011 ) , and sea lamprey ( Binder and McDonald, 2008 ; Binder et al., 2010 ) . However, interactions between environme ntal information and pheromone components in modulating behavior of animals has not been described, and little is known regarding other contexts in which pheromone components function. temperature in e component 3kPZS a compound previously only considered as a mating ph eromone, and stream temperature previously shown to influence navigation of migrating sea lamprey ( Binder and McDonald, 2008 ; Binder et al., 2010 ) , interact to modulate pre - spawn migratory behavior of sea lamprey. D uring cold stream temperatures (< 15 °C) when migrating sea lamprey did not often move upstream , presence of 3kPZS overr ode temperature - induced inactivity and increased upstream movement by over 40%. Chapter 1 presents an example of the hierarchy of contra dictory information and context that animals face when making decisions . Chapter 1 is publishe d in Royal Society Open Science . Upon discovering the overriding effect of 3kPZS on stream temperature in migrating female sea lamprey, I hypothesized that 3kPZS must have undergone a functional shift from a cue during migration to a more directionally proximal and reliable mating signal during the spawning season, given that the mating pheromone function of 3kPZS has been well documented and the mating season for shift of 3kPZS ( i.e. from a cue for migration to a directed and coevolved signal for mating) can 10 still be seen by examining the behavior of female sea lamprey to sources of 3kPZS during three key stages of their life history: early migration (May), late migration (June), and spawning (June/July). Using a field test , I provided behavioral evidence that pre - spawn female sea lamprey shift their preference towards a sub - channel activated with 3kPZS (up to 70%) specifically from ovulation once a pref erence for the 3kPZS activated sub - channel began to appear. Chapter 2 provides insights into the origins of stable communication signalling systems , and is currently submitted to Behavioral Ecology . Chapter 2 is formatted in accordance with the respective Behavioral Ecology guide for authors. Given the life history of the sea lamprey in relation to pheromones ( Figure I - 1), and results from Chapters 1 and 2, it becomes apparent that females continue to become more proximally bias to the odor of mature males, likely to insure synchrony of reproduction. Since vison becomes increasingly degenerated during the spawning phase of adult sea lamprey ( Binder and McDonald, 2007 ) , and a background concentration of 3kPZS remains high within spawning grounds during the mating season ( Xi et al., 2011 ) , I hypothesize that additional components released from males provide further information required for approaching females to locate a mate with accuracy. I predict ed that mixtures of minor components with the main mating pheromone 3kPZS provide additional information to females that allow them to locate the releaser male with high accuracy on spawning grounds. Insects such as moths have been shown to broadcast mixture s of pheromone components that are emitted at specific ratios ( Linn and Roelofs, 1989 ; Reyes - Garcia et al., 2014 ; Wyatt, 2010 ) . Individual components and mixtures that are inconsistent with the natural ratio still often yield behavioral response in moths, yet test subjects often show a peak preference to the ratio that best reconstructs that of the ratio emitted 11 by conspecific se nders ( Cardé et al., 1977 ; Coracini et al., 2001 ; Reyes - Garcia et al., 2014 ) . The specificity of pheromon e components and their mixtures in specific ratios have not been evaluated in vertebrates. In recent years our bioassay guided fractionation system ( Figure I - 2) has yielded several components from sea lamprey that are highly stimulatory in the olfactory epithelia of conspecifics and are released at lower rates than 3kPZS . Recent discoveries include a new polyhydroxysteroid named petromyzestrosterol ( Li et al., 2012 ) , several fatty acid - derived hydroxylated tetrahydrofurans ( Li et al., 201 4 ; 201 5 ) , a new hexahydrophenanthrene sulfate named petromyzonin ( Li et al., 2013a ) , and a new sulphated bile alcohol identified as 3,12 - diketo - 4,6 - petromyzonene - 24 - sulfate, or DkPES ( Li et al., 2013 b ) . Field tests suggest that DkPES increases ovulated female preference for the odorant source when combined with 3kPZS ( Li et al., 2013 b ) , but details of ratios and their influence on behavioral response s remained unknown . In Chapter 3, pheromone in sea lampre I hypothesize d that DkPES functions a proximal pheromone in ovulated females when mixed with 3kPZS, allowing them to efficiently locate the odorant source. In the field, I provided evidence that ovulated females showed the greatest preference to a sour ce of 3kPZS and DkPES when combined at the ratio that best matches that which is naturally occurring in wash - water collected from spermiated males ( SMW , 30:1, 3kPZS:DkPES). Further, I characterize search behavior (sinuosity of swim paths) of females approa ching a range of 3kPZS:DkPES ratios. An increase in search activity occurs when subjects approach the SMW and the 30:1 ratio source . Unique pheromone ratios may provide utility in pheromone - integrated control of invasive sea lamprey in the Great Lakes. Cha pter 3 is the first 12 example of pheromone ratios as mating pheromones in animals outside of insects, and is currently submitted to PLOS ONE. Chapter 3 is formatted in accordance with the respective PLOS ONE guide for authors. Recently, a sulfated steroid identified as petromyzonamine - 24 - monosulfate (PAMS - 24) was also discovered in SMW . In Chapter 4, I hypothesize d that PAMS - 24 also functions as a p roximal pheromone in ovulated females when mixed with 3kPZS , similarly to DkPES . Further, I predicted that PAMS - 24 may dual - function as a component of a territorial pheromone that delineate s nest boundaries among males . Behavior of male sea lamprey to mating pheromone components was virtually unknown in the field up to this point . Beginning in 2013, the U.S. Fish and Wildlife Service ceased their sterile male release program ( Bergstedt and Twohey, 2007 ) allowing male sea lamprey to beco me available for field research. Preliminary field tests using sexually mature male subjects suggested that males were drawn to a source of 3kPZ S, yet avoid a source of SMW. Combining the se results from preliminary field tests with observations that nests can be constructed in high densities, often contain only one male, and nesting males are known to be aggressive to intruder males, evidence was mounting for the existence of a territorial pheromone that is released from sexually mature males ( Teeter, 1 980 ) . I provide d behavioral evidence that SMW , 3kPZS and PAMS - 24 treatments applied to artificial nests in the field were preferred by females , yet PAMS - 24 a nd SMW averted males. Behavioral assays conducted thus far suggest that PAMS - 24 may function as the first identified territorial pheromone component in a vertebrate. Chapter 4 is in preparation for Science and is formatted in accordance with the respective guide for authors. 13 Ch apter A - characterization of these compounds is currently underway . These fatty acids were originally extracted from wash - water conditioned with larval sea lamprey, and thought to be a single compound . Upon further chemical investigation, it was learned that four compounds existed ; two stereoisomers ; 9,(12) - oxy - 10,13 - dihydroxystearic acid (1a) and 10,(13) - oxy - 9,12 - dihydroxystearic acid (1b) consisting as a mixture of (+) - - 1a (971), (+) - - 1b (972). Compounds 972 and 974 were not detected in olfactory epithelia of adult sea lamprey, and so were not tested in the field. Compounds 971 and 973 were olfactory stimulants and were therefore tested in the field during the migratory season at the field site shown in Chapter 1 and 2. Migrating females began to show a preference for 973 alone and 971:97 3 mixtures (1:1, mola r:molar) during late migration (June), but no response was seen when 971 alone was applied. To my surprise during these trials, old subjects that were released during previous trials that spring ( i.e. old subjects that were nd retained a surgically - implanted passive integrated transponder [PIT] tag) began swimming back within the field site each night, locating the odor source (1 m 2 PIT antenna) specifically when a mixture of 971:973 (1:1, molar:molar) w as administered into the center of the 1 m 2 antenna , and remaining on the source (averaging 530 ± 105 sec. inside the 1 m 2 nest emitting the 971:973 mixture ). Over 70 subjects returned to the source of 971:973. Four subjects were hand - grabbed while responding to the 971: 973 mixture . Histological examination of oocytes confirmed these four subjects were ovulated. The four females spent an average of 15 ± 1.6 days at large in the river system before returning to the site. Compounds 971 and 973 were later found to be relea sed by mature male sea lamprey, but field tests presented in the appended Chapter A - 1 were not designed to evaluate a mating 14 function of these fatty acids. Field tests similar to those in Chapters 3 and 4 are underway to further evaluate behavioral activit y of 971:973 in spawning phase sea lamprey, and determine whether stereochemistry of compounds represents yet another context that influences pheromone function in sea lamprey (see Appendix, Chapter A - 1 ). Studies presented in this dissertation descri be several lines of evidence that directly support the overarching hypothesis of this dissertation. Environmental, physiological and social information interact with pheromone mixtures and components in modulating behaviors, and these interactions may not be observable in laboratory conditions. Additional minor pheromone components in combination with 3kPZS may serve as additional information used by females to l ocate a nesting male . Understanding the context in which pheromones function is applicable for research in behavioral ecology, chemical communication, and olfaction, and will provide insights into the origins of chemical signal s . Finally, it is imperative that we understand the multiple contexts in which pheromones function if we are to integrate pheromones into control techniques of invasive sea lamprey in the Laurentian Great Lakes ( Joh nson et al., 2013 ; Li et al., 2007 ) . 15 CHAPTER 1 A PHEROMONE OUTWEIGHS TEMPERATURE IN INFLUENCING MIGRATION OF SEA LAMPREY Brant C . O . , Li K . , Johnson N . S . , Li W. 2015 . A pheromone outweighs temperature in influencing migration of sea lamprey. R. Soc. Open S ci. 2: 150009. http://dx.doi.org/10.1098/rsos.150009 16 ABSTRACT Organisms continuously acquire and process information from surrounding cues. While some cues complement one another in delivering more reliable information, others may provide conflicting information. How organisms extract and use reliable information from a multitude of cues is largely unknown. We examined movement decisions of sea lamp reys ( Petromyzon marinus L.) exposed to a conspecific and an environmental cue during pre - spawning migration. Specifically, we predicted that the mature male - released sex pheromone 3 - keto petromyzonol sulfate (3kPZS) will outweigh the locomotor inhibiting effects of cold stream temperature (< 15°C). Using large - scale stream bioassays, we found that 3kPZS elicits an increase (> 40%) in upstream movement of pre - spawning lampreys when the water temperatures were below 15°C. Both warming temperatures and consp ecific cues increase upstream movement when the water temperature rose above 15°C. These patterns define an interaction between abiotic and conspecific cues in modulating animal decision making, providing an example of the hierarchy of contradictory inform ation. 17 INTRODUCTION Environmental cues ( e.g. abiotic information) and signals ( e.g. an entity evolved by a sender that elicits an evolved behavioral response in a receiver) provide a constant input of information used by organisms in decision making processes ( Davies et al., 2012 ; Laidre and Johnstone, 2013 ) . Exactly when certain information becomes more reliable in the decision making process is likely shaped by natural selection ( e.g. ( Laidre and Johnstone, 2013 ) . How organisms distinguish reliable information from a multitude of environmental, conspecific, and heterospecific cues remains largely unknown. Among abiotic cues, water temperature is critical in influencing the timing of fish migration ( Lucas et al., 2001 ) . Pheromones ( Karl son and Lüscher, 1959 ; Wyatt, 2010 ) are signals that may inform fish of mate readiness and/or habitat suitability ( Wyatt, 2010 ) . The sea lamprey relies on environmental and conspecific cues during reproduction. Specifically, when water temperatures drop below 15 °C, sea lamprey become increasingly less active. Warmer water temperatures correlate with increases in locomotor activity of p re - spawn migrating sea lamprey, likely as an adaptation to reach spawning grounds before full maturation of gametes occurs. ( Binder and McDonald, 2008 ) . In addition, the odor of s tream - resident larvae guides pre - spawning sea lamprey in selecting suitable spawning habitat ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . Later, sexually mature female sea lamprey rely on a sex pheromone released by mature males (3 - keto petromyzonol sulphate or 3kPZS) ( Li et al., 2002 ) t o locate nesting mates ( Johnson et al., 2009 ) . We hypothesized that the conspecific male sex - pheromone (3kPZS) contains more reliable information compared to water temperature regarding spawning conditions upstream. 18 Here we report evidence that both male and female pre - spawn adults move upstream in the presence of 3kPZS within the otherwise locomotor inhibitory temperature range (< 15°C), which is consistent with the hypothesis. While test subjects move upstream, they do not bias towards a channel - side containing the source of 3kPZS during our studies, a previously described behavior in mature female conspecifics ( Johnson et al., 2009 ; L i et al., 2002 ) . 19 METHODS Animals and tagging Procedures involving sea lamprey were approved by the Michigan State University Institutional Animal Care and Use Committee (AUF# 05/09 - 088 - 00). Immature adult sea lamprey were captured by the United States Fish and Wildlife Service and Fisheries and Ocean s Canada from tributaries to Lake Michigan and Lake Huron, in May - June 2009, 2010 and 2012, and transported to the United States Geological Survey Hammond Bay Biological Station (HBBS) for further procedures. Sex and maturity determination, and animal hous ing, followed procedures described ( Johnson et al., 2009 ) . P re - spawn sea lamprey refers to river - migrating adults that are non - feeding, yet not fully sexually matured. Pre - spawn sea lamprey where implanted with a 23 mm - long half duplex passive integrated transponder (PIT) tag (Oregon RFID, Portland, Oregon, U.S.A.) through a 3 mm lateral incision in the mid - abdominal region. Pheromones Use of 3kPZS (synthesized by Bridge Organics, Vicksburg, Michigan, U.S.A.; purity >97%) in the stream was permitted by the United States Environmental Protection Agency (experimental user permit 75437 - EUP - 2). A 10 mg ml - 1 stock solution of synthesized 3kPZS (in 50% methanol) was prepared. 3kPZS stock solution was stored at - 80°C until use in the field. Larval extracts, extracts of water conditioned with larval sea lamprey ( Fine et al., 2006 ; Li et al., 2013 a ) , were used as a positive control to validate the experimental sy stem ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . To collect extracts of water conditioned with larval sea lamprey, over 20,000 larval sea lamprey were held in flowing 500 litre - capacity tanks at HBBS from April - August 2008. Larvae were given a sand substrate for refuge and fed yeast weekly. Tank 20 flows were shut off between ~2000 0800 hours (h) allowing larval odor to concentrate. Larval - conditioned water was passed through vertical columns containing 500 g of methanol - activated Amberlite XAD7HP resin (Sigma - Aldrich, St. Louis, Missouri, U.S.A.) using peristaltic pumps (Masterflex 7553 - 70, Cole - Parmer, Vernon Hills, Illinois, U.S.A.). Loading speed was ~300 ml min - 1 . Three columns were loaded for up to 24 hours at a time. Each column was then eluted with 4 - l of methanol. Eluents were concentrated using a model R - 210 roto - evaporator (Buchi Rotovapor, Flawil, Switzerland) and stored at - 80°C. All larval extracts were f ully thawed, pooled, and thoroughly mixed before further analyses were conducted. Petromyzonamine disulphate (PADS), a component of larval extract ( Sorensen et al., 2005 ) , was used as a benchmark when calculating the volume of larval extract needed to activate the stream with PADS at a concentration of 5x10 - 14 molar (M). The concentration of PADS in the larval extract was determined using high performance liquid chromatography tandem mass spectrometry ( Li et al., 2013 a ) . Field behavio ral tests Subjects were tested at night, when migrating sea lamprey move upstream ( Stier and Kynard, 1986 ) , in experimental sites shown in Figure 1 - 1 . Barriers near each river - mouth prevent wild migrating sea lamprey from entering the upper reaches. Females were tested in a separate field site (Figure 1 - 1a) than males (Figure 1 - 1b) to prevent unwanted reproduction of this invasive species above each barrier. The most upstream section of each site was bifurcated by a natural island, which separated two sub - channels of similar hydrologic and physical qualit ies. Test odor ants were diluted with 30 L of river water in large mixing bins. Bins were kept consistent for each test odor ant to reduce the potential for contamination during dilution. Each solution was pumped into respective sub - channels through separate latex tubes at a rate of 167 21 ml min - 1 (± 5 ml min - 1 ) over the span of three hours using peristaltic pumps (Cole - Parmer). A test odor ant was administered to one sub - channel (activated channel, herein) while an equal volume of methanol was administered into the adjacent sub - channel (control channel, herein), and the activated and control channels were alternated each trial. Trials were conducted between 13 - May and 11 - June 2009, between 3 - May and 8 - June 2010 (females), and between 19 and 31 - May 2012 (males) at night. Stream discharge was estimated every three days, or after every precipitation event, at a fixed location in the stream using a Marsh - McBirney portable flow meter (Flo - Mate 2000, Fredrick, Maryland, U.S.A.) to determine the amount of odor ant stock solution to apply to the stream and maintain consistent concentrations across trials. Stream flow was relatively uniform within each site, and flows were lower on average in the male - only site (950 - 1200 m sec. - 1 in Figure 1 - 1a, and 200 - 320 m sec. - 1 in Fig ure 1 - 1b). Up to two trials were conducted each night, depending upon animal availability. The early trial was conducted from sundown ~ 2030 2320 h, and a late trial was then run from ~ 0010 0310 h (trial times were dependent upon when sundown occurred ). Twenty or 30 PIT - tagged sea lamprey (depending on animal availability) were removed from holding tanks at HBBS and transported to their respective stream acclimation cage at the release point (Figure 1 - 1) between 0300 0500 h the night prior to experim entation. Acclimation/release cages were mesh aluminium (~ 0.25 m 3 ) consisting of a sliding door that was removed manually upon release. Animals were then allowed an acclimation period in the stream for 15+ hours. Mortality during acclimation was < 0.5%. 22 Figure 1 - 1. ( a ) Schematic of the Upper Ocqueoc River, Millersburg, MI, U.S.A. field for female testing only. ( b ) Schematic of Upper Trout River, Rogers City, MI, U.S.A. field site for male testing only. Transecting right ( C R) and left (C L ) PIT antennas recorded the number of test subjects that move into the treatment sub - channel (altern ated each trial). Square - frame O R and O L PIT antennas recorded animals entering near the treatment source. 23 Each trial was three hours long. Stream temperature was recorded at the start of each trial. In the first hour of each trial, the test odor ant was administered to the stream. At the start of the second hour, test animals were released. During the remaining two hours, animal behavior s were monitored while odor ants were administered. The second trial began 30 minutes after the first. Test odor ants were kept consistent for each night of trials. As an example, if 3kPZS was tested during the early trial, 3kPZS w as also tested in the opposite sub - channel during the late trial to prevent the possibility of any unwanted contamination from other test odor ants. No animals were umber prevented any pseudo - replication from test subjects released during previous nights. Movement data were consolidated and stored using a multiplexor (Oregon RFID, Portland, Oregon, U.S.A.). Data were uploaded each trial night using a hand - held Meazura model MEZ1000 personal digital assistant (Aceeca International Limited, Christchurch, New Zealand). Treatments were administered from a fixed point in the centre of square - frame PIT antennas at the upstream end of each sub - channel (Figure 1 - 1). PIT anten nas were tuned to a detection sensitivity of roughly 0.3 m from the edges. Scan frequencies were programmed to 3 scans sec. - 1 . Treatments included: (1) 3kPZS (5x10 - 13 M) administered into one sub - channel (methanol into the adjacent sub - channel), (2) larval extract (5x10 - 14 M PADS) administered into one sub - channel (methanol into the adjacent sub - channel), and (3) 50% methanol (in de - ionized water) into both sub - channels for vehicle controls. Molar concentrations were the estimated final concentrations in th e activated sub - channel of the stream. 24 Data analysis The dependent variable, proportions that moved upstream for 205 m to the channel confluence, were arcsine transformed and examined for violations of assumptions of normality and variance homogeneity using the Univariate Procedure in SAS (SAS Incorporated, Cary, North Carolina, U.S.A.) before conducting statistical analyses. Upon observing no violations of the assumptions, analysis of covariance (ANCOVA, = 0.05) was used to test which explanatory variables influenced the upstream movement of pre - spawn fema le sea lamprey. The explanatory variables tested were: (1) treatment (fixed), (2) date (including year: random), (3) stream temperature (fixed), (4) rate stream temperature decreased per trial (fixed), (5) time when trial was conducted (fixed), and (6) the number of animals released per trial (fixed). The final model was selected based on residual log likelihood (LogLik) and weighted AIC ( w i ) values ( Binder et al., 2010 ) . The difference between the model with the lowest AIC and each additional model presented was calculated using equation ( 1 ): ( 1 ) where i is the difference between the best fitting model ( AIC min ) and each model ( AIC i ). The normalized relative likelihood values were calculated using equation ( 2 ): ( 2 ) where w i is the weighted AIC value (Akaike weight) determined by dividing the relative likelihood of each model ( exp ( - i )) by the sum of relative likelihoods of all models. Differences of least squares means (two - tailed t test, = 0.05) of the proportions of animals moving upstream per trial were examined at the lower (q 1 ), median (q 2 ), and upper (q 3 ) quartiles of temperature across treatments (SAS). Quartiles in Figure 1 - 1 were determined by 25 first dividing the temperature data into two halves at the median (which was itself calculated as a mean of the two middle data). The median of the lower half of the data (lower quartile) and the median of the upper half (upper quartile) were then calculated. To determine whether pre - spawn conspecifics showed a more proximal preference towards treatments, we used logistic regression with a binomial distribution (R version 2.11.1, Vienna, Austria) to examine: (1) total number that moved up stream to the confluence of the sub - channels, (2) total number that entered the sub - channel activated with each treatment, and (3) total number in the treatment channel that entered the square administration point. Details of similar statistical analyses h ave been described ( Johnson et al., 2009 ; Li et al., 2013 b ) . No signs of nonlinearities or overdispersion were observed. All responses were compared to those of methanol control trials. 26 RESULTS Movement upstream Female responses to 3kPZS during 2009 and 2010 were consistent ( F 1, 7 = 3.59 , P = 0.100) and combined in the analysis. Data from males were analysed separately because they were tested in a separate stream with different dimensions. The following variables did not influence the proportion of migrating females moving upstream across trials in our experimental system: the rate of stream temperature decrease per trial (°C hr - 1 ), time of trial (early or late), or the number of immature adult females released per trial (20 or 30) (ANCOVA: F 1,10 = 0.97 , P = 0 .348; F 1,12 = 0.42 , P = 0.530; F 1,11 = 0.01 , P = 0.919; F 1, 12 = 0.40 , P = 0.540, respectively), and were removed from the final model. The final model considered effects of treatment, stream temperature, and their interaction as best explaining variability in the proportion of pre - spawn adult sea lamprey moving upstream (females: F 2, 19 = 9.36 , P = 0.002, males: F 2, 3 = 11.26 , P = 0.040, Figure 1 - 2). Statistical fitness values were lowest in the final model (Females: logLik = - 25.3, AIC = - 21.3 , w i = 0.22; Table 1 - 1 ). This model was then used for analyses of male responses. Post hoc tests compared upstream responses at each temperature quartile. Overall, higher numbers of pre - spawn sea lamprey moved upstream in the presence of 3kPZS compared to meth anol controls, specifically at the lower quartiles of temperature (~15 °C) as seen in Figure 1 - 2 (females: t 19 = 5.16, P < 0.001, males: t 3 = 1.22, P = 0.310, Table 1 - 2). 27 Figure 1 - 2. Upstream movement of pre - spawn female ( a ) and male ( b ) sea lamprey. trials where 3kPZS was administered to the stream (dashed regression line, females R 2 = 0.59, males R 2 regression line, females R 2 = 0.68, males R 2 = 0.57), and = trials where larval extract was administered to the stream (dot - dashed regression line, females R 2 = 0.10, males R 2 = 0.85). q 1 , q 2 , and q 3 = the lower, median, and upper quartiles of temperatures. See Table 1 - 2 for statistical comparison s at each quartile of temperature. 28 Table 1 - 1. Multiple effects, their interactions, and fit statistics for each model examined for best explaining the variation in the proportion of sea lamprey that moved upstream across treatments. Effect(s) Interaction(s) logLik AIC i AIC exp( - i ) w i odor ant; temp odor ant*temp - 25.3 - 21.3 0.00 1.00 0.22 * temp - 12.6 - 6.6 14.70 0.48 0.10 odor ant*period - 4.8 1.2 22.50 0.32 0.07 temp*period - 6.8 - 0.8 20.50 0.36 0.08 odor ant; temp; period odor ant*temp; odor ant*period; odor ant*temp* period - 5.8 - 1.8 19.50 0.38 0.08 temp; period temp*period - 5.8 0.2 21.50 0.34 0.07 odor ant*temp - 7.9 - 3.9 17.40 0.42 0.09 odor ant; period odor ant*period - 4.8 1.2 22.50 0.32 0.07 period - 1.6 4.4 25.70 0.28 0.06 odor ant - 12.6 - 6.6 14.70 0.48 0.10 date 3.7 7.7 29.00 0.23 0.05 *Final model selected with corresponding AIC and logLik values. i AIC and Akaike weights ( w i ) were calculated using equations ( 1 ) and ( 2 ) above. Relative likelihoods were calculated with the following equation: exp( - i ) . Odor ant refers to 3kPZS, methanol (control), or larval extract treatments. Period refers to early (2000 - 2300h) or late (0000 - 0300h) trials. Temp refers to stream temperature (°C). 29 Table 1 - 2. Treatment comparisons at the lower, median, and upper quartiles of stream temperature in Figure 1 - 2. Females methanol vs. 3kPZS 3kPZS vs. LE methanol vs. LE temperature t 19 P t 19 P t 19 P q 1 15.5°C 5.16 <0.001 2.61 0.017 5.96 <0.001 q 2 17.9°C 4.59 <0.001 1.58 0.130 4.98 <0.001 q 3 19.7°C 2.76 0.012 0.36 0.726 3.13 0.006 Males methanol vs. 3kPZS 3kPZS vs. LE methanol vs. LE temperature t 3 P t 3 P t 3 P q 1 14.4°C 1.22 0.310 1.86 0.161 1.16 0.330 q 2 15.6°C 0.66 0.558 0.89 0.439 0.49 0.657 q 3 17.1°C 0.46 0.678 1.57 0.214 1.51 0.229 Odor ant treatments include methanol (control), synthesized 3kPZS (5x10 - 13 M), and larval extract (LE; 5x10 - 14 M PADS). Italicized P - values indicate a significant difference (two - tailed t range (Figure 1 - 2). 30 Movement towards treatment channels Pre - spawn sea lamprey did not prefer the 3kPZS treatment channel over the adjacent vehicle (methanol) channel across a full migratory season, yet both preferred the channel with larval extract over the adjacent vehicle channel. The same was true for within 0.5 - 0.25 m of the treatment source (Table 1 - 3). 31 Table 1 - 3. Behavior responses (percentage) of pre - spawn sea lamprey to treatments. Sex Treatment Trials Released ( N ) Upstream ( n ) Treatment channel ( n ) Treatment source ( n ) Female Control 27 617 55% (342) a 50% (171) a 15% (25) a Female Larval extract 6 140 85% (119) b 86% (102) b 61% (62) b Female 3kPZS 22 489 71% (346) c 42% (146) a 18% (27) a X 2 59.56 67.03 53.21 df 2 2 2 P - value < 0.001 < 0.001 < 0.001 Male Control 7 140 64% (90) a 46% (41) a 76% (31) a Male Larval extract 5 100 83% (83) b 71% (59) b 90% (53) a Male 3kPZS 6 120 72% (86) a 55% (47) a 77% (36) a X 2 10.58 11.94 4.72 df 2 2 2 P - value 0.005 0.003 0.095 Treatments: 3 - keto petromyzonol sulfate (3kPZS, dissolved in 50% methanol), extracted water conditioned with larvae (larval extract, dissolved in 50% methanol), and control (vehicle, 50% methanol). Response variables: percentage swimming upstream (Upstream ), percentage entering the sub - channel containing each treatment (Treatment channel), and the percentage entering within proximity of the treatment source (Treatment source). Different lower - case letters within each response variable within each sex indica 0.05) . 32 DISCUSSION The male pheromone, 3kPZS outweighed the effects of cold temperatures in modulating upstream movement of pre - spawning female sea lamprey. These results suggest that, in addition to functioning as a male - released sex pheromone ( Johnson et al., 2009 ) , 3kPZS also functions as an indicator of the onset of spawning. Pre - spawn migrating sea lamprey showed a positive directional response towards proximity of the larval extract source, which remained consistent with previous studies ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . Upstream movement at the lower quartile of temperature in males was not significant, yet the trend remained. Males, as releasers of 3kPZS, likely rely on a slightly different multitude of cues to establish nesting sites and begin to signal females. Selective pressures favour the use of multiple sources of abiotic and biotic information to promote synchrony and reproductive success in animals that migrate long distances. Environmental cues such as temperature and stream flow ( Binder et al., 2010 ) , and conspecific cues such as larval odor s and information regarding spawning ground size and quality ( i.e. 3kPZS) may form a hierarchy of contradictory information that, when received in specific context, allow indivi duals to optimize timing of aggregation for reproduction. This study compares reliability of a particular abiotic cue and conspecific signal 3kPZS, and provides insights as to why pheromone signals appear to be informative for migratory animals such as the sea lamprey. Data accessibility. Data available in the Electronic Supplementary Material: Raw data. Competing interests . The authors declare no competing interests. 33 ACKNOWLEDGEMENTS We are grateful to personnel of the U.S. Geological Survey Hammond Bay Biological Station for use of facilities, the U.S. Fish and Wildlife Service and Fisheries and Oceans Canada for providing sea lamprey, and to Dolly Trump, Lydia Lorenz, and Andy Meyers for the use of their private land for stream access. We ar e also grateful to all technicians that assisted with field work during this study. Special thanks to Tyler Buchinger and Andrea Jaeger - Miehls for reviewing early versions of this manuscript. Any use of trade, product, or firm names is for descriptive purp oses only and does not imply endorsement by the U.S. Government . This manuscript is contribution number 1929 of the Great Lakes Science Center. Funding statement. This work was supported by the Great Lakes Fishery Commission, Ann Arbor, MI, U.S.A. (to WL). 34 Permission to use published material 35 36 CHAPTER 2 FUNCTIONAL TRANSITION OF A CHEMICAL CUE INTO A CHEMICAL SIGNAL IN SEA LAMPREY 37 ABSTRACT The sensory trap model of signal evolution hypothesizes that signallers adapt to exploit a cue used by the receiver in another context . While exploitation of receiver biases can result in conflict between the sexes, deceptive signalling system s that are mutually beneficial drive the evolution of reliable communication systems. However, female responses in th e non - sexual and sexual context must be uncoupled for communication systems originating through a sensory trap to be reliable. Male sea lamprey ( Petromyzon marinus ) signal with a mating pheromone, 3 - ke t o petromyzonol sulfate (3kPZS), which previous phylogenetic comparisons indicate to be a match to female preference for juvenile odor during migration. Upstream movement of migratory lampreys is partially guided by 3kPZS, but females only move towards 3kP ZS with proximal accuracy during spawning. Here, we use instream behavioral assays paired with gonad histology to document the transition of female preference for juvenile - and male - released 3kPZS that coincides with the functional shift of 3kPZS as a migr atory cue to a mating pheromone. Females became increasingly bias towards the source of synthesized 3kPZS as their maturation progressed into the reproductive phase, at which point, a preference for juvenile odor (also containing 3kPZS naturally) ceased to exist. Uncoupling of female responses during migration and spawning makes the 3kPZS communication system a reliable means of synchronizing mate search. The present study offers a rare example of a functional shift from a non - sexual migratory cue into a ma ting pheromone , and provides insights into the origins of stable communication signalling systems. Key words: pheromone , 3kPZS, conspecific cue , receiver bias, semiochemical . 38 INTRODUCTION The sensory trap model of signal evolution hypothesizes that a male trait evolved as a match to a cue used by females in another context ( Ryan and Cummings, 2013 ) . While exploitation of receiver biases can result in conflict between the sexes, deceptive signals that are mutually beneficial drive the evolution of stable communication systems ( Garcia a nd Ramirez, 2005 ) . However, females must overcome the deception and adjust their responses to fit the sexual context, functionally uncoupling the responses to the trait in the non - sexual and sexual contexts ( Stuart - Fox, 2005 ) . Although many empirical studies implicate a role of sensory traps underlying female preference for male signals (reviewed by ( Ryan and Cummings, 2013 ) , the transition associated with uncoupling responses in the non - sexual and sexual contexts remains poorly documented. Chemical communication is a widely employed sensory modali ty that offers opportunity to examine the evolution of signaling systems. Previous research on the evolution of animal communication systems has focused primarily on visual and vocal stimuli ( Christy, 1995 ; Gasparini et al., 2013 ; Hurd and Enquist, 2005 ; Kelley and Kelley, 2014 ; Stoddard and Prum, 2011 ; Tibbetts, 2002 ) , perhaps because they are often within the range of human perception. The evolution of chemical signals remains poorly understood ( Symonds and Elgar, 2008 ; Wisenden, 2014 ) , but is hypothesized to involve a transition of a chemical cue into a signal ( Wisenden, 2014 ) . For example, stimuli can originate as byproducts of physiological processes such as compou nds ( Davidson et al., 2011 ; Døving et al., 1980 ; Weiss et a l., 2013 ) . Recent phylogenetic comparisons using behavioral bioassays suggest the evolution of the pheromone communication 39 system in sea lamprey ( Petromyzon marinus ) involved the functional transition of chemical cue into a signal ( Buchinger et al., 2013 ) . Sea lamprey use pheromones to modulate key aspects of their life history. Pheromones are discussed here as chemical signals that are produced by an organism and elicit a physiol ogical or behavioral response in conspecifics ( Karlson and Lüscher, 1959 ) . Adult sea lampreys cue onto odors of stream - residing juvenile larvae to navigate towards suitable spawning habitat during their springtime migration from freshwater lakes, or the Atlantic Ocean, into freshwater streams ( Bjerselius et al., 2000 ; Teeter, 1980 ; Wagner et al., 2009 ) . M ales typically move upstream before females ( Applegate, 1950 ) , establish nests ( Manion and Hanson, 1980 ) , and begin to release a bile salt, 3 - keto petromyzonol sulfate (3kPZS), that functions as a sex pheromone ( Johnson et al., 2006 ; Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) . Female preference for 3kPZS is hypothesized to have originated in a non - sexual context ( Buchinger et al., 2013 ) . Female silver lamprey ( Ichthyomyzon unicuspis ), a close relative of sea lamprey, show an upstream migratory response to 3kPZS, but only stream - resident juveniles, not males, release 3kPZS into the water ( Buchinger et al., 2013 ) . Furthermore, general upstream movement of female sea lamprey is initiated by 3kPZS during migration ( Brant et al., 2015 ) . Female preference for 3kPZS was matched by male sea lamprey . Males dramatically upregulate the 3kPZS biosynthetic pathway (up to 8000 - fold increase) in their livers ( Brant et al., 2013 ; Yeh et al., 2012 ) , and release 3kPZS at high rates (~ 0.5mg/h) across specialized cells in the gill epithelia ( Siefkes et al., 2003 ) upon reaching sexual matur ation. Since during migration 3kPZS only elicits general upstream movement ( Brant et al., 2015 ) , while during spawning 3kPZS elicits a highly ta rgeted and proximate preference in females ( Johnson et al., 2009 ; Siefkes et al., 40 2005 ) , female response to 3kPZS in each context had to be uncoupled. Hence, a transition in the function of 3kPZS must occur between the migration and th e spawning periods. Here, we document the functional transition of 3kPZS from a migratory cue to a sexual signal. Using a field study paired with gonadal histology, we track the uncoupling of migratory female responses to juvenile - released 3kPZS and spawn ing female responses to male - released 3kPZS. Female sea lamprey gradually adjusted their proximal attraction to 3kPZS, which coincides with progression of oocyte maturation. Further, to support a new conceptual model, we confirm that 3kPZS is released by s tream - residing larval sea lamprey, and that similar larval - released compounds with similar detection thresholds in the olfactory epithelium ( Fine and Sorensen, 2008 ; Li et al., 1995 ; Li et al., 2002 ) do not yiel d the same behaviors as 3kPZS elicits . Describing the adaptations in male and female sea lamprey associated with 3kPZS signaling provides insights into the evolution of communication in vertebrates. 41 METHODS Field behavioral assay All procedures involving sea lamprey were approved by the Michigan State University Institutional Animal Care and Use C ommittee prior to the start of the study ( Nos. 05/06 - 066 - 00, 03/11 - 053 - 00 and 03/14 - 054 - 00). Pre - spawn migratory adult (immature, herein) se a lamprey we re captured by t he United States Fish and Wildlife Service and Department of Fisheries and Oceans Canada throughout tributaries to Lake Michigan, Lake Superior and Lake Huron , transported to the U nited States Geological Survey - Hammond Bay Biological Statio n , Millersburg, MI, USA ( HBBS) , separated by sex, and held separately in 500 - 1000 L flow through tanks. Sex of each individual was later confirmed during surgical tagging procedures by visual observation of eggs. Males were removed from all treatment anima ls, as only females could be released for testing during this study ( i.e . to avoid repopulation of the invasive species in barrier - controlled tributaries, see Field site for experimental tests ). Ovulated female lamprey (mature, herein) were matured in natu ral conditions using stream acclimation cages. To accomplish this, 10 - 15 immature adults were placed in stream acclimation cages (~ 0.25 m 3 ) that were submerged in the Lower Ocqueoc River near the highway 23 bridge, Millersburg, MI. Acclimation took betwee n 5 - 10 days for subjects to reach sexual maturity. Sexual maturity was determined by applying gentle pressure to the abdomen and observing for expression of ovulated gametes from the cloacal aperture ( Manion and Hanson, 1980 ; Siefkes et al., 2005 ) . Juvenile sea lamprey (larval, herein) for washings experiments were collected in tributaries to Lake Michigan and Lake Huron year - round using electrofishing gear following published methods ( Steeves et al., 2003 ) . 42 Passive integrated transponder (PIT, Oregon RFID, Portland, Oregon, USA) tagging procedures for migratory female subjects followed those procedures described in Johnson e t al. ( 2009 ) , and those for ovulated females followed Li et al. ( 2013 b ) . The procedure typically took less than 30 seconds per subject. Implanted animals were immediately transferred into aerated holding tanks with a constant flow of Lake Huron water for up to 24 hours, until they were stocked into stream release cages. Subjects were monitored throughout the day for signs of distress or mortality. The experimental site for migratory trials consisted of a 250 m - long section of the Upper Ocqueoc River located in Millersburg, Michigan, USA (T35N, R3E, Sec. 27) . The upper reaches of the Ocqueoc River consist of suitable spawning substrate and larval habitat, and were historically infested with nesting sea lamprey ( Applegate, 1950 ) . Wild populations are now physically barred from entering the Upper Ocqueoc River due to a trap - integrated weir located ~20 km downstream, providing a controlled field site with no naturally occurring background pheromones. The most - upstream 45 m of the site is divided by a naturally occurring island, which separates two similar sub - channels . The center island between the two sub - channels was stabilized using sand bags to insure no treatment odorants seeped between sub - channels d uring trials. Treatments were administered into the center of a 1 m 2 PIT antenna placed in the center of the stream at the upstream end of each sub - channel. At the confluence of each sub - channel, a transecting PIT antenna, 0.5 m high x 6 m long, independently monitors each mouth to count the numbers of test subjects that e nter each sub - channel. The detection range outside of each PIT antenna was 0.3 m, and scan frequencies were 3 scans/sec. Downstream 205 m from the confluence of sub - channels, release cages (~ 0.25 m 3 ) were anchored in the center of the main 43 channel (Figure 2 - 1a). Details of the field site for sexually mature female trials are outlined in Li et al. ( 2013 b ) with slight modification seen in Figure 2 - 1b. 44 Figure 2 - 1. The section of the Upper Ocqueoc River, Millersburg, MI, U.S.A (T35N, R3E, Sec. 27) used to examine behaviors of sea lamprey in relation to pheromone treatments. ( a ) Full 250 m - long section and details of the naturally bifurcated sub - channels used for exa mining behaviors of migratory female subjects . Black - dashed lines indicate channel transecting PIT antennas placed at the confluence of the left ( C L ) and right ( C R ) sub - channels, and upstream and downstream of the release point. ( b ) The 45 m - long section used for examining behaviors of sexually mature female subjects . Hollow boxes represent left (T L ) and right (T R ) 1 m 2 PIT antennas where, in the center of each, treatments were administered. Scale bars = 25 m. 45 Permission to administer 3kPZS and si milar compounds/raw odor of lamprey into the stream was obtained from the United States Environmental Protection Agency through experimental user permits 75437 - EUP - 1 and 75437 - EUP - 2 prior to any field testing. 3kPZS was custom synthesized by Bridge Organics (Vicksburg, Michigan, USA; purity >97% ) and stored in powder form at - 80 o C until stock solutions were made. Purity was determined using a Waters ACQUITY LC System coupled with t he Waters Quattro Premier XE tandem quadrupole mass spectrometer (LC MS/MS, Milford, MA, USA) . Nuclear magnetic resonance ( 1 H NMR) was used to confirm the chemical structure of 3kPZS. A 10 mg/mL stock solution of synthesized 3kPZS ( in 100% methanol ) was pr epared and transferred in to five vials of 10 mL aliquots, each. 3kPZS stock solution was stored at - 80 o C until use in the field. Extracted wash - water from larvae (larval extract) was used as a positive control to validate the experimental system. Larval extracts have been shown to induce strong targeted preferences towards the odorant source from migratory female sea lamprey during past studies ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . Methods of extraction followed those outlined by Fine et al. ( 2006 ) . PADS, a component of larval extract with known chemical structure ( Sorensen et al., 2 005 ) , was used as a benchmark compound when calculating the volume of extract to apply to the stream and maintain consistent full stream concentrations. The concentration of PADS was determined using HPLC tandem mass spectrometry (HPLC - MS/MS) following Li et al. ( 2013a ) . Treatments included; (1) synthesized 3kPZS (5x10 - 13 molar, M) administered into one sub - channel (methanol vehicle into the adjacent sub - channel), (2) larval extract (5x10 - 14 M benchmark PADS) ad ministered into one sub - channel (river water vehicle into the adjacent sub - channel), and (3) vehicle control (methanol into both sub - channels). 46 A typical riverine migratory season for sea lamprey lasts from May through late June. To specifically examine wh ether responses to 3kPZS change in female sea lamprey as they approach ovulation, and subsequently the spawning season, we divided field trials into an Early migration, Late migration and Mature season. Mature trials were conducted with sexually mature subjects similarly to previous studies ( Joh nson et al., 2006 ; Siefkes et al., 2005 ) . Early migratory season trials were conducted 08 27 May, and Late trials were conducted 13 27 June, 2013, at night when migratory sea lamprey are most active ( Applegate, 1950 ) . Mature trials were conducted 3 19 July, 20 13 during the day when mature sea lamprey are active and nesting ( Applegate, 1950 ) . Pheromone treatments were diluted with 30 L of river water in large mixing bins. Each solution was pumped into the center of each PIT antenna in respective sub - channels through latex tubes at a rate of 167 ± 5 mL/min over the span of 2.5 hours using peristaltic pumps ( Masterflex 7553 - 70 , Cole - Parmer, Vernon Hills, I llinois, USA). A treatment was administered to one sub - channel ( treatment channel ) while a methanol vehicle was administered into the adjacent sub - channel . The treatment and control sub - channels were alternated each trial to avoid artifacts of channel bias . Stream discharge was estimated following published methods ( Murphy and Willis, 1996 ) throughout the migratory and mat ing season (taken every three days, or after every precipitation event) to determine the volume of treatment stock solution to apply to the stream each trial and maintain consistent full - stream concentrations. Stream discharge ranged from 3.09 1.50 meter s 3 /sec during migratory trials and from 0.85 0.40 meters 3 /sec. during mating trials. Two trials were conducted each night or day. For migratory trials, no more than 20 PIT - tagged sea lamprey were released per trial. For mating trials, no more than 10 ovulated females were released per trial because mature female sea lamprey are more difficult to attain. Subjects 47 were allowed an acclimation period in the stream of 10 15 hours prior to a trial. Subjects for each trial were acclimated in separate relea se cages (2 cages total). Minimal mortality occurred during acclimation of migratory subjects (< 1%). Mortality is more common in sexually mature subjects during acclimation ( 7%), as they are in their final life stage at this time ( i.e. a semelparous species). Therefore, two extra mature subjects were often stocked (12 total) for each trial. If no mortality occurred, extra subjects were removed from the release cage prior to release. Release cages were solid aluminum and stainless steel (~ 0.25 m 3 ), consisting of a sliding door that was removed manually upon release. Migratory trials were conducted from ~ 2020 h (starting at sundown) through ~ 0310 h, and mating trials from ~ 0700 h 1200 h. Each trial was 2.5 hours long. In the first ½ hour, the treatment was administered to the stream allowing the current to carry the compound to the downstream acclimation cage containing test subjects. Subjects were released, and during the remaining two hours, animals were free to swim throughout th e experimental system while treatments were administered. The second trial started 15 30 minutes after the first. Treatments were kept consistent for each day or night of trials ( i.e. if 3 kPZS was tested during the early trial, 3kPZS was also tested duri ng the late trial to prevent the possibility of any unwanted contamination from other treatments that day/night). All equipment was thoroughly rinsed with stream water prior to a new trial. No animals were recovered from the stream after a trial. Unique P IT identification for each subject prevented pseudoreplication during trials. Movement data were consolidated and stored using a multiplexor (Oregon RFID, Portland, Oregon, U.S.A.). Data were uploaded each trial night using a hand - held Meazura model MEZ100 0 personal digital assistant (Aceeca International Limited, Christchurch, New Zealand). 48 Subjects were randomly selected and gonads were dissected for histology during field trials (Supplemental Table S2 - 1). S amples of oocytes were collected from the posterior, medial, and anterior locations of the ovary . Oocytes were fixed in a 4% paraformaldehyde solution and placed in 4 o C until sectioning, hemotoxylin , and eosin staining procedures could be conducted . Histological examination of oocytes and determination of maturity (Days from ovulation, Supplemental Table S2 - 1) followed published procedures ( Yorke and McMillan, 1980 ) . All oocytes on each slide were examined to confirm homogeneity of maturational state throughout each ovary. Days from ovulation were estimated by comparing oocyte morphology and time of year to those estimates of days until ovulation in the literature ( Larsen, 1980 ; Lewis and McMillan, 1965 ; Yorke and McMillan, 1980 ) . Three known larval - released compounds, PZS, PADS, and PSDS, with similar olfactory sensitivity to 3kPZS ( Fine and Sorensen, 2008 ; Li et al., 1995 ; Li et al., 2002 ) , were compared to familiarity to compounds released from conspecifics. Trials were conducted only during Late migration, from 31 May 12 June 2008. All experimental procedures were conducted in the same experimental system (Figure 2 - 1a), with slight modification. Three trials were conducted per night instead of two. Females were pre - exposed to the test article 1 h prior to release, and their movement and distribution in the stream were monitored for 2 h after release. Additional treatments included : ( 1) a mixture of PADS (1x10 - 12 M ), PSDS (5x 10 - 13 M ) , PZS ( 5x10 - 13 M ) , and 3kPZS (5x 10 - 13 M ) , ( 2) a mixture of PADS (1x10 - 12 M ), PSDS (5x 10 - 13 M ) , and PZS ( 5x10 - 13 M ), and ( 3) synthesized 3kPZS (5x10 - 13 M). Vehicle and larval extract control trials were conducted consistently across all behavioral trials. Ratios of PADS, PSDS, and PZS mixtures were kept consistent with those in published literature ( Fine and Sorensen, 2008 ; Sorensen et al., 49 2005 ) . Oocytes were sampled on 03 June ( n = 12) and on 12 June 20 08 ( n = 12, Supplemental Table S2 - 1). Histo logical analyses of oocytes were consistent for all trials (data not shown). PADS, PSDS, PZS and 3kPZS were synthesized by Bridge Organics in 2008 (Vicksburg, Michigan, USA) with purity greater than 95%. Previously, 3kPZS had been considered to be r eleased exclusively from mature male sea lamprey during the mating season to draw mature females towards the nest for courtship ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) . To examine whether larval sea lamprey release 3kPZS specifically during the migratory season when immature migrating sea lamprey would be exposed, we sampled populations of wild larval sea lamprey to analyze larval - conditioned wash - water over that time p eriod. Five collections were completed from 08 December 2012 28 April 2013 ( Table 2 - 3 ). Ice and flooding made collections di fficult for every month, yet our collections were considered representative of the onset of migration through pre - spawn migration in the Great Lakes region ( Manion and Hanson, 1980 ) . Larvae were collected with assistance f rom the U.S. Fish and Wildlife Service in streams using electrofishing gear following published methods ( Steeves et al., 2003 ) . Details of larval collections can be seen in Table 2 - 3 . Details of wash - water collection from larvae can be seen in Supplemental Methods, Supplemental Figure S2 - 1. Data analyses Statistical analyses for all in - stream migratory and sexually mature female behaviors during field studies follow those previously described by our laboratory ( Johnson et al., 2009 ; Li et al., 2013 b ) with slight modification. During 2013 migratory trials, four response variables were examined: (1) the number of subjects that mov ed downstream of the release cage, and did not move back up during the trial ( Down ), (2) the number of subjects that moved upstream from 50 release cages to the confluence of the two sub - channels (205 m, Up ), (3) of the total number of subjects that moved ups tream to the confluence of the two sub - channels, the number of subjects that then entered a sub - channel activated with a pheromone treatment ( Treatment channel ), and (4) of the subjects that entered the treatment channel, the number that entered within 0.5 m of the treatment source ( Treatment source ). In 2008 migratory trials, all response variables were Down - point antenna failure. In sexually mature subjects, our field site was slightly modi fied from published experimental systems ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) . A single response variable was examined in mature subjects: the numb er of subjects that entered within 0.5 m of the treatment source (Treatment source), which was adjacent to another source (1 m 2 PIT antennas, 1.5 m apart). These side - by - side nests represented the left and right sub - channels in statistical analyses when co mparing to migratory trials. Sexually mature females had to be tested in a smaller system than the system used for testing migratory females because subjects were ovulated and therefore physically incapable of moving upstream on such a large scale within t he allotted swim time (2 3 hours). For all behavior data, upon observing no signs of over dispersion or nonlinearities, l ogistic regression (generalized linear model) with a binomial distribution was examined for each response variable in relation to eac h treatment (R version 2.11.1). Pheromone concentration data from monthly larval wash - water collections were standardized by weight (ng/g - larvae/hr). A ll values fr om LC MS/MS analyses associated with a signal - to - noise ratio less than or equal to 10 were co nsidered below the limit of quantitation and removed from the data set. homogeneity of varian ce was used to examine any violations of assumptions normality across variance before further statistical ana lyses were 51 conducted. Data that were not normally distributed or showed heterogeneity across variance were log - tra nsformed. Once homogeneity of variance was observed, an ANOVA and post - hoc ( = 0.05) was conducted for all statistical comparisons (R version 2.11.1, Vienna, Aus tria ). 52 RESULTS Preference for 3kPZS becomes increasingly targeted as females approach ovulation Immature female subjects incrementally increased their directional preference towards the sub - channel activated with 3kPZS as their oocytes approached ovulation. During the early migratory season (08 15 May, 2013) large numbers of females moved upstream (74 - 86%) . Females did not show a bias towards the sub - channel activated with 3kPZS ( Z 5 = 0.34, P = 0.737), while a positive bias was seen towards the sub - channel activated with positive control larval extract ( Z 5 = 2.53, P = 0 .012), see Table 2 - 1 for statistical comparisons. Histology of oocytes from these subjects yielded eggs to be pre - ovulated, with intact follicular cell layer encompassing an adhesive cell layer, pronounced nucleus, and a buildup of fluid of ovulation (Figure 2 - 2a - Early). These subjects are estimated to be more than 10 days from ovulation based on previous studies examining structural and developmental aspects of the ovary in sea lamprey (Supplemental Table 2 - 1). During the late half of the migratory season (13 26 June, 2013), the bias towards 3kPZS matched that of larval extract when significant numbers (up to 75%) of subjects began moving into the sub - chan nel (3kPZS: Z 5 = 2.23, P = 0.026, Larval Extract: Z 5 = 3.07, P = 0.002 ) . High numbers of subjects moved upstream (77 - 90%) and approached the sources of treatments during this time (Table 2 - 1 ). Histology on subjects during late migration showed oocytes to b e closer to ovulation, yet still pre - ovulated, with intact follicular cell layer, a present yet grainy nucleus, and compression of the adhesive cells due to tightening of the follicular cell layer from increasing buildup of fluid of ovulation (Figure 2 - 2a - Late). These subjects are estimated to be within 4 days from ovulation (Supplemental Table S2 - 1). 53 Mature (ovulated) subjects ( tested 07 19 July, 2013) showed a strong preference towards a 3kPZS source as expected ( Johnson et al., 2 009 ; Li et al., 2002 ; Siefkes et al., 2005 ) , where 100% entered the 1 m 2 3kPZS source ( X 2 2 = 57.25, P < 0.001 ). Histology confirmed that oocytes were ovulated in a sub - sample of these test subjects, showing a broken follicular cell layer and exposed adhesive cells on the distal half of each oocyte and no nucleus present. Egg development did not vary between the pos terior, middle, or anterior regions of the ovary in any of our samples (Figure 2 - 2b). Also expected yet previously undocumented, a preference for larval extract treatments did not occur by mature female subjects (Figure 2 - 2b). 54 Figure 2 - 2. Preference response of sea lamprey to conspecific - released compounds. Treatments included methanol vehicle controls (vehicle vs. vehicle), larval extract (LE) at 5x10 - 14 molar benchmark PADS vs. vehicle, and synthesized 3kPZS (3kPZS) at 5x10 - 13 molar vs. vehicle. ( a. ) The proportion of immature female sea lamprey entering the sub - channel containing each treatment during the Early (May, 2013) and L ate (June, 2013) migratory season. Different upper - case letters indicate a significant proportion moved into th e activated treatment sub - channel (logistic regression: X 2 5 = 20.35, P = 0.001). Key histological features examined in immature oocytes included the follicular cell layer ( arrow ) that encompasses an adhesive cell layer, nucleus ( I. in Early , III. in Late ), and build - up of fluid of ovulation ( II. in Early , IV . in Late ). ( b . ) The proportion of Mature ovulated female sea lamprey entering the treatment vs. vehicle source (1 m 2 ) during the mating season in early July. Different upper - case letters indicate a sign ificant proportion moved into the activated treatment source ( X 2 2 = 57.25, P < 0.001). Key features examined in Matur e oocytes included a broken (absent) follicular cell layer with exposed, lysed, adhesive cells ( arrow ), and lack of nucleus ( V. ). Scale bars = 100µm. 55 Table 2 - 1: Additional preference responses (in addition to Figure 2 - 2a) of migratory female sea lamprey to conspecific - released compounds. Treatment Trials Released Down ( n ) Up ( n ) Treatment source ( n ) Vehicle E arly 11 218 17% ( 36 ) A 78% ( 170 ) A 14% ( 12 ) A Larval Extract E arly 4 80 4% ( 3 ) B 86% ( 69 ) AB 43% ( 20 ) BC 3kPZS (5E - 13 M) E arly 4 80 16% ( 13 ) A 74% ( 59 ) A 11% ( 3 ) A Vehicle Late 9 171 6% ( 11 ) B 77% ( 132 ) A 29% ( 18 ) B Larval Extract Late 3 60 5% ( 3 ) B 90% ( 54 ) B 54% ( 21 ) C 3kPZS (5E - 13 M) Late 4 80 5% ( 4 ) B 80% ( 64 ) A 32% ( 13 ) BC X 2 26.05 15.03 27.87 df 5 5 5 P - value < 0.001 0.010 < 0.001 See Figure 2 - 2 for explanation of treatments. Trials were conducted during E arly (May) and L ate (June) 2013 migratory season. Additional preference response Down shows the percentage ( n ) of subjects that moved down from release cages, and did not come back upstream during trials. Up refers to subjects that moved upstream from the release cage and co ntinued to swim 205 m to the confluence of the two sub - channels. Treatment source refers to subjects that entered within 0.5 m of the treatment source after entering the sub - channe l activated with each treatment . Responses that share a letter across treatm ents are not significantly different ( logistic regression ; = 0.05) . 56 Similar compounds did not yield a behavioral response in migratory subjects Behavioral trials testing mixtures of similar compounds PZS, PADS, and PSDS with an addition or subtraction of 3kPZS yielded 3kPZS to be responsible for behavioral responses seen during late migration in our trials (Table 2 - 2 ). Late migratory female respon ses to treatments did not vary between years (vehicle controls: X 2 1 = 1.12, P = 0.289; 3kPZS: X 2 1 = 0.39, P = 0.531; and larval extract controls: X 2 1 = 0.12, P = 0.729) , and so were combined . Greater numbers of subjects again moved upstream towards the sources of 3kPZS and larval extract (68 - 75%), and moved upstream less during vehicle controls ( X 2 4 = 13.24 , P < 0. 001). Both larval extract and all treatments specifically containing 3kPZS increased the numbers of migratory subjects that entered the treatm ent sub - channel by up to 33% from that of vehicle controls ( X 2 4 = 16.24 , P < 0. 001). Both larval extract and all treatments specifically containing 3kPZS increased the numbers of migratory subjects also increased the numbers that entered within 0.5 m of th e treatment source by up to 19% compared to vehicle controls ( X 2 4 = 43.32 , P < 0. 001, Table 2 - 2). 57 Table 2 - 2: Preference responses of migratory female sea lamprey to similar conspecific - released compounds. Treatment Trials Released Up ( n ) Treatment channel ( n ) Treatment source ( n ) Vehicle 8 160 41% ( 65 ) A 42% ( 27 ) A 9% ( 2 ) A Larval Extract 8 160 71% ( 114 ) BC 75% ( 86 ) B 60% ( 52 ) C 3kPZS, PZS, PADS, PSDS 6 120 75% ( 90 ) C 69% ( 62 ) B 28% ( 17 ) B PZS, PADS, PSDS 8 160 42% ( 67 ) A 43% ( 29 ) A 16% ( 5 ) A 3kPZS 6 120 59% ( 71 ) B 68% ( 48 ) B 32% ( 15 ) B X 2 13.24 16.24 43.32 df 4 4 4 P - value < 0.001 < 0.001 < 0.001 Treatments included methanol vehicle controls (vehicle vs. vehicle), larval extract at 5x10 - 14 molar (M) benchmark PADS vs. vehicle, a mixture of PADS (1x10 - 12 M ), PSDS (5x 10 - 13 M ) , PZS ( 5x10 - 13 M ) and 3kPZS (5x 10 - 13 M ) vs. vehicle, the same mixture minus 3kPZS vs. vehicle, and synthesized 3kPZS (3kPZS) at 5x10 - 13 molar vs. vehicle. Response variab les are consistent with those described in Figure 2 - 2, Table 2 - 1. Responses that share a letter across treatments are not significantly different ( logistic regression ; = 0.05) . 58 Larval sea lamprey release 3kPZS Analysis of the monthly wash - water collection revealed that 3kPZS is released from larval sea lamprey in the months leading up to the spawning season. While little variation was seen in PZS release rates through the sampling months, 3kPZS showed an increase in release rates durin g the migratory season. Specifically, 3kPZS release rates were highest from larvae towards the end of March (ANOVA: F 4 = 27.35, P = 0.001). Mean (± 1 SEM) 3kPZS release rate from larvae for all months was 0.107 ± 0.05 ng/g - larvae/hr. Compound PZS was also released from larvae during this time, and did not vary between months ( F 4 = 0.54, P = 0.750). Mean PZS release rate from larvae for all months was 0.422 ± 0.07 ng/g - larvae/hr, which was significantly higher than that of 3kPZS (Two - way t - test: t 8 = 3.41, P = 0.009). Compounds PADS and PSDS were not detected from larvae in our samples during these months (Table 2 - 3). 59 Table 2 - 3: Release rates of larval - released compounds. Date River n Temp (°C) Batch weight (g) 3kPZS release (± 1 SEM) PZS release (± 1 SEM) PADS release PSDS release 8 - Dec - 12 Betsie 168 6.3 210 0.01 (0.01) A 0.25 (0.05) A ND ND 12 - Jan - 13 Betsie 241 3.7 369 0.09 (0.03) A 0.46 (0.26) A ND ND 4 - Mar - 13 White 169 1.7 254 0.03 (0.02) A 0.38 (0.25) A ND ND 31 - Mar - 13 Silver 82 3.8 111 0.31 (0.16) B 0.69 (0.60) A ND ND 28 - Apr - 13 Carp 127 6.0 177 0.10 (0.05) A 0.33 (0.12) A ND ND F 27.35 0.54 NA NA df 4 4 NA NA P - value 0.001 0.750 NA NA Larval sea lamprey were sampled from for tributaries of Lakes Michigan and Huron, located in the northern lower peninsula of Michigan. Compounds examined included: 3 - keto petromyzonol sulfate ( 3k PZS) , petromyzonamine disulfate (PADS), petromyzosterol disulfate (PSDS), and petromyzonol sulfate (PZS) . Mean release rates (ng/g - larvae/hr ) from triplicate samples of each batch of larvae ( n ) are shown for each date of collection. Release rates within each compound that share a letter are not significantly d ifferent ( ANOVA ; = 0.05) . Compounds PADS and PSDS were not detected ( ND ) in our samples during these dates. 60 DISCUSSION Data presented in this study, when taken together with evidence from previous studies, document the transition of 3kPZS from a navigational cue to a mating pheromone in sea lamprey. Migratory sea lamprey use 3kPZS as a navigational cue that initiates non - t argeted upstream movement ( Brant et al., 2015 ) . During spawning, however, male - released 3kPZS function s as a sexual signal that guides females upstream to the proximity of a nest ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2003 ; Siefkes et al., 2005 ) . Phylogenetic studi es indicate that the migratory response to 3kPZS predates the sexual response to 3kPZS ( Buchinger et al., 2013 ) . The pre - existing bias of females to move upstream to 3kPZS likely crea ted selective pressure on males to massively upregulate 3kPZS synthesis ( Brant et al., 2013 ) and amplify 3kPZS as a signal. Although deceptive signals can lead to conflicts of interest between the sexes, the mutual drive of males and females to efficiently find mates likely precludes any conflict associated with chemical communication during spawning. However, females must ad just their responses to 3kPZS during spawning for the communication system to be stable. Here, we document the functional transition of 3kPZS as a migratory cue to a sexual signal by tracking female gonad maturation alongside their behavioral responses to 3kPZS. If a sensory trap leads to a mutually beneficial communication system, female response must be uncoupled between the non - sexual and sexual contexts, enabling females to respond to the male trait in a way appropriate to the sexual context ( Garcia and Ramirez, 2005 ) . For example, male signaling with a terminal yellow band on the caudal fin i n Goodeinae fishes appears to match an existing foraging preferenc e in females, but females of some species have further evolved to use the trait to evaluate male quality. However, females that use the terminal 61 yellow band to select mates have lessened their responses to the yellow band as a foraging cue, effectively unc oupling the responses in the foraging and sexual contexts ( Garcia and Ramirez, 2005 ) . Likewise, female sea lamprey were likely selected to adjust their response to 3kPZS between the migratory and spawning contexts. 3kPZS initiates non - targeted upstream movement of migratory sea lamprey, a behavior fitting the large - scale navigations necessary to locate rivers and tributaries in which to spawn ( Applegate, 1950 ; Manion and Hanson, 1980 ) . D uring spawning, 3kPZS elicits a highly targeted bias towards the 3kPZS source, and resumed search behavior when females exit the odor plume, to locate mates ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ) . Incorrectly responding to 3kPZS in either context is maladaptive for females. Hence, the integration of 3kPZS into the sexual communication system occurred through male adaptations for signaling and female adaptions to appropriately respond to 3kPZS. Interestingly, non - sexual and sexual female preferences in sea lamprey became uncoupled by an increase in female preference in the sexual context, whereas non - sexual and sexual female preferences in Goodeinae fishes became uncoupled by a decreas e in female preference in the non - sexual context ( Garcia and Ramirez, 2005 ) . The transition of 3kPZS from a navigational cue to a mating pheromone in sea lamprey is thus far specific to 3kPZS. Other larval - released compounds, PZS, PADS, and PSDS have been shown to have similar response thresholds in the olfactory epithelium compare d to 3kPZS during electro - physiological analyses ( Li et al., 1995 ; Sorensen et al., 2005 ) , and all appear to share similar biosynthetic pathways ( Brant et al., 2013 ; Venkatachalam, 2005 ) , yet only 3kPZS induced behavioral responses consistent with a shift from chemical cue to chemical signal as femal es mature. Compounds PADS and PSDS were not detected from larvae in our samples, nor were they behaviorally active in our field studies. Behavioral results testing these compounds in 62 the field are consistent with previous field evaluations ( Meckley et al., 2 012 ) , and inconsistent with previous laboratory evaluations ( Sorensen et al., 2005 ) , suggesting disco ntinuity between laboratory and field studies for evaluation of behaviors to pheromones in fishes ( Johnson a nd Li, 2010 ) . Interestingly, PZS was released at significantly higher rates than 3kPZS from larvae, yet females became ritualized to 3kPZS over PZS. The reason 3kPZS was selected over PZS as a pheromone may be attributable to a mechanism of release acro ss gill epithelia in male senders ( Brant et al., 2013 ; Li et al., 2002 ; Siefkes et al., 2003 ) , yet the exact reason remains unknown. In summary, we document a missing behavioral link of the functional transition of 3kPZS from a navigational cue to a pheromone. Results here support the theory that female responses must be uncoupled between the non - sexual and sexual contexts, enabling females to respond to the male trait in a way appropriate to the sexual context ( Garcia and Ramirez, 2005 ) . While our observation of increasingly targeted preferences as female maturation progressed into the spawning phase tracks the uncoupl ing of the responses in each context, the mechanisms underlying the transition remain unknown. We document 3kPZS release by larval sea lamprey, and conclude that 3kPZS was selected to become a conspicuous signal over additional larval - released compounds wi th similar detection thresholds in the olfactory epithelium ( Fine and Sorensen, 2008 ; Li et al., 1995 ; Li et al., 2002 ) . Sea lamprey represent a useful model for understanding the origins of chemical signals in vertebrates (Figure 2 - 3). 63 Figure 2 - 3. Conceptual model of the evolution of pheromone 3 - keto petromyzonol sulfate (3kPZS) in P. marinus . ( a. - product by larvae residing and feeding in streams, along with a suite of other b. ritualized as a navigational cue to migrating adults. ( c. ) Males adapted to massively upregulate - adapted bias to 3kPZS in female receivers. Females adapted to f ine tune their movement towards a 3kPZS source (nest) to synchronize reproduction upon maturation of their gonads. Both parties continued to mutually benefit as 3kPZS evolved from a cue to a signal. 64 ACKNOWLEDGMENTS We thank all personnel at the Uni ted States Geological Survey Hammond Bay Biological Station, Millersburg, Michigan, the United States Fish and Wildlife Service Marquette Biological Station, Marquette, Michigan, and Fisheries and Oceans, Canada, for thei r assistance in animal capture duri ng this project. We also thank Dolly Trump and Lydia Lorenz for the use of their private land to access the field site. Thanks are well deserved of all technicians that participated in field work . Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. This work was supported by the Great Lakes Fishery Commission, Ann Arbor, MI, USA (to Dr. W eiming Li ). 65 CHAPTER 3 MIXTURES OF TWO BILE ALCOHOL SULFATES FUNCTION AS A PROXIMITY PHEROMONE IN SEA LAMPREY 66 ABSTRACT Unique mixtures of pheromone components are commonly identified in insects, and have been shown to increase attractiveness when reconstructed at the natural ratio released by the signaler ( Cardé et al., 1977 ) . In previ ous field studies of pheromones that attract female sea lamprey ( Petromyzon marinus , L.), putative components of the male - released mating pheromone included the newly described bile alcohol 3,12 - diketo - 4,6 - petromyzonene - 24 - sulfate (DkPES) and the well char acterized 3 - keto petromyzonol sulfate ( 3kPZS ). Here, we show chemical evidence that unequivocally confirms the elucidated structure of DkPES, electrophysiological evidence that each component is independently detected by the olfactory epithelium, and behavioral evidence that mature female sea lamprey prefer artificial nests activated with a mixture that reconstructs the male - released component ratio of 30:1 (3kPZS:DkPES, molar:molar), and characterize search behavior (sinuosity of swim paths) of female s approaching ratio treatment sources. Unique pheromone ratios may underlie reproductive isolating mechanisms in vertebrates, as well as provide utility in pheromone - integrated control of invasive sea lamprey in the Great Lakes. Key words : pheromone ratio , sea lamprey, bile alcohol, chemical signal 67 INTRODUCTION A common source of conspecific information used in orientation strategies and mate location across the animal kingdom is provided by pheromones, or unique chemical signatures that are releas ed by animals and influence behavior or development of members of the same species ( Karlson and Butenandt, 1959 ; Wyatt, 2010 ) . Together, the sending and receiving of pheromones result in movement patterns that reduce the distan ce between conspecifics across their odor landscape (attraction) and/or maintain individuals in place (arrestant) to gain advantage in mating or feeding ( Steiger et al., 2011 ; Wyatt, 2010 ) . In insects, pheromones that function in sex and aggregation are often comprised of multiple components at specific ratios ( Cardé et al., 1977 ; Coracini et al., 2001 ; Reyes - Garcia et al., 2014 ) . For pheromones of unique ratios to function as species - specific attractants, a level of discrimination against individual components must occur at the sensory and behavioral level. In fishes, olfactory systems have been shown, via electrophysiological cross - adaptation experiments, to detect and discriminate between compounds with separate receptors ( Hara, 1992 ; Keller - Costa et al., 2014 ; Lipschitz and Michel, 1999 ) . Detection and discrimination of individual compounds in phylogeneti cally similar groups of fishes has been proposed as an adaptation involved in speciation ( Keller - Costa et al., 2014 ) . However, behavioral evidence is rarely presented to evaluate the specificity of pheromone ratios in vertebrates ( Wyatt, 2010 ) . The olfactory epithelium of the s ea lampr ey ( Petromyzon marinus , L . ) has been shown to detect and discriminate multiple compounds that are structurally similar ( Li et al., 2014 ; Li et al., 1995 ; Sorensen et al., 2005 ) . Using pheromones for key aspects of their life history ( Bjerselius et al., 2000 ; Johnson et al., 2009 ; Teeter, 1980 ; Wagner et al., 2009 ) , sea lamprey present a useful 68 vertebrate model for studies regarding pheromone ratio specificity . Sea lamprey begin their single reproductive season by migrating into freshwater tributaries to the Atlantic Ocean (native range), or the Laurentian Great Lakes (invasive range), that are activated with compounds released by stream - residing conspecific la rvae ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . Males often move upstream in greater numbers earlier in the migratory season through April May, and establish nests in suitable spawning habitat in early June ( Johnson et al., 2015b ; Manion and Hanson, 1980 ) . Upon reaching sex ual maturation in streams, males release a pheromone that includes a main component, 3 - keto petromyzonol sulfate (3kPZS ), across gill epithelia ( Li et al., 2002 ; Siefkes et al., 2003 ) . Synthesized 3kPZS alone draws significant numbers of mature females upstream towards the source ( Johnson et al., 2009 ; Siefkes et al., 2005 ) . However, 3kPZS as a single component is often less attractive than the whole male odor ( Johnson et al., 2009 ) . Upon analyzing whole male odor (termed spermiated male washings, or SMW, he rein), the chemical structure of a new sulfate - conjugated compound, 3,12 - diketo - 4,6 - petromyzonene - 24 - sulfate (DkPES) was elucidated ( Li et al., 2013 b ) . Mature females were shown to increase their preference for mixtures of 3kPZS and DkPES compared to 3 kPZS alone. However, many critical issues have not been addressed regarding the identity and function of DkPES ( Li et al., 2013 b ) , including chemical synthesis of DkPES and confirmation of the elucidated structure, discrimination of DkPES from 3kPZS by the olfactory epithelia, the orientation mechanisms used by sea lamprey to locate a mixture of these two compounds, and the effective range of ratios between the two pheromone components required for attraction. This study reports the next step in identi fying the function of DkPES as a pheromone component in sea lamprey. Here, we confirm the structure elucidated for DkPES with its synthesized copy, show the olfactory epithelium of sea lamprey discriminates DkPES from 69 3kPZS, demonstrate that females are at tracted to a reconstructed ratio similar to that seen in extracts from mature male sea lamprey ( Li et al., 2013 b ) , and further define the orientation strategies of mature females to the mixture of DkPES and 3kPZS. Based on the results, we provi de a conceptual model outlining the importance of specific pheromone ratios in guiding chemo - orientation strategies in sea lamprey. 70 RESULTS Spectra of synthesized DkPES match those of purified DkPES The compound DkPES ammonium salt applied in this study was synthesized by Apeloa Kangyu Pharmaceutical Co. ( Dongyang, Zhejiang, China ) according to the structure deduced in the preceding study ( Li et al., 2013 b ) . Differential scanning col orimetry (DSC) analysis ( Giron and Goldbronn, 1995 ; Van Dooren and Müller, 1984 ) of the synthetic product indicated that its purity was 97.0% (Supplementary Information Figure S 3 - 1 ). The chemical structure of DkPES was further confirmed by comparison of high resolution mass and NMR spectra of the purified compound ( Li et al., 201 3 b ) and the synthesized copy . The pseudo - molecular formula determined by high resolution mass spectrometry ( m/z 449.2012 [M NH 4 ] ) of the synthetic product was C 24 H 33 O 5 S , which was in good agreement with the calculated mass ( m/z 449.1998, 1.4, ppm 3.1, Supplementary Information Table S 3 - 1). The purified and the synthesized compounds showed overlapped NMR spectra (Figure 3 - 1), indicating that they are identical in their chemical structure. 71 Figure 3 - 1. Spectral comparisons of synthesized and natural (purified) DkPES. (A) Comparison of 1 H NMR spectra of natural and synthesized DkPES obtained from 600 MHz NMR spectrometry (Varian Inova) in methanol - d 4 . (B) Comparison of 13 C NMR spectra of natural and synthesized D kPES obtained from 600 MHz NMR spectrometry (Varian Inova) in methanol - d 4 . 72 DkPES is discriminated from 3kPZS in sea lamprey olfactory epithelia Electro - olfactogram (EOG) recording showed that both synthesized 3kPZS and DkPES are high ly stimulatory for the olfactory epithelia of adult sea lamprey (Figure 3 - 2A). At each concentration tested, the amplitude of t he olfactory response elicited by 3kPZS was larger than that elicited by DkPES. The threshold of detection for 3kPZS was 10 - 13 Molar ( M ) and for DkPES was 10 - 10 M. The d ifferen ce in the magnitude and slope in the c oncentration - response relationships for 3kPZS and DkPES suggested that the olfactory receptor mechanisms for these two compounds may differ ( Kang and Caprio, 1997 ) . This hypothesis was further supported by a set of cross - adaptati on experiments ( Kang and Caprio, 1997 ) , in which preadaptation of the olfactory epithelium to one compound did not suppress the olfactory responses to the other compound ( ANOVA: F 320 = 7.33, P = 0.02; Figure 3 - 2B) . In particular, when the sensory epithelium was subjected to prolonged perfusion (pre - adaptation) with 3kPZS, the normalized EOG response to DkPES was larger than that to 3kPZS (Figure 3 - t - test, t = 3.78, P = 0.004). Vice versa, when the epithelium was pre - adapted to DkPES, the EOG response to 3kPZS was larger than the response to DkPES (Figure 3 - 2B, t = 2.81, P = 0.019 ). These data indicate d that the olfactory epithelium distinguished DkPES and 3kPZS by at l east two different receptors. 73 Figure 3 - 2. Olfactory detection and discrimination of DkPES and 3kPZS by adult sea lamprey . (A) Semi - logarithmic plots of electro - olfactogram (EOG) responses t o different concentrations of DkPES and 3kPZS . The response amplitude was corrected for the blank response amplitude and normalized against the response amplitude of a standard odorant, L - arginine at 10 - 5 Molar. ( B) Qualitative differences in DkPES and 3kPZS a ssessed by cross - adaptation. Percentage unadapted response is the response amplitude to an odorant (treatment) when the olfactory epithelium was pre - adapted to an odorant, expressed as a percentage of the response amplitude to the same testing odorant when the olfactory epithelium was not pre - adapted. 3 kPZS adapted and DkPES adapted indicate when the olfactory epithelium was pre - adapted to 3kPZS and DkPES, respectively. D ifferent letters indicate significant differences in responses amplitude with respect to the corresponding adaptation ( DkPES: t = 3.78, P = 0.004; 3kPZS: t = 2.81, P = 0.019 ). Vertical bars represent one standard error, n = 6. 74 Females are attracted to the 3kPZS and DkPES mixture at ratios similar to those identified in SMW extracts A ratio of 3kPZS:DkPES observed in spermiated male wash - water (SMW) extracts from a group of mature males was found to be 30:1 (3kPZS:DkPES) in our previous work ( Li et al., 2013 b ) , and thus the 30:1 ratio was examined along with other ratios during th ese trials. The point in the stream at which the molar concentration of our treatments reached that of our target whole - stream concentration of 5x10 - 13 M for 3kPZS (as indicated by S in Figure 3 - 3A) was calculated based on rhodamine dye concentrations as p reviously described ( Johnson et al., 2009 ) an d determined to be roughly 15 m downstream from the experimental nests (Figure 3 - 3A ) . Swim tracks overlain onto rhodamine plume maps indicate d that subjects showed a clear preference for the 30:1 3kPZS:DkPES mixture compared to the 1:1, 10:1, and 20:1 mixtures (Figure 3 - 3B and Supplemental Figure S 3 - 2). Sinuosity, calculated as the total track length (starting at point S ) divided by the shortest distance between the start and end of the track, was highest when subjects were exposed to SMW. Mean sinuosit y of swim tracks was lowest during 1:1 mixture treatments, and highest during SMW treatments ( ANOVA and post - hoc HSD: F 5, 74 = 3.19, P = 0.012 ). Finally, mean sinuosity did not differ between SMW and 30:1 treatments (HSD: P = 0.766), yet was lower for 3kPZS treatments compared to SMW treatments (HSD: P = 0. 023, Figure 3 - 3C). Full swim tracks for all treatments can be seen in Supplemental Figure S3 - 1 . 75 Figure 3 - 3 . Behavioral responses of mature female sea lamprey to DkPES in a natural spawning stream. (A) Schematic of t he 45 m - long section of the Upper Ocqueoc River used for field bioassays. The point at which treatment concentrations reached that of our target whol e - stream molarity was calculated based on rhodamine concentration, and is indicated at S . The plume map and estimated treatment concentrations (Molarity) are shown, and were mapped following described methods ( Johnson et al., 2009 ) . Locations of passive integrated transponder (PIT) antennas are shown. Downs tream release cages are indicated with solid black boxes. ( B ) Swim tracks of individual test subjects during 3kPZS (5x10 - 13 M) vs. 3kPZS (5x10 - 13 M) control treatments and 3kPZS (5x10 - 13 M) vs. ratio 30:1 (5x10 - 13 M 3kPZS:1.7x10 - 14 M DkPES) treatments are shown starting at point S . ( C) Mean sinuosity (track length/shortest connecting line) of tracks for each treatment (± 1 SEM) was calculated from point S up to adjacent nests (~ 15 m) during treatments: 3kPZS (5x10 - 13 M), spermiated male washings (SMW, applied at 5x10 - 13 M 3kPZS benchmark), ratio 1:1 (5x10 - 13 M 3kPZS:5x10 - 13 M DkPES), ratio 10:1 (5x10 - 13 M 3kPZS:5x10 - 14 M DkPES), ratio 20:1 (5x10 - 13 M 3kPZS: 2.5x10 - 14 M DkPES), and ratio 30:1. Treatments that share a letter are not significantly different ( ANOVA and post - hoc HSD: F 5,74 = 3.19, P = 0.012) . The number of responding subjects ( n ) are indicated within each column. 76 From passive integrated transponder (PIT) telemetry data, the proportion of mature female subjects that moved upstream was not different across treatment levels (Logistic regression: X 2 5 = 5.85, P = 0.322 , Supplemental Table S3 - 2 ). This is likely due to th e fact that all treatments contained 3kPZS, a pheromone component known to induce upstream movement of mature female sea lamprey ( Johnson et al., 2009 ) . The proportion of subjects entering the treatment nest (within 0.5 m of the treatment source) during SMW or 30:1 treatments did not differ from on e - another (Two - tailed t - test: t = 1.52, P = 0.132), and both showed highest entry rates ( X 2 5 = 25.51, P < 0. 001, Figure 3 - 4A) compared to all other treatments. However, the mean time spent within the nest by mature females differed between SMW and the 30:1 treatments. Specifically, subjects showed higher retention (minutes) inside nests when SMW was administered ( F 11, 123 = 3.55 , P < 0.001 ) compared to all other treatments. No differences were observed in retention time across all other treatments ( see Figure 3 - 4B for all statistical comparisons ) . 77 Figure 3 - 4 . Pair - wise comparison of male pheromone components at various ratios for their induction of p reference response in mature female sea lamprey. Treatments included: 3kPZS (5x10 - 13 M, n = 16) vs. 3kPZS (5x10 - 13 M, n = 17), spermiated male washings (SMW, applied at 5x10 - 13 M 3kPZS benchmark, n = 26) vs. river water (RW, n = 1), ratio 1:1 (5x10 - 13 M 3kPZS: 5x10 - 13 M DkPES, n = 16) vs. 3kPZS (5x10 - 13 M, n = 9), ratio 10:1 (5x10 - 13 M 3kPZS:5x10 - 14 M DkPES, n = 9) vs. 3kPZS (5x10 - 13 M, n = 11), ratio 20:1 (5x10 - 13 M 3kPZS: 2.5x10 - 14 M DkPES, n = 8) vs. 3kPZS (5x10 - 13 M, n = 7), and ratio 30:1 (5x10 - 13 M 3kPZS:1.7x10 - 14 M DkPES, n = 12) vs. 3kPZS (5x10 - 13 M, n = 3). Dashed vertical grey lines separate pair - wise comparisons. (A ) Percentage of subjects that entered each treatment nest. Horizontal grey line indicates 50%. Trea tments that share a letter are not significantly different (Logistic regression: X 2 5 = 25.51, P < 0.001). (B) Mean (± 1 SEM) retention (min.) of subjects inside respective treatment nests. Treatments that share a letter are not significantly different ( ANOVA and post - hoc F 11,123 = 3.55, P < 0.001 ) . 78 DISCUSSION This study confirms the structure of DkPES and defines the function of DkPES and 3kPES as a pheromone mixture in sea lamprey. Our previous work suggested that certain ratios of DkPES and 3kPZS may be important for orientation of receiver females to the mal e odorant source, and mature females showed a behavioral preference to the mixture at multiple ratios ( Li et al., 2013 b ) . In the present study, we unequivocally confirmed the structure of DkPES through chemical synthesis, showed that DkPES was discrimi nated from 3kPZS in the olfactory epithelia, and elucidated orientation strategies of receiver females exposed to mixtures of DkPES and 3kPZS in a range of ratios. From the data on locomotion patterns of mature females released in a sea lamprey spawning s tream, b oth the behavioral pref erence and the sinuosity analyse s indicate that a range of mixtures is likely to be effective at drawing receiver females towards the source, provided that DkPES remains the more minor component in the mixture ( Li et al., 20 13 b ) . Thus, we postulate that minor components such as DkPES allow receivers to gauge distance to the source, as DkPES is likely not detectable until the receiver is within proximity of a signaler (Figure 3 - 5). Further , we postulate that the mixture with peak behavioral activity (30:1, 3kPZS:DkPES) remains within a range similar to that released by mature males. Our previous study observed an average 30:1 ratio of 3kPZS:DkPES (10 - 10 M 3kPZS:3.3x10 - 12 M DkPES) in wash - water extracts collected from a gr oup of mature males ( Li et al., 2013 b ) . In the future, it will be useful to test individual male variation in ratios, and examine variables that may influence these ratios. The increasing behavioral preference for a mixture of pheromone components that approaches that of the natural ratio released from a signaler sea lamprey is consistent with 7 9 research on insects such as moths to mixtures of pheromone components ( Linn and Roelofs, 1989 ; Reyes - Garcia et al., 2014 ; Wyatt, 2010 ) . While several components and variations of mixtures can yield attractive behaviors in moths, individual moth species often show a peak p reference to the ratio that best reconstructs that of the natural ratio emitted by conspecific senders ( Cardé et al., 1977 ; Coracini et al., 2001 ; Reyes - Garcia et al., 2014 ) . Behavioral data here suggests a similar relationship in sea lamprey. Unique ratios that vary between members of similar Lepidopteran species have been theorized to act as reproductive iso lating mechanisms ( Cardé et al., 1977 ) . While there appears to be some overlap in sex pheromone components used among phylogenetically similar lamprey species ( Buchinger et al., 2013 ) , specific ratios of sex pheromones have not yet been identified in other lampreys. The detection thresholds for the two pheromone components varied between our previous and current studies. The threshold of detection of i solated DkPES in our previous study was within a range of 10 - 7 10 - 8 M in our subjects ( Li et al., 2013 b ) , while the current study showed a EOG detection threshold of 10 - 10 M. Similarly, compound 3kPZS showed an EOG limit of detection of 10 - 10 M in the previous study ( Li et al., 2013 b ) , and 10 - 13 M in the current study. In a separate study (Siefkes 2002), the detection threshold for 3kPZS was determined as 10 - 12 M. This variation may be attributable to sea lamprey conditions during the time of research. In our experience, longevity and sensitivity can vary based upon stream temperatures, date of capture, time held in traps by government agencies, holding conditions before transport to researchers, and length of transport. Additionally, physiolog ical processes degrade as the spawning season progresses and sea lamprey approach natural senescence ( Manion and Hanson, 1980 ) , which likely impacts their detection threshold of compounds and ability to respond. Finally, variability 80 in detection thresholds may be due to an updated and more sensitive electro - olfactogram (EOG) recording system in our laboratory since our previous study. In summary, our data suggest that mature females detect DkPES and 3kPZS with independent receptors w hile making orientation decisions based on the odorant plume ( Johnson et al., 2012 ) . The sinuous pattern of movement and casting into and out of the odorant plume seen here are consistent with behaviors observed in birds and fishes when tracking odors ( DeBose and Nevitt, 2008 ) . Our data suggest that female sea lamprey encounter a plume structure , discriminate between 3kPZS and minor components such as DkPES at the olfactory level, and utilize stream flow using odor - conditioned rheotaxis ( Johnson et al., 2012 ) , to gauge location of a conspecific signaler (Figure 3 - 5). We hypothesize that additional components or their mixtures with 3kPZS and DkPES function to arrest females inside the nest boundaries for courtship with signaler males. Taken together, examination of chemical communication in sea lamprey may provide insights into the independent detection mechanisms of pheromone mixtures at various ratios ( Kang and Caprio, 1997 ; Keller - Costa et al., 2014 ) , the underlying role of these ratios as reproductive isolating mechanisms in vertebrates ( Keller - Costa et al., 2014 ) , and provide a useful tool for the integrated pest management of invasive sea lamprey ( Johnson et al., 2013 ; Teeter, 1980 ) . 81 Figure 3 - 5 . A c onceptual model for a female sea lamprey encountering the conspecific male pheromone. Nesting males release compound 3kPZS that induces upstream movement in mature (ovulated) females ( Johnson et al., 2009 ; Li et al., 2002 ) . Background 3kPZS concentrations remain very high throughout spawning grounds ( Xi et al., 2011 ) , indicated by (3kPZS) under stream flow. Females are able to detect and discriminate between 3kPZS and minute compounds such as DkPES at the olfactory level, while increasing the sinuosity of their swim path around the odor plume ( Johnson et al., 2012 ) , as they approach the source. 82 METHODS Test Subjects All procedures using sea lamprey presented in this manuscript were approved by t he Michigan State University Institution al Animal Care and Use Committee prior to any experimentation (AUF# 05/09 - 088 - 00 and 03/ 11 - 053 - 00 ). Adult sea lamprey used for EOG recording s were collected in spring 2012 by commercial fishing companies in Lake Huron and transported to United States Geological Survey and Great Lakes Science Center Hammond Bay Biological Station (HBBS), and later shipped to Michigan State Univers ity, East Lansing, Michigan . Migrating adult sea lamprey were captured by the United States Fisheries and Wil dlife Service and Department of Fisheries and Oceans Canada from tributaries to Lake Michigan and Lake Huron in May June 20 12, following animal use and care protocols established by those agencies. Live lamprey were transported to HBBS and held in 500 - 10 00 L - capacity flow through tanks until use. Males were kept separate from females in tanks and never used for this study . S ea lamprey are a non - native invader of the Laurentian Gre at Lakes and their tributaries. Hence, males and females could not be releas ed together upstream of sea lamprey barriers to reduce the risk of repopulation of the river system. To produce sexually mature ovulated female (MF) test subjects , lamprey were transferred to acclimation cages constructed of polyurethane mesh and PVC pipe (0.5 m 3 ) located in the lower Ocqueoc R iver, Millersburg, Michigan , to allow natural maturation in situ . Sea lamprey were monitored daily for signs of sexual maturation . Briefly, MFs were identified by first checking for secondary sexual characteristics ( Applegate, 1950 ) , then applying gentle pressure to the abdomen and checking for expression of ovulated oocytes from the cloacal aperture. 83 Purity analysis of DkPES Quality of synthetic DkPES in chemical structure and purity were examined using a series of chemical analyses by high resolution mass spectrum (HR - ESI - MS) , 1 H, 13 C NMR spectra and DSC analyse s . 1 H - NMR and 13 C NMR experiments were performed on the synthetic DkPES ammonium salt using a Varian Inova 600 MHz NMR spectrometer at MSU Max T. Rogers NMR facility. Samples (ca. 5.0 mg) was prepared in CD 3 OD and subjected to NMR analysis. The results were compared with original DkPES and displayed in Figure 3 - 1. HR - ESI - MS of synthetic DkPES ammonium salt was performed on a TQ - S TOF LC mass sp ectrometer (Waters Corporation, Milford, M A , USA). 2 O = 1:1, v/v) was in jected by auto - sampler. The mobile phase consisted of water as (A), and methanol (B). The isocratic gradient (30%A and 70%B) was used a eluant . The UHPLC effluent was introduced into the mass spectrometer with electrospray ionization in the negative mode . The ESI - MS parameter was set as capillary voltage, 2.60 kV; extractor voltage, 5 V; source temperature, 150 o C; desolvation temperature, 500 o C; desolvation gas flow, 800 L/h (N 2 , 99.9% purity). Argon (99.9999% purity) was introduced as the collision gas i nto the collision cell at a flow rate of 0.15 mL/min. Data were collected in centroid mode with a scan range of 50 1000 m/z . Data process performed on MassLynx 4.1. DSC were used to evaluate the absolution purity of synthetic DkPES without standard. The DSC experiment was carried out on DSC Q2000 (TA Instrument, New Castel, DE, USA). Sample (ca. 3.2 mg) was set in a Tzero TM pan with lid ( TA Instrument, New Castel, DE, USA ) and stored on oven by auto - sampler. The temperature ramp was set as start from 20 o C to 250 o C , heating rate is 1.0 o C/min . The absolute purity of DkPES was analy zed by software TA Instruments universal analysis 2000 provided by the manufacturer. 84 Synthesized Pheromone Components Compound 3kPZS was custom synthesized by Bridge Organics Co. ( purity = 97% , Vicksburg, MI) as a white powder salt. DkPES was chemically synthesized by Apeloa Kangyu Pharmaceutical Co. ( purity = 97 % , Dongyang, Zhejiang, China ) as a white powder salt. The synthetic compound exhibits the same spectral characteristi cs and biological activity as the published natural compound ( Li et al., 2012 ) . A 1 mg/mL stock solution of each compound (in 50 % methanol :deionized water ) was prepared . S tock solution s were stored at - 2 0°C until use. Electro - olfactogram (EOG) Recording Electro - olfactogram recording s were obtain ed from lamprey in spring 2012. Our procedures for EOG a re detailed in Li et al. ( 2013 b ) . Briefly, sea lamprey were anesthetize d with 3 - aminobenzoic acid ethyl ester (100 mg/L; MS222, Sigma - Aldrich Chemical Co.), immobilized with an intra - muscular injection of gallamine triethiodide (3 mg .kg - 1 in 0.9% saline), and placed in a partially inundated V - shaped Plexiglas cradle. Gills w ere continuously irrigated with aerated water containing 50 mg/L MS222. The olfactory rosette was surgically exposed and olfactory responses to stimuli were recorded by borosilicate electrodes filled with agar 0.04% in saline 0.9 % connected to solid - state electronics with Ag/AgCl pellets in 3mol/L KCl. Electrodes where placed between olfactory lamella (recording electrode) and external skin (reference electrode). Olfactory responses were filtered and amplified by a NeuroLog system model NL102, filtered wit h a low - pass 60Hz, model NL125 (Digitimer Ltd., Hertfordshire , England ) , digitized by a Digidata 1550 ( Molecular Devices LLC., Sunnyvale, California ), and stored on a PC running AxoS cope 10.4 software ( Molecular Devices LLC. ). For concentration - response curves, stimuli were serially diluted in charcoal filtered water from a 10 - 3 M stock solution of synthesized 3kPZS and DkPES, respectively. Responses were 85 measured in mV, blank subtracted and normalized to those of L - arginine at 10 - 5 M. Response thresholds were determined from the concentration with no significant difference with the blank t - test, = 0.05). Cross - adaptation experiments followed the protocol of Huertas et al. ( Huertas et al., 2007 ) . Dilutions of 3kPZS and DkPES that evoked the same EOG amplitude (10 - 9 M 3kPZS and 10 - 7 was continually exposed to unadapted 3kPZS solution for at least 1 min and the response to a sample 2x10 - 9 - of 3kPZS and DkPES at 10 - 9 M and 10 - 7 response. The amplitudes of self - adapted control and adapted responses were then showed as a percentage of the appropriate unadapted response. The olfactory epithelium was then exposed to charcoal filtered water for 10 min, and the process repeated using DkPES as the adapting solution and 3kPZS and the mixture as stimuli. The sequence of ada ptation of 3kPZS and DkPES were randomized. Overall treatment effects were tested with an analysis of variance (ANOVA), and s ignificant differences between responses of test odorants when the epithelium was adapted to particular odorants were tested by Stu t - test ( = 0.05 ). Passive Integrated Transponder (PIT) Tagging Procedures P assive integrated transponder (PIT) tag ging procedures for MFs followed Johnson et al. ( Johnson et al., 2009 ) . Each PIT tag was fitted into a latex sleeve and attached to the mid - dorsal region of each MF using a suture on both sides (Size 3 - 0, Ethicon Inc., Cornelia, GA ). Subjects were also fitted with unique color combinations of ribbon tags (Hallprint, Hindmarsh Valley, South AU) through each dorsal fin to identify in dividuals for visual tracking during trials. 86 Tagged animals were immediately transferred into aerated holding tanks with a constant flow of Lake Huron water for up to 24 hours, until they were stocked into stream acclimation cages. Field Bioassays Trials were ran from 12 June 27 July 2012, a time - frame representative of a typical spawning season for sea lamprey ( Applegate, 1950 ; Manion and Hanson, 1980 ) , in a 45 m - long stretch of the Upper Ocqueoc River, Millersburg, MI. The field site is consistent with our previous study ( Li et al., 2013 b ) , w ith slight modification (Figure 3 - 3A). T wo 1 m 2 nest antennas were placed side - by - side to one another on the upstream end of site, laid flat on the stream bed, 1.5 m apart . These antennas monitored the proportion of subjects that entered a particular containing treatments. Downstream 45 m , two aluminum - mesh release cages (0.25 m 3 ) equipped with sliding release doors were positioned in the center of the stream channel. A PIT antenna roughly 0.5 m - high x 6 m - long was positioned 5 m upstream of the release cages to monitor individuals that exit the cage and move upstream (Figure 3 - 3 A). Stream temperatures were recorded at the start of and end of each trial. Stream discharge was estimated every three days, or after every precipitation event, at a fix ed location in the stream using a Marsh - McBirney portable flow meter (Fl o - Mate 2000, Fredrick, MD ) to determine the amount of treatment stock solution to apply to the stream and maintain consistent concentrations across trials. Treatments were diluted with 20 L of river water in large mixing bins on shore . Bins were kept consistent for each test treatment, and rinsed in the stream several times before each new trial, to reduce the potential for contamination during mixing . Each treatment solution was then p umped from bins into the stream at the center of each nest antenna at a rate of 167 mL/min (± 3 mL/min ) using p eristaltic pumps (Cole - Parmer). Trials were a total of two hours long. In th e first half - hour of each trial the treatments were 87 administered to the stream while s ubjects remained in the release cage. At the start of the following 1.5 hours , subjects were released and their movements were monitored with PIT antennas until the trial ended . No animals were recovered from the stream af ter a trial. Copper wire was wrapped around each antenna frame twice during the construction of PIT antennas for a more focused read range. Antennas were wired to a multiplexor in the field for consolidation and storage of data (Oregon RFID, Portland, OR) . Antennas were tuned to a detection sensitivity of roughly 0.3 m from the frame edges. Scan frequencies of each antenna were programmed to three scans/sec. Data for each trial were uploaded each day using a hand - held Meazura model MEZ1000 personal digital assistant (Aceeca International Limited, Christchurch, New Zealand). Details of Treatments Treatments included: (1) 3kPZS ( 5x10 - 13 M ) vs. 3kPZS (5 x10 - 13 M), (2) spermiated male washings (SMW, applied at 5 x10 - 13 M 3kPZS benchmark) vs. river water (RW) , (3) ratio 1:1 (5 x10 - 13 M 3kPZS:5 x10 - 13 M DkPES) vs. 3kPZS (5 x10 - 13 M), (4) ratio 10:1 (5 x10 - 13 M 3kPZS:5 x10 - 14 M DkPES) vs. 3kPZS (5 x10 - 13 M), (5) ratio 20:1 (5 x10 - 13 M 3kPZS: 2.5 x10 - 14 M DkPES) vs. 3kPZS (5 x10 - 13 M), and (6) ratio 30:1 (5 x10 - 13 M 3kPZS:1. 7x10 - 14 M DkPES) vs. 3kPZS (5 x10 - 13 M). were alternated each trial. Up to two trials were conducted each day depending upon the availability of mature animals . The early trial was conducted from ~0700h 0900 h, and a late trial was then run from ~ 0930h 1130 h. Ten PIT - tagged MFs were transferred to respective acclimation/release cages for each trial betw een 2000 2200 h the night prior to experimentation. Subjects were then allowed an acclimation period in the stream for a minimum of 9 hours. 88 Swim Track Mapping Swim tracks were mapped during trials following Johnson et al. ( Johnson et al., 2009 ) , with slight modification stated here. The stream section seen in Figure 3 - 3 A was fixed with after that until reaching release cages. Each transecting string was divided into tenths (of the total stream width). Given that each test subje ct was marked with a unique color combination of ribbon tags , we were able to visually observe and record individuals onto scale maps by hand as they swam upstream. Observers followed each subject until reaching the nests, using transecting strings as refe rence markers. Only subjects that were observed exiting the release cage were followed. Preference responses for the rest of the subjects that exited unseen by observers were recorded via PIT antennas. To map the odor plume, rhodamine dye was administered to the stream at our treatment pumping rate (167 mL/min) when streamflow was 537 L/sec (which fell into the range of our average stream flow across all trials of 598 ± 76 L/sec). Stream samples were taken following Johnson et al. ( Johnson et al., 2009 ) with one modification stated here. Rhodamine concentrations we re detected and recorded at each sample point ( i.e. every tenth of the stream widths, marked along transecting strings) with a hand - held DataBank datalogger and Cyclops - 7 Optical Rhodamine Dye Tracer (Turner Designs, Sunnyvale, CA), instead of hand - grabbin g and analyzing water samples with a laboratory spectrophotometer ( Johnson et al., 2009 ) . All swim tracks were traced onto a digital map using a tablet computer (Lenovo X201 Tablet). Swim tracks and the odor plumes were both mapped to scale, independently, and tracks were later overlain onto plumes in a double - blind design for all track figur es ( Supplemental Figure S3 - 1 ). 89 Statistical Analyses of Behavioral Data Rhodamine concentrations were used to estimate the downstream point (transecting line) at which average treatment molarity reached that of our instream target concentration and became detectable from bank - to - bank. From this transect, indicated as S in Figure 3 - 3A , sinuosity of each swim track was calculated by dividing the track length by the length of a straight line connecting the start and end of each track. Transect S was chosen for a sinuosity calculation start point because it was the point at which lamprey would begin exposure to a gradual increase in odorant concentration and plume edges that would allow subjects to cast into and out of the plume structure. Since s inuosity values are proportions ( i.e. a value of 1 is a straight line), values were square root - homogeneity of varian ce was used to examine variance of newly transformed dat a. Once homogeneity of variance was observed, an ANOVA and post - hoc ( = 0.05) was conducted for statistical comparisons of retention data across treatments using R - software (R version 2.11.1 Vienna, Austria ) For all PIT telemetry data, l ogistic regression with a binomial distribution was ex amined for each response variable across treatments (R version 2. 11.1 ). No signs of nonlinearities or over dispersion were observed in the models. All behavioral statistics reported are two - tailed analyses ( = 0.05). Two main binary response variables wer e examined from PIT data: ( 1) the distribution of subjects that swam upstream from release cages and did not move back down ( Up : 1 = hit on upstream release antenna and continued towards nests, 0 = did not hit on upstream release antenna) and (2 ) of those animals that hit on the upstream antenna , the distribution that entered the containing the test treatment ( Enter nest : 1 = hit treatment nest, 0 = hit control nest ) . Since 3kPZS was administered to both nests during control trials, one 90 ne st was randomly assigned f or statistical purposes nest was randomly chosen to be the right nest, and alternated every trial to follow the same pattern of the other treatments . dentification number prevented any pseudo - replication from test subjects released during trials. When a subject entered a nest, observers recorded the amount of time spent inside the 1 m 2 area (retention, min.) until respective subjects moved on. All retention data was examined for violation of assumptions of normality and homogeneity across variance before further stati stical analyses were conducted. Retention data that were not normally dist ributed or showed heterogeneity across variance were log - tra nsfor homogeneity of varian ce was used to examine variance of newly transformed dat a. Once homogeneity of variance was confirmed, an ANOVA and post - hoc ( = 0.05) was conducted for statistical comparisons of retention across treatments (R version 2.11.1). 91 ACKNOWLEDGEMENTS This work was supported by the Great Lakes Fis hery Commission, Ann Arbor, MI. We are grateful to personnel of the U.S. Geological Survey Hammond Bay Biological Station for use of facilities, the U.S. Fish and Wildlife Service and Fisheries and Oceans Canada for providing sea lamprey, and to Dolly Trump and Lydia Lorenz for the use o f their private land for stream access. Special thanks to field technicians; Ethan Buchinger, Elizabeth Racey, Brian Grieve, Zak Smillie, and Kyle Hill, for their help with tracking and mapping sea lamprey movement . 92 SUPPLEMENTAL INFORMATION Table S 3 - 1. High resolution mass spectrum report for synthesized DkPES ammonium salt (HR - ESI - MS) 93 Table S 3 - 2. Percentage of sexually mature female sea lamprey that moved upstream 45 m (Up) to side - by - side nest antennas activated with pheromone treatments. Treatment nest Adjacent nest Trials Released Up 3kPZS (5E - 13M) 3kPZS (5E - 13M) 12 112 29% (33) A SMW River water 8 81 33% (27) A 1:1 3kPZS (5E - 13M) 9 90 28% (25) A 10:1 3kPZS (5E - 13M) 7 69 29% (20) A 20:1 3kPZS (5E - 13M) 6 66 23% (15) A 30:1 3kPZS (5E - 13M) 7 81 19% (15) A X 2 5.85 df 5 P - value 0.322 Treatments included: 3kPZS (5E - 13 M) vs. 3kPZS (5E - 13 M), spermiated male washings (SMW, applied at 5E - 13 M 3kPZS benchmark) vs. river water, ratio 1:1 (5E - 13 M 3kPZS:5E - 13 M DkPES) vs. 3kPZS (5E - 13 M), ratio 10:1 (5E - 13 M 3kPZS:5E - 14 M DkPES) vs. 3kPZS (5E - 13 M), ratio 20:1 (5E - 13 M 3kPZS: 2.5E - 14 M DkPES) vs. 3kPZS (5E - 13 M), and ratio 30:1 (5E - 13 M 3kPZS:1.67E - 14 M DkPES) vs. 3kPZS (5E - 13 M). 94 Figure S 3 - 1 . All tracks and plumes for all treatments during field trials. Treatments and ratios are described in Figure 3 - 3 . 95 CHAPTER 4 A TERRITORIAL PHEROMONE THAT DEFINES NEST BOUNDARY IN THE SEA LAMPREY 96 ABSTRACT Territorial pheromones, substance s that advertise territory ownership and avert intruders of the same species , have been hypothesized but in no species have they been identified. We foun d that individual spawning male sea lamprey can host many mature females, but rarely males, in their nests. Intruding males were immediately attacked and cast out of the nest. From spawning male washings a novel steroid, P AMS - 24 , was identified and shown to be a distinct odorant in adults. PAMS - 24 r elease was exclusive to spawning males, and was released with a mixture of 3kPZS (a known mating pheromone in sea lamprey) in a ratio of 100:1 (3kPZS:PAMS - 24, molar:molar). Spawn ing male wash - water and mixtures of synthetic 3kPZS: PAMS - 24 (100:1) applied to artificial nests in a spawning habitat lured females and often averted males. Some large males, after increasing search activity in the vicinity, entered 100:1 treated nests. We conclude that PAMS - 24 advertises nest ownership and reduces intrusion as a partial territorial pheromone . 97 INTRODUCTION A territory is an area occupied by an animal, often exclusively, through overt defense or advertisement ( Wilson, 1970 ) . Many animals defend territories to secure resources such as mates, but do so at the risk of aggressive intrusions . Selection should favour the evolution of signals that advertise territories and minimize such risk. Indeed chemic al signals are used throughout the animal kingdom to convey territory ownership. However, a agonistic behavior ( Hölldobler and Wil son, 1977 ) , has not been identified. We sought to identify the structure and function of a territorial pheromone in mature male sea lamprey ( Petromyzon marinus ) that establish and defend territories for nesting ( Manion and Hanson, 1980 ; Teeter, 1980 ) . When the spawning season commences in early summer, mature males (at the onset of spermiation) congregate in gravel patches of spawning streams, build nests (~ 0.5 m 2 ), and reproduce within the final weeks of their life ( Manion and Hanson, 1980 ) . Nests can be evenly spaced, yet as close as a few meters to one - another. Once nesting, the male re leases a sex pheromone 3 - keto petromyzonol sulfate (3kPZS) ( Li et al., 2002 ; Yun et al., 2003 ) that draws females to the nest ( Johnson et al., 2009 ; Li et al., 2002 ; Siefkes et al., 2005 ; Yu n et al., 2003 ) . Mature female sea lamprey are polyandrous ( Applegate, 1950 ; Manion and Hanson, 1980 ) , often moving among multiple active nests and mates. In stark contrast, mature males defend their nests from male intruders, often times casting them from their nest ( Applegate, 1950 ; Manion and Hanson, 1980 ; Teeter, 1980 ) . 98 RESULTS We hypothesize that sea lamprey have evolved an effective mechanism that communicates occupancy (territory) inside a nest within the nest cluster to avoid constant intrusion by rival males and broadcast mate - readiness to gravid females. To determine if the male territorial signal is chemical, we tracked movement patterns of mature adults in spawning ground s in the presence o f male odors (Supplementary Figure S4 - 1). Mature females moved to the exact source of spermiated male washings (SMW), as expected ( Teeter, 1980 ) , while mature males often avoided SMW (Supplementary Figure S4 - 2). We then examined whether male sea lamprey protect a nest - sized territory in their natural habitat (Ocqueoc and Cheboygan Rivers, MI, USA). The close proximity of the responses shown in Supplementary Figure S4 - 2 (0.5 - 1 m 2 ) were similar to the average size of natural nests (0.65 ± 0.19 m 2 , n = 18). Of the 51 occupied nests and 116 nesting sea lamprey observed in those nests in 2009, only 10 male - male interactions were observed: three intruders avoided the resident male while the other seven intruders were attacked and cast from the nest (Supp lementary Movie S4 - 1). In all observed male - male conflicts, the resident male cast the intruder male from the nest. The number of females accompanying a male on a nest ranged from zero (Supplementary Figure S4 - 3a) up to seven (Supplementary Figure S4 - 3b). Larger males courted with a greater number of females at one time compared to smaller males (Supplementary Figure S4 - 4a). Nests were maintained almost always by a single male (Supplementary Figure S4 - 4b). Having shown that SMW may indeed contain a territo rial odor , we fractionated SMW using a bioactivity - guided strategy. Solid phase extract of SMW was subjected to liquid chromatography over silica gel and detected by Thin Layer Chromatograph (TLC), subsequently yielded nine fractions. Fraction 8 (30 mg) wa s found to contain two main molecular masses ( m/z 99 at 471 and 623) by ESI mass spectra (full scan). The compound with m/z 471 was confirmed as the known sex pheromone 3kPZS high resolution mass spectrometry. The other compound was further purified by succes sive chromatography, yielding a compound 1 (Figure 4 - 1a, 1.4 mg, purity ca. 99.5% by HPLC analysis). The high - resolution negative ESI - MS at m/z 623.4108 [M H] (calculated for C 34 H 59 N 2 O 6 S, 623.4049) implied a molecular formula of C 34 H 60 N 2 O 6 S. The 1 H NMR spectrum (Supplementary Table S4 - 1) suggested that 1 was a steroid with characteristic appearance of a side chain similar to cholesterol. Comparison of the carbon resonance of 1 with those of PADS and squalamine (Supplementary Figure S4 - 5) showed 1 as (3 ,5 ,7 ,24 R ) - 1 - [3 - [[24 - sulfooxy - cholestan - 3 - yl]amino] - propyl] - 2 - pyrrolidinone (Figure 4 - 1A) , herein named petromyzonamine - 24 - monosulfate (PAMS - 24), after petromyzonamine disulfate (PADS) ( Sorensen et al., 2005 ) . The assignment of PAMS - 24 structure was unequivocally confirmed by the identical mass and overlapped NMR spectra shown by the purified 1 and the compound synthesized according to the deduced structure, indicating both the natural identified and synthesized compounds possess the c hemical structures of PAMS - 24. We further reasoned that, if PAMS - 24 was indeed a territorial pheromone, its release should be exclusive to nesting male s and it must be a potent odorant for adults. As expected, extensive LCMS analyses failed to detect PAMS - 24 in wash water from larvae, females or immature males. Analysis of wash - water from the head region and tail region of mature males held in a bisected chamber (Siefkes 2003) indicated PAMS - 24 was released exclusively from head region (Supplementary Figure S4 - 6a, b), likely through gills, consistent with the release route of 3kPZS which is released exclusively from the head region of sexually mature male s by specialized gill cells ( Brant et al., 2013 ; Siefkes et al., 2003 ) . The release rate of PAMS - 24 was 100 between 0.9 29 ng/g - lamprey/hr , and the release ratio of 3kPZS to PAMS - 24 remained consistent at 100:1 (M:M) across a range of mature male weights (Su pplementary Figure S4 - 6c ) . Electro - olfactogram (EOG) recording showed both purified and synthetic PAMS - 24 were highly stimulatory for the olfactory epithelium of the adult sea lamprey, with virtually identical concentration - response dynamics (Figure 4 - 1b) . The threshold of detection was 10 - 12 molar (M), similar to that of 3kPZS ( Siefkes and Li, 2004 ) . Cross adaptive analyses, in which the olfactory epithelium was pre - adapted to one chemical and then EOG response to the second chemical was recorded, showed that PAMS - 24 and 3kPZS do not suppressed the res ponsiveness of each (Supplementary Figure S4 - 7). These data indicate PAMS - 24 is a potent odorant distinguished from 3kPZS by the adults. 101 Figure 4 - 1. ( A ) Structure of PAMS - 24 ( 1 ) including key 1 H - 1 H COSY (bold lines) and HMBC (arrows) correlations of PAMS - 24 recorded in DMSO - d 6 using a Brüker NMR spectrometer ( 1 H NMR, 900 MHz; 13 C NMR, 225 MHz). ( B ) Semi - logarithmic plot of normalized electro - olfactogram (EOG) amplitudes recorded in sea lamprey in response to different concent rations of PAMS - 24 purified from spermiating male lamprey washings and PAMS - 24 in synthesized form. Data are the means ± SEM ( n = 6), and are blank corrected and normalized to the amplitude of response to 1E - 5 M L - arginine. B 102 A territorial pheromone is thought to both advertise and to deter ( Hölldobler and Wilson, 1977 ) . We predicted that PAMS - 24 should both advertise to mature females and deter mature males, since a nest is constructed primarily by males which then signal to females for courtship ( Teeter, 1980 ) . In a system where mature females were released 45 meters downstream of two adjacent nests (1 meter 2 each) positioned 1.5 meters a part (Figure 4 - 2a), one nest perfused with 3kPZS at 5x10 - 13 M and the other with 3kPZS:PAMS - 24 at 5x10 - 13 M: 5x10 - 15 M (100:1), females moved upstream and preferred the 100:1 activated nest (Supplementary Table S4 - 2a). While 3kPZS alone is known to induce upstream movement in sea lamprey ( Johnson et al., 2009 ) , t he natural ratio likely functions as a proximal cue to females of the location of a mate - ready male ( i.e. as a sex pheromone). Responses of mature females to PAMS - 24 alone could not be evaluated, as mature females will not move upstream without the presenc e of 3kPZS in the system ( Johnson et al., 2009 ) (Supplementary Table S4 - 2a). It is possible that the mass plume of 3kPZS emanating from a stream area of high nest density induces upstream movement ( Johnson et al., 2009 ) , whereas PAMS - 24, released at a much lower rate and detected at lower sensitivity, adds an additional layer of information to advertise an individual nest. Similar to mature females, mature males also swam upstream towards a source of MMW, synthesized 3kPZS, or 3kPZS:PAMS - 24 at 5x10 - 13 M: 5x10 - 15 M (100:1) (Supplementary Table S4 - 2b). Males preferred a nest perfused with synthesized 3kPZS (Figure 4 - 2b ), which is also consistent with females ( Johnson et al., 2009 ; Siefkes et al., 2005 ) . However, mature males often sharply avoided a nest (0.5 meters 2 ) perfuse d with MMW and 3kPZS:PAMS - 24 at 5x10 - 13 M: 5x10 - 15 M (100:1) which is inconsistent with females. We then eliminated the main mating pheromone component contained in MMW, 3kPZS, as a possible candidate for a territorial pheromone. 103 Having demonstrated that avoidance to PAMS - 24 is exclusive to mature males, and now that we understood the scale of avoidance specifically in mature males, we aimed to evaluate whether mature male subjects become more active in their search behavior (frequ ency of turning) when approaching a source of PAMS - 24, similar to behaviors of other animals as they ( Bradbury and Vehrencamp, 2011 ) . Within 5 m downstream of the treatment sources, mature males increased their sharp turning (> 9 0° turn) behavior specifically during treatments containing the natural ratio of 3kPZS to PAMS - 24 (MMW and 100:1 treatments , Figure 4 - 2c). Since males still approach the boundary of a nest perfused with treatments containing PAMS - 24, we then summarized the numbers of swim - tracks that specifically entered or avoided each treatment nest. Of the subjects that were manually tracked in the stream, 50% avoided the natural 100:1 ratio ( i.e. the other 50% choosing to enter, Figure 4 - 3). We concluded that males choo se to enter a natural ratio of 3kPZS:PAMS - 24 or MMW as frequently as they avoid these treatments . Further, responses to MMW and 3kPZS alone show a clear inverse relationship where mature male subjects preferred to enter 3kPZS and avoid SMW. Finally, since we observed an equal number of males entering and avoiding the natural ratio of 100:1, and since larger males are more likely to win fights ( Manion and Hanson, 1980 ; Teeter, 1980 ) , we examined whether males that chose to enter were larger in body weight (g) tha n those that avoided. Larger males chose to enter the 100:1 ratio compared to 3kPZS alone ( t 15 = 1.75, P = 0.04 , Figure 4 - 4 ). 104 Figure 4 - 2. Details of field behavioral studies testing mature male sea lamprey. ( A ) The 18.5 m - long section of the Upper Trout River used for field bioassays. Downstream release cages are shown as solid black boxes Plumes were mapped and concentrations of treatments (Molarity) were estimated using rhodamine dye concentrations following Jo hnson et al. ( Johnson et al., 2009 ) . T he point at which treatment concentrations reached that of our target whole - stream concentration is indicated ( S ). ( B ) Swim tracks of mature male subjects during trials where 3kPZS (5x10 - 13 M) was applied to both nests, mature male washings (MMW, applied a t 5x10 - 13 M benchmark 3kPZS) was applied to the left nest vs. vehicle, a 100:1 3kPZS:PAMS - 24 ratio (3kPZS 5x10 - 13 M:PAMS - 24 5x10 - 15 M was applied to the left nest vs. 3kPZS at 5x10 - 13 M in adjacent nest, and when 3kPZS alone (5x10 - 13 M) was applied to one nest and Vehicle was applied to the adjacent nest. ( C ) Mean number of sharp turns ( X > 90° ± 1 SEM) of mature male subjects as they approached within 5 m of the treatment sources. Treatment 1:1 was 3kPZS:PAMS - 24 (5x10 - 13 M: 5x10 - 13 M) vs. 3kPZS (5x10 - 13 M). Treatments that share a letter are not significantly different (ANOVA and post - hoc F 4,74 = 3.19, P = 0.012). Sample sizes ( n ) are shown within columns. 105 Figure 4 - 3. The proportion of mature male sea lamprey that approached the source of treatments and either directly entered (Enter) the nest (0.5 meter 2 ), or sharply avoided (Avoid) the boundary of the nest. Treatments are explained in Figure 4 - 2 . Responses within each treatment were evaluated with logistic regression (GLM: X 2 4, 64 = 15.66, P = 0.0285). Responses that share a letter are not significantly different ( = 0.05). 106 Figure 4 - 4. Mean (± 1 SEM) weight (g) of mature male sea lamp rey that either entered (Enter) or sharply avoided each treatment. Treatments are explained in Figure 4 - 2. Responses within each treatment was evaluated with a t - test. Responses that share a letter are not significantly different ( = 0.05). 107 DISCUSSION Taken together, we show that mature male sea lamprey, during their single reproductive season, release a novel compound that may function as a partial territorial pheromone. We identify the compound as PAMS - 24, and show this compound to be detected and discriminated in the olfactory epithelium of conspecific adults. PAMS - 24 induced sexually dimorphic behavioral responses in spawning phase conspecifics. The evolution of more proximal territorial pheromone components such as PAMS - 24 in sea lam prey was likely shaped within the context of the sea lamprey mating system. Mating pheromone 3kPZS is currently considered the most abundant compound released from mature males ( Brant et al., 2013 ; Li et al., 2002 ; Yun et al., 2003 ) causing background stream concentrations of 3k PZS to remain high during the reproductive season ( Xi et al., 2011 ) . Since male - male conflict is relatively rare given the high density of nests in some stream stretches, and olfactory senses are paramount to vision during at this time ( Binder and McDonald, 2007 ; Johnson et al., 2006 ) , a territorial pheromone is the most parsimonious explanation regarding regulation of nest boundaries among males. Proximity information is attainable du e to the ratio of compounds released from males ( i.e. PAMS - 24 is released in much less quantity compared to 3kPZS). The sul f ated steroid is similar to petromyzonamine disulfate (PADS) ( Sorensen et al., 2005 ) , a major compound released by larval - stage sea lamprey, indicating a possible precursor to PAMS - 24 synthesis in adults. While no PAMS - 24 was detected in larval sea lamprey durin g our preliminary analysis, a more detailed study should be conducted to determine whether larvae can synthesize and release PAMS - 24, and if so, whether PAMS - 24 functions as a navigational cue in migrating pre - spawn adults similarly to the whole smell of l arvae ( Bjerselius et al., 2000 ) . 108 territory, and olfaction is known to play a critical role in communicating territory ownership throughout the animal kingdom by means of territorial scent marking ( Bradbury and Vehrencamp, 2011 ) . The frequency of sharp turning activity (> 90°) doubled as males approached nests activated with MMW and 100:1 treatments. In the hermit crab Pagurus berhardus , individuals exposed to wash - water conditioned with fighting crabs have been shown to elicit startled behaviors ( e.g. hide in shell, turn away), while no noticea ble responses were observed from treatments with plain sea water or wash - water conditioned with non - fighting conspecifics ( Briffa and Williams, 2006 ) . Yet to our knowledge, no territorial pheromone has been identified to date. In summary, our chemical, electrophysiological, and behavioral data presented here support PAMS - 24 as a partial component of a territorial pheromone in sea lamprey. P heromones are an environmentally friendly means of integrated population control of the sea lamprey, whose invasion of the Laurentian Great Lakes sparked arguably one of the largest control strategies of an invasive fish in the w orld ( Commission, 2001 ) . The integration of pheromones such as PAMS - 24 into the sea lamprey control program could bring about a long - term reduction in sea lamprey populations that cannot be accomplished with current methods ( Christie and Goddard, 2003 ) . 109 METHODS Sea lamprey T he Michigan State University Institution al Animal Care and Use Committee approved all procedures using sea lamprey presented in this manuscript prior to any experimentation (AUF# 05/09 - 088 - 00). Since we examined PAMS - 24 release in both sexes of sexually mature adult sea lamprey, we establish her e that mature male (MM) refers to spawning - phase males that are spermiated (observed expression of spermatozoa when gentle pressure was applied to the abdomen) and mature female (MF) refers to spawning - phase ovulated females (observed expression of oocytes with gentle pressure to abdomen). Sea lamprey were captured by the USFWS and Department of Fisheries and Oceans Canada from tributaries to Lake Michigan and Lake Huron in May June 20 13, following animal use and care protocols established by those agenci es. Sea lamprey were held in 500 - 1000 L - capacity flow through tanks with constant aeration until use. Males were kept separate from females . S ea lamprey are a non - native invader of the Laurentian Gre at Lakes and their tributaries. Hence, males and females could not be released together upstream of sea lamprey barriers to reduce the risk of reproduction and repopulation of the river system (Supplemental Figure S4 - 1) . To produce MM and MF test subjects , immature subjects were transferred to acclimation cages constructed of polyurethane mesh and PVC pipe (0.5 m 3 ) located in the lower Ocqueoc River, Millersburg, Michigan, USA, to allow natural maturation in stream water. Sea lamprey were monitored daily for sexual maturation . MMs were identified by first checki ng for secondary sexual characteristics ( Chung - Davidson et al., 2013 ) , then applying gentle pressure to the abdomen and checking for expression of spermatozoa from the genital papilla. MFs were 110 identified by applying gentle pressure to the abdomen and observing a steady expression of oocytes from the cloacal aperture. Wild nesting sea lamprey observations Wild nesting sea lamprey were observed at s ites along the Ocqueoc River , Millersburg, MI an d downstream of the Cheboygan D am , Cheboygan MI, USA, from 16 - 26 June 2009 and 5 8 June 2010. Once a nest wa s located (see Supplementary Figure S4 - 3 for an example of a the back ( Chung - Davidson et al., 2013 ) . Nests were then approached from downstream. While standing downstream and to the side of each nest, observations were made for 15 minutes. The number of each sex within a nest was immediately recorded. During the rest of the observation, the frequency of visits by intruder males and male - male aggression was recorded. After each observation, depth of nest (m), water temperature (°C), water velocity (m/sec.), and the area of each nest (m 2 ) was recorded. Numbers o f males and females observed on nests can be seen in Supplementary Figure S4 - 4. Depth of nests observed ranged from 0.26 1.8 m, water temperature ranged from 17 25 °C, water velocity ranged from 0.17 0.59 m/sec., and area of each nest ranged from 0.0 8 2.0 m 2 . Extraction of sea lamprey - conditioned water Extracted odors were collected from MM sea lamprey following methods outlined in Fine et al. ( Fine et al., 2006 ) , with slight modification. Briefly, MM - conditioned water was collected throughout June and July, 2012. Wash - water was passed through vertical columns containing 2 kg of methanol - activated Amberlite XAD7HP resin (Sigma - Aldrich, St. Louis, Missouri, U.S.A.) using peristaltic pumps (Masterflex 7553 - 70, Cole - Parmer, Vernon Hills, Illinois, U.S.A.). Loading speed was ~300 ml /min . Three columns were loaded for up to 24 hours at a time. Eac h 111 colum n was then eluted with 4 L of methanol followed by 4 L of acetone. The organic s olvent s were removed under reduced pressure at 40 o C using a model R - 210 roto - evaporator (BuchiRotov apor, Flawil, Switzerland) to produce 4.2 L of extract (stored at - 80 o C ). T he extract s (containing a large amount of water) were further concentrated by lyophilization to yield brown residues . The residues were suspended in methanol and then s uccessively filtered through iltrate was collected and c oncentra ted by a roto - evaporation under reduced pressure at 40 o C to obtain 3.1 g dark residue. Roughly 60 mature males were used for wash - water collection and extraction. Separation of PAMS - 24 Crude extract from MMs was subjected to liquid chromatography over silica gel (150 g ; gradient elution from 95% CH 3 Cl/MeOH to 100% MeOH, ca. 8 L total volume ). Thin Layer Chromatograph (TLC) analysis indicated that 9 fractions were produced. Fraction 8 (30 mg) was further purified using Sephadex LH - 20 eluted with CHCl 3 - MeOH (1:1) and MeOH (100%) which subsequently yielded the compound PAMS - 24 (1.4 mg). Structural analysis 1D and 2D NMR spectra of PAMS - 24 were recorded on a Bruker Avance 900 MHz Spectrometer or an Agilent 500 MHz sp ectrometer. Mass spectra were performed on a TQ - S TOF LC mass spectrometer (Waters Corporation, Milford, Massachusetts, USA). Si gel (70 - 230 and 230 - 400 mesh, Merck, Darmstadt, Germany), RP - 18 reverse - phase Si gel (Merck), and Sephadex LH - 20 (Merck) were u sed for open column chromatography. TLC was conducted on glass plates precoated with GF 254 Si gel (Merck). Spots were visualized under UV light at 254 nm and stained by spraying plates with 5% anisaldehyde acid alcoholic solution (Sigma - Aldrich, St. Louis, Missouri, USA). 112 Determination of PAMS - 24 concentrations in extracts and wash - water samples To examine the release and distribution patterns of PAMS - 24, a novel ultra - high performance liquid chromatography tandem mass spectrometry ( UHPLC MS/MS ) method was developed to realize baseline separation and short analysis time. UHPLC - MS/MS analyses were carried out on a Waters Acquity ultra - performance liquid chromatography system (Waters, Milford, MA, USA) with a Xevo Quattro Premier XE tandem quadrupl e mass spectrometer (Waters, Manchester, UK) equipped with ESI source. The compound PAMS - 24 was separated on a C18 Acquity column (2.1 × 50 mm, 1.7 m) with the column temperature at 45 o C. The mobile phase for elution was a gradient established between so lvent A (10mM TEA in water) using the gradient started at 60% A/40% B and followed by a linear increase of B to 90% at 7 min. An isocratic elution of 90% methanol wa s maintained from 7.0 to 8.0 min, and the initial conditions were restored between 8.0 and 8.1 min and retained 2 min for equilibration. Samples were maintained in the autosampler at 4 o C throughout the analysis. The ESI source operated in negative ion mod e, and its main working parameters were set as follows: capillary voltage, 3.50 kV; extractor voltage, 5 V; source temperature, 130 o C; desolvation temperature, 350 o C; desolvation gas flow, 600 L/h (N2, 99.9% purity). Argon (99.9999% purity) used as the c ollision gas was introduced into the collision cell at a flow rate of 0.20mL/min. Data were collected in centroid mode with a scan range of 50 - 1000 m / z . Multiple reaction monitoring (MRM) measurement (623.32 > 96.68) was performed using individually optimized cone voltage (10 V) and collision energy (49 eV). The dwell time established for each transition was 0.2 s, and interscan delay was set at 20ms. Data acquisition was carried out by Masslynx 4.1 software (W aters Corp., Milford, MA, USA). 113 To examine release of PAMS - 24 across life stages and sex, lamprey conditioned wash - water was collected from larvae, IMs, MFs, and MMs. Larval sea lamprey (N = 229, 111.5 g batch - weight) were collected by the USFWS from Silv er Creek in Tawas, MI, USA, transferred to USGS - HBBS, divided into three equal groups by weight, and placed into three separate washing chambers (~ 37 g - larvae/chamber). Chambers were supplied with equal volumes of sand substrate, aeration, and fresh Lake Huron water maintained to the temperature of the stream (2.7 3.2 °C). A fourth control chamber was maintained the same as the other three chambers excluding the addition of any larvae. Each chamber contained 22.5 l of water and 6500 g of sifted sand and was 32 cm - tall x 26 cm - wide x 50 cm - long in dimension. Larvae were then allowed to acclimate in their respective chambers for 24 h with normal flow. Water supplies were then removed for collection of larval odors for 24 h. Triplicate 10 ml larval - condition ed water samples were then taken from each chamber and placed at - 20 °C until later analyses by the UHPLC MS/MS method . For IM washing collection, eight individuals were selected during the migratory season. Each individual was placed in a separate 20 l - ca pacity container containing 5 l of D.I. water for 1 hr. A control was also conducted using the same method without the addition of a male to the container. A portable aerator was used for a constant supply of oxygen to each container during washing . The te mperature of the D . I . water used for washings was acclimated to the same temperature as the stream ( 16 18 ° C ). Triplicate 10 ml samples were then collected from each container for later analyses by UHPLC MS/MS . To examine whether female sea lamprey rele ase PAMS - 24, 15 female sea lamprey were matured in acclimation cages in the Lower Ocqueoc River. MFs were placed in 30 L of D.I. 114 water for 20 hr with proper aeration. Triplicate 1 l samples were then collected and frozen at - 20 °C for later analyses with U HPLC MS/MS . To identify whether PAMS - 24 was released from the head - region of mature males, bisected chambers were used to collect wash - water from the anterior and posterior region from each subject following Siefkes et al. ( Siefkes et al., 2003 ) , with slight modifications . W e collected 7 L of posterior washings and 7 L of anterior washings from 7 individual MMs , independently, and froze samples at - 20 °C for later analyses with UHPLC MS/MS . For confirma tion that PAMS - 24 was released primarily from the head - region of MMs, wash - water extracts from the head and tail regions of 20 mature males were mixed into one batch and analyzed for PAMS - 24 concentrations. The head and tail wash - water was passed through s eparate beds of 500 mg Amberlite XAD 7HP resin contained in respective glass columns (Ace Glass Inc., Vineland, New Jersey, USA). Load speeds were m aintained between 150 and 200 mL /mi n. Extracts were eluted with 3 L of methanol and solvent s were removed un der red uced pressure at ~ 40 o C by roto - evaporation (Buchi ) yielding 0.5 L of head extract and 0.5 L of tail extract. Triplicate 1 mL samples of head and tail extract samples were analyzed with UHPLC MS/MS . To examine the release rate of PAMS - 24 and known mating pheromone 3kPZS from individual MMs, six male sea lamprey were matured in acclimation cages in the Lower Ocqueoc River, Millersburg, MI, USA. Once all individuals were confirmed to be mature ( i.e. development of secondary sexual characteristics, ex pression of gametes with gentle pressure to abdomen) each individual was placed in a separate 5 L of D.I. water for 10 min. A control was also conducted using the same method minus the addition of a male to a container. A portable aerator was used for a co nstant supply of oxygen to each container during washing . The 115 temperature of the D . I . water used for washings was acclimated to the same temperature as the stream ( temperature range during sampling was 1 8 - 22° C ). Triplicate 10 ml samples were then collected from each container for analyses with UHPLC MS/MS. Each male was weighed, and release rates were calculated (ng/g - lamprey/hour) for both PAMS - 24 and 3kPZS. All values from UHPLC MS/MS analyses associated with a signal - to - noise ratio 10 were considered below the lower limit of quantitation and automatically removed from the data set. All comparisons were examined for violation of assumptions of normality and homogeneity across variance before further statistical analyses were conducted . Data tha t were n ot normally distributed or showe d heterogeneity across variance were log - homogeneity of variance was used to examine variance of newly transformed data. Once homogeneity of variance was observed, an ANOVA and post - hoc Tuk or t - test = 0.05) was conducted in R ( version 2.11.1 ) for final statistical comparisons of compound release data. PIT tagging and marking for field behavioral studies P assive integrated transponder (PIT) tag ging procedures for MM and MF followed procedures previously described ( Johnson et al., 2009 ) . To prevent the expulsion of gametes from a surgical incision, each PIT tag was fitted into a latex sleeve and attached to the mid - dorsal region of each animal using a suture on both sides (Size 3 - 0, Ethicon Inc., Cornelia, Georgia, USA). MM and MF were also fitted with unique color combinations of ribbon tags (Hallprint, Hindmarsh Valley, South AU) through each dorsal fin to identify individuals for visual observations and tracking during mating trials. These procedures typically took less than 30 seconds /animal . PIT tagged animals were immediately transferred into aerated holding tanks with a constant flow of Lake Huron w ater for up to 24 hours, until they were stocked into stream 116 acclimation cages. Tagged individuals were monitored throughout the day for signs of distress or mortality. Electro - olfactogram (EOG) responsiveness to PAMS - 24 Electro - olfactogram recording s were obtain ed from lamprey in summer 2013. Our procedures for EOG a re detailed in previous works by Li et al. ( 2 013 b; 2012 ) . Briefly, sea lamprey were anesthetized and secured in a partially inundated trough, whereby gills could be continuously irrigated with water. The olfactory rosette was surgically exposed. Respons es to stimuli were recorded by borosilicate electrodes placed between olfactory lamellae (signal electrode and external skin - reference electrode). Olfactory responses were filtered and amplified by a NeuroLog filter and pre - amplifier (Digitimer Ltd., Hertf ordshire , England ) , integrated by a digidata system ( Axon Instruments, Inc.), and stored on a PC running Axoscope software (Axon Instruments, Inc.). Stimuli were serially diluted from a 1x10 - 3 M stock solution of isolated and synthesized PAMS - 24, PADS, and PSDS, respectively. Responses were blank subtracted and normalized to those of L - argininge at 1x10 - 5 M. Synthesized odorant treatments The well documented mating pheromone 3kPZS was custom synthesized by Bridge Organics Co. (Vicksburg, Michigan, U.S.A.; purity >97% ) in 2007 . A 10 mg ml - 1 stock solution of synthesized 3kPZS (in 100% methanol) was prepared and transferred into five vials of 10 - ml aliquots. 3kPZS stock solution was stored at - 2 0°C until use. PAMS - 24 was chemically synthesized by Bridge Organ ics Co (Vicksburg, Michigan, U.S.A.; purity >95%) in 2012. The synthetic compound exhibits the same spectral characteristics and biological activity as the natural compound. These data confirm that the observed olfactory response activity was due to the pu rified steroid derivative, not to a trace compound co - purified with the natural PAMS - 24. 117 Experimental designs for preliminary behavioral tests using telemetry Details of e xperimental site s for behavioral field tests are described in Supplementary Figure S4 - 1 , and Figure 4 - 2. The upper reaches of each field site were historically accessible to wild sea lamprey. These sites are now physically blocked by a low - head dam (sea lamprey barrier) near the mouth of each respective river. MFs were tested in a separa te river system from MMs to insure that no unwanted reproduction would occur upstream of sea lamprey barriers by mixing sexes. For MFs, t wo 1 m 2 nest antennas were placed adjacent to one another on the upstream end of the site, flat on the stream bed, and 1.5 m apart . These antennas monitored the proportion , two aluminum - mesh release cages (0.25 m 3 ) equipped with sliding release doors were positioned in the center o f the stream channel. Two PIT antennas roughly 0.5 m - high x 6 m - long were positioned approximately 5 m upstream and 5 m downstream of the release cages to monitor individuals that exit the cage and move upstream or downstream, respectively (Supplementary F igure S4 - 1a). For MMs, t wo 0.5 m 2 nest antennas were placed adjacent to one another on the upstream end of the site, flat on the stream bed, and 1 m apart . These antennas monitored the proportion of ain treatments. Downstream 18.5 m , two aluminum - mesh release cages (0.25 m 3 ) equipped with sliding release doors were positioned in the center of the stream channel. Two copper wire PIT antennas roughly 0.5 m - high x 3 m - long were positioned approximately 0.5 m upstream and 0.5 m downstream of the release cages to monitor individuals that exit the cage and move upstream or downstream, respectively (Supplementary Figure S4 - 1b). 118 For both sexes, t rials were 2 hours long. Stream te mperatures were recorded at the start of and end of each trial. Stream discharge was estimated every three days, or after every precipitation event, at a fixed location in the stream using a Marsh - McBirney portable flow meter (Flo - Mate 2000, Fredrick, Mary land, U.S.A.) to determine the amount of treatment stock solution to apply to the stream and maintain consistent concentrations across trials. Test treatments were diluted with 20 l of river water in large mixing bins on shore . Bins were kept consistent fo r each test treatment, and rinsed in the stream several times before each new trial, to reduce the potential for contamination during mixing . Each treatment solution was then pumped from bins into the stream at the center of each nest antenna so that tes t subject could swim within 0 0.5 m of the source upon entering the nests ( i.e. close proximity to the source) at a rate of 167 mL/min (± 5 mL/min ) over the span of 2 hours using p eristaltic pumps (Cole - Parmer). Peristaltic pumps, powered with a gas gene rator, allowed concentrations of compounds in the stream to remain consistent throughout each trial. In th e first half - hour of each trial the test treatments were administered to the stream while test subjects remained in the release cage. At the start of the following 1.5 hours test subjects were released , and their movements were monitored with PIT antennas until the trial ended . No animals were recovered from the stream after a trial. Copper wire was wrapped around each antenna frame twice during the co nstruction of PIT antennas for a more focused read range. Antennas were wired to a multiplexor in the field for consolidation of data (Oregon RFID, Portland, OR, USA). Antennas were tuned to a detection sensitivity of roughly 0.3 m from the frame edges. Scan frequencies of each antenna were programmed to three scans/sec. Data for each trial were uploaded each day using a hand - 119 held Meazura model MEZ1000 personal digital assistant (Aceeca International Limited, Christchurch, New Zealand). Details of treatments for behavioral tests using telemetry Test treatments for MF and MM trials are described in Figure 4 - 2. For both sexes, trials were ran from 25 June 01 August 2013, and from 27 June 10 August 2014, which is inside the range of the natural spa wning season for sea lamprey ( Applegate, 1950 ) . Treatments applied were alternated back - and - forth for each trial. Up to two trials were conducted each day depending upon the availability of mature animals . The early trial was conducted from ~0700h 0900 h, and a late trial was then run from ~ 0930h 1130 h. Ten PIT - tagged mature subjects were transferred to respective acclimation/release cages for each trial betw een 2000 2200 h the night prior to experimentation. Subjects were then allowed an acclimation period in the stream for 9 + hours. Statistical analysis of PIT data ( shown in Supplementary Table S4 - 2) Each test s - replication from test subjects released during previous trials for all sites . Three main binary response variables were examined for telemetry experiments, including: ( 1) the distribution o f subjects that swam downstream from the release cage and did not come back during trials ( Down: 0 = did not hit on downstream antenna , 1 = hit on downstream antenna ) (2) the distribution of subjects that swam upstream from release cages and did not move back down ( Up: 0 = did not hit on upstream release antenna , 1 = hit on upstream release antenna ), (3 ) of those animals that hit on the upstream antenna , the distribution that entered the nest containing the test treatment ( Treatment nest: 0 = missed treatm ent nest , 1 = enter ed treatment nest ) . Since 50% methanol (vehicle) was administered to both nests during negative control trials, one nest 120 was randomly assigned f or statistical purposes by flipping a coin . The as randomly chosen to be the right nest, and alternated every trial to follow the same pattern of other actual odorant treatments . Statistical analyses for all behavioral tests using PIT equipment at both sites followed Li et al. ( 2013 b) . Briefly, logistic regression with a binomial distribution was examined for each response variable, followed by a post - hoc t - test for comparisons across treatments within each response variable, using R - software (R version 2.11.1, Vienna, Austria). No sig ns of nonlinearities or over dispersion were observed in the models. All behavioral statistics reported are two - Swim track and plume mapping Swim tracks were mapped from random individuals during trials examining mature sea lamp rey following Johnson et al. ( Johnson et al., 2009 ) . Briefly, stream sections seen in Supplementary Figure S4 - 1 were fitted with transecting strings, every 1 m downstream of transecting string (stream width) was divided into tenths of the total width and labelled. Since each test subject was marked with a unique color combination of ribbon tags , we were able to visually observe and follow individuals as they swam upstream. The stream segments were mapped following Johnson et al. ( Johnson et al., 2009 ) . Briefly, swim tracks of moving subjects were recorded by hand. We followed the subject until it reached a nest. Only subjects that were observed exiting the release cage wer e followed. The rest of the subjects that exited unseen by observers were recorded by PIT telemetry. To map the odor plume and estimate in stream concentrations of treatments, rhodamine dye was administered to the stream and samples were taken following Jo hnson et al. ( Johnson et 121 al., 2009 ) with slight modification. Rhodamine concentrations were detected and recorded at each sample point with a hand - held DataBank datalogger and Cyclops - 7 Optical Rhodamine Dye Tracer (Turner Designs, Sunnyvale, CA, USA). Statistical analysis of swim tracks and plume mapping All swim tracks that were mapped during trials were traced onto a digital field site map using a tablet computer (Lenovo X201 Tablet). Each track was overlain onto an odor plume map. To better examine the behaviors of MM as they approached treatments in side - by - sid e nests, we evaluated each swim track within 5 meters of the treatments sources ( see Figure 4 - 2). Since we the number of sharp turns made by each individual (> 90°) within the 5 m range. Once homogeneity of variance was observed, the mean number of sharp turns within 5 m of each source across treatments were evaluated with ANOVA and post - hoc t - test (R - Software ® for Windows, R Foundation for Statisti cal Computing , Vienna, Austria). Next we compared tracks that enter each treatment compared to those that avoided (Figure 4 - 3). Avoidance (or alternatively enter) responses were evaluated using logistic re gression and a binomial distribution, similarly to PIT data (0 = avoid, 1 = enter, = 0.05). Finally based on our nest observations of larger males often beating smaller males in fights, we aimed to test whether larger males choose to enter natural male odors or treatments of 3kPZS:PAMS - 24 at natural mixtures over sma ller males ( i.e. larger males are more likely to win a contest). To do this, we compared the mean sizes of males that entered respective treatments ( see Figure 4 - 4) to the mean size of males that avoided using a one - way t - test ( = 0.05). All data presented were examined for violation of assumptions of normality and homogeneity across variance before further statistical analyses were conducted . Data tha t were not normally distributed or showe d 122 heterogeneity across variance were log - transformed. The variance was used to examine variance of untransformed or newly transformed data. 123 ACKNOWLEDGEMENTS We would like to extend a special thank you to all personnel at the United States Geological Survey Hammond Bay Biological Station, Millersburg, Michigan, the United States Fish and Wildlife Service Marquette Biological Station, Marquette, Michigan, and Fisheries and Oceans, Canada, for providing sea l amprey . We also thank Dolly Trump and Lydia Lorenz , as well as Andy Meyer and family, for the use of their private land to access the field site s . Thanks are well deserved of all technicians that participated in field work, Ethan Buchinger, Elizabeth Racey , . This work was supported by the Great Lakes Fishery Commission, Ann Arbor, MI, USA (to Dr. W eiming Li ). 124 SUPPLEMENTAL INFORMATION SUPPLEMENTAL TABLES AND FIGURES 125 Table S4 - 1. 1 H (900 MHz, J in Hz) and 13 C NMR (225 MHz) spectroscopic data for PAMS - 24 in DMSO - d 6 Table S4 - 1 is curtesy of Dr. Ke Li, Michigan State University. The symbol * indicates an overlap with H 2 O in DMSO - d 6 , symbol indicates an overlap with DMSO - d 6 , and a indicates a ssignments are interchangeable . 126 Table S4 - 2 Numbers of sexually mature female ( a. MF) and mature male ( b. MM) sea lamprey that approached conspecific odorants during field tests, as recorded by telemetry. a. MF (Figure S4 - 1a) Treatment Nest Control Nest Trials Released Down Up Enter treatment nest Enter control nest Vehicle Vehicle 7 71 15% (11) A 15% (11) B 0% (0) 0% (0) 3kPZS Vehicle 5 42 12% (5) A 50% (21) A 100% (13) B 0% (0) MMW Vehicle 5 44 11% (5) A 50% (22) A 100% (17) B 0% (0) 3kPZS 3kPZS 14 127 10% (13) A 45% (57) A 52% (24) A 48% (22) 1:1 3kPZS 8 71 17% (12) A 31% (22) A 47 % (8) A 53% (9) 100:1 3kPZS 6 55 0% (0) A 31% (17) A 78% (7) B 22% (2) X 2 33.37 30.46 80.16 df 5 5 5 P - value < 0.001 < 0.001 < 0.001 b. MM ( Figure S4 - 1b) Treatment Nest Control Nest Trials Released Down Up Enter treatment nest Enter control nest Vehicle Vehicle 7 59 24% (14) A 32% (19) A 60% (3) A 40% (2) 3kPZS Vehicle 8 73 16% (12) A 27% (20) A 79% (11) AB 21% (3) MMW Vehicle 12 128 20% (25) A 21% (27) A 80% (8) B 20% (2) 3kPZS 3kPZS 9 90 12% (11) A 19% (17) A 43% (3) A 57% (4) 1:1 3kPZS 12 127 24% (31) A 13% (17) A 63% (5) A 38% (3) 100:1 3kPZS 8 96 14% (13) A 36% (35) B 57% (13) A 43% (10) X 2 6.79 29.67 14.06 df 5 5 5 P - value 0.237 < 0.001 0.050 These responses were recorded with passive integrated transponder telemetry. Treatments are described in Figure 4 - 2. Response variables were monitored with telemetry and include: Down - the percentage of test subjects that moved downstream 5 m or more from release cages, Up - the percentage of subjects that moved upstream at least 5 m or more from the release cage, Enter treatment nest - the percentage of subjects that entered each respective treatment nest, Enter control nest - the percentage of animals that entered the adjacent control nest. Each response variable was evaluated with logistic re gression. Responses that share a letter are not significantly different ( = 0.05). 127 Figure S4 - 1. In - stream behavioral field sites for observing behaviors of sea lamprey to pheromone components. ( A ) The 45 m - long section of the Upper Ocqueoc River, Millersburg, MI, USA, used for observing movement patterns of sexually mature female sea lamprey to male - released pheromones. ( B ) The 18.5 m - long section of the Upper Trout River, Rogers City, MI, USA, used for observing movement patterns of sexually mature male sea lamprey in relation to mature male pheromones. At the upstream end of each site, odorants were applied into the cente r of a square passive integrated transponder (PIT) antennas (hollow boxes, 1 m 2 and 1.5 m apart in A, 0.5 m 2 and 1 m apart in B ) . Transecting PIT antennas were placed within each site (grey rectangles) to observe the proportion of subjects moving out of re lease cages (solid boxes), upstream, and hitting on a nest PIT antenna. 128 Figure S4 - 2. In - stream swim tracks of mature sea lamprey to treatments of mature male conditioned wash - water (MMW). Treatments are maintained at consistent stream concentrations based on an in - stream concentration of 5E - 13 M benchmark 3kPZS. Mature female ( MF ) subjects were tested in Figure S1A, where MMW was applied to the right ( a. ) and left ( b. ) nests respectively. Mature male (MM) subjects were tested in Figure S1B, where MMW was also applied to the left ( a .) and right ( b. ) nests respectively. Swim tracks were mapped by manually tracking and recording the path of each subject. Transecting strings (dashed lines) were strung every 1 m downstream of the source to aid in swim track mapping. Plumes (outlined in grey) were mapped using rhodamine dye following Johnson et al. ( Johnson et al., 2009 ) . Scale bars = 1 m. 129 Figure S4 - 3. Wild nesting sea lamprey. ( A ) A single male maintains a nest in the lower Cheboygan River, below the Cheboygan Dam, Cheboygan, MI, USA. ( B ) One male accompanied by 7 females in a nest in the Lower Ocqueoc River, Millersburg, MI, USA. 130 Figure S4 - 4. Observations of natural sea lamprey nests . Observations suggest that larger males (MM) are often ac companied with greater numbers of mature females (MF) per nest ( a : Linear regression: F 1,57 = 3.67, P = 0.060), while mature males rarely join other mature males in a nest ( b : Linear regression: F 1,57 = 11.95, P = < 0.001). 131 Figure S4 - 5. Comparison of carbon resonances of PAMS - 24 with squalamine ( Wehrli et al., 199 3 ) and PADS ( Hoye et al., 2007a ) . 13 C NMR data of PAMS - 24 and squalamine were acquired in DMSO - d 6 and displayed in black and blue, respectively. 13 C NMR data of PADS were acquired in methanol - d 4 and displayed in green. 132 Figure S4 - 6. Concentrations of washings and release rates of PAMS - 24 compared to 3kPZS. ( a. ) M eans +/ - 1 SEM of PAMS - 24 concentrations in washings for larval sea lamprey, immature males (IMs, n = 8), mature males head region only (MM - H) and same mature males tail region only (MM - T, n = 7 ), and ovulated females (MF, n = 15). * t 14 = 2.14, P <0 .001. ( b. ) Mean +/ - 1 SEM of PAMS - 24 concentrations in extracted (SPE) and concentrated washing from mature males head region only (MM E - H ) compared to the same mature males tail region only (MM E - T, n > 20 ) . ( c . ) Release rates of PAMS - 24 and 3kPZS by weight (g) of 6 mature males sampled from a natural spawning stream . The natural release ratio of the two is roughly 1:0.01, 3kPZS:PAMS - 24. PAMS - 24 (Linear: R² = 0.3305) and 3kPZS (Linear: R² = 0.2731) regression lines are shown. 133 Figure S4 - 7. Qualitative differences in a) 3kPZS and PAMS and b) PAMS, PADS and PSDS assessed by cross - adaptation in immature male lamprey. Data are expressed as a percentage of the unadapted response. (SAC), self - adapted control; odorant 1 v. odorant 2, odorant 1 against an adapting solution of odorant 2. Values are means ± S.E. ( n = 6). 134 CONCLUSION TO DISSERTATION 135 Studies presented in this dissertation provide several lines of evidence that directly support the overall hypothesis that pheromone modulation of sea lamprey behavior is dependent upon multiple environmental, physiological, and social factors . The multiple contexts presented by these studies in which sea lamprey respond to pheromones will contribute to our understanding of how pheromones modulate the life history of other vertebrates. Field tests in Chapters 1 and 2 suggested that females became ritualized to 3kPZS as a navigational cue that can overr ide the locomotor - inhibiting effects of cold stream temperatures, induce upstream movement in migrating adults, and transition from a migratory cue to a mating signal as females approach ovulation. Given that the raw odor of larvae or spermiated males cons istently induced greater behavioral activity towards the source during positive control trials throughout these studies, I predict additional compounds and mixtures of optimal ratios, along with additional environmental, physiological, chemical, and social factors that mediate these behavioral responses to pheromones, will continue to be identified with the field assays presented here. Understanding the complexities of pheromone communication in sea lamprey may offer ample opportunities to learn about how signals evolve ( Symonds and Elgar, 2008 ) . The diversity of compounds that are only beginning to surface, coupled with variation in release rates among components, variations in olfactory sensitivity to these components, and the multiple contexts ( i.e. environmental, physiological, chemical and social) that modulat e behavioral responses to these compounds present an opportunity for a unique signal design. Pheromones emitted into water form a complex and conspicuous odor plume comprised of multiple components and stereoisomers that add layer upon layer of information to the signal. The complexity of a signal such as one provided by pheromones is not easily comparable to research of more readily accessible signals within the lines of human perception ( i.e. sound and visual cues), making 136 pheromones unique when thinking of how signals evolve. Additional compounds relea sed from mature males that, when combined with 3kPZS, were preferred by ovulated females in field tests may have been selected to become conspicuous to allow females to gauge the distance and sexual status o f a potential mate. Similarly, male responses to component PAMS - 24 may have been selected for as a means to govern the resources held by a nesting male as a territorial pheromone. Field tests used in Chapters 1, 2, 3, 4 and A - 1 allowed me to observe int eractions between environmental and social factors and pheromone mediated behaviors in sea lamprey, which may not have been possible in laboratory contexts ( Johnson and Li, 2010 ) . Indeed, behaviors of fishes to pheromones in laboratory contexts often remain inconsistent with similar studies conducted in the field ( Johnson and Li, 2010 ; Meckley et al. 2012) . The bioassay guided fractionation system developed and implemented in Chapters 3, 4, and A - 1 is effective at identifying new pheromone components in sea lamprey ( Li et al., 2015 ; 2013a ; 2013 b ; 2014 ; 2012 ; Li et al., 2002 ) , and may be useful for identifying compounds in other aquatic species that are hypothesized to use pheromones ( Breithaupt and Thiel, 2010 ; Buchinger et al., 2014 ; Døving et al., 1980 ; Liley, 1982 ) . 137 CONTROL IMPLICATIONS The multiple environmental, physiological, and social factors that influence pheromone mediated behaviors in sea lamprey must be understood in order to integrate pheromones into control techniques of invasive sea lamprey in the Laurentian Great Lakes ( Johnson et al., 2013 ; Li et al., 2007 ) . Currently, removal of adult sea lamprey from tributaries of the Great Lakes is conducted mainly with barrier integrated traps (low - head barrier dams consisting of traps placed at the dam - face, typically in the corners). Control of juvenile larvae is primarily conducted by integrating barriers ( to prevent re - introduction to upper tributaries ) with lampricide treatments . The lampricide consists of two main compounds: TFM (3 - trifluoromethyl - 4' - nitrophenol), and Bayluscide (2', 5 - dichloro - 4' - nitrosalicylanilide) which selectively kill larval lamprey ( Christie and Goddard, 2003 ; Smith and Tibbles, 1980 ) . The Great Lakes Fishery Commission (GLFC), established by treaty between the U.S. and Canada in 1954, contracted the U.S. Fish and Wildlife Service (USFWS), Fisheries and Oceans Canada (DFO), and the U.S. Geological Survey (USGS) as agents t o operate these control techniques. Since that time, the current population in the upper three Great Lakes (Michigan, Huron, Superior) has been reduced to 10% of the historical abundance during the peak of the infestation ( Commission, 2011 ) . However, relaxing control efforts for even a short time period (1 year) results in an immediate and disastrous bounce - back of sea lamprey populations ( Christie and Goddard, 2003 ) . Control of sea lamprey in the Great Lakes is, and will continue to be, an ongoing effort to protect the estimated $7 billion/year f ishery. Taken together, the data presented in this dissertation characterize new pheromone components and factors that influence the behavioral activity of these compounds in 138 the field, which in turn may be integrated into current control techniques and co ntribute to the reduction of the remaining estimated 350,000 adult sea lamprey in the Great Lakes. Management - scale research is currently underway to examine whether pheromones can be integrated into the control program. Recent field tests using 3kPZS as a trap lure for invasive sea lamprey showed the application of 3kPZS (10 - 12 M) into barrier integrated traps alternated among eight Great Lakes tributaries over a 3 year period increased trapping efficiency (averaging a 10% increase across years) compar ed to unbaited traps ( Johnson et al., 2013 ) . However, the capture rates of 3kPZS baited traps remained variable, often resulting in minimal differences in capture rates, if any, compared unbaited traps in certain systems at certain times ( Johnson et al., 2013 ) . These inconsistencies were hypothesized to be caused by an additional suite of uncontrollable factors that exist in management - scale contexts which influenc e pheromone mediated behavior in sea lamprey ( Johnson and Li, 2010 ) . M any fishes including sea lamprey re ly on a hierarchy of environmental, physical, and social information to make behavioral decisions for the ultimate goal of synchronizing reproduction ( Binder and McDonald, 2008 ; Binder et al., 2010 ; Brant et al., 2015 ; Døving et al., 1980 ; Johnson and Li, 2010 ; Kemp et al., 2011 ; Moore et al., 2012 ; Quinn and Adams, 1996 ) , and this dissertation targets these factors. Bot h Chapter 1 and 2 of this dissertation unveil new factors that influence behavioral responses of sea lamprey to 3kPZS, each of which are important to consider in management - scale scenarios. In Chapter 1, the overriding effect of 3kPZS increased upstream mo vement of migrating sea lamprey during otherwise less active conditio ns (cold stream temperatures, < 15 °C). Given these results, upstream application of 3kPZS during colder stream temperatures in the springtime may increase the proportion of upstream migrating sea lamprey in a specific tributary ( Brant et al., 2015 ) , and therefore increase trap encounter rates a respective barrier integrated 139 traps. Over time, repeated 3kPZS treatments in specific tributaries ma y re - distribute and consolidate larval populations for targeted lampricide treatments that are more effective and economical. In Chapter 2, field tests showed female sea lamprey began to shift their preference for 3kPZS from a general upstream response to a more proximal directional response as they approach ovulation. 3kPZS baited traps that are not placed at the base of a barrier, but rather remain distributed throughout a migratory stretch of stream, may be effective at catching female sea lamprey at thi s time ( i.e. during later June migration, or when females are less than 3 days from ovulation). A 3kPZS - baited trap that is placed along a migratory route may falsely indicate a 3kPZS activated sub - channel to migrating females, and therefore yield increase d trap encounter and entry rates as females begin to direct their orientation towards the 3kPZS source. Pheromone - baited traps placed at barriers during the spawning season must out - compete natural nests, presenting a new obstacle that must be contended with. Areas below barrier - integrated trapping sites often consist of suitable sea lamprey spawning habitat with adequate flow, substrate, and cover ( Manion and Hanson, 1980 ) . 3kPZS alone may not provide enough information to increase 3kPZS - baited trap catches of ovulated females during the mating season ( Johnson et al., 2015a ) because females are intercepted with spawning opportunities before arriving at the traps. Recent management - scale field tests comparing 3kPZS with the whole male pheromone (spermiated male washings; SMW) indicated that SMW baited traps are more effective compared to 3kPZS baited traps primarily in streams with low population density and minimum spawning habitat ( Johnson et al., 2015a ) . However, 3 kPZS or SMW baited traps did not often differ in their catch rates compared to one - another or unbaited traps in heavily infested streams with higher nest densities ( Johnson et al., 2015a ) . These results are not surprising given the results of previous studies. 3kPZS is the main component released from males ( Brant et al., 140 2013 ; Li et al., 2002 ; Siefkes et al., 2003 ) , and remains in high concentrations downstream from spawning grounds ( Xi et al., 2011 ) . Applying 3kPZS alone to barrier integrated traps would be expected to induce upstream movement towards the 3kPZS source ( Johnson et al., 2009 ) , yet become ineffective within proximity of the 3kPZS - activated trap due to background pheromone noise and spawning activity near the bar rier. Traps baited with SMW would also be expected to remain ineffective in increasing trap catches in heavily infested scenarios. The application of SMW into an artificial nest during field studies often induces an arrestant - like behavior in ovulated fema les within the source ( Johnson et al., 2009 ; Li et al., 2013 b ; Siefkes et al., 2005 ) . Therefore ovulated females that are exposed to a SMW - activated trap would be expected to arrest and begin to establish nests around the immediate trapping area leaving fewer females to encounter the trap mouth. A potential solution to this problem is to apply a mixture of components released from mature males that provides enough information to a female to induce a search behavior for the source ( i.e . increase trap encounter and entry rates), while remaining incomplete enough from the natural mixture of SMW as to not induce an arrestant or nesting behavior in females. In Chapter 3, the new component DkPES is described as a minor proximity pheromone that, when combined with 3kPZS, increased the entry of mature females into the nest compared to a nest with 3kPZS alone. DkPES is one of several additional minor components that may complete the chemical message of the male pher omone enough to induce a directional preference for a trap mouth, yet still lack additional components/mixtures/stimuli that induce arrestment on a nest. PAMS - 24, and additional minor component described in Chapter 4, also increased female preference into a nest when combined with 3kPZS at the natural ratio compared to nest containing 3kPZS alone. Based on data presented in Chapter 3 and 4 and results from previous 141 management - scale field tests ( Johnson et al., 2015a ; Johnson et al., 2013 ) , partially complete mixtures consisting of main component 3kPZS, and minor components DkPES and PAMS - 24 at consistent ratios may remain effective as a trap lure compared to traps baited with 3kPZS alone or whole ma le odor (SMW) in heavily infested streams. Finally, field data collected with mature males in Chapter 4 suggest that PAMS - 24 may function as a partial territorial pheromone. The addition of PAMS - 24 alone at high concentrations may deter males from entering areas that are not ideal for trapping or l ampricide treatments. Traps that are activated with a partial mating pheromone consisting of 3kPZS and minor components such a DkPES and PAMS - 24 at natural ratios may deter male intrusion into certain traps, while increasing female entry. Targeting females in the population will remove substantial reproductive potential from the population, as each female can produce over 100,000 eggs ( Applegate, 1950 ) . Understanding the multiple factors in the field that influence the way sea lamprey respond behaviorally to pheromones brings us closer to understanding how pheromones can be optimized for use in controlling sea lamprey in the Great Lakes, which will u ndoubtedly aid in removal of invasive sea lamprey and further protect the fishery ( Commission, 2011 ) . 142 APPENDI X 143 APPENDIX CHAPTER A - 1: FATTY ACIDS WITH STEREOCHEMISTRY FUNCTION AS A PHEROMONE IN SEA LAMPREY 144 ABSTRACT We report the identification, olfactory sensitivity, and preliminary behavioral activity of a tetrahydrofuran diol fatty acid with stereochemistry in sea lamprey using a novel bioassay guided fractionation system. S tereoisomers 9,(12) - oxy - 10,13 - dihydroxystearic acid ( 1a ) and 10,(13) - oxy - 9,12 - dihydroxystearic acid ( 1b ) consisting as a mixture of (+) - 1a (973) , ( ) - 1a (971) , (+) - 1b (974) , and ( ) - 1b (972) were identified from behaviorally active fractions that were extracted from juvenile conspecific odor. Electrophysiological experiments yielded stereoisomers (+) - 1a (973) , ( ) - 1a (971) to be detected by adult sea lamprey ( threshold of detection of 10 - 11 m olar [M]) while likely sharing an olfactory receptor . Stereoisomers (+) - 1b (974) , and ( ) - 1b (972) were not detected . Preliminary field tests in a river that forms a natural Y - maze design showed pre - spawn migrating female sea lamprey to preferred to move i nto a branch of a stream activated with 973 ( 5x10 - 13 M), yet spawning phase (ovulated) females preferred to enter a mixture of 971:973 (1:1, 5x10 - 13 M:5x10 - 13 M). Field tests suggest that stereochemistry may influence pheromone perception in sea lamprey, a previously undescribed phenomena in a vertebrate. Further behavioral studies are required to determine the stereo - specificity and behavioral function of thes e stereoisomers. We concluded that the bioassay guided fractionation system presented here is efficient at identifying new pheromones. 145 INTRODUCTION Pheromones, or unique chemical signals released by a species that influence the physiology or behavi or of other members of the same species ( Karlson and Lüsche r, 1959 ) , are hypothesized to function in organisms across the animal kingdom. However identifying a new pheromone and further characterizing its function among a species is rare, and often requires integrated research in analytical and theoretical chem istry, behavioral ecology, electrophysiology, and behavior. P heromones have been implicated to modulate critical life history event s in many freshwater fishes ( Liley, 1982 ; Liley and Stacey, 1983 ; Sorensen, 1992 ; Stacey and Cardwell, 1995 ; Stacey et al., 1996 ) , yet new compounds of these pheromones are rarely identified. To date, the identified fish pheromones are bile acid d erivatives ( Li et al., 2002 ; Sorensen et al., 2005 ) , prostaglandins ( Sorensen et al., 1988 ; Sorensen et al., 1989 ) , and amino acids ( Yambe et al., 2006 ) , which possess diverse structures with a hydrophilic function al group and are relatively stable in aquatic environments. The sea lamprey ( Petromyzon marinus ) is a jawless ver tebrate that relies in part on pheromones to modulate migration and mating ( Bjerselius et al., 2000 ; Li et al., 2002 ) . Unlike salmon which track the odor of natal streams learned early in life, migratory adult sea lamprey are attracted to odorants of conspecific larvae ( Teeter, 1980 ) , supporting the theory that a larval odor influences selection of spawning streams by migratory adult s (a phenomena first recognized when trapping records indicated that migratory adults prefer streams with high densities of larvae ( Moore and Schleen, 1980 ) ) . A tagging study further supported the idea that sea lamprey do not travel home to natal streams for reproduction, implicating an innately discerned attractant ( Bergstedt and Seelye, 1995 ) . Several compounds have been identified from the odor of conspecific larvae, and 146 shown to induce some behavioral activity in the labor atory ( Sorensen et al., 2005 ) , but these behaviors have not been replicated in the field ( Meckley et al., 2012 ) . Therefore, we hypothesized that larval sea lamprey residing in natal streams release a com ponent as migratory pheromone to guide sexually immat ure adults to suitable streams. Here we develop a novel Bioassay - guided fractionation system of extracted larval wash - water, which resulted in the isolation of an active component, consisting of two cons titutional isomers. These compounds , identified here as 9,(12) - oxy - 10,13 - dihydroxystearic acid ( 1a ) and 10,(13) - oxy - 9,12 - dihydroxystearic acid ( 1b ) , were each found to be an enantiomeric mixture , resulting in compounds (+) - 1a (973) , ( ) - 1a (971) , (+) - 1b (974) , and ( ) - 1b (972) . Herein, we report the bioassay - guided isolation, structure elucidation, enantiomeric separation, olfactory activities, and preliminary field study of components and mixtures of these compounds. Our data suggests fatty acids with st ereochemistry function as pheromones in an aquatic vertebrate, a novel molecular template as a chemical signal . 147 RESULTS Field bioassay guided fractionation yielded behaviorally active components Chromatography on silica gel of larval washing s eluted with gradient chloroform and methanol gave nine fractions ( 1 9 ) , which were applied to EOG assays ( Figure A - 1 - 1a) , field studies, and high performance liquid chromatography - mass spectrometry analysis. The raw odor of larval sea lamprey is known to elicit a strong preference response in migrating adult conspecifics ( Bjerselius et al., 2000 ; Wagner et al., 2009 ) . We therefore used extracted larval odor as a positive control in all field tests. Following the fractionation of raw larval odor, we combined fractions into four pools (Figure A - 1 - 1 b ) . We first aimed to examine the behavioral activity, if any, induced by these four pools. Using a 250 m - long section of the Upper Ocqueoc River (Figure A - 1 - 2 ), we discovered that all for pools combined replicated the response seen from the raw larval extra ct in drawing significant numbers of migrating sea lamprey into the activated sub - channel (Table A - 1 - 1) . Additionally, Pool 3 replicated the response seen in all four pools combined (Table A - 1 - 1a). We combined the nine fractions in Pool 3 into three separa te sub - pools (Sup - pool 3.1 = FR24, 25, 26, Sub - pool 3.2 = FR27, 28, 29, 30, and Sub - pool 3.3 = FR31 and 32) for field testing in the 2011 migratory season. Behavioral trials yielded Sub - pool 3.1 as best replicating the response seen from larval extract con trol trials ( Table A - 1 - 1b ). Followed by bioassay guidance, the components in Sub - pool 3.1 were analyzed by mass spectrometry and compared with known compounds. We confirmed the presence of compound 1 ( m/z 329 , Figure A - 1 - 3 ), petr omyzonin ( m/z 308) ( Li et al., 2013 a ) , and petromyroxol ( m/z 273) ( Li et al., 2014 ) in Sub - pool 3.1. 148 Figure A - 1 - 1. Bioassay guided fractionation p inpoint ed active compounds. ( A ) Olfactory responses to larval fractions F1 - F9 measured by EOG . ( B ) Mass spectra of pools 1 to 4 of larval sea lamprey fractions, components of 1 ( m/z 329), petromyzonin ( m/z 308), and petromyro xols ( m/z 273) are occurring in pool 3. 149 Figure A - 1 - 2 . Schematic of the field site in a 250 m - long section of the Upper Ocqueoc River, Millersburg, MI, USA, used for behavioral testing of component 1. The downstream release point is shown, along with the Up and Down passive integrated transponder (PIT) antennas used to monitor subjects moving upstream or downstream, respectively, after release. The upper 45 m of the section is naturally bifurcated by a n island. Proportions of subjects entering each sub - channel was monitored by respective Treatment channel PIT antennas. The proportion of subjects entering the treatment source was then monitored by 1 m 2 Treatment source antennas, where treatments were adm inistered into the stream. 150 Table A - 1 - 1 . Responses of migratory sea lamprey to (a) fraction pools (1 - 4) and (b) sub - pools (3.1 - 3.3) from larval odor. 2010 Treatment Trials Released Up Treatment channel Treatment nest Vehicle 8 157 69% (109) A 47% (52 ) A 15% (8) AD Larval Extract 4 80 75% (60) AB 82% (49) B 51% (25) BC Pools 1 - 4 5 100 58% (58) A 74% (43) BC 63% (27) B Pool 1 4 80 85% (68) B 51% (35) A 9% (3) D Pool 2 4 80 64% (51) A 33% (17) A 12% (2) AD Pool 3 15 300 64% (192) A 67% (129) C 26% (34) A Pool 4 4 79 81% (64) AB 61% (39) AC 31% (12) AC X 2 27.94 45.28 48.54 df 6 6 6 P - value < 0.001 < 0.001 < 0.001 2011 Treatment Trials Released Up Treatment channel Treatment nest Vehicle 5 99 42% (42) A 43% (18) A 28% (5) A Larval Extract 5 100 72% (72) B 89% (64) B 61% (39) B Sub - pool 3.1 6 124 52% (64) A 64% (41) C 24% (10) A Sub - pool 3.2 4 79 56% (44) A 45% (20) A 15% (3) A Sub - pool 3.3 2 40 53% (21) A 33% (7) A 43% (3) A X 2 19.19 44.82 22.98 df 4 4 4 P - value < 0.001 < 0.001 < 0.001 Tri als were conducted over the 2010 and 2011 migratory season s in the Upper Ocqueoc River, Millersburg, MI . Treatments Vehicle (50% MeOH) and Larval Extract ( extracted raw larval odor applied to one sub - channel at a volume achieving 5E - 14 M PADS benchmark and vehicle applied to the adjacent sub - channel ) were controls. In 2010, pools in Figure A - 1 - 1b were tested . In 2011, three sub - pools from active Pool 3 were tested, including ; Sub - pool 3.1 (Fraction 5), Sub - pool 3.2 ( Fraction 6 ), and Sub - pool 3.3 (F raction 7). Responses include Up (percentage moving 200 m upstream to the confluence of the two sub - channels), Treatment channel ( of the subjects that moved up, percentage that enter the sub - channel containing each treatment), and Treatment nest (of the subjects that entered the treatment channel, the percentage that swam through a 1 m 2 nest fixed to the center of the stream bed at the upstream end of the respective sub - channel). Each response across treatments was evaluated with logistic regression . Values that share a letter within each response variable are not significantly different ( = 0.05). 151 Figure A - 1 - 3 . Stereochemical - - 1b (972), (+) - 1a (973), and (+) - 1b (974). 152 Bioassay - guided isolation, identification, and quantitative analysis of active components See S upplemental Information Supplemental Results for details of chemical i solation, identification, and quantitation of the components of compound 1 (Figure A - 1 - 3). The details of these results were prepared by Dr. Ke Li, Michigan State University. - and (+) - 1a elicit olfactory responses See Supplemental Information Supplemental Results for details of olfactory sensitivity to the components of compound 1 (Figure A - 1 - 3). The details of these results were prepared by Dr. Mar Huertas, Michigan State University. - 1a (971) and (+) - 1a (973) show stereo - selective and temporally variable behavioral activity in pre - spawn sea lamprey Synthesized components - 1b ( 972 ) and (+) - 1b ( 974 ) did not yield EOG responses and so were not tested in the field. Components - 1a (971) and (+) - 1a (973) showed stereo - selective behavioral activity that varied between Ear ly and L ate migratory season. Upstream movement of pre - spawn migrating female sea lamprey towards treatments remained high ( 67 - 93%) during field trials across the season (Table A - 1 - 2 ). However, p re - spawn migrating sea lamprey did not show a bias towards c - 1a ( denoted as 971 during field tests ) activated sub - channels or the treatment source at any point , while s ubjects began to show a preferent ial bias towards the sub - channel activated with (+) - 1a ( denoted as 973 during field tests ) applied at 5E - 13 M specifically during Late (June) migration, at which time 70% entered the 973 sub - channel (Logistic regression: X 2 4 = 24.2, P < 0.001). Larval extract controls drew significant numbers of subjects into the treatment channel and treatment source during both Early ( May, Table A - 1 - 2 a) and Late (June, Table A - 1 - 2 b) migration, serving as an effective positive control. 153 Table A - 1 - 2 . Responses of migratory female sea lamprey to new tetrahydrofuran diol compounds 971 and 973 during Early May and Late Jun e 2013 and 2014 migratory season s . Early migration - May Treatment Trials Released Down Up Treatment channel Treatment sourc e Vehicle 19 378 13% (48) A 84% (318) A 49% (156) A 20% (31) A Larval Extract 8 160 5% (8) B 89% (142) A 60% (85) BC 42% (36) B 971+973 8 160 17% (24) AC 71% (114) B 52% (59) AC 17% (10) A 971 4 80 28% (22) C 71% (57) B 37% (21) A 19% (4) A 973 4 80 8% (6) AB 93% (74) A 59% (44) AC 27% (12) AB X 2 27.74 30.61 11.58 17.18 df 4 4 4 4 P - value < 0.001 < 0.001 0.021 0.002 Late migration - June Treatment Trials Released Down Up Treatment channel Treatment source Vehicle 17 331 18% (60) A 69% (228) A 45% (103) A 26% (27) A Larval Extract 7 140 11% (16) B 86% (120) B 63% (76) BC 61% (46) B 971+973 8 160 27% (43) C 67% (107) A 64% (68) BC 13% (9) C 971 8 160 14% (22) A 67% (107) A 54% (58) AB 19% (11) A 973 7 139 12% (17) A 72% (100) A 70% (70) C 27% (19) A X 2 19.99 27.91 24.20 41.48 df 4 4 4 4 P - value 0.001 < 0.001 < 0.001 < 0.001 Trials were conducted over the 2013 and 2014 migratory season in the Upper Ocqueoc River, Millersburg, MI, USA. Treatments included vehicle controls (50% MeOH), Larval Extract controls (Larval extract applied to one sub - channel and vehicle applied to the a djacent sub - channel, 971:973 (1:1, 1E - 12 molar M total ), 971 alone (5E - 13 M), and 973 alone (5E - 13 M). Responses include Down (percentage moving downstream of the release cages and not coming back up during the 2.5 hour - long trial), Up (percentage moving 2 00 m upstream to the confluence of the two sub - channels), Treatment channel (percentage that enter the sub - channel containing each treatment), and Treatment source (of the subjects that entered the treatment channel, the percentage that then passed through a 1 m 2 source antenna fixed to the center of the stream bed at the upstream end of the respective sub - channel). Responses were evaluated with a generalized linear model and binomial distribution. Responses that share a letter are not significantly differe nt ( = 0.05). 154 - 1a (971) and (+) - 1a (973) draw and retain mature females at the source Over 2500 migratory sea lamprey were released into the stream during field trials for Table A - 1 - 2 , each with a uniquely identified PIT tag. Old subjects from past trials that were remaining in the system were also continuously monitored with our PIT antenna system along - 1 a (971) and (+) - 1a (973) , and found them to spent significantly more time on a mixture of the two compounds compared to each compound alone, yet not as much time as was spent of 3kPZS nests (Figure A - 1 - 4 ). A sample of subjects ( n = 4) - 1 a (971 ) and (+) - 1a (973) mixture (1:1) were hand - grabbed from the center of the nest at night for histological confirmation of maturation. All samples where fully ovulated according to published studies ( Yorke and McMillan, 1980 ) . While it was unfortunate that only four specimens could be hand - grabbed in the field during this time , we determined that four subjects had spent an average of 15 ± 1.6 days at large in the river system before returning to the site. 155 Figure A - 1 - 4. Mean ± 1 SEM retention ( sec. ) of female sea lamprey inside artificial nests (within 0.5 m of the source) while pheromone treatments were administered. Tr eatments included 3kPZS at 5x10 - 13 - 1a ( 971 , 5x10 - 13 M) and (+) - 1a ( 973 , 5x10 - 13 M) at a 1:1 ratio (totaling 1x10 - 12 M), 971 alone (5x10 - 13 M), 973 alone (5x10 - 13 M), and vehicle control (methanol). Trials were conducted in the Ocqueoc River, Millersburg, MI (Figure A1 - 2). Columns with different lower - case letters are significantly different, ANOVA and post hoc Tukey HSD: ( F 4, 123 = 15.58, P = < 0.0001). 156 DISCUSSION Our findings collectively reveal fascinating novel molecular diversity of sea lamprey metabolites and illuminate a new molecular template; dyhydroxylated THF fatty acid. Field assays suggest that this compound shows stereo - selective behavioral function in sea lamprey. In our previous investigation, a pair of diastereomers possessing dyhydroxylated THF moiety, petromyroxols ( Li et al., 2015 ) and iso - petromyroxols ( Li et al., 2015 ) , have been characterized from water conditioned with larvae of sea lamprey. Howe ver, their role in modulating behavior of sea lamprey are still ambiguous due to a lack of access to materials required for further investigation of biological activity. From the view of chemical structure and nature, THF - diol fatty acids represent a new pheromone template that is unlike hormones, which often function as known pheromones excreted as reproductive by - products ( Shorey, 2013 ) . The derivatives of THF - diols are a class of typical amphipath ic molecules containing both hydrophobic (THF moeity) and hydrophilic (secondary alcohol and carboxylic acid) functional groups , indicating a water insoluble property and tendency to form a monolayer over an aqueous sub - phase resulting in fast dispersion o n the air and water interface . Therefore, THF - diols fatty acids possess the potential to act as pheromones that can travel over long distances on water . On the other hand, each individual compound has four chiral centers and two variable elongation chain s , providing chemical diversity , possessing 16 possible absolute configurations , and tens of homologs. Combining the diversity of homologs and spatial configurations, THF - diol derivatives can provide hundreds of messenger combinations to deliver complex info rmation. 157 Results of field assays suggest that component 1 shows stereo - selective behavioral activity in sea lamprey, which also appears context - dependent on a temporal scale of their migration. Only stereoisomer (+) - 1a (973) induced preference like responses in migrating female sea lamprey, which only occurred in the late half of the migratory season (June). Since migrating female subjects showed a preference to the whole larval extract throughout the entire migratory season, w e cannot yet conclude w hether the new stereoisomer (+) - 1a (973) is a migratory pheromone, or a chemical cue that allows aggregation around spawning grounds at the onset of the reproductive season. Data suggest that additional components or unique mixtures exist in raw larval ext ract that function as the main migratory pheromone, and these additional compounds are currently being targeted for elucidation by our group. Stereo - selectivity of compounds that function as semiochemicals have been described in insects ( Chapman et al., 1978 ; Coracini et al., 2001 ; Klun et al., 1973 ) , but to our knowledge, has not been described in a vertebrate. Synthesized stereoisomer (+) - 1a (973) alone appears to be attractive in pre - spawn females and a - 1a (971) and (+) - 1a (973) appears to induce behavioral activity in spawning phase females. Exact mixtures of all stereoisomers may be required for maximum attraction. In Lepidoptera, a single geometrical isomer has been shown to be weekly attractive in receivers to the source, while only a re - construction of all geometrical isomers yie lded a maximum attractiveness ( Klun et al., 1973 ) . We may be observing a similar phenomenon in sea lamprey. Unfortunately we did not know the ratio of each isomer during field trials, and so combined mixtures in a 1:1 ratio for field testing. Further investigation is required to determine the exact ratio s of each isomer, and determine whether ratio influences behavioral activity towards these components. 158 METHODS Animals All procedures involving sea lampreys were conducted in conformity with the Public Health Service (PHS) Policy on Humane Care and Use of Laboratory Animals, incorporated in the Institute for Laboratory Animal Research Guide for Care and Use of Laboratory Animals, and were approved by the Michigan State University Institutional Animal Use and Care Committee (Animal use form number: 0 3/11 - 053 - 00). Larval sea lampreys were captured in tributaries of the Laurentian Great Lakes by the US Fish and Wildlife Service and Fisheries and Oceans Canada according to approved scientific collection permits from those government agencies, transported to the US Geological Survey Hammond Bay Biological Station, Millersburg, Michigan, USA, and held in flow through raceways (1.83 m wide by 0.61 m deep by 15.24 m long). A dult sea lampreys used for olfactory functional studies were collected in December and January by commercial fishing agencies from Lake Huron and transported to the US Geological Survey Hammond Bay Biological Station where they were then delivered to Michigan State University, East Lansing, Michigan, USA. Sea lampreys were held in flow thro ugh tanks (254 L) supplied with well - water chilled to roughly 8 ° C. Electro - olfactogram recording was conducted in February 2013 at Michigan State University . Equipment and Materials 1D and 2D NMR spectra of petromyroxol were recorded on an Agilent 500 MHz spectrometer. Mass spectra were performed on a TQ - S TOF LC mass sp ectrometer (Waters Corporation, Milford, Massachusetts, USA). Silica gel (70 - 230 and 230 - 400 mesh, Merck, Darmstadt, Germany), RP - 18 reverse - phase silica gel (Merck), and Sephadex LH - 20 (Me rck) 159 were used for open column chromatography. TLC was conducted on glass plat es pre - coated with GF254 silica gel (Merck). Spots were first visualized under UV light at 254 nm and then stained by spraying an acidic methanol solution of 5% anisaldehyde (Sigma - Aldrich, St. Louis, Missouri, USA). Extraction of larval sea lamprey - conditioned water Every four to five days over the course of ca. five months, water conditioned with sea lamprey larvae (ca. 4000 L for each cycle) was passed through four beds ea ch containing 1 kg of Amberlite XAD 7HP resin at a rate of roughly 200 mL/min/bed (800 mL/min, total) and subsequently eluted with methanol (4 L). The organic solvents were removed under reduced pressure at 40 ° C by a rotary evaporator. The resulting resid ue was pooled and stored at - 80 ° C until further processing. The total extract ( ca. 42 L and containing a large amount of water) was thawed and concentrated by lyophilization. This residue was suspended in methanol and s m fi lter paper. The filtrate was concentrated again under reduced pressure at 40 o C to yield 2.0 g of a dark residue. Field Study Details of the e xperimental site are consistent with those described by Brant at al. ( Brant et al., 2015 ) , with slight modification. Briefly, a section of the Upper Ocqueoc River in Millersburg, MI, USA, was used to monitor the mov ements of migrating female sea lamprey to pheromone treatments. A two - choice system was constructed using two sub - channels that were naturally bifurcated by an island. Downstream, 205 m from the confluence of the two sub - channels, two release cages were pl aced in the center of the stream. Release cages were constructed of mesh aluminum (0.25 m 3 ) equipped with sliding release doors. Transecting c opper wire passive integrated transponder ( PIT ) antennas were placed 5 m upstream and downstream 160 of release cages, and at the confluence of each sub - channel. Approximately 45 m upstream of each sub - channel, a 1 m 2 PIT antenna frame was laid flat on the stream bed. Treatments were administered into the center of the 1 m 2 frame. Antennas were tuned to a detection sensit ivity of roughly 0.3 m from the frame edges. Scan frequencies of each antenna were programmed to three scans/sec. Data for each trial were uploaded each day using a hand - held Meazura model MEZ1000 personal digital assistant (Aceeca International Limited, Christchurch, New Zealand). PIT tagging P assive integrated transponder (PIT) tag ging procedures followed Johnson et al. ( Johnson et al., 2009 ) . Briefly, a 23 mm - long half duplex PIT tag ( Ore gon RFID, Portland, Oregon, USA ) was surgically implanted into each experimental animal through a 3 4 mm lateral incision in the mid - abdominal region . These procedures typically took less than 30 seconds /animal . PIT tagged animals were immediately transferred into aerated holding tanks with a constant flow of Lake Huron water for up to 24 hours, until the y were stocked into stream release cages. Tagged individuals were monitored throughout the day for signs of distress or mortality. Details of trials In behavioral field trials, our primary objective was to test whether individual compounds that showed EOG activity were also capable of influencing movements of migrating sea lamprey similarly to behaviors elicited from larval extract controls. For Early migration , trails were conducted 8 29 May 2013 and 27 May 09 June 2014. For Late migration , trials were conducted 30 May 23 Ju ne 2013 and 12 21 June 2014. Treatments included: 1) Vehicle control methanol administered into both sub - channels, 2) Larv al Extract , or extracted raw larval odor, into one sub - channel while vehicle control is applied to the adjacent sub - channel, 3) a 1:1 mixture of - 1a (971) and (+) - 1a (973) (5x10 - 13 M:5x10 - 13 M) applied to one sub - channel and 161 vehicle control applied to the adjacent sub - channel, 4) - 1a (971) alone at 5x10 - 13 M into one sub - channel and vehicle into the adjacent sub - channel, and 5) (+) - 1a (973) alone at 5x10 - 13 M into one sub - channel and vehicle into the adjacent sub - channel. Treatment and vehicle sub - channels were alternated each trial. Test treatments were diluted with 2 5 L of river water in large mixing bins on shore . Bins were kept consistent for each test treatment, and rinsed in the stream several times befor e each new trial, to reduce the potential for contamination. Each treatment solution was then pumped from bins into the stream at the center of each 1 m 2 PIT antenna ( Treatment source ) at a rate of 167 mL/min (± 5 mL/min ) over the span of 2 hours using p er istaltic pumps (Cole - Parmer). Stream temperatures were recorded at the start of and end of each trial. Stream discharge was estimated every three days, or after every precipitation event, at a fixed location in the stream using a Marsh - McBirney portable fl ow meter (Flo - Mate 2000, Fredrick, Maryland, U.S.A.) to determine the amount of treatment stock solution to apply to the stream and maintain consistent concentrations across trials. Trials were conducted at night when migrat ing sea lamprey are most active ( Stier and Kynard, 1986 ) . Trials were 2.5 hours long. In th e first half - hour of each trial the test treatments were administered to the stream while test subjects remained in the release cage. At the start of the following 2 hours , test subjects were released and their movements were monitored with PIT antennas until the trial ended . No animals were recovered from the stream after a trial , and unique PIT tag IDs prevented any pseudoreplication from occurring. Up to two trials were conducted each night, depending upon animal availability. The early trial was conducted from sundown (roughly 2030h ) to roughly 2 300 h and the last trial was from roughly 2330h 0 200 h ( exact trial times were dependent upon when sundown occurred) . Twenty PIT - tagged subjects were released per trial. Subjects were removed from holding tanks at HBBS and transported to 162 their respective stream release cage at the release point betw een 0300h and 0500 h the night prior to experimentation. Subjects were then allowed an acclimation period in the stream for 15+ hours. Statistical analysis of PIT data Trials were divided into Early and Late migration for statistical analyses because during our preliminary three years of behavioral testing in the stream we began noticing that behaviors of migrating immature sea lamprey showed some discontinuity as the mating season approached. We theorized that female s ea lamprey adapted to adjust their response threshold to pheromones as the migratory season shifts into the mating season to insure their synchrony of reproduction with males. For this reason, we specifically began examining our response variables to treat ment compounds within the first half of the migratory season ( Early migration - May) versus the latter half ( Late migration - June) to better examine whether any changes in preference to treatments occurs. L ogistic regression with a binomial distribution w as examined for each response variable, followed by a post - hoc t - test for comparisons across treatments within each response variable, using R - software (R version 2.11.1, Vienna, Austria). No signs of nonlinearities or over dispersion were observed in the models. All behavioral statistics reported are two - tailed analyses Four main binary response variables were examined during field trials: ( 1) the distribution of subjects that swam downstream from the release cage and did not come back up past release cages during the trial ( Down: 0 = did not hit on Down antenna , 1 = hit on Down antenna only ) (2) the distribution of subjects that swam upstream from release cages and did not move back down ( Up: 0 = did not hit on Up antenna , 1 = hit on Up antenna and continued upstream ), (3 ) of those subjects that hit on the Up antenna , the distribution that entered the sub - channel containing the test treatment ( Treatment channel : 0 = entered vehicle c hannel , 1 = 163 enter ed treatment channel ) , and (4) of those subjects that entered the treatment channel, the distribution that entered within 0.5 m of the treatment source (Treatment source: 0 = missed treatment source antenna, 1 = entered source). Since met hanol (vehicle) was administered to both nests during Vehicle control trials, one nest was randomly assigned f or statistical purposes alternated every trial to follow the same pattern of other actual odorant treatments . 164 SUPPLEMENTAL RESULTS After successive chromatographic purification, Component 1 was isolated as a colorless oil and displayed a pseudo - molecular - ion at m/z 329 [M H] in the negative Q - TOF - ESI mass spectrum. The molecular formula C 18 H 34 O 5 was determined by HRESIMS at m/z 329.23 32 [M H] (calcd. for C 18 H 33 O 5 , 329.2328 ), indicating two degrees of unsaturation. The doubling signals in 13 C NMR spectrum suggested that component 1 contained two constitutional isomers 1a and 1b . The 1 H NMR spectrum (Supplementary Fig. 3) showed signals for four oxygenated methines and one methyl were observed, as well as resonances for few alphatic methylene protons. The 13 C NMR and HSQC experiments revealed the presence of 18 carbon signals corresponding to four oxygenated methines at C 82.7, 80.2, 74.1, and 73.4, one quaternary carbon at C 178.1, one methyl at C 14.0, and remaining 12 alphatic methylenes. The presence of tetrahydrofuran ring was indicated by 1 H 1 H COSY and HMBC correlations (Supplementary Fig. 6 - 8) ( Li et al., 2014 ) . The 1 H NMR, 13 C NMR chemical shifts and 1 H 1 H COSY correlations allowed us to assign two hydroxyl groups. The remaining carbon signals were attributable to a carboxyl group and two aliphati c chains. To anal yze the chain length of component 1 , a GC - MS analysis was apply to the trimethylsilyl ether derivative. A peak with retention time of 20.6 was observed in the mass chromatogram. The ESI mass spectrum for the peak contain ed prominent fr agme nt ions at m/z 173 and 317. Component 1 could be identified as 9,(12) - oxy - 10,13 - dihydroxystearic acid ( 1a ) and 10,(13) - oxy - 9,12 - dihydroxystearic acid ( 1b ) , which, after comparing 1 H, 13 C - NMR and GC - MS data , was consistent with the literature ( Li et al., 2014 ; Markaverich et al., 2002 ) . 165 The relative configuration of component 1 was established through analysis of the NOESY correlations and coupling constants as well as comparison of NMR data with literature values ( Capon et al., 1998 ; Li et al., 2014 ; Warren et al., 1980 ) . NOESY correlations between H - 9/H - 10, H - 9/H - 11a, H - 10/H - 11a, H - 13/H - 11a, and H - 12/H - 11b placed protons H - 9, H - 10, H - 11a, and H - 13 on the same face of the tetrahydrofuran ring, and H - 12 and H - 11b on the another face of the tetrahydrofuran ring. Meanwhile, component 1 displayed a high degree of similarity with the analogue posses sing H - 9/H - 12 trans relationship and apparent differences with H - 9/H - 12 cis analogue by comparison of the chemical shifts for oxygen - bearing methylen es in 1 H and 13 C NMR spectra ( Capon et al., 1998 ; Warren et al., 1980 ) , which was in good agreement with the deduction by NOESY correlations. This conclusion was confirmed by the performance of a single - crystal X - ray diffraction of (+) - 1a . Therefore, the relative configurations of compounds 1a and 1b can be assigned as 9 R* , 10 R* , 12 S* , 13 S* - 1a and 9 S* , 10 S* , 12 R* , 13 R* - 1b . To determine the absolute configurations of compounds 1a and 1b , authentic standards of enantiomers were prepared using a highl y stereoselective synthesis. The compounds (+) - 1a and ( ) - 1a showed identical mass . GC - MS spectra and 1 H and 13 C NMR spectra, and compounds (+) - 1b and ( ) - 1b are consistent . The MS, NMR, and GC - MS spectra of natural occurring component 1 are consistent with those of a mixture of (+) - 1a , ( ) - 1a , (+) - 1b , and ( ) - 1b . The absolute configurations of the stereogenic centers of these analogues were e stablished using Mosher ester methodology ( Hoye et al., 2007b ; Ohtani et al., 1991 ) . Treatment of ( ) - 1a with ( R ) - ( ) - - methoxy - - (trifluoromethyl)phenylacety l chloride [( R ) - MTPA - Cl] in pyr idine yielded a mixture of 6,9 - bis - ( S ) - Mosher ester and 9 - mono - ( S ) - Mosher esters o f ( ) - 1a ) - 1b with ( S ) - (+) - MTPA - Cl afforded an analogous mixture of the bis - and mono - ( R ) - Mosher esters. Each of these mixtures was separated by semi - preparative HPLC to obtain the mono - ( S ) - and the mono - ( R ) - Mosher ester, respectively. 166 The ratios among these analogues in samples of extracted larval wash - water were estimated using chiral UPLC - MS/MS. A characteristic fragment [M + H 2H 2 O] + ( m/z 295.0) was chosen as the daughter ion to increase the specificity (Supplementary Table 2). The stereoisomers were separated in 35 min using a polysaccharide based column (Chiralpak AD - particles and an isocratic elution of ethanol: n - hexa ne: f ormic acid (85:15:0.1, v/v). The stereoisomers (+) - 1b , (+) - 1a, ( ) - 1b - 1a appeared at retention time of 17.0 - 18.0, 18.0 - 19.5, 30.5 - 31.5, and 31.5 - 33.0 min, respectively. Once separated, the enantiomers were subjected to APCI (positive ion mode) and collision - induced dissociation tandem mass spectrometry (CID - MS/MS) using the MRM mode. The quantitation method of stereoisomers was validated based on the FDA guidelines . The percentage of compounds (+) - 1b , (+) - 1a, ( ) - 1b - 1a in natural occurring component 1 was estimated as 14.4%, 18.8%, 28.6%, and 38.2%, respectively. The concentration of (+) - 1b , (+) - 1a, ( ) - 1b - 1a has been evaluated from larval sea lamprey wash water and shows a relevant concentration to function as a pheromone. - and (+) - 1a elicit olfactory responses - 1a and (+) - 1a showed strong olfactory responses during EOG analyses, with a threshold of detection of 10 - 11 - 1 b and (+) - 1 b - 1 b and (+) - 1 b , if any, were assumed to be non - specific. The detection threshold for petromyzonin was 10 - 11 M (n = 6) ( Li et al., 2013 a ) . This threshold was only one order of magnitude higher than the threshold (10 - 12 M) for 3k PZS, a known lamprey pheromone ( Li et al., 2002 ) . 167 Responses to b inary mixtures of stereoisomers at 10 - 7 M were not higher than the responses of single stereoisomers, indicati ng a lack of synergism in the mixtures. However mixtures of either - 1b and (+) - 1b stereoisomers with (+) - 1 a show lower responses than individual tests of (+) - 1 a , indicating a possible antagonist effect. Cross adaptation studies indicated that EOG - 1 a and (+) - 1 a were not significantly different to their self - adapted control (Supplementary Fig. 11), suggesting that olfactory responses for this two compounds are likely mediated by the same olfactory receptor. Cross adaptation a - 1 b and (+) - 1 b were inviable since they did not elicit olfactory responses to allow comparison between stereoisomers. - 1 a and (+) - 1 a . - 1 a can displace 50% of (+) - 1 a 10 - 7 M olfactory responses (effective concentration 50, EC 50 ) at 5.8 10 - 8 M, whereas stereoisomer (+) - 1 a - 1 a 10 - 7 M olfactory response at 4.1 10 - 9 M. The fact that (+) - 1 a - 1 a at a lower - 1 a to (+) - 1 a implies that (+) - 1 a is a stronger competitor for the receptor binding site. 168 PUBLICATION LIST Publication List A - 1. Recent publications during the characterization of sea lamprey pheromones. Brant CO, Li K, Johnson NS, Li W (2015) A pheromone outweighs tempe rature in influencing migration of sea lamprey. Royal Society Open Science 2 (5), 150009. Li K, Brant CO, Bussy U, Pinnamaneni H, Patel H, Hoye TR, Li W (2015) iso - Petromyroxols: Novel dihydr oxylated tetrahydrofuran enantiomers from sea Lamprey ( Petromyzon marinus ). Molecules 20 (3), 5215 - 5222 Li K, Huertas M, Brant CO, Chung - Davidson Y - W, Bussy U, Hoye TR, Li W (2015) (+) - - petromyroxols: antipodal tetrahydrofurandiols from larval se a lamprey ( Petromyzon marinus L.) that eicit enantioselective olfactory responses. Organic Letters 17(2): 286 - 289. Li K, Brant CO, Huertas M, Li W. (2013). Petromyzonin, a hexahydrophenanthrene sulfate isolated from the larval sea lamprey ( Petromyzon mar inus L.). Organic Letters 15:5924 - 5927. Brant CO, Chung - Davidson Y - W, Li K, Scott AM, Li W. (2013). Biosynthesis and release of pheromonal bile salts in mature male sea lamprey. BMC Biochemistry 14:30. Choi J, Jeon S, Johnson NS, Brant CO, Li W. (2013). Odor - conditioned rheotaxis of the sea lamprey: modeling, analysis and validation. Bioinspiration & Biomimetics 8(4):046011. Li K, Brant CO, Siefkes MJ, Kruckman HG, Li W. (2013). Characterization of a n ovel bile alcohol sulfate released by sexually mature male sea lamprey ( Petromyzon marinus ). Plos ONE 8(7): e68157. Li K, Siefkes MJ, Brant CO, Li W. (2012). Isolation and identification of petromyzestrosterol, a polyhydroxysteroid from sexually mature male sea lamprey ( Petromyzon marinus L.). Steroids 77(7):806 - 810. Li K, Wang H, Brant CO, Ahn SC, Li W. (2011). Multiplex quantification of lamprey specific bile acid derivatives in environmental water using UHPLC - MS/MS. Journal of Chromatography B 879(32):3879 - 3886. Xi X, Johnson NS, Brant CO, Yun S - S, Chambers KL, Jones AD, Li W. (2011). Quantification of a male sea lamprey pheromone in tributaries of Laurentian Great Lakes by liquid chromatography - tandem mass spectrometry. Envir onmental Science and Technology 45(15):6437 - 6443. Johnson NS, Yun S - S, Thompson H, Brant CO, Li W. (2009). A synthesized pheromone induces upstream movement in female sea lamprey and summons them into traps. PNAS 106:1021 - 1026. 169 REFERENCES 170 REFERENCES Applegate VC, 1950. Natural history of the sea lamprey ( Petromyzon marinus ) in Michigan. Washington D. C. : United States Department of the Interior Fish and Wildlife Service. Bergstedt RA, Seelye JG, 1995. Evidence for lack of homing by sea lampreys. Transactions of the American Fisheries Society 124:235 - 239. Bergstedt RA, Twohey MB, 2007. Research to support sterile - male - release and genetic alteration techniques for sea lamprey control. J Gt Lakes Res 33:48 - 69. doi: 10.3394/0380 - 1330(2007)33[48:rtssag]2.0.co;2. Binder TR, McDonald DG, 2007. Is there a role for vision in the behaviour of sea lampreys ( Petromyzon marinus ) during their upstream spawning migration? Canadian Journal of Fisheries and Aquatic Scien ces 64:1403 - 1412. Binder TR, McDonald DG, 2008. The role of temperature in controlling diel activity in upstream migrant sea lampreys ( Petromyzon marinus ). Canadian Journal of Fisheries and Aquatic Sciences 65:1113 - 1121. Binder TR, McLaughlin RL, McDonald DG, 2010. Relative importance of water temperature, water level, and lunar cycle to migratory activity in spawning - phase sea lampreys in Lake Ontario. Transactions of the American Fisheries Society 139:700 - 712. Bjerselius R, Li W, Teeter JH, Seelye JG, Joh nson PB, Maniak PJ, Grant GC, Polkinghorne CN, Sorenson P W, 2000 . Direct behavioral evidence that unique bile acids released by larval sea lamprey ( Petromyzon marinus ) function as a migratory pheromone. Canadian Journal of Fisheries and Aquatic Sciences 57 :557 - 569. Bradbury J, Vehrencamp S, 2011. Principles of Animal Communication, Sunderland, MA: Sinauer Associates. 697 p. Brant CO, Chung - Davidson YW, Li K, Scott AM, Li W, 2013. Biosynthesis and release of pheromonal bile salts in mature male sea lamprey. BMC Biochemistry 14:30. Brant CO, Li K, Johnson NS, Li W, 2015. A pheromone outweighs temperature in influencing migration of sea lamprey. Royal Society Open Science 2:150009. Breithaupt T, Thiel M, 2010. Chemical communication in crustaceans: Springer Sci ence & Business Media. Briffa M, Williams R, 2006. Use of chemical cues during shell fights in the hermit crab Pagurus bernhardus. Behaviour 143:1281 - 1290. doi: 10.1163/156853906778691577. Buchinger TJ, Li W, Johnson NS, 2014. Bile Salts as Semiochemicals in Fish. Chemical Senses. doi: 10.1093/chemse/bju039. 17 1 Buchinger TJ, Wang H, Li W, Johnson NS, 2013. Evidence for a receiver bias underlying female preference for a male mating pheromone in sea lamprey. Proceedings Biological sciences / The Royal Society 28 0:20131966. doi: 10.1098/rspb.2013.1966. Capon RJ, Barrow RA, Rochfort S, Jobling M, Skene C, Lacey E, Gill JH, Friedel T, Wadsworth D, 1998. Marine nematocides: Tetrahydrofurans from a southern Australian brown alga, Notheia anomala. Tetrahedron 54:2227 - 2 242. doi: 10.1016/s0040 - 4020(97)10432 - x. Cardé R, Cardé A, Hill A, Roelofs W, 1977. Sex pheromone specificity as a reproductive isolating mechanism among the sibling speciesArchips argyrospilus andA. mortuanus and other sympatric tortricine moths (Lepidopt era: Tortricidae). J Chem Ecol 3:71 - 84. Chapman OL, Klun JA, Mattes KC, Sheridan RS, Maini S, 1978. Chemoreceptors in Lepidoptera - stereochemical differentiation of dual receptors for an achiral pheromone. Science 201:926 - 928. doi: 10.1126/science.201.435 9.926. Christie GC, Goddard CI, 2003. Sea Lamprey International Symposium (SLIS II): Advances in the integrated management of sea lamprey in the Great Lakes. J Gt Lakes Res 29:1 - 14. Christy JH, 1995. Mimicry, mate choice, and the sensory trap hypothesis. A merican Naturalist:171 - 181. Chung - Davidson YW, Priess MC, Yeh CY, Brant CO, Johnson NS, Li K, Nanlohy KG, Bryan MB, Brown CT, Choi J, Li WM, 2013. A thermogenic secondary sexual character in male sea lamprey. J Exp Biol 216:2702 - 2712. doi: 10.1242/jeb.0857 46. Clemens BJ, Mesa MG, Magie RJ, Young DA, Schreck CB, 2012. Pre - spawning migration of adult Pacific lamprey, Entosphenus tridentatus, in the Willamette River, Oregon, USA. Environmental biology of fishes 93:245 - 254. Commission GLF, 2001. Strategic visio n of the Great Lakes Fishery Commission for the first decade of the new millennium: Great Lakes Fishery Commission. Commission GLF, 2011. Strategic Vision of the Great Lakes Fishery Commission 2011 2020. Great Lakes Fishery Commission: Ann Arbor, MI, USA. Coracini M, Bengtsson M, Reckziegel A, Löfqvist J, Francke W, Vilela E, Eiras A, Kovaleski A, Witzgall P, 2001. Identification of a four - component sex pheromone blend in Bonagota cranaodes (Lepidoptera: Tortricidae). Journal of economic entomology 94:911 - 9 14. Davidson YWC, Huertas M, Li WM, 2011. A review of research in fish pheromones. In: Breithaupt T, Thiel M, editors. Chemical communication in crustaceans New York: Springer. p. 467 - 482. Davies NB, Krebs JR, West SA, 2012. An introduction to behavioural ecology: John Wiley & Sons. 172 DeBose J, Nevitt G, 2008. The use of Odors at Different Spatial Scales: Comparing Birds with Fish. J Chem Ecol 34:867 - 881. doi: 10.1007/s10886 - 008 - 9493 - 4. Døving KB, Selset R, Thommesen G, 1980. Olfactory sensitivity to bile aci ds in salmonid fishes. Acta Physiol Scand 108:123 - 131. doi: 10.1111/j.1748 - 1716.1980.tb06509.x. Fine JM, Sisler SP, Vrieze LA, Swink WD, Sorensen PW, 2006. A Practical Method for Obtaining Useful Quantities of Pheromones from Sea Lamprey and Other Fishes f or Identification and Control. J Gt Lakes Res 32:832 - 838. doi: 10.3394/0380 - 1330(2006)32[832:APMFOU]2.0.CO;2. Fine JM, Sorensen PW, 2008. Isolation and biological activity of the multi - component sea lamprey migratory pheromone. J Chem Ecol 34:1259 - 1267. Fo rsythe P, Scribner K, Crossman J, Ragavendran A, Baker E, Davis C, Smith K, 2012. Environmental and lunar cues are predictive of the timing of river entry and spawning site arrival in lake sturgeon Acipenser fulvescens. J Fish Biol 81:35 - 53. Garcia CM, Ram irez E, 2005. Evidence that sensory traps can evolve into honest signals. Nature 434:501 - 505. Gasparini C, Serena G, Pilastro A, 2013. Do unattractive friends make you look better? Context - dependent male mating preferences in the guppy. Proceedings of the Royal Society B: Biological Sciences 280:20123072. Giron D, Goldbronn C, 1995. Place of DSC purity analysis in pharmaceutical development. Journal of Thermal Analysis and Calorimetry 44:217 - 251. Hara TJ, 1992. Mechanisms of olfaction. Fish chemoreception: Springer. p. 150 - 170. Hölldobler B, Wilson EO, 1977. Colony - specific territorial pheromone in the african weaver ant oecophylla longinoda (latreille). Proceedings of the National Academy of Sciences 74:2072 - 2075. Hoye TR, Dvornikovs V, Fine JM, Anderson KR , Jeffrey CS, Muddiman DC, Shao F, Sorensen PW, Wang J, 2007a. Details of the structure determination of the sulfated steroids PSDS and PADS: New components of the sea lamprey ( Petromyzon marinus ) migratory pheromone. The Journal of organic chemistry 72:75 44 - 7550. Hoye TR, Jeffrey CS, Shao F, 2007b. Mosher ester analysis for the determination of absolute configuration of stereogenic (chiral) carbinol carbons. Nat Protoc 2:2451 - 2458. doi: 10.1038/nprot.2007.354. Huertas M, Hubbard PC, Canário AV, Cerdà J, 20 07. Olfactory sensitivity to conspecific bile fluid and skin mucus in the European eel Anguilla anguilla (L.). J Fish Biol 70:1907 - 1920. 173 Hurd PL, Enquist M, 2005. A strategic taxonomy of biological communication. Anim Behav 70:1155 - 1170. doi: http://dx.doi.org/10.1016/j.anbehav.2005.02.014 . Johnson N S , Tix J, Hlina B, Wagner CM, Siefkes M, Wang H, Li W, 2015a. A Sea Lamprey ( Petromyzon marinus ) Sex Pheromone Mixture Increases Trap Catch Relativ e to a Single Synthesized Component in Specific Environments. J Chem Ecol 41:311 - 321. doi: 10.1007/s10886 - 015 - 0561 - 2. Johnson NS, Buchinger TJ, Li W, 2015b. Reproductive Ecology of Lampreys. In: Docker MF, editor. Lampreys: Biology, Conservation and Contro l: Springer Netherlands. p. 265 - 303. Johnson NS, Li W, 2010. Understanding behavioral responses of fish to pheromones in natural freshwater environments. Journal of Comparative Physiology A 196:701 - 711. Johnson NS, Luehring MA, Siefkes MJ, Li WM, 2006. Mat ing pheromone reception and induced behavior in ovulating female sea lampreys. North Am J Fish Manage 26:88 - 96. doi: 10.1577/m05 - 018.1. Johnson NS, Muhammad A, Thompson H, Choi J, Li W, 2012. Sea lamprey orient toward a source of a synthesized pheromone us ing odor - conditioned rheotaxis. Behavioral Ecology and Sociobiology 66:1557 - 1567. Johnson NS, Siefkes MJ, Wagner CM, Dawson H, Wang H, Steeves T, Twohey M, Li W, 2013. A synthesized mating pheromone component increases adult sea lamprey ( Petromyzon marinus ) trap capture in management scenarios. Canadian Journal of Fisheries and Aquatic Sciences 70:1101 - 1108. Johnson NS, Yun SS, Thompson HT, Brant CO, Li WM, 2009. A synthesized pheromone induces upstream movement in female sea lamprey and summons them into t raps. Proc Natl Acad Sci U S A 106:1021 - 1026. doi: 10.1073/pnas.0808530106. Kang J, Caprio J, 1997. In vivo responses of single olfactory receptor neurons of channel catfish to binary mixtures of amino acids. Journal of N europhysiology 77:1 - 8. Karlson P, B utenandt A, 1959. Pheromones (ectohormones) in insects. Annu Rev Entomol 4:39 - 58. doi: 10.1146/annurev.en.04.010159.000351. Karlson P, Lüscher M, 1959. Pheromones: a new term for a class of biologically active substances. Nature 183:55 - 56. doi: 10.1038/183 055a0. Keller - Costa T, Canário AV, Hubbard PC, 2014. Olfactory sensitivity to steroid glucuronates in Mozambique tilapia suggests two distinct and specific receptors for pheromone detection. The Journal of experimental biology 217:4203 - 4212. Kelley LA, Kel ley JL, 2014. Animal visual illusion and confusion: the importance of a perceptual perspective. Behavioral Ecology 25:450 - 463. 174 Kemp PS, Russon IJ, Vowles AS, Lucas MC, 2011. The influence of discharge and temperature on the ability of upstream migrant adul t river lamprey (Lampetra fluviatilis) to pass experimental overshot and undershot weirs. River research and applications 27:488 - 498. Klun JA, Chapman OL, Mattes KC, Wojtkowski PW, Beroza M, Sonnet PE, 1973. Insect sex pheromones: minor amount of opposite geometrical isomer critical to attraction. Science, USA 181:661 - 663. Laidre ME, Johnstone RA, 2013. Animal signals. Curr Biol 23:R829 - R833. doi: 10.1016/j.cub.2013.07.070. Larsen LO, 1980. Physiology of Adult Lampreys, with Special Regard to Natural Starva tion, Reproduction, and Death after Spawning. Canadian Journal of Fisheries and Aquatic Sciences 37:1762 - 1779. doi: 10.1139/f80 - 221. Lewis J, McMillan D, 1965. The development of the ovary of the sea lamprey ( Petromyzon marinus L.). Journal of morphology 1 17:425 - 466. Li K, Brant CO, Bussy U, Pinnamaneni H, Patel H, Hoye TR, Li WM, 2015b. iso - Petromyroxols: Novel Dihydroxylated Tetrahydrofuran Enantiomers from Sea Lamprey ( Petromyzon marinus ). Molecules 20:5215 - 5222. doi: 10.3390/molecules20035215. Li K, Bra nt CO, Huertas M, Hur SK, Li W, 2013a. Petromyzonin, a Hexahydrophenanthrene Sulfate Isolated from the Larval Sea Lamprey ( Petromyzon marinus L.). Organic Letters 15:5924 - 5927. doi: 10.1021/ol402478r. Li K, Brant CO, Siefk es MJ, Kruckman HG, Li WM, 2013b . Characterization of a novel bile alcohol sulfate released by sexually mature male sea lamprey ( Petromyzon marinus ). PLoS One 8:8. doi: e6815710.1371/journal.pone.0068157. Li K, Huertas M, Brant C, Chung - Davidson Y - W, Bussy U, Hoye TR, Li W, 2014 . (+) - and - Petromyroxols: Antipodal Tetrahydrofurandiols from Larval Sea Lamprey ( Petromyzon marinus L.) That Elicit Enantioselective Olfactory Responses. Organic Letters. doi: 10.1021/ol5033893. Li K, Siefkes MJ, Brant CO, Li WM, 2012. Isolation and identificati on of petromyzestrosterol, a polyhydroxysteroid from sexually mature male sea lamprey ( Petromyzon marinus L.). Steroids 77:806 - 810. doi: 10.1016/j.steroids.2012.03.006. Li W, Sorensen PW, Gallaher DD, 1995. The olfactory system of migratory adult sea lampr ey ( Petromyzon marinus ) is specifically and acutely sensitive to unique bile acids released by conspecific larvae. The Journal of General Physiology 105:569 - 587. doi: 10.1085/jgp.105.5.569. Li W, Twohey M, Jones M, Wagner M, 2007. Research to guide use of pheromones to control sea lamprey. J Gt Lakes Res 33:70 - 86. doi: 10.3394/0380 - 1330(2007)33[70:rtguop]2.0.co;2. 175 Li W , Scott AP, Siefkes MJ, Yan HG, Liu Q, Yun SS, Gage DA, 2002. Bile acid secreted by male sea lamprey that acts as a sex pheromone. Science 296:138 - 141. doi: 10.1126/science.1067797. Liley N, 1982. Chemical communication in fish. Canadian Journal of Fisheries and Aquatic Sciences 39:22 - 35. Liley N, Stacey N, 1983. Hormones, pheromones, and reproductive behavior in fish. Fish physiology 9. Linn C, Roelofs W, 1989. Response specificity of male moths to multicomponent pheromones. Chemical Senses 14:421 - 437. Lipschitz D, Michel W, 1999. Physiological Evidence for the Discrimination ofl - Arginine From Structural Analogues by the Zebrafish Olfactory S ystem. Journal of neurophysiology 82:3160 - 3167. Lucas MC, Baras E, Thom TJ, Duncan A, Slavik O, 2001. Migration of freshwater fishes. Manion PJ, Hanson LH, 1980. Spawning behavior and fecundity of lampreys from the upper three Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences 37:1635 - 1640. Markaverich BM, Alejandro MA, Markaverich D, Zitzow L, Casajuna N, Camarao N, Hill J, Bhirdo K, Faith R, Turk J, 2002. Identification of an endocrine disrupting agent from corn with mitogenic activity. Bioch em Biophys Res Commun 291:692 - 700. Meckley TD, Wagner CM, Gurarie E, 2014. Coastal movements of migrating sea lamprey ( Petromyzon marinus ) in response to a partial pheromone added to river water: implications for management of invasive populations. Canadia n Journal of Fisheries and Aquatic Sciences 71:533 - 544. doi: 10.1139/cjfas - 2013 - 0487. Meckley TD, Wagner CM, Luehring MA, 2012. Field evaluation of larval odor and mixtures of synthetic pheromone components for attracting migrating sea lampreys in rivers. J Chem Ecol 38:1062 - 1069. doi: 10.1007/s10886 - 012 - 0159 - x. Moore A, Bendall B, Barry J, Waring C, Crooks N, Crooks L, 2012. River temperature and adult anadromous Atlantic salmon, Salmo salar, and brown trout, Salmo trutta. Fisheries Management and Ecology 19:518 - 526. Moore HH, Schleen IP, 1980. Changes in spawning runs of sea lamprey ( Petromyzon marinus ) in selected streams of Lake S uperior after chemical control. Can J Fish Aquat Sci 37:1851 - 1860. doi: 10.1139/f80 - 227. Ohtani I, Kusumi T, Kashman Y, Kakisa wa H, 1991. High - field FT NMR application of mosher method: The absolute configurations of marine terpenoids. Journal of the American Chemical Society 113:4092 - 4096. doi: 10.1021/ja00011a006. 176 Oisi Y, Ota KG, Kuraku S, Fujimoto S, Kuratani S, 2013. Craniofa cial development of hagfishes and the evolution of vertebrates. Nature 493:175 - 180. Quinn TP, Adams DJ, 1996. Environmental changes affecting the migratory timing of American shad and sockeye salmon. Ecology:1151 - 1162. Reyes - Garcia L, Cuevas Y, Ballesteros C, Curkovic T, Löfstedt C, Bergmann J, 2014. A 4 - component sex pheromone of the Chilean fruit leaf roller Proeulia auraria (Lepidoptera: Tortricidae). Ciencia e Investigación Agraria 41:187 - 196. Ryan MJ, Cummings ME, 2013. Perceptual biases and mate choic e. Annual Review of Ecology, Evolution, and Systematics 44:437 - 459. Shorey HH, 2013. Animal communication by pheromones: Academic Press. Siefkes M J , Li W, 2004. Electrophysiological evidence for detection and discrimination of pheromonal bile acids by the olfactory epithelium of female sea lampreys ( Petromyzon marinus ). Journal of Comparative Physiology A 190:193 - 199. Siefkes MJ, Scott AP, Zielinski B, Yun SS, Li WM, 2003. Male sea lampreys, Petromyzon marinus L., excrete a sex pheromone from gill epithelia . Biol Reprod 69:125 - 132. doi: 10.1095/biolreprod.102.014472. Siefkes MJ, Winterstein SR, Li WM, 2005. Evidence that 3 - keto petromyzonol sulphate specifically attracts ovulating female sea lamprey, Petromyzon marinus . Anim Behav 70:1037 - 1045. doi: 10.1016/ j.anbehav.2005.01.024. Skov C, Aarestrup K, Baktoft H, Brodersen J, Brönmark C, Hansson L - A, Nielsen EE, Nielsen T, Nilsson PA, 2010. Influences of environmental cues, migration history, and habitat familiarity on partial migration. Behavioral Ecology:arq1 21. Smith BR, Tibbles JJ, 1980. Sea lamprey ( Petromyzon marinus ) in Lakes Huron, Michigan, and Superior: history of invasion and control, 1936 - 78. Canadian Journal of Fisheries and Aquatic Sciences 37:1780 - 1801. doi: 10.1139/f80 - 222. Sorensen P W , Hara T, S tacey N, Goetz FW, 1988. F prostaglandins function as potent olfactory stimulants that comprise the postovulatory female sex pheromone in goldfish. Biology of Reproduction 39:1039 - 1050. Sorensen PW, 1992. Hormones, pheromones and chemoreception. Fish Chemo reception: Springer. p. 199 - 228. Sorensen PW, Fine JM, Dvornikovs V, Jeffrey CS, Shao F, Wang JZ, Vrieze LA, Anderson KR, Hoye TR, 2005. Mixture of new sulfated steroids functions as a migratory pheromone in the sea lamprey. Nat Chem Biol 1:324 - 328. doi: 1 0.1038/nchembio739. Sorensen PW, Stacey NE, Chamberlain KJ, 1989. Differing behavioral and endocrinological effects of two female sex pheromones on male goldfish. Horm Behav 23:317 - 332. 177 Stacey N, Cardwell J, Hormones as sex pheromones in fish: widespread d istribution among freshwater species. Proceedings of the Fifth International Symposium on the Reproductive Physiology of Fish 1995. Symposium. p. 244 - 248. Stacey N, Cardwell J, Murphy C, 1996. Hormonal pheromones in freshwater fishes: Preliminary results o f an electroolfactogram survey. Fish Pheromones: Origins and Modes of Action:47 - 55. Steeves TB, Slade JW, Fodale MF, Cuddy DW, Jones ML, 2003. Effectiveness of Using Backpack Electrofishing Gear for Collecting Sea Lamprey ( Petromyzon marinus ) Larvae in Gre at Lakes Tributaries. J Gt Lakes Res 29, Supplement 1:161 - 173. doi: http://dx.doi.org/10.1016/S0380 - 1330(03)70485 - 7 . Steiger S, Schmitt T, Schaefer HM, 2011. The origin and dynamic evolution o f chemical information transfer. Proc R Soc B - Biol Sci 278:970 - 979. doi: 10.1098/rspb.2010.2285. Stier K, Kynard B, 1986. Movement of sea - run sea lampreys, Petromyzon marinus , during the spawning migration in the Connecticut River. Fish Bull 84:749 - 753. St oddard MC, Prum RO, 2011. How colorful are birds? Evolution of the avian plumage color gamut. Behavioral Ecology:arr088. Stuart - Fox D, 2005. Deception and the origin of honest signals. Trends Ecol Evol 20:521 - 523. Symonds MRE, Elgar MA, 2008. The evolution of pheromone diversity. Trends Ecol Evol 23:220 - 228. doi: 10.1016/j.tree.2007.11.009. Teeter J, 1980 . Pheromone communication in sea lampreys ( Petromyzon marinus ): I mplications for population management. Can J Fish Aquat Sci 37:2123 - 2132. doi: 10.1139/f80 - 254. Tibbetts EA, 2002. Visual signals of individual identity in the wasp Polistes fuscatus . Proc R Soc Lond B 269: 1423 - 1428. Tinbergen N, 1952. " Derived" activities; their causation, biological significance, origin, and emancipation during evolution. T he Quarterly Review of Biology 27:1 - 32. Van Dooren A, Müller B, 1984. Purity determinations of drugs with differential scanning calorimetry (DSC) a critical review. International journal of pharmaceutics 20:217 - 233. Venkatachalam K, 2005. Petromyzonol sulf ate and its derivatives: the chemoattractants of the sea lamprey. BioEssays 27:222 - 228. Vrieze LA, Bergstedt RA, Sorensen PW, 2011. Olfactory - mediated stream - finding behavior of migratory adult sea lamprey ( Petromyzon marinus ). Canadian Journal of Fisherie s and Aquatic Sciences 68:523 - 533. doi: 10.1139/F10 - 169. 178 Vrieze LA, Bjerselius R, Sorensen PW, 2010. Importance of the olfactory sense to migratory sea lampreys Petromyzon marinus seeking riverine spawning habitat. J Fish Biol 76:949 - 964. doi: 10.1111/j.10 95 - 8649.2010.02548.x. Wagner CM, Twohey MB, Fine JM, 2009. Conspecific cueing in the sea lamprey: do reproductive migrations consistently follow the most intense larval odour? Anim Behav 78:593 - 599. doi: 10.1016/j.anbehav.2009.04.027. Warren RG, Wells RJ, Blount JF, 1980. A novel lipid from the brown alga Notheia anomala. Aust J Chem 33:891 - 898. doi: 10.1071/CH9800891. Wehrli SL, Moore KS, Roder H, Durell S, Zasloff M, 1993. Structure of the novel steroidal antibiotic squalamine determined by 2 - dimensional NMR spectroscopy. Steroids 58:370 - 378. doi: 10.1016/0039 - 128x(93)90040 - t. Weiss I, Rössler T, Hofferberth J, Brummer M, Ruther J, Stökl J, 2013. A nonspecific defensive compound evolves into a competition avoidance cue and a female sex pheromone. Nature co mmunications 4 :2767. Wilson EO, 1970. Competitive and aggressive behavior. Social Science Information 9:123 - 154. Wisenden BD, 2014. The Cue Signal Continuum. Fish Pheromones and Related Cues: John Wiley & Sons, Inc. p. 149 - 158. Wyatt TD, 2010. Pheromones a nd signature mixtures: defining species - wide signals and variable cues for identity in both invertebrates and vertebrates. J Comp Physiol A - Neuroethol Sens Neural Behav Physiol 196:685 - 700. doi: 10.1007/s00359 - 010 - 0564 - y. Xi XD, Johnson NS, Brant CO, Yun SS, Chambers KL, Jones AD, Li W, 2011. Quantification of a Male Sea Lamprey Pheromone in Tributaries of Laurentian Great Lakes by Liquid Chromatography - Tandem Mass Spectrometry. Environ Sci Technol 45:6437 - 6443. doi: 10.1021/es200416f. Yambe H, Kitamura S, Kamio M, Yamada M, Matsunaga S, Fusetani N, Yamazaki F, 2006. L - Kynurenine, an amino acid identified as a sex pheromone in the urine of ovulated female masu salmon. Proceedings of the National Academy of Sciences 103:15370 - 15374. Yeh C - Y, Chung - Davidson Y - W, Wang H, Li K, Li W, 2012. Intestinal synthesis and secretion of bile salts as an adaptation to developmental biliary atresia in the sea lamprey. Proceedings of the National Academy of Sciences 109:11419 - 11424. Yorke MA, McMillan DB, 1980. Structural as pects of ovulation in the lamprey, Petromyzon marinus . Biol Reprod 22:897 - 912. Yun SS, Scott AP, Li WM, 2003. Pheromones of the male sea lamprey, Petromyzon marinus L.: structural studies on a new compound, 3 - keto allocholic acid, and 3 - keto petromyzonol sulfate. Steroids 68:297 - 304. doi: 10.1016/s0039 - 128x(02)00178 - 2.