Morphometric and physical characteristics distinguishing adult Patagonian lamprey, Geotria macrostoma from the pouched lamprey, Geotria australis

The pouched lamprey, Geotria australis Gray, 1851, has long been considered monotypic in the Geotriidae family with a wide southern temperate distribution across Australasia and South America. Recent studies have provided molecular and morphological evidence for a second Geotria species in South America; Geotria macrostoma (Burmeister, 1868). The aim of this study was to determine morphometric and physical characteristics of adult G. macrostoma that further differentiate this re-instated species of Geotriidae from G. australis. The diagnostic features discriminating immature adult G. macrostoma from G. australis when entering fresh water, are distinct differences in dentition, oral papillae and fimbriae counts and differences in coloration. In addition, G. macrostoma display greater growth of the prebranchial region and oral disc and has a deeper body depth and higher condition factor. All current ecological knowledge of the genus Geotria is based on Australasian populations, which may not be applicable to G. macrostoma. To ensure the conservation and protection of the Patagonian lamprey as a re-identified species, further investigations are needed to understand its life history, biology and ecology throughout its range.


Introduction
Lampreys are ancient jawless fishes with a lineage dating back around 500 million years [1]. In their review, Docker and Potter [2] reported that lampreys, which have an antitropical distribution, are represented by over 40 species. The majority of those species are restricted to the a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 analysis and preparation of the manuscript. The funder did not have any role in salaries for authors [CFB, PW, EW] but did partially cover salary for JK. The specific roles of these authors are articulated in the 'author contributions' section. Further funding was provided by IEASA (Integración Energética Argentina, www.ieasa.com.ar), Agencia Nacional de la Promoción de la Ciencia y la Tecnología (ANPCyT, Argentina, grant no. PICT2015-3490 to CRR, www.agencia.mincyt.gob.ar), CONICET (Consejo Nacional de Investigaciones Científicas y Técnicas, http://www.conicet.gov.ar), as well as a visiting professor grant in 2003 from MNHN (Paris) to CBR and a research grant in 2010 from the Canadian Museum of Nature to study at the NHM (London) to CBR. third, and the dorsal fins change shape and become closer. In addition, the labial teeth on the oral disc become smaller and widely spaced, the transverse lingual lamina changes from tridentate to bidentate, the relative size of the prebranchial region and the length and width of the oral disc increases, a large gular pouch develops in males [10,27,29] and the females develop a 'rope', a raised ridge in front of the first dorsal fin [6].
Morphological characters have traditionally been used to identify lamprey species [15,35,36]. Although G. australis undergoes pronounced ontogenetic changes during its freshwater spawning run, Riva-Rossi et al. [4] identified two morphological characters that distinguished Patagonian G. macrostoma from G. australis at the immature adult stage. Namely, the second dorsal and caudal fins are connected by a low skin fold and are contiguous in G. macrostoma, whereas there is a distinct space between the posterior end of the second dorsal fin and the origin of the caudal fin in G. australis. The gap between the second dorsal fin and the caudal fin in G. australis has disappeared by sexual maturity, but currently this has only been documented for females in Australia [32,37]. However, this may not be a gender specific difference and could reflect the paucity of records of mature spawning G. australis. In addition, the position of the cloaca is always posterior to the origin of the second dorsal fin in G. macrostoma, whereas it is anterior to, or under, the origin of the second dorsal fin in G. australis, as is also the case with ammocoetes [13].
Detailed morphological investigations beyond obvious macroscopic descriptors are important, however, for defining and characterizing populations, understanding biodiversity and evolution of the species, as well as supporting ecological knowledge and conservation [36,38,39]. The objective of this study on Geotria species, was, therefore, to quantify morphometric characters, such as the lengths of various body parts, and physical characteristics, which are traditionally used by lamprey taxonomists to differentiate lamprey species [15]. These characters are then used to determine the features that distinguish adults of G. macrostoma and G. australis and thereby further characterize the re-assigned species of Geotriidae.

Materials and methods
This study was carried out in accordance to the ethical regulations of CONICET (Consejo Nacional de Investigaciones Científicas y Tecnológicas) for biomedical and biological research with laboratory and farm animals and those obtained in nature (Resolution D 1047 Annex II of the year 2005), and in accordance with the NIWA Animal Ethics Committee approval AEC189. Lamprey capture and handling procedures were approved by specific permits issued by the Ministerio de Agricultura, Ganaderia y Pesca from the Río Negro Province (Resolution 007), by the Instituto Provincial del Agua, Administración General de Recursos Hídricos from the Chubut Province (Resolution 24/19DGAguas-IPA), by the Ministerio de Producción, Comercio e Industria, Subsecretaría de Coordinación Pesquera from the Santa Cruz Province (Resolution MPCI 438818/18 del Provincia de Santa Cruz), from the Ministry for Primary Industries, Fisheries New Zealand, (Special Permit 666/2), and from the South Australian Minister for Primary Industries and Regional Development (Section 115 Ministerial Exemption ME9903055). Lamprey anesthesia in Argentina was performed using a mild dose (30 mg/mL) and for mature adult lamprey only, an overdose (100 mg/mL) of benzocaine (Parafarm, CABA, Argentina). For genetic analyses presented in Riva-Rossi et al. [4], a sample of 25 individuals (15 from the Santa Cruz River and 10 from the Chubut River) were euthanized with an overdose of benzocaine, stored at -20˚C, and transported to the laboratory in Puerto Madryn. In New Zealand and Australia lamprey were anaesthetized with 0.05 ml/L stock solution of AQUI-S (AQUI-S, Lower Hutt, New Zealand). As G. australis is considered a threatened species [40] all individuals were released alive back to the river of capture.
Although Clemens [5] used adult to encompass non-feeding, pre-spawning lamprey that have commenced their upstream migration and are in various stages of sexual maturation, sexual maturation in adult Geotria takes around 16 months. Therefore, to clarify the life stages examined in the present study, adult Geotria are termed immature when collected on entry to fresh water and show no external sexual dimorphism. Adults that show sexual dimorphism but have not fully matured (i.e. are not ready to spawn) are termed immature male or female. Clemens [5] terms post-spawning lamprey in the process of dying senescent. However, postspawning Geotria remain active for months, exhibiting parental care and guarding of their eggs, therefore, lamprey that show sexual dimorphism at or post-spawning are termed mature adults.

Sample collection
A total of 164 immature adult G. macrostoma were collected from two sampling sites in Argentina during their upstream migration (Fig 1). Between February and March 2019, 125 G. macrostoma were captured using fyke nets in the lower Santa Cruz River, Argentina (50.05˚S, 69.01˚W). In May 2019, 39 G. macrostoma were collected by electrofishing and by hand from the lower Chubut River, Argentina (43.45˚S, 65.94˚W). In Australasia, 155 immature adult G. australis were collected from three sampling sites during their upstream migration (Fig 1). In August 2019, 84 G. australis were captured by hand from two rock weirs located immediately above the tidal zone in the Waikawa River, Southland, New Zealand (46.59˚S 169.14˚E). In September 2019 and August 2020, 42 G. australis were captured using cage traps in fishways on the Murray Barrages in the lower Murray River, Australia (35.53˚S, 138.81˚E). In August 2020, 29 G. australis were captured by hand from a rock weir located approximately 84 km inland in the Mokau River, New Zealand (38.32˚S 174.58˚E). While no fresh run immature adults of G. australis from Chile were collected as part of this study, eight immature adult G. australis collected from Temuco, Chile, in October 1963 and measured by Neira [10] were included in the dataset (Fig 1).
As G. australis undergoes a wide range of morphometric changes during development it was important to examine lampreys from each location at a known point of their freshwater spawning run. To measure fresh run immature adult lamprey upon entry to fresh water, whilst still displaying the blue-green coloration, sampling locations for the Waikawa River, Chubut River, Santa Cruz River and Murray River were in the lower reaches, at or just above tidal influence. The Waikawa River, Chubut River and Murray River individuals were examined within four days of capture during their autumn through spring migration run. The Santa Cruz River lamprey enters fresh water during summer and logistical challenges prevented measuring the entire sample immediately. Twenty-five individuals were measured immediately and released back to the river with the remaining 100 held alive at the Piedra Buena hatchery, Argentina, in a flow through outdoor tank fed directly from the Santa Cruz River. These lamprey were measured six weeks after capture. It is important to note that although the month of river entry varies between the Waikawa, Santa Cruz, Chubut and Murray populations, the lamprey are all entering fresh water in the same state as fresh run immature adults, this is verified by the blue coloration, which they retain during their entire oceanic phase but lose soon after entering fresh water [32]. To determine if time in fresh water and corresponding morphometric growth led to the characters of G. australis and G. macrostoma overlapping, lampreys from the Mokau River, New Zealand, were also examined. Migratory G. australis had already lost their blue-green coloration prior to reaching the inland fishing location in the Mokau River. Based on observations of Baker et al. [6], this indicated that the lamprey had been in fresh water for at least four weeks. Mokau River lamprey were held in the NIWA laboratory for four additional weeks before measurements were taken, which was when early development of the gular pouch became evident. In addition, the lamprey collected from Temuco, Chile, had already lost their blue-green coloration prior to reaching the inland fishing location and, therefore, are comparable to the Mokau River lamprey, having been in fresh water for at least four weeks.
For comparison with the immature adults, measurements from sexually mature adults of both species were included in the dataset. Three post-spawning G. macrostoma were analyzed. One male was captured in a fyke net set in the lower Santa Cruz River during January 2020 (and released back to the river), while the other specimens had been collected during April 2016 from both the lower Santa Cruz River (male) and during October 2017 from the upper Santa Cruz River (female). The latter two specimens were fixed in 10% neutral buffered formalin and deposited in the Ichthyology Collection of the Instituto de Diversidad y Evolución Austral (IDEAus-CONICET), Puerto Madryn, Argentina. Four sexually mature post-spawning G. australis from Neira [10] were also included. These lamprey consisted of a male from the Chillán River (captured February 1974), a male from the Maullín River (captured February 1979), a female from the Andalién River (captured May 1981) and another female from Concepción Bay, Talcahuano (captured February 1977). In addition, photographs of a post-spawning male and female G. australis from Canterbury, New Zealand, have been included in figures for descriptive comparisons with mature G. macrostoma.

Morphometric and physical characters
In total, 22 morphometric characters were measured or calculated, of which, 18 were measured to the nearest 0.01 mm using an electronic caliper (Fig 2). Measurements largely followed Renaud [15] and Neira [10], but eye height as opposed to eye length was measured and the length of the cloacal slit was included in tail length. In addition to the 17 measurements  [32], the width of the oral disc was measured after it had been splayed out on a glass plate. Precloacal length (pa) was calculated by adding trunk length (Lt) with prebranchial length (d-b 1 ) and branchial length (b 1 -b 7 ). The oral disc of all lampreys measured was photographed to examine its dentition. Two further morphometric ratios were calculated, oral disc length as a percentage of prebranchial length (d/ d-b 1 x 100), and tail length as a percentage of the length from the origin of the second dorsal fin to the tip of the caudal fin (Lta/ Ld 2 -c x 100). Gender was not assigned as sexual dimorphism is not apparent externally upon entry to fresh water and most individuals were released alive. In accordance with Renaud [15], morphometrics were taken on the left side of the lamprey (head pointing left) and were measured as the shortest distance point to point (Fig 2).
The number of oral fimbriae of 20 G. australis from the Waikawa River and 29 G. macrostoma from the Santa Cruz River were counted from scaled digital images using Image J software (https://imagej.nih.gov/ij/download.html). The total length of 10 oral fimbriae from each of the 49 specimens was measured; five from the antero-lateral region and five from the posterior region of the oral disc. As fimbriae were not excised from the specimens as by Lethbridge & Potter [41], their length was measured from the base of the exposed fimbriae to the tip of the longest finger-like processes. The fimbriae lengths of each lamprey were expressed as a percentage of its oral disc width.

Condition factor
The condition factor of G. australis and G. macrostoma was calculated by using Fulton's (K) index [42]. For this, the weight of each lamprey was measured to the nearest 1 g and condition factor was calculated using the formula W/TL 3 x 10 6 , where W is wet weight in g and TL is total length in mm.

Data analysis
As lamprey length varies between individuals, the various morphometrics were standardised by expressing them as a percentage of total length; except with eye height, fin height and length, and branchial and trunk depth. Note that the absence of a space between the posterior end of the second dorsal fin and origin of the caudal fin (d 2 -c) of G. macrostoma meant this metric was not available for inclusion in analyses.
All morphometrics were log 10 transformed and subjected a priori to the Shapiro-Wilks and Levene's tests to determine if they met the assumptions of normality and homogeneity of variance, respectively. One-way Analysis of Variance's (ANOVAs) were performed on log 10 transformed data for each of the remaining 22 morphometric characters from the five lamprey populations (Waikawa River, Mokau River, Murray River, Chubut River and Santa Cruz River; α � 0.05). When significant differences occurred for a given character, Tukey's Honest Significant Difference (HSD) test was used to determine pairwise differences among lamprey populations. Oral fimbriae were subjected to factorial ANOVAs to identify whether fimbriae size relative to location on the disc differed between the two Geotria species. A t-test was carried out to determine if the total number of oral fimbriae was significantly different between the two species.
Standard Discriminant Function Analysis (DFA) was used to determine the morphometrics that could best discriminate between lamprey populations [43]. DFA creates a predictive model for assigning groups, determining which morphometric measurements were the best predictors of lamprey populations based on the percentage of correctly classified individuals. Pearson Product Moment Correlations were first performed to identify co-linearity in the morphometrics measured. Highly correlated variables (>0.85) were removed and only the two variables with a correlation greater than 0.75 were used in the DFA analyses; snout to eye % TL and oral disc length % prebranchial length (0.81), as their inclusion strengthened the discriminant model. The DFA analysis was carried out on log 10 transformed data from the five immature lamprey populations (Waikawa River, Mokau River, Murray River, Chubut River and Santa Cruz River), the mature adult lamprey from the Santa Cruz River and the immature and mature adult lamprey from Chile [10]. Wilks's λ was used to compare the differences among groups, which ranges from 0 (perfect discrimination) to 1.0 (lack of discrimination). The eigenvalues, percentage of variance, and cumulative percentage of variance were calculated in this analysis. Discriminant functions or canonical roots were considered useful for explaining the data if the eigenvalues were greater than 1. The standardized coefficients of the canonical roots were determined for estimating the relative contribution of each variable to each of the roots, thus, indicating the power of discrimination for each of the selected variables. Finally, the matrix of structure factors was calculated to determine the intra-group absolute correlations between each of the variables and the canonical root. The larger standardized coefficients and the larger correlations between each variable and the root were utilized to explain the data.
For initial analyses the immature Santa Cruz lamprey were separated into two groups to determine if any morphometric factor changed during the six-week holding period. As no discrimination between the fresh run and older groups was evident, the data were pooled for subsequent analyses.
Morphometric measurements taken from 14 lamprey specimens held at the Muséum National d'Histoire Naturelle (Paris) (measured in May-June 2003) and at the Natural History Museum (London) (measured January 2010), including types and non-types, were also used in descriptive analyses to determine if they exhibit the morphometric descriptors of G. macrostoma.

Coloration, fin shape and their relative position
As with G. australis, G. macrostoma enters fresh water with two dorsal longitudinal blue-green stripes and a blue-green coloration on the upper half of its body, whereas the ventral surface is silver/white (Fig 3). A characteristic of G. macrostoma, which has not been documented for G. australis, is the iridescent blue coloration, almost fluorescent in appearance, that is present on the edges of the eyes, over the pineal gland and along the trailing edges of the dorsal fins and entire edge of the caudal fin, as well as on the fleshy tip of the tail (Figs 3 and 4). In contrast, the skin over the pineal gland is creamy/white in G. australis with no iridescent blue markings on the edges of the eyes and fins. After six weeks in fresh water, the Santa Cruz River lamprey no longer displayed the fin markings and most had lost their vibrant blue-green coloration, fading to dull brown (Fig 4).
One of the diagnostic characteristics of immature G. macrostoma is the contiguous second dorsal and caudal fins (Fig 4; [4]). In Australian G. australis, the loss of a space between the posterior end of the second dorsal fin and the origin of the caudal fin has only been documented in mature females not males [32,37]. In sexually mature New Zealand G. australis, the two fins are contiguous in both sexes (Fig 5). At sexual maturity, the dorsal fins of G. australis lose their peaked triangular shape and the apex becomes rounded (Fig 5). The rounding of both dorsal fins, and particularly of the second dorsal fin is less pronounced in mature G. macrostoma (Fig 4). Mature female G. macrostoma lack the raised dorsal ridge (rope), which

Dentition and oral fimbriae
In general, the number of cusps on each lamina of G. australis and G. macrostoma were the same. Both Geotria species possess four cusps on the supraoral lamina, three or two large cusps on the transverse lingual lamina, four or five cusps on both longitudinal lingual laminae, and 8-11 small cusps on the infraoral lamina (Fig 6). There is one enlarged darkly pigmented oral papilla on either side of the oral disc in both species (Fig 6). However, the shape of the teeth and changes undergone between immature and mature G. macrostoma differ markedly from those of G. australis.  [27,32,44]. In immature G. macrostoma key differences are: the infraoral lamina is either reduced or absent (Fig 6A-6D) whereas it is retained in G. australis throughout its spawning run [32]; the two central pointed teeth in the supraoral lamina are markedly smaller with the outer cusps not displaying the characteristic spatulate shape, but instead consist of smaller elongated cusps; the remaining teeth in the oral disc are not spatulate as in G. australis, particularly the lateral circumorals (adjacent to the oesophageal opening); and the single row of ridge-like radial plates in the posterior section of the disc in G. australis are reduced in number or absent in G. macrostoma.
In G. australis, aside from the two outer cusps of the transverse lingual lamina, the cusps of the other laminae and teeth on the oral disc become smaller and blunter as lamprey reach sexual maturity [32,44]. This reduction in tooth size is also seen in sexually mature G. macrostoma (Fig 7). Although the infraoral lamina disappears early in the spawning run, the supraoral lamina has become markedly reduced to two small triangular cusps by sexual maturity (Fig 7). In addition, the growth of the oral disc is unprecedented for any lamprey species with a width of over 100 mm in mature G. macrostoma; more than double the documented disc width for G. australis.
Another difference between the species is the number and size of the oral fimbriae. G. macrostoma possessed between 67 and 76 oral fimbriae, which was significantly more than the 53 to 64 recorded in G. australis (P < 0.0001; Table 1). The length of the oral fimbriae as a percentage of the oral disc width was also significantly smaller in G. macrostoma than in G. australis (P < 0.0001 ; Fig 8). For G. macrostoma, no significant difference was found in the size of oral fimbriae between the anterior-lateral and posterior regions (Fig 8). In contrast, the anterior-lateral oral fimbriae were significantly smaller than those from the posterior region of G. australis (P < 0.0001; Fig 8). Although only three mature adult G. macrostoma could be examined, erosion/loss of the finger-like processes was evident in all individuals with the oral fimbriae presenting as fleshy nodes (c.f. Figs 6 & 7). In contrast, mature adult G. australis retain their oral fimbriae finger-like processes throughout adult life [6,32].

Morphometrics and condition
The ANOVA's found all 21 morphometric variables and condition factor were significantly different among the five lamprey populations (Fig 9). A number of characters differed significantly between the two species at the time of their entry to fresh water. The prebranchial region and oral disc of G. macrostoma was larger (Fig 9), resulting in the prebranchial and snout to eye lengths, and length and width of the oral disc being significantly greater than in all G. australis populations (P < 0.05; Fig 9). As the cloaca of G. macrostoma is positioned further behind the origin of the second dorsal fin than in G. australis [4], G. macrostoma has a significantly smaller tail length, and tail length forms a significantly smaller proportion of the region from the origin of the second dorsal fin to the end of the caudal fin (P < 0.05; Fig 9). The precloacal region, branchial depth and trunk depth of G. macrostoma were also significantly larger than that of G. australis (P < 0.05). Differences between the species are also reflected in the  total length, weight and condition factor (Fig 9). G. macrostoma was heavier relative to total length and exhibited a significantly higher condition factor than G. australis (P < 0.05). The branchial and trunk lengths, eye height, and the height and length of the two dorsal fins did not show a clear pattern between the two Geotria species (Fig 9). However, the space between the two dorsal fins was significantly greater in G. australis than G. macrostoma (P < 0.05; Fig 9).
Within species differences were also evident, with G. macrostoma populations more heterogeneous than G. australis. G. macrostoma from the Chubut River exhibited significantly larger branchial and trunk depths and space between the two dorsal fins than those from the Santa Cruz River (P < 0.05; Fig 9). G. macrostoma from the Chubut River was also significantly longer and heavier, and with a significantly higher condition factor than lamprey from the Santa Cruz River (P < 0.05). G. macrostoma from the Santa Cruz River exhibited significantly larger oral disc length and width, snout to eye and prebranchial regions from those in the Chubut River (P < 0.05). Although the gular pouch had started to develop in Mokau River lamprey, differences in the morphometrics measured were minimal, with only the branchial region, total length and weight being significantly larger than Waikawa River lamprey (P < 0.05; Fig  9). Australian G. australis also exhibited larger trunk depths and a smaller sized oral disc compared to New Zealand populations (P < 0.05).  The Discriminant Function Analysis (DFA) consisted of eight variables: length of the first dorsal fin, height of the second dorsal fin, maximum trunk depth, oral disc length as a percentage of prebranchial length, and four characters standardised by total length; snout to eye length, tail length, space between the first and second dorsal fins and precloacal length. Of the eight variables in the DFA, all except precloacal length were found to be significant characters (P < 0.01). The DFA clearly separated immature G. australis and G. macrostoma, and mature adult individuals of G. macrostoma from the Santa Cruz River were different from all immature populations and mature adult G. australis from Chile (Wilks'λ = 0.0059, F 56,1712 = 48.62, P <0.00001). Three canonical roots were generated by the DFA with eigenvalues higher than 1. Root 1 had an eigenvalue of 7.65 explaining 57.8% of the total variance, while root 2 had an eigenvalue of 3.59 explaining 27.1% of the total variance and root 3 had an eigenvalue of 1.10 explaining 8.3% of the total variance, accounting for a cumulative proportion of 93% (Fig 10;  Table 2).
Of the seven statistically significant variables in the DFA, the largest standardized coefficient in the first canonical root was the snout to eye length as a percentage of total length (0.655), followed by maximum trunk depth (0.435) and oral disc length as a percentage of prebranchial length (0.402; Table 2). Length of the first dorsal fin was the most important variable influencing root two (0.918), followed by height of the second dorsal fin (0.313) and tail length as a percentage of total length (0.253). Maximum trunk depth was the most important variable influencing root three (-0.750) followed by oral disc length as a percentage of prebranchial length (0.466) and height of the second dorsal fin (0.283). Accordingly, the structure matrix showed that the snout to eye length as a percentage of total length and oral disc length as a percentage of prebranchial length were the variables that showed the highest correlation with the first canonical axis, whereas length of the first dorsal fin and maximum trunk depth were the most closely related variables to the second and third axis, respectively ( Table 2).
The DFA model discriminated well between lamprey populations with 84.6% of original grouped cases correctly classified (Fig 10; Table 3). Incorrect classifications were only found within species, with no G. macrostoma classified as G. australis and vice versa ( Table 3). The Mahalanobis distances were significantly different between all lamprey populations (P < 0.025). Comparing the six immature populations, Mahalanobis distances also indicated a stronger separation between the G. australis and G. macrostoma species (25.4-57.8) than different populations of the same species (1.0-16.5 for New Zealand, Australia and Chile G. australis, and 7.7 for Santa Cruz and Chubut G. macrostoma). Mature adults from the Santa Cruz River also displayed a strong separation from mature G. australis from Chile (195.4).
Overall, DFA results indicated that G. macrostoma from the Santa Cruz and Chubut rivers had significantly larger snout to eye lengths relative to total length, larger trunk depths and larger oral discs relative to the length of the prebranchial region than G. australis from New Zealand, Australia and Chile. G. macrostoma from the Chubut River also had a significantly larger trunk depth, and a longer first dorsal fin than those from the Santa Cruz River (Figs 9 &  10). The difference and variation in size of the snout to eye and oral disc regions in G. macrostoma is illustrated by Figs 6 and 11. By sexual maturity, G. macrostoma had significantly larger snout to eye lengths relative to total length than G. australis. In addition, the length of the first dorsal fin of sexually mature G. macrostoma and G. australis was significantly shorter than in immature adults.
The key morphometrics discriminating G. macrostoma from G. australis are also highlighted by comparing immature lampreys in the present study with sexually mature G .   Fig 9. Plots of the 21 morphometric characters and condition factor. For each character, the mean ± 95% confidence interval is displayed. All characters were significantly different among the five lamprey populations (NZW: Waikawa River, New Zealand; NZM: Mokau River, New Zealand; Au: Murray River, Australia; ArS: Santa Cruz River, Argentina; ArC: Chubut River, Argentina). Error bars that do not overlap depict significant differences between lamprey populations (P < 0.05). Abbreviations: TL: total length; CF: condition factor; W: width. All measurements are in millimetres and weight is measured in grams. Note d 2 -c (space between the posterior end of the second dorsal fin and origin of the caudal fin) is not graphed as it is absent in G. macrostoma. https://doi.org/10.1371/journal.pone.0250601.g009

PLOS ONE
Discrimination of Geotria species using morphometric analyses australis from Western Australia measured by Potter et al. [32]. As raw data from Potter et al. [32] were unavailable, analyses are limited to descriptive comparisons of means and 95% confidence limits. The prebranchial length of mature male and female G. australis was, on average, respectively 13.2 and 12.1% of their total length in Australia [32] and 15.5% of their total length (sexes combined) in Chile, whereas the mean prebranchial length of G. macrostoma (sexes combined) measured 12.0-12.9% of their total length upon entry to fresh water ( Table 1). As G. macrostoma becomes sexually mature, the prebranchial length measures up to 23.7% of their total length ( Table 1).
The oral disc of G. macrostoma also undergoes greater growth through to maturity than that of G. australis. Upon entry to fresh water, the oral disc length of G. australis averaged 2.8-3.8% of their total length (Table 1), whereas the oral disc length of G. macrostoma averaged 5.3-6.3% of their total length (Table 1). Once G. australis reaches sexual maturity, Potter et al. [32] found the average oral disc length had increased to around 5% of their total length, with  Chilean adults reaching 8.7%, whereas the oral disc length of the three mature G. macrostoma was recorded as 9.7%, 10.9% and 16.1% of their total length (Table 1). Condition factor was also a significant measure distinguishing G. macrostoma. Potter et al. [32] found across seven years (1976)(1977)(1978)(1979)(1980)(1981)(1982) the mean condition factor of male and female G. australis captured during the first four months of their spawning run was between 0.79 and 0.99, respectively. These values are similar to that recorded for G. australis in the present study (Fig 9). However, the mean condition factor for G. macrostoma was significantly higher at 1.15 and 1.28 for Santa Cruz and Chubut River lamprey, respectively (Fig 9).
A comparison of morphometric characters between immature and mature G. macrostoma and museum specimens of G. australis, including its holotype, and those of G. allporti Günther, 1872, G. saccifera Regan, 1911, and Velasia chilensis Gray, 1851 (synonyms of G. australis) was also undertaken (Table 1). Although specimens are labelled velasia (immature) or adult, this was based on external characteristics and the exact timing since entry to fresh water is unknown. Six of the eight immature individuals possessed ridges of epithelium flanking the labial teeth (Fig 12) and in five of the six mature adults these were absent (Fig 13); with the other mature adult still showing remnants (Table 1). Of the eight specimens classified as immature, in seven individuals the prebranchial length as a percentage of total length exceeded that recorded in fresh run New Zealand, Australian and Chilean G. australis, fitting within the range in G. macrostoma (Table 1). Similarly, for the snout to eye length as a percentage of total length, five of eight specimens exceed that recorded in fresh run G. australis in the present study (Table 1). At the immature stage, the tail length as percentage of total length of G. australis should be greater than in G. macrostoma based on the cloaca being located further forward. However, four of the eight immature G. australis specimens had relative tail lengths smaller than that measured in any of the six lamprey populations (Table 1). Therefore, it is likely that the variability recorded in the morphological characters across the museum specimens demonstrates the growth of the oral disc, prebranchial and snout to eye lengths in G. australis paired with the shrinking of total length during the protracted spawning run (Table 1). However, the key morphometrics (oral disc length, and prebranchial and snout to eye lengths as percentages of total length) of the two G. australis specimens from the Negro River, Argentina, fell within the range of those seen in G. macrostoma in the present study and outside of the range recorded for G. australis (Table 1). In particular, the oral disc length (absolute and as a percentage of total length and prebranchial length) were larger than in all other immature G. australis museum specimens.
The number of oral fimbriae and papillae were also diagnostic in differentiating the specimens. In the two individuals from the lower Negro River, Argentina, the fimbriae and papillae counts were 67-71 and 23-24, respectively (Table 1). In comparison, the six G. australis that included the holotypes of G. australis, G. saccifera, Velasia chilensis, and non-type adults (North Island, New Zealand and South Australia, Australia), the fimbriae and papillae counts were 50-60 and 16-19, respectively ( Table 1). The range of oral fimbriae in the two lower Negro River specimens matches that seen in immature G. macrostoma from the Santa Cruz River (Table 1).
A comparison of the three sexually mature G. macrostoma with mature museum specimens highlights the extensive growth of the key morphometrics (oral disc length, and prebranchial and snout to eye lengths) through the spawning run, relative to that seen in G. australis (Table 1). Of the six mature museum specimens, oral disc absolute length and prebranchial length as a percentage of total length only overlapped with the values in one and two mature G. macrostoma specimens, respectively (Table 1). For the snout to eye length as a percentage of total length and oral disc length as a percentage of total length, no mature G. australis specimen exhibited values as high as those seen in mature G. macrostoma.

Discussion
This study has highlighted a range of morphometric and physical characteristics that discriminate between Geotria in Patagonian rivers and those found in Australasia and Chile (Table 4). According to Renaud [15], the taxonomy of lampreys is based primarily on the dentition in the adult stage, and the present study identified several characteristics that can be used to discriminate the two Geotria species during the immature adult stage (Table 4). In G. macrostoma, the infraoral lamina is greatly reduced or absent, the supraoral lamina does not display the spatulated outer cusps, the remaining teeth of the oral disc are pointed rather than spatulated (particularly the lateral circumorals), and the posterior ridge-like radial plates present in G. australis are absent. G. macrostoma also exhibits an iridescent blue/green coloration to the edges of the eyes, fins and over the pineal gland; however, this coloration fades soon after entry to fresh water. Although not diagnostic, G. macrostoma displays greater growth of the prebranchial region and oral disc and has a deeper body depth and higher condition factor.
The results of this study add to the molecular and morphological differences documented by Riva-Rossi et al. [4] between Patagonian lamprey and those from Chile and Australasia, to validate that G. macrostoma is a distinct species. There are numerous diagnostic characters that distinguish the two Geotria species as they enter fresh water as immature adults; the position of the cloaca and the contiguous second dorsal and caudal fins (described by Riva-Rossi et al. [4]), and four related to the dentition and coloration (found in the present study). The fact that the cloaca of adult G. macrostoma is located posterior to, rather than under the origin or in front of the second dorsal fin as in G. australis, parallels the difference found between ammocoetes of Geotria from Argentina and those from Chile, Australia and New Zealand [13]. The ammocoetes measured by Neira et al. [13] from the Limay River in Argentina most likely represented G. macrostoma and those from the other regions G. australis. Therefore, the difference in the cloaca position in ammocoetes will distinguish between the two Geotria species at the larval stage, for which there are fewer well-defined morphological distinctions among lamprey species compared to adults. The oral fimbriae and papillae were also diagnostic characteristics between G. australis and G. macrostoma. Although both species exhibited a large gap devoid of oral papillae along the posterior aspect of the oral disc perimeter, the number of papillae in G. macrostoma (23)(24) exceeded that of G. australis in the present study (16)(17)(18)(19) and those previously reported for Australasia. In seven G. australis from Australia, Khidir and Renaud [46] reported counts of 16-19 oral papillae, while Maskell [27] reported a range of 15-22, usually 18, for the species in New Zealand. G. macrostoma also possessed higher numbers of oral fimbriae than G. australis. The counts of 67-76 recorded in the present study are similar to the 72-74 reported for the holotype of G. macrostoma by Burmeister [17]. The lower number of oral fimbriae recorded in New Zealand G. australis in the present study (53-64) fit within the range documented for Australian populations, where Khidir and Renaud [46] reported counts of 55-65 and Lethbridge and Potter [41] reported counts of 50-68. The size of the oral fimbriae relative to location on the disc also differed between the Geotria species. In line with the findings of Lethbridge and Potter [41] for Australian G. australis, the fimbriae in the posterior of the disc in New Zealand G. australis were significantly larger than those in the anterior and lateral regions. In contrast, G. macrostoma did not exhibit detectable differences in fimbriae size between disc regions.
Collectively, the present data and investigations by Nardi et al. [16] and Riva-Rossi et al. [4] support the resurrection of G. macrostoma as a distinct species found in Argentina, with historical records of only two individuals from Uruguay [19] and one adult individual in the San Juan River, a coastal stream flowing into the Chilean side of the Strait of Magellan, on the extreme southern tip of South America [30]. Based on a sole specimen, the presence of G. macrostoma in southern Chile is questionable. Further assessments of ammocoetes from rivers flowing into the Magellan Channel and/or the Pacific Ocean are warranted to confirm its presence in Chilean Patagonia. Presently, data indicate that G. macrostoma is likely to be an endemic species to the Patagonian region of Argentine including their South Atlantic islands.
Petromyzon macrostomus was first described by Burmeister [17] with Berg [19] re-assigning the species as Geotria macrostoma (Burmeister, 1868), adding to the description with a second specimen collected near the island of Flores, off Montevideo, Uruguay. The original description was based on a single specimen of 400 mm total length collected on 26 Sept. 1867 from a street in Buenos Aires. The oral disc is 60 mm in length (oral disc length as percentage of total length, 15) and 80 mm in width. The oral fimbriae number 72-74. A 40 mm long gular pouch reaches the first branchial opening. The snout to eye length is 70 mm (snout to eye length as percentage of total length, 17.5). There are two triangular-shaped dorsal fins separated from each other by 20 mm. The cloaca is under the anterior part of the second dorsal fin and 70 mm from the tip of the caudal fin (tail length as percentage of total length, 17.5). The Patagonian lamprey in this study (Chubut and Santa Cruz rivers) fits with the original description of Geotria macrostoma, thereby confirming the former's identity. Examination of historical descriptions and material suggest museum specimens from the Negro River, Argentina should also be re-identified as G. macrostoma.
The marked morphological changes that occur in Geotria during the protracted spawning run has led to a longstanding unresolved taxonomy of the genus. Comparison of morphometric measures between adult G. macrostoma and G. australis revealed that both Geotria species are characterized by significant morphological changes through the spawning run with the growth of the oral disc and snout to eye lengths (relative to total length) of sexually mature G. macrostoma exceeding that seen in G. australis and all other lamprey species [15,32]. Morphological differences both among and within lamprey species have been linked to differences in behavior, which may affect ecological processes. For example, Pacific lamprey with a shorter distance between the first and second dorsal fins (that were closer to sexual maturity) were more likely to use refuges during passage at Bonneville Dam [47] and had lower passage success in an experimental vertical slot fishway [48]. The present study has identified marked differences between the two Geotria species, especially regarding growth of the oral disc. In particular, in sexually mature G. macrostoma the finger-like processes of the oral fimbriae are lost/eroded whereas G. australis retain the full structure throughout its adult life [6,32]. Lethbridge and Potter [41] suggested that the fimbriae are linked to creating an effective seal on surfaces, not only for feeding and migration but also during nest construction and spawning. As the teeth of G. australis reduce in size and sharpness during maturation, Lethbridge and Potter [41] speculated that lamprey may become more dependent upon the fimbriae to aid attachment during reproduction. Although conjecture, the loss/erosion of the oral fimbriae finger-like processes and the excessive growth of the oral disc in mature adult G. macrostoma could indicate ecological differences between the Geotria species during migration and breeding. Further investigations are needed to fully understand how morphological differences between the Geotria species affect behavior and ecology.
Within the two Patagonian lamprey populations examined in the present study, differences in characters recorded may be the result of differences in the parasitic oceanic phase. G. macrostoma collected from the Chubut River were significantly larger and heavier with deeper trunks and less growth of the head region than lamprey in the Santa Cruz River. These differences could relate to the time spent at sea and timing of entry to fresh water. In the Chubut River, adult G. macrostoma enter the river during fall, similar to that observed in Australasian G. australis [6,7,49]. In contrast, in the Santa Cruz River, G. macrostoma initiated river entry during summer (December through February), similar to Chilean G. australis [10] and Northern Hemisphere lamprey species [50,51]. Further investigations are necessary to understand the migration patterns of G. macrostoma and G. australis across South American rivers and how these relate to morphometric variation within each species.
Alternatively, the morphological discrimination between the two G. macrostoma populations may be the result of population structure within the species. Using morphological characters and heart fatty acid signatures, Lança et al. [52] suggested three separate stocks existed in sea lamprey (Petromyzon marinus) populations in Portugal possibly based on seabed topography and geographical separation of oceanic host species off the western Iberian Peninsula. From morphological characteristics, Vatandoust et al. [53] also suggested two independent populations of Caspian lamprey (Caspiomyzon wagneri) had formed across two major rivers flowing into the Caspian Sea basin. Based on the findings of the present study, further molecular and morphological investigations of G. macrostoma from within its range are warranted to verify if population structure is occurring within the species.

Conclusions
The present study and investigations by Nardi et al. [16] and Riva-Rossi et al. [4] support the resurrection of G. macrostoma as a distinct species inhabiting the major Patagonian basins.
Upon entry to fresh water, key morphometric and physical characteristics that discriminated adult G. macrostoma from G. australis were several differences in the dentition and oral papillae and fimbriae. In addition, the edges of the eyes, fins and over the pineal gland possess an iridescent blue/green coloration only in G. macrostoma. Similar to G. australis, G. macrostoma is also characterized by significant developmental changes through its spawning run, and by sexual maturity the growth of the oral disc exceeded that recorded in any lamprey species. Presently, many anadromous lamprey species are threatened or in decline from a number of anthropogenic pressures. Currently, all ecological knowledge of G. australis is based on Australasian populations, which may not be applicable to G. macrostoma. To ensure the conservation and protection of the Patagonian lamprey further investigations are needed to understand its life history, biology and ecology throughout its range.