Eugenol and Lippia alba essential oils as effective anesthetics for the Amazonian freshwater stingray Potamotrygon wallacei (Chondrichthyes, Potamotrygonidae)

This study assessed the potential of eugenol and the essential oil of Lippia alba (EOLA) in providing suitable anesthetic induction and recovery times, and their consequent effects on the blood and respiratory physiology, as well as the gill architecture of an Amazonian freshwater stingray, Potamotrygon wallacei, shortly after reaching the recovery and 48 h later. Juveniles of P. wallacei were exposed to increasing concentrations of eugenol (75, 100, 125, and 150 µL L−1) and EOLA (150, 175, 200, and 225 µL L−1) in an immersion bath. Anesthetic induction was found to be faster with the use of eugenol compared to EOLA. On the other hand, the stingrays anesthetized with eugenol displayed a longer recovery time than those exposed to EOLA. The highest concentrations of eugenol caused moderate to severe histological changes in the gills. No significant changes were found for hematocrit and plasma energy metabolites in the stingrays anesthetized with all concentrations of both eugenol and EOLA shortly after reaching the recovery from the ansthesia, when compared to those recovered after 48 h. Investigations regarding the potential use of these natural anesthetics are unprecedented for freshwater stingray species, and 200 μL L−1 EOLA is recommended as the most suitable anesthetic for use in juveniles of P. wallacei.

was found to be faster with the use of eugenol compared to EOLA. On the other hand, the stingrays anesthetized with eugenol displayed a longer recovery time than those exposed to EOLA. The highest concentrations of eugenol caused moderate to severe histological changes in the gills. No significant changes were found for hematocrit and plasma energy metabolites in the stingrays anesthetized with all concentrations of both eugenol and EOLA shortly after reaching the recovery from the ansthesia, when compared to those recovered after 48 h. Investigations regard-

Introduction
Regardless of the experience of researchers, physiologists, and other professionals in activities that involve the restraint and immobilization of elasmobranch fishes, the required ability to perform scientific assessments, surgical procedures, or collection of biological samples is not a trivial practice. Due to the high risk of accidents, the use of anesthetics is a fundamental procedure to provide safe handling and mitigate the consequent stress to the animal. The use of anesthetics during research practices allows the manipulation of these animals more safely since the presence of stingers with toxins on their tails generates dangerous injuries and an intense nociceptive response in the handler (Haddad Jr. et al. 2004;Silva et al. 2015Silva et al. , 2020Kimura et al. 2018).
Anesthetics are also used in fish during management procedures to reduce the secondary responses to stress (Marshall et al. 2012;Husen and Sharma 2014). Cortisol is the main parameter for assessing the primary stress response in teleost fish (Wendelaar-Bonga 1997;Balasch and Tort 2019), while in elasmobranchs, this role is attributed to 1-α hydroxycorticosterone, a steroid hormone produced by interrenal cells (Idler and Truscott 1966;Anderson 2012;Lambert et al. 2018). Although Ruiz-Jarabo et al. (2019) recently revealed the glucocorticoid action of this hormone in an elasmobranch species (Scyliorhinus canicula) submitted to air exposure, the measurement of this hormone is still complicated (Wheaton et al. 2018), limiting its use in routine stress monitoring practices. Thus, other parameters like corticosterone and plasma metabolite levels like glucose are also used as physiological indicators of stress in this animal group (Brinn et al. 2012;Marshall et al. 2012;Ariotti et al. 2021).
In recent years, the search for anesthetics of natural origin as an alternative to replace chemical anesthetics has intensified, for both teleost (Hoseini et al. 2019;Souza et al. 2019a) and marine elasmobranch species (Silbernagel and Yochem 2016;. Eugenol is the major component (80-90%) of the oil initially extracted from the clove Syzygium aromaticum, a plant species that originated in Indonesia and spread worldwide due to its plentiful medicinal and pharmacological properties (Lorenzini and Matos 2008;Ascenção and Filho 2013). Considering its natural source, low cost, and anesthetic efficiency, eugenol has become one of the main options for bath anesthesia and sedation in teleost fish (Hoseini et al. 2019;Souza et al. 2019a). In anesthetic trials using eugenol for the marine ray Zapteryx brevirostris, safe induction and good recovery times were observed ). On the other hand, the exposure of marine sharks, i.e., Triakis semifasciata and Mustelus californicus, to AQUI-S 20E®, which contains 10% pure eugenol in its formulation, resulted in longer induction and recovery times in these animals compared to a teleost species investigated in the same studies (Silbernagel and Yochem 2016;Lécu et al. 2017). These previous findings indicate that the observed responses to the same anesthetic type may differ in the distinct elasmobranch species investigated so far.
The essential oil of Lippia alba (EOLA) comes from an aromatic shrub belonging to the Verbenaceae family. It is a tropical plant species found throughout South and Central America and Africa and with different chemotypes (Hennebelle et al. 2008). The use of increasing concentrations of EOLA linalool chemotype proportionally decreased the induction time to anesthesia in the seahorse Hippocampus reidi (Cunha et al. 2011), the Nile tilapia Oreochromis niloticus (Hohlenwerger et al. 2016), and the silver catfish Rhamdia quelen (Cunha et al. 2010;Toni et al. 2014). However, to date, there are no studies demonstrating the anesthetic efficiency of EOLA in elasmobranchs. Although there are a great number of studies involving anesthesia in teleost fishes, most of them looks to the establishment of induction and recovery times, not taking in consideration the possible adverse effects and the consequent physiological responses evoked in these animals to restore homeostasis.
Potamotrygon wallacei Carvalho et al. (2016) (cururu stingray) is the smallest species of Amazonian freshwater stingray known so far (maximum disc width of 310 mm) and endemic to the middle Negro River Basin. This species is resistant to acidic waters with low conductivity and dissolved oxygen levels (Oliveira et al. 2017;Carvalho et al. 2016), and its rusticity, small size, and a high level of individual polychromatism are additional attributes that make this species the second most exported potamotrygonid species in the Amazon region to the international ornamental fish trade (Carvalho et al. 2016).
Thus, this study aimed to assess the potential of eugenol and EOLA in providing suitable anesthetic induction and recovery times, as well as their consequent effects on the respiratory physiology, gill morphological features, and survival of P. wallacei, immediately after reaching recovery from the ansthesia, and after 48 h of recovery to both anesthetics.

Material and methods
Capture and acclimatization of animals Juveniles of P. wallacei (disc width between 120 and 140 mm; Araújo 1998) were obtained by professional fishermen between December 2017 and February 2018 through night catches in the Igarapé Daraquá (0°30′23.4504″S, 63°12′51.3″W), located in the Mariuá Archipelago, near the Municipality of Barcelos, Amazonas, Brazil. Immediately after capture, the stingrays were individually kept in plastic monoblocs containing freshwater from the capture site and taken to the basecamp, where they were temporarily stored in 1-m 3 net tanks (n = 6-8 per tank) fixed in the river, containing sand at the bottom for greater animal comfort. Subsequently, the stingrays were transported to the city of Barcelos, then transferred to different monoblocs containing 10 L of fresh and oxygenated water from the Rio Negro and transported by boat (~ 24 h) to the city of Manaus. From there, they were transported to the Laboratory of Experimental Physiology and Behavior of Aquatic Animals (LEFCAQ), at the Federal University of Amazonas (UFAM), and acclimated for 15 days (Ariotti et al. 2021) in plastic pools (400 L) provided with constant aeration and water renewal from an artesian well. During this period, the water quality parameters in the acclimatization pools were monitored by using an Orion Five Star multiparametric device (Thermo Scientific Inc., USA), and the following values (mean ± SD) were obtained: electrical conductivity = 8.23 ± 0.2 μS cm −1 , dissolved oxygen = 5.03 ± 0.7 mg L −1 , pH = 5.20 ± 0.4 units, and temperature = 26.01 ± 0.5 °C. The stingrays were fed daily with 2-3% of body weight with live earthworms and shrimps, and small pieces of fresh fish. Unconsumed food and fecal residues were removed by siphoning daily, and freshwater was restored from the same source.
The use of juvenile specimens was defined based on ICMbio regulations (IN N o . 204/2008) that determine neonates and this size class for commercialization at the international ornamental fish trade.
Obtaining eugenol and Lippia alba essential oils The eugenol used in the tests (99-100%; Biodynamics, Ibiporã, PR, Brazil) was acquired from the local market. The essential oil of L. alba (EOLA) was obtained from leaves from fresh plants cultivated at the campus Frederico Westphalen from the Federal University of Santa Maria (UFSM), Rio Grande do Sul, Brazil, by hydrodistillation in Clevenger apparatus for 3 h, as recommended by the European Pharmacopoeia (2007). Linalool was the major chemotype component (Souza et al. 2019b). Aliquots of EOLA (20 mL) were stored in separate amber glass jars, sealed, and stored at -4 °C until the experiments were initiated.

Pilot studies
Considering the lack of data in the literature on the use of eugenol in freshwater stingrays, pilot studies were conducted to refine the range of variation of acceptable eugenol levels for P. wallacei. In these experiments, juveniles (n = 3 for each concentration) with a disc width of 110.7 ± 36.5 mm and weight of 109.5 ± 30.7 g were exposed to increasing concentrations of eugenol, i.e., 50,75,100,125,150,200, and 300 µL L −1 , established from previous studies conducted with marine stingray and shark species (Stamper 2004;Grusha 2005;Takatsuka et al. 2019).
Similar procedures were applied with the use of EOLA, in which juvenile stingrays (n = 3 for each concentration) were exposed to increasing concentrations of 50, 100, 150, 200, 300, and 500 μL L −1 . Since there are no reports of studies addressing the use of EOLA as an anesthetic substance for any elasmobranch species, either marine or freshwater, this range was defined considering the data registered for different teleost species (Cunha et al. 2010;Toni et al. 2014;Hohlenwerger et al. 2016). Due to the weak solubility of eugenol and EOLA in water, these substances were diluted in ethanol (1:10, v/v) immediately before use and kept at room temperature.
To perform these experiments, stingrays were individually transferred from the acclimatization pools to glass aquariums (68.0 × 38.5 × 29.5 cm) containing 40 L of fresh solution with the different tested concentrations for both anesthetics. The stingrays were fasted 24 h before the start of the experiments to avoid regurgitation and possible branchial obstruction. The water used in the experimental aquariums was the same as used in the acclimatization pools and was fully renewed at the end of each tested concentration.
The different stages of anesthesia proposed for marine elasmobranch species (Stamper 2004) were applied to the cururu stingray for the definition of the anesthetic plan (Table 1). During anesthetic exposure, the ventilatory (spiracular) frequency, swimming capacity, and tonus and reflex muscle activity of each animal were evaluated and recorded in a minute-by-minute interval. When the animal stopped moving and showed a marked reduction of spiracular frequency, it underwent stimulation by means of slight touching with a glass stick in the anterior and posterior regions of the disc edge. When it no longer responded to stimuli with the glass stick, and to make sure that it was in a stage of deep anesthesia, the animal had its body gently turned to the supine position (belly upwards) with the aid of the glass stick ( Fig. SI-1 (Supplementary information)). In those cases, the confirmation that the stingray had reached a plane of deep anesthesia (stage III, Table 1) was assumed from the observed instant it was unable to reverse to the normal body position. This was considered a fundamental procedure since the stingray routinely remains at the bottom of the aquarium and the loss of balance in the water column often used in teleosts to define the anesthetic event (Sneddon 2012;Zahl et al. 2012) is not applicable to these animals and may even induce errors during the determination of the real anesthetic times of induction and recovery.
Immediately after reaching the level of deep anesthesia and recording the time, the stingray was transferred to another aquarium with the same size and volume containing anesthetic-free water. During the recovery phase, the same physiological and behavioral parameters of induction were recorded. Recovery was considered complete when the animal showed stable and guided swimming, avoiding the walls of the aquarium. For this evaluation step, the five recovery stages established by Grusha (2005) for the marine stingray Rhinoptera bonasus were adopted to juveniles of P. wallacei (Table 2).

Anesthesia protocols
Based on the results obtained in the pilot studies, the anesthetic concentrations of 75, 100, 125, and 150 µL L −1 and 150, 175, 200, and 225 μL L −1 were defined for eugenol and EOLA, respectively. Twelve stingrays were exposed to each concentration for both anesthetics, where six stingrays (n = 6) were sampled at each concentration shortly after reaching the recovery, and the other six animals (n = 6) were assessed 48 h after the exposure to anesthesia. Considering the two anesthetics evaluated, a total of 96 stingrays were used in the final experiments. They were transferred individually from the acclimatization pools to the experimental glass aquariums (the same as used in pilot studies) with the different tested anesthetics concentrations for the assessment of induction times. Immediately after, the same animal was transferred to another aquarium with free-anesthetic water and the respective recovery time interval was recorded. Once the animal reached the stage V of recovery (Table 2), it was removed from the aquarium, immobilized, and gently placed in dorsal decubitus with a moistened cotton towel covering its body to avoid skin dryness. Thus, a blood sample was obtained by puncturing the branchial vessel with a 3-mL syringe containing 10% EDTA as the anticoagulant without the use of additional anesthetic. Immediately after, the stingray was euthanized by sectioning the spinal cord, and biometric measures (weight, length, and disc width) and the necropsy for removing the gill apparatus (for further histological evaluation) was performed.
The stingrays sampled after 48 h for evaluating the physiological responses, in addition to monitoring the resumption of normal food intake and survival after exposure to the concentrations of both anesthetics, were submitted to the same protocols mentioned above, and no anesthetics were used before the blood sampling and euthanizing procedures. It is important to emphasize that the time interval involved from the moment that the stingray was removed from the recovery aquarium to the end of blood sampling and euthanasia procedures was less than 2 min (Lawrence et al. 2018) for all the tested concentration for both anesthetic agents.

Spiracular frequency
The ventilatory responses of P. wallacei juveniles exposed to both anesthetics were estimated through the frequency of spiracular movements (SF; beats min −1 ). Measurements were made according to Graham et al. (1990) by direct counting of the number of spiracle movements in the first 30 s of each 1-min Cessation of voluntary swimming, total loss of reactivity to visual and tactile stimulation, slight decrease in spiracular rhythm and muscle tone II Slight narcosis Excitation phase, erratic and uncontrolled swimming, partial loss of muscle tone, irregular spiracular rhythm, reactive only to strong stimulus, responds to position changes Deep narcosis Stingray ceases to respond to changes in position, decreased spiracular rhythm and muscle tone, reactive to strong stimulus III Slight anesthesia Total loss of swimming balance and muscle tone; even greater decrease in spiracular rhythm; suitable for minor surgical procedures Deep anesthesia Total absence of reaction, even to strong stimulus; very low spiracular rhythm; total loss of all reflexes IV Medullar collapse Respiratory failure and eventual death interval. Twelve stingrays (n = 12) had their SF assessed in the anesthetic-exposed (onset and deep anesthesia) and recovery treatments of eugenol and EOLA, and six stingrays (n = 6) in the basal (not exposed and not handled) group that were resting in the acclimation pools before to be used in the experiments. Two consecutive counts of spiracular movements were performed in the same animal at each minute interval for a greater accuracy, and the average was considered for quantification of ventilatory rate.

Gill histology
The gills of each individual stingray were fixed in 10% buffered formaldehyde, and part of the second branchial arch filaments (Duncan et al. 2010(Duncan et al. , 2011 was separated and gradually dehydrated in alcohol series (70 to 100%), diaphonized with xylol, and embedded in paraffin at 60 °C. Then, 5-μm sections were obtained using a Leica microtome (model RM 2125RT, Leica Microsystems, Wetzlar, Germany), stained with hematoxylin-eosin and assembled for further analysis and photographic documentation, using a Leica DM500 optical microscope with an integrated Leica ICC50 W camera (Leica Microsystems, Wetzlar, Germany). The observations made by Karan et al. (1998) for common carp exposed to copper sulfate were followed to identify and classify histopathological changes in the gills of P. wallacei.

Statistical analysis
All data were previously checked for homoscedasticity and normality by the Levene and Shapiro-Wilk tests, respectively. Data regarding the spiracular frequency and water quality parameters were submitted to one-way ANOVA and Tukey posteriori test, or Kruskal-Wallis one-way ANOVA on ranks followed by Dunn's posteriori test for non-homoscedastic data. Hematocrit and plasma metabolite values obtained from stingrays just after the recovery for both eugenol and EOLA were compared to those after 48 h through a Student's t-test. The statistical analyses were performed with the software SigmaPlot version 11.0, and the significance level was accepted at p < 0.05.

Anesthesia with eugenol
During the pilot studies, stingrays exposed to 50 µL L −1 of eugenol took more than 10 min to anesthetize and more than 20 min to recover, and two of the three stingrays exposed to this concentration died during the recovery phase. In stingrays exposed to 175 µL L −1 of eugenol, remarkable burns were observed at the edge of the disc (Fig. SI-2 (Supplementary information)), while at 200 µL L −1 , the animals ceased spiracular movements and from the three stingrays exposed to this concentration, two died due to pronounced respiratory depression. Most stingrays showed hyperactive behavior at the first contact with the four evaluated concentrations of eugenol (75, 100, 125, and 150 µL L −1 ). Initially, there was an increase in the rate of spiracular movements (0-1 min), followed by eye and spiracle closure, increased movement of fins and sudden swimming movements. Moments after (1-2 min), the stingrays began to calm down, reaching stage I (sedation). Soon after (2-4 min), they reached stage II (excitation), again showing sudden swimming movements and irregular breathing, but at greater intensity. After this period, stingrays again decreased their fin movements and exhibited increased respiratory effort, with increased amplitude of spiracular opening movements. The marked depression of the ventilatory frequency, associated with the total absence of disc movements and reflex activity (when stimulated with the glass stick), demonstrated that the animal had reached stage III (anesthesia), on average between 4 and 5.27 min after the first contact with eugenol (Table 3).
The first contact reactions were more intense at 125 and 150 µL L −1 . Stage I (sedation) was fast or almost imperceptible, followed by stage II with hyperexcitation, uncontrolled swimming, then rapid and deep ventilatory depression, culminating in some cases with the cessation of spiracular movements. These animals presented marked hyperemia in the region of the mouth, belly, and genitalia. The death of one stingray was recorded during the recovery phase of anesthesia at 100 µL L −1 , and two others during exposure to 150 µL L −1 of eugenol. No significant differences were found in the induction times between the concentrations of 100, 125, and 150 µL L −1 of eugenol, except that 75 µL L −1 required more time to induce anesthesia. The recovery time from the anesthesia was, on average, between 15 and 17.6 min, with no significant differences between all tested eugenol concentrations.
Anesthesia with Lippia alba essential oil (EOLA) During the pilot studies, the stingrays exposed to 50 and 100 μL L −1 EOLA (n = 3) did not reach anesthesia after 10 min of exposure. At 300 μL L −1 , both tested animals (n = 2) showed neurological sequelae after exposure (escape from offered food and irregular swimming movements 12 h after recovery), while the exposure to 500 μL L −1 EOLA led to the death of two exposed stingrays (n = 2) due to profound ventilatory failure during the induction phase. Taking the results for these highest concentrations of EOLA in consideration, and for ethics reasons, the use of a third animal in the tests was unnecessary and, thus, not applied.
The stingrays took, on average, between 4.7 and 12.4 min to reach the different anesthetic stages (Table 3) when exposed to 150,175,200, and 225 μL L −1 , and presented a mild reaction at the first contact with the anesthetic, with slight excitatory phase (stage II) and no evidence of ventilatory failure or mortality in any of the tested anesthetic concentrations. The mean recovery times for EOLA were between 8 and 13.3 min for all tested concentrations. Stingrays anesthetized with 200 μL L −1 showed significantly faster recovery than those anesthetized with 150 μL L −1 (p < 0.05) ( Table 3).
In order to evaluate the resumption of normal food intake by the stingrays after the contact with both anesthetics, feeding was provided 12 h after the anesthetic return and then again after 12 h. For stingrays anesthetized with eugenol at the concentration of 75 μL L −1 (n = 6), three of them resumed eating within the first 12 h and the others only 24 h after anesthesia. At the other tested eugenol concentrations, all animals only resumed eating 24 h after. On the other hand, all stingrays exposed to all tested concentrations of EOLA resumed eating within the first 12 h.

Spiracular frequency
Under basal conditions, the SF of the stingrays maintained in the acclimation pools ranged from 36 to 60 beats min −1 . Significantly higher SF values (p < 0.05) were observed in the groups of stingrays submitted to different concentrations of eugenol during the onset of anesthesia when compared to basal group, with exception of those exposed to 125 µL L −1 (Fig. 1A-D). No significant differences were observed between stingrays assessed in the deep anesthesia and recovery phases at the different concentrations of eugenol (p > 0.05), even compared to the basal group (Fig. 1). Stingrays evaluated at the onset of anesthesia with EOLA only showed significantly higher SF (p < 0.05) than the basal group at 150 μL L −1 . On the other hand, significantly lower values of SF were observed during the deep anesthesia stage compared to the onset of anesthesia in all the tested concentrations, except for 225 μL L −1 . Also, stingrays exposed to 150 μL L −1 and 225 μL L −1 EOLA presented higher SF values (p < 0.05) in the recovery stage compared to the deep anesthesia ( Fig. 2A-D).
During anesthetic induction at the tested concentrations with eugenol, the SF remained in a higher range, when compared to those recorded for EOLA-anesthetized animals. The SF time-response was not coincident to the tested concentrations of eugenol at the anesthetic induction, with a sudden decrease in the last minutes, and a fast return of stable SF values to 30-50 beats min −1 during the recovery phase (Fig. 1E). On the other hand, the decrease of SF in the stingrays exposed to EOLA was gradual with a regular distribution throughout the induction phase and proportional to the tested concentration, i.e., higher concentrations induced a more pronounced decrease in SF and a longer return time to recovery. Also, as the stingrays recovered, the SF gradually returned to values of 30 to 50 beats min −1 , and in all tested concentrations the range of variation between animals was lower than in those anesthetized with eugenol (Fig. 2E).

Blood parameters
No significant differences were found (p > 0.05) for hematocrit and plasma metabolite levels assessed in stingrays exposed to all concentrations of eugenol shortly after reaching the recovery stage and after 48 h of recovery (Table 4). Similarly, these parameters did not change (p > 0.05) in stingrays exposed to all concentrations of EOLA when the group sampled shortly after reaching the recovery stage was compared with those sampled after 48 h of recovery (Table 5).

Gill histology
No histopathological changes were observed in the gill primary and secondary lamellae in juveniles of P. wallacei exposed to anesthetic-free water. These structures have a simple pavement epithelium and mucous cells supported by cuboid pillar cells, and groups of ionocytes (chloride cells) at their base (Fig. 3). A qualitative description of gill lesions found in stingrays anesthetized with different concentrations of eugenol and EOLA is shown in Table 6.
Histopathological changes in the gills of the stingrays anesthetized with eugenol, such as hyperplasia and hypertrophy of epithelial cells and ionocytes (Fig. 4A), were reasonably frequent. In contrast, aneurysms in various sizes were accompanied by irreversible damage to lamellar architecture, such as apoptosis of ionocytes, filament epithelial detachment, and rupture of the epithelium (Fig. 4B-C-D) ( Table 6). On the other hand, stingrays anesthetized with EOLA had a higher frequency of morphological alterations of the primary response, such as hyperplasia and hypertrophy of epithelial cells and ionocytes ( Fig. 5A-B). Vascular alterations, such as capillary dilation and lamellar aneurysms, were observed in stingrays anesthetized with 150 to 200 µL L −1 (Fig. 5C). However, irreversible damage such as apoptosis of ionocytes was only observed in the group anesthetized with 225 μL L −1 (Table 6). Melanomacrophages were observed in the central blood vessel of the primary gill lamellae of stingrays anesthetized with 100 µL L −1 of eugenol and 175 μL L −1 of EOLA (Fig. 5D).

Discussion
The present study examined the potential use of eugenol and EOLA as anesthetic substances for application in juveniles of an Amazonian freshwater stingray species, Potamotrygon wallacei. To date, the use of both anesthetics is unprecedented for representatives of Potamotrygonidae family, and only a few investigations have assessed the effects of anesthetic substances obtained from herbal products on blood and plasma biochemistry (Frick et al. 2009); no studies . Basal, stingrays not handled and not exposed (n = 6); OA, onset of anesthesia (n = 12); DA, deep anesthesia (n = 12); and RE, recovery 48 h later (n = 12). Different letters indicate significant difference (p < 0.05) between stages of EOLA (no statistical analysis was applied). Basal, stingrays not handled and not exposed (n = 6); OA, onset of anesthesia (n = 12); DA, deep anesthesia (n = 12); and RE, recovery 48 h later (n = 12). Different letters indicate significant difference (p < 0.05) between stages have focused on use of EOLA in any elasmobranch fish. This also applies for the evaluation of potential histological alterations on the gills of these fish when they are in contact with these substances in the water. This limitation is, in part, due to complicated logistics in acclimatization and handling procedures to keep elasmobranchs in captivity, which makes the collection of physiological data after anesthesia a non-trivial task.
The similarity of the physical-chemical characteristics of the water observed during the acclimatization of the stingrays, as well as in the induction and recovery times in the experimental aquariums, are in line with the amplitude of these parameters documented in the areas of natural occurrence of the cururu stingray (Oliveira et al. 2017). This indicates that the behavioral (swimming capacity and reflex activity) and physiological responses (SF and blood parameters), as well as the histological changes observed in the gills of P. wallacei during anesthesia and recovery, are attributed only to the effects of tested anesthetics, and not due to changes in the water quality used in the experiments.
As well as in teleosts, depression of ventilatory frequency and cardiac rate are expected responses in elasmobranch species under anesthesia (Hove and Moss 1997). For instance, a transitory decrease in ventilatory frequency (measured trough spiracular opening and closing movements) in the freshwater stingray P. motoro was observed after oral administration of 50 mg kg −1 of ketamine (Raines and Clancy 2009). Conversely, in the spotted bamboo shark (Chiloscyllium plagiosum), the intravenous administration of propofol (2.5 mg kg −1 ) resulted in stable Table 4 Blood parameters obtained for juveniles of Potamotrygon wallacei (cururu stingray) shortly after reaching recovery (n = 6) and after 48 h from anesthesia (n = 6) with different concentrations of eugenol (µL L −1 ). No significant differences were observed (p > 0.05) for any parameter in all tested concentrations of eugenol when stingrays submitted to anesthesia were compared to those after 48 h of recovery 1.48 ± 0.28 1.01 ± 0.08 1.17 ± 0.17 0.99 ± 0.13 1.54 ± 0.18 1.10 ± 0.14 1.07 ± 0.17 1.11 ± 0.13 Cholesterol (mmol L −1 ) 2.17 ± 0.38 1.83 ± 0.19 0.91 ± 0.10 1.68 ± 0.26 2.38 ± 0.29 1.61 ± 0.26 0.87 ± 0.10 1.72 ± 0.29 Triglycerides (mmol L −1 ) 0.40 ± 0.10 0.67 ± 0.13 0.26 ± 0.04 0.56 ± 0.11 0.61 ± 0.12 0.66 ± 0.12 0.23 ± 0.03 0.52 ± 0.14 Total proteins (g.L −1 ) 8.43 ± 1.32 8.80 ± 1.02 6.91 ± 0.88 5.32 ± 1.14 7.52 ± 1.18 9.34 ± 0.99 6.74 ± 0.57 4.75 ± 1.19 Urea (mmol L −1 ) 1.61 ± 0.11 1.87 ± 0.31 2.13 ± 0.45 2.46 ± 0.52 1.46 ± 0.116 1.85 ± 0.29 2.00 ± 0.29 2.56 ± 0.69 Lactate (mmol L −1 ) 0.68 ± 0.11 0.53 ± 0.07 0.67 ± 0.14 0.59 ± 0.11 0.73 ± 0.17 0.66 ± 0.07 0.78 ± 0.15 0.70 ± 0.13 respiratory and heart rates during anesthesia (Miller et al. 2005). These results indicate that cardiorespiratory responses may diverge in different species and according to the dosage and anesthetic type. In our study, the SF values obtained for the basal group of juveniles of P. wallacei before exposure to eugenol and EOLA are in line with those recorded for the same species (Pastório 2014) under normoxic conditions (40 to 60 beats min −1 ). On the other hand, it was observed that the first contact with eugenol resulted in a hyperactivity response, which was reflected in the maintenance of a high SF (> 60 beats min −1 ) during the onset of the induction phase in almost all tested concentrations. This pattern, associated with an excessively long excitatory phase (stage II), may have resulted in stress and physical exhaustion, making the anesthetic induction with this herbal compound aversive for stingrays. These ventilatory responses are in line with the behavioral manifestations of hyperactivity and intense excitability observed in these eugenolexposed animals, which were not attenuated during the phase of deep anesthesia. In this line of evidence, Barbas et al. (2021) demonstrated for juveniles of tambaqui (Colossoma macropomum), an Amazonian teleost species, that 65 mg L −1 eugenol induced the fish to immobilization (with loss of muscular tonus and reflex) in a time-dependent manner. However, electroencephalographic studies revealed that eugenol failed to depress the central nervous system (CNS), resulting in an intense neuronal excitatory condition. Thus, the evidence of the behavioral responses observed in juveniles of P. wallacei when exposed to eugenol corroborates the above-cited results (Barbas et al. 2021), suggesting that stingrays had been immobilized (with a loss of muscle tonus) by the action of eugenol, without depressing their CNS function. These features led us to carefully consider the use of this herbal product as a safe anesthetic and draw our attention to its possible toxic action for animals when submitted to anesthetic practices. In a comparative perspective, the sequence of behavioral patterns shown by juveniles of P. wallacei under exposure to eugenol and EOLA was like those observed by Grusha (2005) for the marine stingray R. bonasus  californicus treated with an isoeugenol-based anesthetic (AQUI-S 20E®), a eugenol isomer, showed a prolonged induction time and more than one excitatory phase (Silbernagel and Yochem 2016). In this study, the stingrays anesthetized with eugenol showed long recovery times, a response that differs from that observed for viola ray, Z. brevirostris, anesthetized with 85 mg L −1 eugenol . Conversely, juveniles of P. wallacei showed considerably shorter recovery times (4.2 min) after anesthesia, apoptotic chloride cells (Ac); C filament epithelial detachment (Ed); D lamellar aneurysm (arrowheads) with epithelial cell hyperplasia (Eh) and epithelial rupture (asterisk); central blood vessel (Bv); ionocytes (I); secondary lamellae (Sl). Hematoxylin-eosin stain which are coincident to those observed in teleost fish (Marking and Meyer 1985;Keene et al. 1998), suggesting that eugenol can act differently in freshwater stingrays when compared to marine elasmobranchs. The exposure of P. wallacei juveniles to 50 μL L −1 eugenol revealed that this concentration was not suited to deep anesthesia due to the long time to reach light or deep anesthesia (> 10 min). The overexposure of P. wallacei to eugenol probably resulted in excessive absorption of this substance, culminating in mortality during the recovery phase. On the other hand, the abrupt entry into the anesthesia phase and deaths . Hematoxylin-eosin stain recorded at concentrations above 75 μL L −1 eugenol increase the risk of anesthesia, cause deep respiratory depression and the anesthetic induction process must be stopped immediately as soon as deep anesthesia is observed (phase III).
When juveniles of P. wallacei were exposed to EOLA at concentrations of 50 and 100 μL L −1 , they did not reach mild or deep anesthesia after 10 min; therefore, these concentrations are not recommended to the practice of anesthesia for these stingrays. On the other hand, at concentrations over 300 μL L −1 of this essential oil, neurological sequelae and subsequent deaths were observed after exposure, which may be related to a decrease in blood flow at the cerebral level during anesthesia. In bamboo sharks, Miller et al. (2005) observed signs of neurological alterations associated to intravenous administration of propofol. In the other concentrations of EOLA evaluated in this study, a uniform decline in the respiratory pattern was observed during anesthesia, with no obvious stressing effects, such as the initial high SF and sudden falls in ventilation observed with eugenol. This fact resulted in a better condition for physiological recovery after anesthesia, which was evidenced by the faster return to regular feeding and routine swimming after recovery from the anesthesia.
Considering that eugenol has a higher density (1.06) than that of EOLA (0.80), this physical property could be a relevant aspect when referring to bottom-dwelling elasmobranchs, like P. wallacei. Even when previously diluted with ethyl alcohol for its use (Neiffer and Stamper 2009), eugenol may be easily left at the bottom of the experimental aquarium, exacerbating its irritating effects when coming into direct contact with the eyes (causing evident eye retraction), skin (burns on the edges of the disc, and areas of focal necrosis on the body surface), and gills. The overall features of gill histology assessed in the present study confirm the detrimental effects of this anesthetic on the architecture of this tissue. The necrotizing characteristics promoted by eugenol in the gills led to impaired protection responses in the primary epithelium, promoting irreversible damage, such as ionocyte (chloride cell) apoptosis and lamellar necrosis. In contrast, despite promoting tissue irritation, EOLA triggered less irreversible damage and greater reversible changes in the protective response of the epithelium, such as hyperplasia and cellular hypertrophy. This increase in the proliferation and size of cells can lead to the fusion of lamellae, a response mechanism to protect the lamellar epithelium from direct contact with toxic agents; this increases the distance from the gill surface used for the diffusion of blood to inhibit the intake of substance from the surrounding environment (Heath 1987;Karan et al. 1998).
Another feasible explanation could be related to the inflammatory chemical mediators released from the injured tissue responsible for vasodilation and increased local vascular permeability. In this process, the increase in blood flow through capillary dilation, while allowing the arrival of defense cells, might predispose to edema and the formation of aneurysms (Pober and Sessa 2015). In our study, we observed the incidence of melanomacrophages in the blood vessels of branchial filaments of juveniles of P. wallacei exposed to both anesthetics. Melanomacrophages are cells with phagocytic activity having a considerable number of dark pigments inside, like melanin, lipofuscin, and hemosiderin. These cells are commonly found in the stroma of lymphoid tissues of teleost fishes, such as the kidney and liver, and are associated with the immune response in organs like the gills, liver, kidney, and spleen (Agius and Roberts 2003;Steinel and Bolnick 2017) and related inflammatory processes (Manrique et al. 2014). However, other functions associated with the natural physiological mechanisms of cell renewal, waste cleaning and collagen recycling (Gutierre et al. 2017), lipid deposition during fasting (Neyrão et al. 2019), and iron recycling from hemoglobin molecules (Wolke 1992;Passantino et al. 2005;Sales et al. 2017) have also been documented. Studies on elasmobranch species have been sporadic (Borucinska et al. 2009;Moraes et al. 2016) and inconclusive. In Potamotrygon motoro, the presence of these cells was recognized and associated with the immune response, acting as the site of antigen processing (Moraes et al. 2016), and with environmental pollution triggered by gold and oil exploitation in the occurrence area of this species in Peruvian territory (Ramos-Espinoza et al. 2017). However, in the present study, the arrival of melanomacrophages in the gills of P. wallacei exposed to concentrations of 100 µL L −1 eugenol and 175 μL L −1 EOLA cannot be attributed entirely to the toxic effect of these anesthetics. The presence of this cell type surrounding the blood vessels of the gill filaments seems to be more associated with natural physiological processes, because they were only observed at intermediate concentrations, and since despite the most expressive changes evidenced in the epithelium at the highest concentrations of both anesthetics, this cell type was not found. Thus, further investigations are required to better understand the role of melanomacrophages in routine physiological and immunological responses in the gills and other tissues of freshwater stingrays.
The hematocrit values obtained in juveniles of P. wallacei were within the range recorded for this species under different environmental and experimental conditions (Brinn et al. 2012;Pastório 2014;Oliveira et al. 2016) and similar to those observed in marine stingray species, such as D. americana (Cain et al. 2004), A. rostrata (Routley et al. 2002;Speers-Roesch et al. 2012), Hypanus sabinus (Lambert et al. 2018), and Zearaja maugeana (Morash et al. 2020). In all these studies, a commonly observed characteristic was the lack of changes in this parameter in response to different types of stressors. Brinn et al. (2012) mentioned that the variables related to erythrogram did not function as suitable physiological indicators in the assessment of stress responses in juveniles of P. wallacei (= Potamotrygon cf. histrix) submitted to river transport. Our results corroborate this hypothesis and indicate that, at least for the tested concentrations, both eugenol and L. alba essential oil did not promote noticeable changes in this blood parameter in cururu stingrays, even after 48 h of recovery. In addition, the lack of changes in the observed values of the plasma energy metabolite profile analyzed immediately after the recovery from anesthetic action and 48 h later, especially glucose levels, is a clear indicative that the extent of contact with the different anesthetic concentrations of both eugenol and EOLA was not enough to promote changes in these blood physiological parameters, and hence do not interfere in the overall homeostasis of P. wallacei juveniles. These findings are expected in part since elasmobranchs have a reduced reliance on carbohydrates, being more dependent on ketone bodies and amino acids as energy substrates (Ballantyne 1997;Speers-Roesch et al. 2006), and are in line with the results found for a shark species (Mustelus antarcticus) exposed to anesthesia (Frick et al. 2012), as well as for juveniles of P. wallacei submitted to different types of stressors, like progressive hypoxia (Pastório 2014) and river boat transport (Brinn et al. 2012). Interestingly, in another long-term transport experiment simulating the export of P. wallacei juveniles, the addition of EOLA prevented the increase in plasma glucose levels at arrival (32 h later), which was observed in stingrays transported without this essential oil (Ariotti et al. 2021).
The plasma lactate levels assessed in juveniles of P. wallacei exposed to all concentrations of both anesthetics fell within the range of other elasmobranch species minimally stressed (controls/baseline) or after recovery from an acute stressful condition (Cliff and Thurman, 1984;Wise et al. 1998;Frick et al. 2009Frick et al. , 2012Barragán-Méndez et al. 2019;Scarponi et al. 2021). On the other hand, the inability of these anesthetics to produce changes in the plasma lactate levels contrast with previous studies conducted with other elasmobranch species anesthetized with eugenol or its isomer, the isoeugenol. In Australian tiger sharks Cephaloscyllium laticeps, sedated with an anesthetic derived from clove oil (AQUI-S® isoeugenol, New Zealand Ltd.), the mean plasma lactate levels were significantly higher than those of untreated sharks (Frick et al. 2009), and like P. wallacei, no significant differences in plasma glucose were observed between the treatment groups. Usually, sharks alter blood parameters (like hematocrit, pH, lactate, and glucose levels) only when submitted to vigorous and intense physical exercise (Skomal 2007;Frick et al. 2010;Brooks et al. 2012). In juveniles of Carcharhinus taurus, lactate levels remained significantly higher up to 12 h after capture stress (Kneebone et al. 2013). In a similar approach, Fuller et al. (2020) found that individuals of the Atlantic sharpnose shark (Rhizoprionodon terraenovae) showing higher lactate levels in blood after capture stress also exhibited lower clinical patterns of behavior required for safe release.

Conclusions
In the current study, although the blood responses were similar between the two anesthetics, the noticed oscillating effects on spiracular frequency (ventilation rate) during anesthetic administration with eugenol, allied with a long excitatory stage during anesthetic induction and a faster return, in addition to the extent of lesions observed in both external disc and gill lamellas, makes the use of this substance somewhat risky for juveniles of P. wallacei. On the other hand, concerning the better ventilatory pattern, the unchanging levels of blood lactate, combined with a lower degree of histopathological changes observed in the gill apparatus, this study recommends the use of EOLA for anesthesia of P. wallacei stingrays during short-term anesthetic procedures. In this sense, a concentration of 200 μL L −1 is indicated since it required a shorter time to achieve anesthesia and provided a subsequent safe recovery.