The Class Chondrichthyes (sharks, rays, and chimeras) is the most evolutionary distinct radiation of vertebrates, for which extinction risk has been determined for the entire clade (Stein et al., 2018; Dulvy et al., 2021). In this context, in the Mediterranean Sea, among 88 chondrichthyan species (Otero et al., 2019), more than half (at least 53%) are threatened because of overfishing (Dulvy et al., 2016). In this sense, understanding the overall process of reproduction would be useful for assessing the population status of these species (Marongiu et al., 2021) by investigating male–female interactions, physiology, biochemistry, and anatomy (e.g., Storrie et al., 2008; Jordan et al., 2021).

In all vertebrates, reproduction is regulated by the brain-pituitary-gonadal (BPG) axis, which promotes gametogenesis and subsequent gamete maturation (Sherwood and Lovejoy, 1993; Awruch, 2013; Dufour et al., 2020). Moreover, it represents an important anatomical interface that brings together the environmental cues, such as photoperiod and temperature, as well as the central nervous system and the development of gonads (Maruska and Gelsleitcher, 2011). Nowadays, the BPG axis has become an important research topic in conservation biology due to its role in species adaptation and adaptability to environmental factors in the context of global, climatic, and anthropogenic changes, showing some of the strong links between ecosystems and biodiversity health (i.e., “One Health” concept) (Dufour et al., 2020). However, the understanding of this complex regulatory system shows several basic gaps and needs further exploration along vertebrate classes (Maruska and Gelsleitcher, 2011; Awruch, 2013; Kanda, 2019; Santiago-Andres et al., 2021). To date, the research on this system is more complete in mammals, where the gonadotropin-releasing hormone (GnRH) acts on gonadotroph cells in the pituitary and stimulates the release of gonadotropin, luteinizing hormone (LH), and follicle-stimulating hormone (FSH), known as gonadotropins (GTHs) (e.g., Cattanach et al., 1977; Kumar et al., 1997; Abel et al., 2000; Ma et al., 2004).

In Chondrichthyes, several studies have been conducted to explain the neural control of reproduction of the BPG system, identifying the GnRH analog as a positive factor in the release of GTHs or in the direct maturation of the gonads stimulating steroidogenesis (e.g., Dobson and Dodd, 1977; Jenkins and Dodd, 1980; Callard et al., 1988; Fasano et al., 1989; Callard and Koob, 1993; Sherwood and Lovejoy, 1993; D'Antonio et al., 1995; Demski et al., 1997; Forlano et al., 2000; Maruska and Gelsleitcher, 2011). Distinct populations of GnRH neurons are found in the basal forebrain or preoptic area (GnRH1), the midbrain tegmentum (GnRH2), and the terminal nerve ganglia (GnRH3) (Powell et al., 1986; Lovejoy et al., 1991, 1992; Sherwood and Lovejoy, 1993; D'Antonio et al., 1995; Demski et al., 1997; Forlano et al., 2000; Masini et al., 2008; Gaillard et al., 2018; Ogawa et al., 2022). GnRH1 neurons project axons to the rostral lobe (RL), to the median lobe, and to the neurointermediate lobe of the hypophysis (NIL) (Demski et al., 1997; Forlano et al., 2000). Despite that, they do not reach the isolated ventral lobe (VL), which contains the highest level of GTHs (Dodd, 1983; Callard et al., 1989; Callard and Koob, 1993; Quérat et al., 2001). Indeed, GnRH may reach and stimulate VL through an intraventricular route or general circulation via the carotid arteries (Dodd, 1983). However, it must be noted that other lobes of the pituitary may display gonadotropic activity, albeit to a much smaller amount (Sumpter et al., 1978a,b). Generally, in contrast to the GnRH, which is the principal positive factor in the BPG axis (Millar, 2005; Kah et al., 2007), catecholamines (CAs) can be considered one of the main negative factors in vertebrate reproduction, providing an additional brain pathway for the integration of various species-specific, internal, and environmental cues. The degree of CAs' inhibition and their role in the regulation of puberty, seasonal reproduction, ovulation, and sperm production might differ throughout vertebrate classes as well as within smaller evolutionary groups such as teleost fish or even mammals (Dufour et al., 2005, 2010). Comparative data about the CAminergic system in vertebrates were formerly collected using the formaldehyde-induced fluorescence (FIF) method, which revealed the position of CAs in cells (Eränkö, 1967). In contrast, recent research on this system has primarily focused on the immunohistochemistry against tyrosine hydroxylase (TH), as this enzyme is widely considered crucial in determining the CAminergic phenotype. This has facilitated the transfer of understanding of the CAminergic system from anamniotes to amniotes (Smeets and González, 2000). Notably, CAs are synthesized from the aromatic amino acid tyrosine by a series of enzymes, the first and rate-limiting enzyme being TH, which initiates the biosynthetic pathway that produces dioxyphenylalanine (L-DOPA) (Nagatsu et al., 1964). However, not all TH immunoreactive (TH+) neurons are involved in CA synthesis. Indeed, neurons in several areas of the brain of various vertebrates are TH+ but immunonegative for another enzyme of the biosynthetic pathway, such as the aromatic amino acid decarboxylase (AADC), which is required for the biosynthesis of dopamine (DA) from L-DOPA (Ikemoto and Panksepp, 1999; Karasawa et al., 2007; Ugrumov et al., 2022). There is evidence that some of these cell groups may contain L-DOPA as a terminal neurotransmitter (Meister et al., 1988; Okamura et al., 1988). In some brain areas, monoenzymatic TH+ or AADC+ neurons cooperate in the production of DA by transferring L-DOPA from the first type to the other one (Ugrumov et al., 2022). It is also possible to categorize neuron DA or/and noradrenaline immunoreactive, which are TH (TH-), AADC, and dopamine beta-hydroxylase (DBH) immunonegative (Smeets and Steinbusch, 1990; Smeets and González, 2000). Nevertheless, the interpretation of such findings remains uncertain (Smeets and González, 2000).

For what concerns the reproductive processes, the inhibitory role of CAs was demonstrated both pharmacologically and anatomically in teleosts, where CAs are released by recruited neurons of the preoptic nucleus, which are TH+ and dopaminergic (DAminergic). They project axons directly to the region of the pituitary where GTH-secreting cells are located (Kah et al., 1984, 1987; Vacher et al., 2003; Dufour et al., 2005, 2010).

To date, although Chondrichthyes are an essential group for analyzing ancestral brain organization (Rodríguez-Moldes et al., 2020), most of the research in their CAminergic brain areas concerns their mapping only, revealing a rather consistent pattern (Wilson and Dodd, 1973b; Meredith and Smeets, 1987; Northcutt et al., 1988; Stuesse et al., 1990, 1991, 1992, 1994; Carrera et al., 2005, 2012).

In this group, it is known that the evolutionary topographic homolog of the TH+ preoptic nucleus is the TH+ periventricular preoptic nucleus (PO), and nearby another large group of neurons, the TH+ suprachiasmatic nucleus (SCN), is located (Stuesse et al., 1991, 1994). These brain nuclei were also analyzed using the FIF method (Wilson and Dodd, 1973b; Molist et al., 1993; Rodríguez-Moldes et al., 1993) and immunohistochemistry against dopamine, which was specifically carried out on the thorny skate Ambliraja radiata (Meredith and Smeets, 1987). Notably, due to its topography, SCN may be involved in the primary visual pathways (Repérant et al., 1986; Northcutt, 1990), revealing DAminergic somata (Meredith and Smeets, 1987) or fairly FIF+ somata and projecting its CAminergic axons to the NIL (Wilson and Dodd, 1973b; Molist et al., 1993). Moreover, the existent descriptions of CAminergic PO share the presence of FIF+, TH+, and DA+ neurons, which are mostly scattered under the floor of the third ventricle (Wilson and Dodd, 1973b; Meredith and Smeets, 1987; Rodriguez-Moldes and Anadon, 1987; Stuesse et al., 1990, 1991, 1994; Molist et al., 1993; Rodríguez-Moldes et al., 1993; Carrera et al., 2012). Furthermore, PO is characterized by the presence of periventricular cerebrospinal fluid-contacting (CSF-c) neurons, which are generally FIF+ and TH- (Molist et al., 1993). In some species, in the PO, both non-CSF-c and CSF-c neurons are reported to be DAminergic (Meredith and Smeets, 1987; Molist et al., 1993; Rodríguez-Moldes et al., 1993). At the same time, it is suggested that CSF-c neurons could accumulate DA from the ventricular route without any biosynthesis from the TH, constituting the so-called organum vasculosum praeopticum (Meredith and Smeets, 1987; Molist et al., 1993).

Among Chondrichthyes, catsharks have several characteristics that make them an ideal elasmobranch model for comparing morphology or physiology through vertebrate evolution (Grunow et al., 2022). In this sense, the small-spotted catshark Scyliorhinus canicula (Linnaeus, 1758) represents one of the most studied species among oviparous sharks because it is abundant, easy to capture (Coolen et al., 2008), and has a well-described annual reproductive cycle (e.g., Dodd, 1972, 1983; Kousteni and Megalofonou, 2019).

In S. canicula, while the GnRH system is widely described (e.g., Jenkins and Dodd, 1980; D'Antonio et al., 1995; Gaillard et al., 2018), the organization of CAminergic neuronal populations has been studied in detail in embryos, juveniles, and adults only from a qualitative point of view (Molist et al., 1993; Carrera et al., 2005, 2012; Quintana-Urzainqui et al., 2012). To the best of our knowledge, the most complete information about the role of this system is known in the SCN. Its CAminergic component projects axons to the NIL (Wilson and Dodd, 1973b; Molist et al., 1993; Carrera et al., 2012), and it may exercise inhibitory control over the release of the melanocyte-stimulating hormone, assessing the skin paling of the shark (Wilson and Dodd, 1973a,b; Wilson et al., 1974; Molist et al., 1993). Regarding the CAminergic PO cells, their role has been suggested for a steroid-feedback mediation, because estradiol is generally uptaken in the preoptic area (Jenkins and Dodd, 1980). Moreover, TH+ PO cells were observed in both the embryos and the juveniles (Carrera et al., 2012) and correspond to those detected in the adult, in which they were CAminergic, as determined by the FIF and anti-TH immunoassay (Molist et al., 1993). In the prehatching embryo, there is a reduction in the number of TH+ CSF-c neurons as the organism develops through the juvenile stage. CSF-c neurons form a subset that can be distinguished from non-CSF-c neurons, located ventrally to the preoptic recess in the PO (Carrera et al., 2012). Moreover, CSF-c neurons are not TH+ in adults, but FIF+ only, and they are further distinguished from the non-CSF-c population, which is mostly TH+ and FIF+ (Molist et al., 1993). Anyway, due to FIF's fluorimetric spectrum, some authors considered the above-described PO populations as CAminergic (Rodriguez-Moldes and Anadon, 1987) and other ones presumably DAminergic (Molist et al., 1993). Furthermore, whereas the majority of TH+ fibers come from neurons in the SCN, some may potentially originate from neurons in the preoptic region (Molist et al., 1993; Carrera et al., 2012). These fibers contain a sufficient level of CAs at the NIL level to be detected by the FIF method (Molist et al., 1993). On the other hand, the specific origin of CAminergic fibers in the blood vessel walls of the RL of hypophysis is unknown (Wilson and Dodd, 1973b).

According to the established role of the CAminergic TH+ preoptic nucleus in teleost fish and other vertebrates (Dufour et al., 2005, 2010, 2020), as well as the observations made in S. canicula (e.g., Molist et al., 1993; Carrera et al., 2012), we hypothesize that the putative homolog of the TH+ preoptic nucleus in this species (Stuesse et al., 1994) may change in the morphometry and number of neurons throughout the reproductive cycle. For our purpose, we studied TH+ neurons in the preoptic region of S. canicula females, from immature to egg-laying specimens, considering the TH+ PO as a positive control in ovarian development and the TH+ SCN as a negative one, through two different approaches. Specifically, we sampled and grouped S. canicula at different sexual maturity stages (immature, maturing, mature, and mature egg-laying) (Follesa et al., 2019) and investigated the (i) regional distribution and morphometric differences among TH+ PO somata and TH+ SCN somata and (ii) changes in morphometry and number of neuronal bodies among the stages. Such important information describe a potential model of a putative CAminergic system, presumably involved in the S. canicula female BPG system.

In contrast to the wealth of knowledge in mammals, the understanding of BPG axis regulation in other vertebrate classes is ambiguous (Kanda, 2019), in particular in Chondrichthyes, in which many hypotheses explaining the pathways regulating the control of reproduction have not been proven (Awruch, 2013). In this study, we focused on TH+ neurons, possibly involved in the control of reproduction in S. canicula females, as a reference in BPG system research for oviparous sharks. We hypothesized anatomical changes in the morphometry and the number of TH+ periventricular preoptic nucleus (PO) neurons, many of which are CAminergic (Wilson and Dodd, 1973b; Rodriguez-Moldes and Anadon, 1987; Molist et al., 1993; Carrera et al., 2012), along the sexual maturity stages. On the other hand, the adjacent TH+ suprachiasmatic nucleus (SCN) was considered a negative control.

From the qualitative and quantitative observations provided for the first time in this species, it was found that the TH+ PO and the TH+ SCN neurons were two clearly distinct populations. The cytoarchitecture of these two brain nuclei confirmed the pattern previously described in the juveniles of the same species by Carrera et al. (2012) as well as in other species like Raja undulata (Molist et al., 1993) and Squalus acanthias (Stuesse and Cruce, 1992). To the best of our knowledge, till now, a very similar distinction among the morphology and the distribution of these two neuronal populations was detected in two other elasmobranchs: the horn shark Heterodontus francisci (Stuesse et al., 1991) and the thorny skate Ambliraja radiata (Meredith and Smeets, 1987), whose neuronal characterization was made by immunohistochemistry against dopamine.

Considering the morphometric analysis of the TH+ somata in the preoptic area, the fusiform neurons of TH+ SCN were similar in shape to the somata of two other studied sharks, H. francisci (Stuesse et al., 1991) and S. acanthias (Stuesse and Cruce, 1992). In H. francisci, the oval-like TH+ somata may be congruous with the TH+ PO somata morphology of S. canicula females observed in this work. Furthermore, in the current study, any of the TH+ CSF-c neurons were detected, and this may be in line with the previous literature (Molist et al., 1993; Carrera et al., 2012), since the smallest immature individual here analyzed was not from a direct post-hatching phase, characterized by a number reduction of TH+ CSF-c neurons (Carrera et al., 2012). Moreover, following the descriptions of Molist et al. (1993), CSF-c neurons are not TH+ in adults, but FIF+ only, and they are further distinguished from the non-CSF-c population, which is mostly TH+ and FIF+, suggesting that they may accumulate CAs rather than producing them from the TH biosynthesis.

The morphology and number of detected TH+ PO neurons changed according to the sexual maturity stages, which might have occurred due to the changes in the activity of TH+ neurons according to the degree of ovarian maturity. The gradual decrease of TH+ PO somata size from immature females (F1 stage) toward those carrying egg cases (F3b stage) could suggest that the average neurons observed in the F1 stage may be less excitable than those found in more sexually developed individuals. In addition, the duplication of TH+ neurons in maturing females (F2 stage) may be characterized by the increased immunoreaction of the smallest and presumably more reactive TH+ cell units. Our results may be explained according to the “size principle” (Shepherd, 1994), which states that smaller motoneurons are more easily induced to fire action potentials than comparable units of larger size. These properties may be extendible to other neurons too (Henneman et al., 1965; Somjen et al., 1965; Fromm and Evarts, 1981). Notably, the smaller neurons may have a lower excitability threshold and may be recruited before the larger ones, showing lower membrane capacitance and higher membrane resistance. This would allow for a greater change in membrane potential for a given synaptic current (Henneman, 1957; Stuart and Enoka, 1983). In addition, differences in presynaptic terminal density might also contribute to the excitability of the cell units, and this was recently associated with size (Rana et al., 2019). If the “size principle” is applied to our results and TH immunoreaction is representative of CAminergic neurons in the PO of S. canicula, it may indicate that CAs are potentially produced by a wider amount of cell units and released at higher rates during the early steps of gametogenesis, such as those in the developing stage. Subsequently, the number of TH+ PO neurons in mature females (F3a) was more likely to show the same pattern found in the F1 stage, decreasing by about 60% compared to those estimated in the F2 stage. This suggests a loss in the involvement of TH+ PO's cell units in the production of neurotransmitters.

Our results, combined with previous knowledge, may reveal a fuller picture of the BPG axis in Chondrichthyans. In the maturing S. canicula females, characterized mainly by previtellogenic oocytes, the TH+ PO is more excitable and constituted by more TH+ cells, while LH content in the VL of the hypophysis is low, and androgens and estradiol levels in the plasma are almost undetectable (Sumpter and Dodd, 1979). Subsequently, in the mature phase, when oocytes are large and yolked and plasmatic vitellogenin production increases, as well as estradiol, testosterone (T), and LH levels (Craik, 1979; Sumpter and Dodd, 1979), the TH+ PO's cells become significantly less. This information may encounter similarities in the European Eel Anguilla anguilla, one of the most phylogenetically ancient teleosts, which remains, in its life cycle, at the prepubertal previtellogenic stage for a very long time, until the decrease of CAminergic action by the preoptic neurons in the BPG axis (Dufour et al., 2005, 2010, 2020). This drop, simultaneously with GnRH and T enhancement, causes an increase in the synthesis of LH as well as in vitellogenin plasma levels and stimulation of ovarian vitellogenesis (Dufour et al., 2003, 2005; Vidal et al., 2004). The CAminergic inhibition on the onset of puberty is also known in other teleostean species, such as the viviparous mosquitofish Gambusia affinis (Bhat and Ganesh, 2019), the gray mullet Mugil cephalus (Aizen et al., 2005), and the spadefish Chaetodipterus faber (Marcano et al., 1995), in which administration of dopamine's antagonists blocked D2-like receptors' action with subsequent gonadal maturation.

In continuity with the analyses, we examined mature egg-laying females (F3b stage), whose number of TH+ PO neurons was similar to that observed in mature females (F3a stage). At the same time, the average somata size was not significantly different from any of the previous stages. One possible reason for this observation could be that a few percentages of the smallest cells might have become immunonegative, causing, in the F3b stage, a little change in the average cell size, which slightly increases. As somata get larger, they may become less reactive, resulting in a modest decrease in the rate at which CAs are released. The changes here described are not documented among Chondrichthyes, but they are rather similar to what occurs in Osteichthyes, in which a further fall of CAaminergic action triggers the last steps of the oocyte's maturity and the eventual spontaneous spawning (Dufour et al., 2005, 2010). This pattern has also been described in the most phylogenetically ancient Chondrosteans as the white sturgeon Acipenser transmontanus (Pavlick and Moberg, 1997; Dufour et al., 2020) and in the teleosts as cyprinids (Peter et al., 1986, 1991; Lin et al., 1988; Fontaine et al., 2013), silurids (De Leeuw et al., 1986), percomorphs (Yaron et al., 2003), pleuronectiforms (Guzmán et al., 2011) and cyprinodontiforms (Bhat and Ganesh, 2019). However, the intensity of the CAminergic inhibitory tone at the time of spawning varies according to the studied species, from a drastic barrier in gray mullet Mugil cephalus (Aizen et al., 2005) to a milder control in salmonids and in the rainbow trout Oncorhynchus mykiss (Linard et al., 1995; Saligaut et al., 1999; Vacher et al., 2000).

Furthermore, in teleost fish, via binding to estrogen receptors, estradiol amplifies CAminergic inhibitory signals in the preoptic nucleus, controlling the expression of TH in neurons (e.g., Linard et al., 1995, 1996; Dufour et al., 2010, 2020). However, the probable mechanism by which estrogen affects TH production in CAminergic PO neurons in S. canicula is unknown. At the same time, these cells, along with preoptic GnRH cells, have been hypothesized to be estradiol-binding, presumably playing a role in steroid-feedback mediation (Jenkins and Dodd, 1980; Wright and Demski, 1993). According to the scant knowledge available on the estradiol cycle, the ovary and the epigonal organ are the primary sources (Jenkins and Dodd, 1982), which contribute to the development of secondary sexual characteristics (Dodd and Goddard, 1961). Notably, immature females have low or undetectable estradiol levels, which rise to high levels in mature females (Sumpter and Dodd, 1979), where heightened concentrations may be required for egg capsule synthesis (Jenkins and Dodd, 1982). Specifically, as estradiol levels rise, the quantity of GTHs in the VL decreases (Jenkins and Dodd, 1982). As a result, oocyte atresia in the ovary and the existence of produced egg capsules are linked to a drop in estradiol levels (Sumpter and Dodd, 1979).

For what concerns the analysis of the TH+ SCN, considered in this study as a negative control, they showed detectable changes in the morphometry and the neuronal number between the immature F1 females and the other more advanced stages. Indeed, TH+ SCN in the F1 stage was characterized by smaller, rounder, less elliptical, and less immunoreactive neurons than those in the other stages. These changes may underline some ontogenetic aspects related to the juveniles' visual system, as the neuronal shape (Roundness and AR) turnovers through the developing F2 stage toward the adult F3a and F3b. In addition, these neurons may be more excitable in F1 than the other ones measured in the more mature specimens, according to the above-mentioned “size principle”. On the other hand, our results did not show significant differences in cell number between the developing F2 stage and the mature ones, showing the same pattern. In S. canicula, the suprachiasmatic nucleus (SCN) is both a primary (Repérant et al., 1986; Northcutt, 1990) and a secondary visual region (Wilson and Dodd, 1973b; Wilson et al., 1974). Additionally, its CAminergic component may exert inhibitory control over the release of melanocyte-stimulating hormone in NIL, as well as integrating impulses from light reflected off the environmental background and regulating the skin paling of the catshark (Wilson and Dodd, 1973a,b; Wilson et al., 1974; Molist et al., 1993; Carrera et al., 2012). The changes found in the TH+ SCN in the current study could be attributable to internal factors, such as the animal's growth, or undefined environmental influences, such as light intensity related to bathymetry or seasonal variations. Changes in the number of neurons in the suprachiasmatic nucleus, for example, occur in mammals in response to changes in environmental light intensity, although various characterizations, distinct from the TH immunoreaction, were utilized (Cambras et al., 2005; Porcu et al., 2022).

Results from this study may provide some hints for future surveys, such as verifying if there is an anatomical connection between the GnRH neuronal population of the preoptic area and the TH+PO. In the stingray Dasiatys sabina adult males, the finding of a seasonal increase in the number of GnRH+ neurons of the preoptic area during the reproductive period, which may contribute to the control of reproduction as the periodic expression of GnRH (Forlano et. al., 2000), could allow us to hypothesize that the same neuronal population found in S. canicula plays the same role. In addition, given the proximity, TH+ PO projections may reach the GnRH-releasing neurons of the preoptic area described by D'Antonio et al. (1995) to exert control.

In Chondrichthyes, a direct release of CAs on the GTH cells of the VL of hypophysis should be excluded, as any general nervous connection has not been demonstrated (Dodd, 1983). Alternatively, in S. canicula, we could hypothesize that CAs reach the VL through blood circulation due to the presence of CAminergic fibers in the blood vessel walls of the RL of hypophysis (Wilson and Dodd, 1973b). Furthermore, the distribution and expression of estrogen receptors are unknown, but they may play an important role in the neuroendocrine control of reproduction in elasmobranchs, similar to their known role in the control of the BPG system in teleost fish and other vertebrates (reviewed by Dufour et al., 2005, 2010, 2020).

Our data suggest that anatomical changes in the TH+ PO may play a role in controlling the reproductive cycle of the small-spotted catshark. Our findings on a target species, S. canicula, provide new valuable insights into the understanding of Chondrichthyans' BPG axis, contributing to increase the awareness and knowledge of biodiversity evolution and ecosystems at an ecological scale. Other important clues can be aimed at determining changes in the morphometry and in the number of PO TH+ neurons in S. canicula male populations.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethical approval was not required for the study involving animals in accordance with the local legislation and institutional requirements because sixteen female S. canicula specimens were collected in the waters around Sardinia (central-western Mediterranean), observing the minimum number of samples required for the research. Individuals were obtained from commercial hauls conducted under the Data Collection Framework (European Union Regulation 199/2008) and during the Mediterranean International Trawl Survey (MEDITS, Spedicato et al., 2019). The catsharks obtained through fishing captures were euthanized on-site via decapitation, in accordance with guidelines and protocols approved by the European Community and Italian legislation governing the protection of animals used for scientific purposes (Directive 2010/63/UE L 276 20/10/2010, implemented by Italian Legislative Decree 26/2014). The animals were neither housed nor subjected to experimental procedures.

RP: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing—original draft. CP: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing—original draft. GM: Data curation, Formal analysis, Methodology, Software, Visualization, Writing—review & editing. SS: Data curation, Formal analysis, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing—review & editing. MF: Funding acquisition, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing.