Trophic Ecology of Slender Snipe Eel Nemichthys scolopaceus Richardson, 1848 (Anguilliformes: Nemichthyidae) in the Central Mediterranean Sea

Abstract

The slender snipe eel Nemichthys scolopaceus Richardson, 1848 is cosmopolitan in tropical and temperate seas, inhabiting the mesopelagic and bathypelagic zone between 200 and 1000 m depth. It is known to be an active predator in the DSL (Deep Scattering Layer) and the NBA (Near Bottom Aggregation), feeding mostly on decapod and euphausiid crustaceans, and playing a central role in carbon fluxes through meso- and bathypelagic ecosystems. Despite its potential importance in the deep trophic web ecosystem, the trophic ecology of Nemichthys scolopaceus is not well known. The aim of this study was to start to fill this knowledge gap. A total of 35 specimens of N. scolopaceus caught through bottom trawling in the Mediterranean Sea were analyzed in the laboratory for stomach content composition. As expected, mainly decapod crustaceans were found, in particular Plesionika martia, Pasiphaea multidentata, Funchalia woodwardi, and Robustosergia robusta species. The degree of digestion of prey in the stomachs was high in all cases. Our findings seem to confirm the specialist diet of Nemichthys scolopaceus based on shrimp-like crustaceans.

1. Introduction

The slender snipe eel Nemichthys scolopaceus Richardson, 1848 is a cosmopolitan snipe eel in the order Anguilliformes, reported in the Atlantic, Indian, and Pacific Oceans and in the Mediterranean Sea [1]. N. scolopaceus is known to inhabit the mesopelagic and bathypelagic depth strata [2,3], mainly between 200 and 1000 m [4]. In contrast to other Nemichthyidae species, which do not seem to be strong vertical migrants, N. scolopaceus, appears to carry out wide vertical displacements in the water column to capture prey [5,6]. According to Gartner et al. [7], N. scolopaceus was frequently observed in both the DSL (Deep Scattering Layer) and NBA (Near Bottom Aggregation) layer, over the North Carolina and Virginia continental slopes, in the western North Atlantic Ocean.

This species shows marked sexual and ontogenetic dimorphism. At the juvenile stage, all individuals have non-occlusible, beak-like jaws armed with very small villiform teeth. As soon as sexual maturity is reached, males are subjected to jaws radically shortening and complete loss of teeth [8] while females keep a morphology similar to juveniles. These notable changes in the physical features of mature males are indicative of their focus on concentrating their energy on reproduction rather than foraging, most likely due to slender snipe eels being semelparous [9]. Moreover, the male N. scolopaceus will not be able to catch their prey, concentrating their energy on reproduction to the detriment of their life [8]. With regard to the general size and shape of this fish, as reported by [8], N. scolopaceus has an elongate body, reaching up to 1.5 m in total length, with a long, very thin tail. The profound physical transformations described above may explain its rarity in the wild, especially in the Mediterranean Sea [10]. For food, the male leaves its deep-sea habitat (generally below 500 m) and becomes more susceptible to trawling, while the adult female remains in the deep-sea habitat and continues feeding with its mouth open.

Nemichthyid eels may have several predation strategies, including ambushing prey that passes near them while they maintain a vertical body position [11], or actively chasing prey [7,12]. In the Gulf of Maine (North-Atlantic), N. scolopaceus feed exclusively on crustacea, mostly decapoda and euphausiacea, avoiding potential prey that are very abundant in that habitat (an “oceanic rim” ecosystem [13], such as fishes, cephalopoda, and other crustacea (e.g., amphipoda, brachiura, and copepoda). This prey selectivity, linked to the high biomass of decapod crustaceans and large euphausiids in this oceanic zone and their ecological importance, and the fact that N. scolopaceus is one of the few mesopelagic fish predators of shrimps and euphausiids in the area, underlines their central role in carbon fluxes through meso- and bathypelagic ecosystems [6]. On the other hand, a metabarcoding analysis of gut contents carried out by Bucklin et al. [14] highlighted that the slender snipe eel foraged mostly on shrimp-like crustaceans, and only occasionally preyed on smaller zooplankton (e.g., copepods), pelagic mollusks, and fish. Really, this “apparent” preference for pelagic macrocrustaceans would instead appear to be the result of an evolution of the morphological structure of its mandibles that would allow it to trap only prey with “hairy” antennae such as pelagic decapods and euphausiids ([6] and references therein). Due to this feeding method, N. scolopaceus females are thought to be very active in searching for and capturing their prey, performing continuous day–night trophic migrations following their prey. This emphasizes its role in the benthic–pelagic food web as an active vehicle of biomass and energy from the deep demersal zone to the upper mesopelagic layers.

The slender snipe eel Nemichthys scolopaceus has been poorly studied in general, especially from the trophic point of view. In particular, in the Mediterranean Sea, the food habits of this Nemichthydae have never been investigated. Only three recent studies focused on the diet of N. scolopaceus, all carried out in the northwest Atlantic [6,14,15]. The present paper aims to contribute to filling the gap of knowledge about the trophic ecology of N. scolopaceus, particularly in our study area, and to compare our findings with other studies carried out in other zones of the global ocean. Our specific objective was to determine if this species feeds exclusively (or almost exclusively) on large shrimp-like crustaceans in the Central Mediterranean Sea as just highlighted [6,14,15].

2. Materials and Methods

2.1. Study Area and Samples Collection

At almost all sampling points, vertical temperature and salinity profiles were performed using a multiparametric probe (SBE 911plus, Seabird Electronics, Bellevue, WA, USA). In late spring and summer, a marked thermocline was observed between 10 and 30 m (26.70–17.85 °C) and between 5 and 20 m (26.30–17.65 °C) along the Tyrrhenian coast of Sicily and southern Calabria, respectively. The temperature near the seabed (300–700 m) was always between 14.0 and 14.4 °C. During the autumn sampling in the Sardinia Channel, the spatial variability of salinity was significantly reduced compared to a previous survey carried out in late spring–early summer. Water temperature ranged between ∼24 °C and ∼14.8 °C, while salinity ranged between ∼37.65 and ∼38.7 at both the surface and the bottom, respectively. This region was also characterized by a homogeneous upper thermal layer.

Briefly, and in agreement with [16], three different water masses characterize the thermohaline vertical structure: Tyrrhenian Surface Water (TSW), Atlantic Water (AW), and Tyrrhenian Intermediate Water (TIW). The less salty Atlantic Water (AW), entering from the Strait of Gibraltar, characterizes the surface circulation (0–200 m) of the Southern Tyrrhenian Sea. Along its path, the AW enters the area close to the northern coast of Sicily, where it forms a large cyclonic vortex. As the summer season approaches, the anticyclonic cells of the coastal recirculation appear to shift westward. The AW flowing through the Strait of Sicily turns right and enters the Tyrrhenian basin, flowing along the Sicilian continental slope.

A bathymetric survey carried out along the margin of the continental slope between 400 and 800 m of the Gulf of Patti has highlighted the complex of incisions and the various submarine valleys that represent the area that tend to flow towards the Stromboli canyon system. The seabed sediment composition through all hauls was mostly mud; in areas such as the Gulf of Patti and the Sardinia Channel, the seabed geomorphology was represented by submarine canyons, while the rest of the sampling points were characterized by a relatively flat seabed.

A total of 35 N. scolopaceus specimens were collected through bottom trawling (

Table 1. Collection of metadata for Nemichthys scolopaceus samples from all the cruises.

Campaign	Fishing Area	Local Date	Lat. N	Long. E	Sampling Depth (m)	Local Time (Start–Finish)	Specimens Caught (Per Time)	Specimen ID Number
Commercial fishing	Gulf of Patti (ME)	May–June 2022	38°14.346′	15°7.818′	300–700	-	11	7CB, 13CB, 14CB, 15CB, 18CB, 21CB, 22CB, 23CB, 24CB, 25CB, 26CB
		May–June 2023				-	12	34CB, 35CB, 36CB, 37CB, 38CB, 39CB, 40CB, 41CB, 42CB, 47CB, 48CB, 49CB
NAUCRATES22	Pa-Ca Canyon	12/10/2022	38°25.91′	10°42.24′	309–305	12:53–17:35	1	28CB
		19/10/2022	38°13.65′	10°37.32′	323–358	22:20–1:47	1	31CB
		14/10/2022	38°15.00′	10°37.32′	370–311	12:30–16:33	2	32CB, 33CB
MEDINEA22	Pa-Ca Canyon	14/10/2022	38°03.60′	10°47.28′	332–349	19:07–23:07	2	29CB, 30CB
MEDITS24	GSA10	21/06/2024	39°28.24′	15°44.27′	583–626	07:13–08:13	2	44CB, 45CB
		22/06/2024	39°13.18′	16°02.19′	83–86	06:51–07:21	1	50CB
		22/06/2024	39°06.21′	15°55.49′	634–619	14:46–15:46	1	46CB
		25/06/2024	38°10.56′	14°51.02′	26–18	12:00–12:30	1	51CB
		26/06/2024	38°07.70′	13°38.44′	568–529	16:15–17:15	1	43CB

), 23 of which as bycatch of commercial fishing in the Gulf of Patti (Southern Tyrrhenian Sea, Central Mediterranean) during the summer 2022–2023. The remaining 12 specimens were collected in 2022 and 2024 during three scientific fishing oceanographic cruises (NAUCRATES and MEDINEA in 2022, MEDITS in 2024) off the Calabria and Campania coast in the Southern Tyrrhenian Sea, and between Sardinia Island and the Tunisian coast in the Sardinia Channel, respectively (Figure 1). The bottom trawls used in the commercial fishing were between 30 and 50 m long, 12–20 m wide, and had a vertical opening of 1.5–2.5 m. The trawl was made up of two panels with a minimum square mesh size of 40 mm, which is the minimum mesh size allowed by the European Union. The MEDINEA22 and NAUCRATES22 scientific cruises were studying the fishing selectivity of bottom trawling for the giant red shrimp Aristeomorpha foliacea and the blue and red shrimp Aristeus antennatus as targets, and because of this, the trawl net used in the MEDINEA22 cruise was similar to the commercial trawls (e.g., same square mesh size). However, during the NAUCRATES22 cruise, a metal grid was added to the trawl net (inserted before the cod end) to evaluate its effectiveness in reducing the bycatch of juvenile shrimps. The sampling gear used during the MEDITS cruise was different and consisted of a four-panel bottom trawl (model IFREMER GOC73, Plouzané, France) capable of operating between 10 and 800 m deep. The net had a minimum mesh side of 10 mm at the cod end, which meant that it had a mesh opening of 20 mm and therefore as low a selectivity as possible. This trawl had a total length of 40 m, a horizontal opening of 22 m, and a vertical opening of 2 m. In most cases, N. scolopaceus were captured near the seabed, at depths between 18 and 700 m. The specimens were frozen on board at −20 °C.

2.2. Laboratory Analysis

In the laboratory, fish were thawed and weighed and morphometric measurements were taken. Whole wet weight (WW, ±0.01 g) of intact fish (i.e., no missing tail or jaw parts) was taken after blotting (Sartorius-ENTRIS32O2-1S, Gottinga, Germany) the fish body to remove excess water. Total length (TL) was measured for fish with intact tails and jaws to the nearest 0.5 cm. Moreover, because of the peculiar morphology of the snipe eel, with its long and very thin body that can cause fragmentation of the tail or jaws, the distance between the eye and gill cover was measured (EGD: Eye–Gill cover Distance, ±0.1 mm, Figure 2). This measurement was possible to record for all fish regardless of missing tails or jaws. Prior to our study, the relationship between EGD and TL was unknown. Our study therefore provides an alternative method of estimating the TL of slender snipe eels that have damaged tails or jaws.

The stomachs were extracted and preserved in 70% ethanol with seawater. Subsequently, the stomachs were blotted to remove excess water, weighed (WW, ±0.1 g (Mettler-Toledo AG204, Columbus, OH, USA), and carefully opened using scissors and forceps while viewed under a stereomicroscope (ZEISS STEMI SV 8, Oberkochen, Germany; Leica WILD M10, Wetzlar, Germany).

The degree of digestion of prey items was evaluated according to the following scale: 1 = no evident signs of digestion, prey whole and complete; 2 = prey partly digested with missing portions; 3 = prey from enough to strongly digested, only pieces remaining; and 4 = almost totally digested, only traces remaining [6]. Prey identification was carried out by species if possible, or to the lowest feasible taxonomic level, by the following taxonomic manuals: [17] for decapod crustaceans; [18] for fish otoliths; and [19,20] for fish scales. Prey items were preserved in 70% ethanol. After prey extraction, the empty stomach was weighed to obtain the weight of the stomach content (subtracting this weight from the full stomach weight). Prey size was determined by comparing the anatomic parts found in the stomachs (rostrum, telson, uropods, antennal segments, abdominal segments, eyes and eye stalk, scaphocerites) with those of intact specimens stored in our laboratory collection and using the size–weight relationship reported in references [21,22,23,24,25,26]. In particular, the size and weight of prey items were reconstructed using proportions to scale said anatomic parts and obtain the total length for fishes and the carapace length for crustaceans, which were then used by applying the formula W = a(Length)^b; we used a and b values from the papers cited above. To determine the size and weight for prey items of which we had determined only the genera, we calculated the estimated size and weight for all the species from that genera present in the Mediterranean Sea and then obtained a mean value (e.g., for Funchalia sp. we obtained the size and weight for Funchalia woodwardi and Funchalia villosa, and then obtained the mean value between these two species). The weights obtained were used to calculate the relative percentage indexes.

To establish whether the relationships between TL and WW and TL and EGD were statistically significant, a regression analysis was carried out using the R Software (version 4.5.1). WW and EGD were plotted as functions of TL. For each parameter, an allometric regression model was fitted to evaluate the WW/TL relationship, and a linear one was fitted to evaluate the EGD/TL relationship. EGD was plotted to understand if this measurement could be used as a more reliable morphometric feature than TL in the length–weight relationships of this organism.

2.3. Trophic Indexes

Trophic indexes were calculated for individuals with stomachs containing food, with the aim to evaluate the dietary composition of N. scolopaceus. First of all, % empty was the number of empty stomachs divided by the total number of stomachs × 100. The degree of stomach fullness was estimated by the Stomach Content Index (%SCI) as follows:$\%SC{I}_x={\displaystyle \frac{{\sum }_n^{1}W}{TW-\left({\sum }_n^{1}W\right)}}\times 100$, where the numerator reports the sum of the weights of the prey found in the stomach of the predator $x,$ and the denominator reports the value resulting from the difference between the weight of the predator and that of the prey ingested by it. The average value of %SCI was subsequently calculated.

The importance of each prey i was assessed by calculating the following food indices. Percentage in number: $\%Ni={\displaystyle \frac{ni}{nt}}$× 100, where ni is the total number of individuals of prey i found in the stomachs, while nt is the total number of prey; weight percentage: %Wi =${\displaystyle \frac{wi}{wt}}$ × 100, where wi is the total weight of individuals of prey i found in the stomachs, while wt is the total weight of all prey; percentage frequency: $\%Fi={\displaystyle \frac{nsi}{ns}}$ × 100, where nsi is the number of stomachs containing prey i, while ns is the total number of stomachs containing prey. The Relative Importance Index [27,28,29] was also estimated for each prey i, which takes into account and integrates the values of the previous food indices (%N, %W, %F): $\%IRIi=\left(\%Ni+\%Wi\right)\ast \%Fi$. The percentage contribution of the Relative Importance Index for each prey i (%IRIi) was then estimated for each prey i (%IRI_i) relative to the total prey: $\%IRIi={\displaystyle \frac{IRIi}{\sum IRI}}$ × 100.

The feeding strategy of N. scolopaceus was evaluated using the graphic method of Costello [30] modified by Amundsen [31], which plots on a two-dimensional graph the specific abundance, for number and for weight, of each prey category with respect to the frequency (%F) of finding the prey in the stomach contents. This method allows us to deduce information on the importance of the prey and the feeding strategy used by the predator. In this graphical representation, the specific abundance of each prey i (Pi) is calculated as follows: $Pi={\displaystyle \frac{\sum Si}{\sum Sti}}$× 100, where Si is the total abundance (as weight or number) of prey i; Sti is the total stomach content only of the samples in whose stomachs prey i is present.

The trophic level (TROPH; [32]) of N. scolopaceus was estimated by implementing the weight contribution and the trophic level of each prey species to the diet with the formula TROPH = 1 + ∑DCj ∗ TROPHj, where TROPHj represents the fractional trophic level of the prey j and DCj represents the weight contribution of the prey j to the total weight of all the prey items found. The TROPH can range from 2.0 for herbivores/detritivores, to 5.0 for piscivores/carnivores [33,34,35]. Trophic level values of each prey item were evaluated according to [34].

3. Results

3.1. Fish Metrics

According to Feagans-Bartow and Sutton [6], data presented in this article refer to juveniles and adult females. No males were collected during the sampling. The 35 (31 intact and 4 damaged) N. scolopaceus specimens examined for stomach contents ranged from 500 to 1325 mm TL, having a mean value of 807.97 mm (standard deviation: ± 189.95 mm). The individuals were assigned, according to their TL, to size classes of 100 mm. Figure 3 shows only the size classes of the 31 intact specimens.

The WW ranged from 2.94 to 76 g (mean WW: 31.29 g; standard deviation ± 23.23 g), while the EGD varied from 5.8 to 25.5 mm (mean EGD: 16.5 mm; standard deviation ± 5.45 mm).

Of a total of 35 individuals, 31 were used to build the regression lines, and 4 were excluded because they were damaged (broken beak or tail). WW showed a steep increase described by the equation y = 2∙10⁽⁻⁶⁾ x^2.3623, where 2∙10⁽⁻⁶⁾ corresponds to the value a, and 2.3623 corresponds to the value b (Figure 4). This model yielded an R² value equal to 0.3363, indicating that 33.63% of WW variability is explained by TL (p < 0.004). On the other hand, EGD showed a trend modeled by the equation y = 0.0181x + 1.723 with an R² value equal to 0.4351, meaning that TL accounts for roughly 43.51% of its variability (p < 0.0001) (Figure 4).

3.2. Trophic Ecology

Of the 35 stomachs analyzed, 4 were empty, while 31 contained prey items (% empty = 11.43%). The mean degree of stomach fullness was calculated and the resulting value was between 1 and 2 (1.48 mean), which reflected the results obtained from the Stomach Content Index (%SCI); in fact, these values ranged from 0.0028% to 14.56% with a mean value of 1.65%. This confirms the overall high degree of digestion of prey items. A total of 34 prey were found in stomachs of N. scolopaceus, with a mean number of 1.1 prey per specimen. The total weight of prey was 123.66 g. The list of N. scolopaceus food items is reported in

Table 2. Prey of Nemichthys scolopaceus with trophic indexes found in the study area.

Phylum/	Family	Species/Taxa							%IRIi
Subphylum/			Frequency		Number		Weight
Class/Order			n	%	n	%	n	%
CRUSTACEA/
Malacostraca/
Decapoda
	Pandalidae	Plesionika martia	2	6.45	2	5.88	11.51	9.32	2.15
		Plesionika sp.	1	3.23	1	2.94	10.3	8.34	0.8
	Pasiphaeidae	Pasiphaea multidentata	1	3.23	1	2.94	5.75	4.65	0.54
	Sergestidae	Robustosergia robusta	1	3.23	1	2.94	0.97	0.79	0.26
	Penaeidae	Funchalia woodwardi	1	3.23	1	2.94	0.29	0.23	0.22
		Funchalia sp.	3	9.68	3	8.82	0.53	0.43	1.96
		Decapoda unid.	5	12.9	5	14.71	16.3	13.19	7.88
		Crustacea unid.	6	19.35	6	17.65	19.55	15.82	14.17
CHORDATA
VERTEBRATA
Teleostei
Stomiiformes	Stomiidae	Stomias boa	1	3.23	1	2.94	0.89	0.72	0.26
Gadiformes	Macrouridae	Macrouridae unid.	1	3.23	1	2.94	13.66	11.06	11.85
			2
Unidentified			12	38.71	12	35.3	43.8	35.45	59.91
Totals			31		34	100	123.55	100	100

, together with the dietary index values for each prey. Among the 31 stomachs containing food items, 12 contained only unidentifiable traces because of the high degree of digestion. In the remaining ones, we observed mainly decapod crustaceans, in particular Plesionika martia (A. Milne-Edwards, 1883), Pasiphaea multidentata (Esmark, 1866), Funchalia woodwardi (Johnson, 1868), Robustosergia robusta (Smith, 1882), the genera Funchalia and Plesionika, and unidentified crustaceans. Moreover, we found traces of teleosts, including one otolith of Stomias boa (Risso, 1810) and scales from Macrouridae. Some sand was found in one stomach.

The degree of digestion observed was in all cases high (2–4; prey partially, highly, or almost completely digested). We did not find any prey whole and complete. Unidentified prey was the most numerically abundant (%N = 35.3) and frequently found (%F = 38.71) in stomachs of N. scolopaceus, followed by unidentified crustaceans (%N = 17.76; %F = 19.35) and decapods (%N = 14.71; %F = 12.90). The majority of weight was also from unidentified prey (%W = 35.45%), unidentified crustaceans (%W = 15.82%), and decapods (%W = 13.19%). However, fish prey (Macrouridae %W = 11.06%), Plesionika martia (%W = 9.32%), and Plesionika sp. (%W = 8.34%) were also relatively important on a weight basis. Calculation of %IRI further highlighted the relevance of Unidentified Crustacea (%IRI = 14.17%), Unidentified Decapoda (%IRI = 7.88%), Plesionika martia (%IRI = 2.15%) and Funchalia sp. (%IRI = 1.96%), and Teleostei Macrouridae (%IRI = 11.85%).

The estimated value of the index of trophic level (TROPH) for N. scolopaceus was 4.34.

3.3. Feeding Strategy

Our results showed that the diet of N. scolopaceus was dominated by unidentified prey from both a numerical (Figure 5a) and biomass (Figure 5b) point of view, while other prey items, if considered at species levels, did not stand out in terms of either abundance or weight. Nevertheless, the predominance of crustaceans was highlighted by considering them at the group level, regardless of single species, genera, and family (hence including unidentified crustaceans; Figure 6).

4. Discussion

Regarding the relationships between length and weight, the results obtained by the regression models underscore a distinct scaling behavior for WW and EGD. In the WW allometric model, a moderately steep regression line is visible. N. scolopaceus is skinnier and longer than most bony fishes, and this could explain why the regression line is not as steep as expected for a length–weight relationship, and why b does not reach the cubic value. The R² value obtained is quite low, underscoring the relatively low reliability of this model. On the other hand, the relationship between EGD and WW appears to be more reliable with a modest R² value, but it may still be unprecise for this reason. EGD and WW showed a trend modeled by the equation y = 0.1718x^1.6192, where a = 0.1718 and b = 1.6192, and yielded an R² value equal to 0.4351, which means that EGD accounts for 43.51% of WW variability. Lastly, EGD is linear to TL because both are linear measurements, and especially because it is tied to skeletal development and less subject to environmental conditions. WW can easily fluctuate with regard to feeding status, and in this species the full stomach condition can greatly influence the total weight of the animal, as it did with some of our specimens which had a noticeable bulge, indicating the presence of big, heavy prey. The WW could also be influenced by the reproductive stage, which is not known at the moment of this study, unfortunately; moreover, as no adult male individuals were caught in the surveys, we were not able to assess differences in the regression, if any. Moreover, some influence on WW may have been caused by the passage from the ship freezers to the lab ones, translating to a possible mass reduction due to defrosting. Thus, further studies are required to explain the relations between all the morphometric features of the slender snipe eel in a better and more appropriate way.

Of the 35 specimens of N. scolopaceus examined in this study, 31 had stomachs containing prey and 4 were empty, resulting in a empty percentage of 11%, which was lower than that detected by [6] (30%), and in turn even less than other mesopelagic predatory fishes [12] (60%). Since the deep-sea fishery resources of this area consist mostly (about 80%) of decapod crustaceans [36], it can be hypothesized that the availability of prey could be one of the reasons for the low empty index found. However, the high percentage of stomachs containing prey might be due to the slow digestion rate of this nemichthyid [6,37,38]. In addition, the long digestion time did not permit an evaluation of the feeding [6]. Moreover, the degree of digestion of prey found in the stomachs in this study was in all cases high (2–4; prey partially, highly, or almost completely digested). We did not find any prey whole and complete. Prey might have been ingested more than 24 h before fish sampling [6,12,37,38]; another reason may be the relatively large size of prey found in this study (e.g., Pasiphaea multidentata with a carapace length of 30.7 mm and one of the two Plesionika martia specimens which measured 24.6 mm).

According to some recent studies [6,15], the stomachs of N. scolopaceus mainly contain decapod crustaceans, similar to our study. However, we did not detect any traces of euphausiids, which represent a fair portion of the diets of N. scolopaceus in the northwest Atlantic [6,15]. Nevertheless, the unidentified prey may have contained euphausiids since small prey digest faster. Euphausiids are important components of the pelagic food web, occurring in the diet of many fish species of high commercial value [39,40,41] and becoming part of an energy flux which also involves mesopelagic fishes [42]. Some studies indicate that the distribution of M. norvegica may be controlled by phytoplankton blooms resulting from upwelled nutrient rich water, like the Algero-Provencal Basin [43], northwest Mediterranean Sea [44,45,46], the Ligurian Sea [47,48,49], and the Strait of Messina, where it can be observed in enormous swarms. This krill species prefers waters with a temperature between 3° and 15 °C for its reproduction, which might explain its abundance in the Straits of Messina, sometimes found stranded from November to April [50]. Thus, a more plausible explanation for its absence in the diet of N. scolopaceus could concern its ecology and geographical distribution. Indeed, in the southern Tyrrhenian Sea and in general in the oligotrophic waters of the Aeolian archipelago, M. norvegica is almost absent, or present in low percentages compared to other euphausiid species [51], as well as in other similar areas of the Mediterranean [52].

Further studies will be necessary to investigate this aspect of the food habits of this nemichthyid in the Mediterranean Sea. Moreover, in the stomach contents of N. scolopaceus we found traces of teleosts, also found by [14]. However, given that in the studies of [6,15] no traces of fish were found, and given that we found only one otolith (Stomias boa) and three scales (Macrouridae), we could hypothesize secondary ingestion; in other words, the possibility that the otolith and the scales found in the stomachs of N. scolopaceus originated in turn from the stomach contents of the crustaceans eaten by N. scolopaceus itself. In support of this hypothesis, at least two of the prey found in the eels’ stomach contents, the decapods Pasiphaea multidentata and Plesionika martia, are known to be active predators of small fishes, as well as scavengers [53,54]. Moreover, in the study of [14], no whole fish specimens were found in the stomach contents of N. scolopaceus, but rather the presence of teleosts was detected by DNA metabarcoding [14], which could not exclude the possibility of secondary ingestion. Further reinforcement for this hypothesis is the predation strategy of N. scolopaceus. In fact, this eel is highly selective for shrimp-like crustaceans, which is reflected in its morphology [5,6,15,55]. The thin and elongated jaws of juveniles and adult females of N. scolopaceus are covered with innumerable and very small villiform teeth, which can stick to crustaceans’ antennae and help to entrap them by entanglement of their antennae [5,6,15,55]. According to [6,15], this peculiar feeding method, together with non-occlusible jaws, could contribute to making this nemichthyid extremely selective for shrimp-like crustaceans, avoiding other potential prey present in the environment, such as fishes, cephalopods, and other crustaceans (e.g., amphipods, copepods). Further studies are needed to shed light on this aspect as well, in addition to the apparent absence of euphausiids in our study. Hence, as shown in Figure 5, N. scolopaceus appears to be a specialist when it comes to feeding strategies, and the examined population displayed a high between-phenotype component (high BPC in both the (a) and (b) plots). This means that this eel has little to no competition and overlap for the resources used, and the single individuals specialize in different resource types [31]. This explanation is consistent with what we found, as the snipe eels had stomachs full of only one prey on average, and moreover they almost always contained a different type of prey. On the other hand, in Figure 5, plot (b), by weight, some prey items can be found in the bottom left side of the graph; this could represent some sort of occasional feeding on Funchalia sp. and Stomias boa, as they presented a low value of occurrence and prey-specific abundance. It is also true, however, that the high digestion degree detected by us could lead to a bias where the prey items we found are overemphasized [31], but in our opinion is nevertheless valid and reflects the reality, given the support of similar studies in the rest of the world [6,15]. The specialist strategy could fit well with the fact that the TROPH level of N. scolopaceus is quite high (4.34), meaning that it feeds on organisms that belong to the third trophic level, and is close to being an apex predator of the submarine canyon ecosystem where it has a high availability of its favorite prey at its disposal.

Prey found in the stomach contents of Nemichthys scolopaceus in our study were benthic (Plesionika martia, Macrouridae) and benthopelagic (Pasiphaea multidentata) organisms linked to muddy bottoms between 300 and 700 m of depth [12,16,19,21,22,23,24,26,53,54], or mesopelagic (Robustosergia robusta, Funchalia woodwardi, Stomias boa) organisms that can be found near the bottom [7,12,17,22,26] at the same depths. Most specimens of slender spiny eels in the present study were caught exactly at these depths, near the bottom, where the crustacean and fish species detected in their stomachs can frequently be found. Therefore, the type of prey found in the present study can only confirm the statement of [7] that, in the northwest Atlantic, Nemichthys scolopaceus is part of the Near Bottom Aggregation. The finding of some sand in the stomach of one specimen of N. scolopaceus examined was a further sign that this eel fed frequently near the bottom. Furthermore, the prey species observed in stomach contents in the present paper, coupled with the sampling depth of most specimens (300–700 m), support the results of [2,3], according to which N. scolopaceus inhabit the mesopelagic and bathypelagic depth strata, mainly between 200 and 1000 m [4]. On the other hand, two specimens (of the thirty-five sampled) of Nemichthys scolopaceus in the present study were caught at shallow depths during the daytime, in one case at 83–86 m a few hours after sunrise (around 7:00) and in the other case at the very shallow depths of 18–26 m, around noon (12–12:30). These observations confirm that N. scolopaceus, despite living habitually between 200 and 1000 m (or deeper) [4], is able to carry out wide displacements in the water column to catch prey [5,6]. Nevertheless, in the case of specimens caught at 18–26 m, it is possible that these are dying individuals that are unable to return to mesopelagic depths after night migration. With regard to the feeding chronology of Nemichthys scolopaceus, considering what has already been said about the difficulty of evaluating it through the digestion degree of prey and sampling time because of the slow digestion of this Nemichthiydae eel [6,15], it is not possible to contradict or confirm the results of [5,6]. According to these studies, N. scolopaceus is able to feed round the clock, migrating upwards at night, chasing midwater shrimps that in turn rise up, and also catching them during daytime at higher depths, exploiting their reduced activity (and reduced ability to avoid predation) due to “recovering” from the previous night’s migration. Therefore, the presence of N. scolopaceus near the bottom at 300–700 m of depth during daytime is reasonable and in line with previous studies [2,3,4,5,6,7,15].

In

Table 3. Comparison of sample size and of main prey items between this study and the more relevant ones.

	This Paper	Faegans-Bartow, Sutton, 2014 [6]	Bucklin et al., 2024 [14]
			(Visual Method Only)
Examined specimens	35	164	7
Macrocrustacea	N = 20 (%N = 58.82%)	N = 131 (%N = 100%)	N = ~1 (%N = ~50%)
Amphipoda	N = 0 (%N = 0%)	N = 0 (%N = 0%)	N = ~1 (%N = ~20%)
Teleostei	N = 2 (%N = 5.88%)	N = 0 (%N = 0%)	N = 0 (%N = 0%)

we present a more immediate comparison between this paper and [6,14,15], which represent the main works about this topic at the present time. First of all, we wanted to point out the differences between the number of sampled specimens in the different areas. In this paper, we examined 35 specimens, caught only by bottom trawling in a time span of three years (2022, 2023, and 2024) and by different activities such as commercial fishing and scientific surveys. On the other hand, the survey [6] was performed using midwater trawling between 435 and 670 m depth and, despite the lower number of hauls (15 in total), they caught an enormous number of N. scolopaceus individuals (1487) in only five days, but examined the gut contents of only 164 of them. Finally, the survey [14] was performed using both midwater trawling and MOCNESS nets, deployed at different depths, but they only caught seven individuals of N. scolopaceus in total in a time span of two years.

This great difference is probably due to the main target of the surveys in question and to the equipment used. The surveys from which we retrieved our samples were more oriented towards demersal resources and often limited by time, except for commercial fishing: in fact, this latter recorded the highest number of collected specimens used in this work (23 out of 35). In addition, they usually go fishing in the canyon off the Gulf of Patti; not to mention that in MEDINEA22 and NAUCRATES22, we also performed sampling in a canyon, catching the same number of individuals as the MEDITS campaign, despite being smaller campaigns themselves. In [6], and therefore in [15] too, the authors probably had no such limitations, and they centered their sampling in a narrow depth range and specifically in the canyons south of Georges Bank. Moreover, as stated earlier, in this oceanic zone, a high biomass of decapod crustaceans and large euphausiids was found [6]. This abundance of specific prey, coupled with the fact that N. scolopaceus is one of the few mesopelagic fish predators that feed on shrimps and euphausiids in the area [6], likely contributes to increasing the numerical abundance of this Nemichthyidae eel, making it possible to sample a high number of specimens. Instead, in [14], the authors sampled in an area where canyons were absent and, moreover, by midwater trawling. As a probable result, they caught fewer N. scolopaceus individuals, just like almost all of the sampling area covered by the MEDITS campaign by bottom trawling, in which we caught only six individuals despite of a total of 70 hauls carried out. Therefore, the reason behind the relatively small number of individuals in this study is probably related to the seafloor morphology of the single sampling points.

As said before, the Nemichthys scolopaceus specimens examined in this paper were, in most cases, caught as bycatch in commercial fishing, specifically the very commercially relevant shrimp trawl in the Gulf of Patti. Therefore, a better understanding of N. scolopaceus could, in turn, broaden knowledge on the ecology of species targeted by commercial fishing, providing useful information for the management of fishery resources. In particular, this Anguilliformes appears to be linked with several species of commercial shrimps, such as the red shrimp Aristaeomorpha foliacea, which share the same habitat and in whose fishing N. scolopaceus was accidentally captured, and others like the genus Plesionika, which was also found in this Nemichthyidae eels’ stomach contents in the current study.

Additionally, in our opinion, it would be important to encourage professional fishermen to collaborate with the study of fish ecology, especially regarding bycaught species that do not have commercial value, but conversely could have a very high scientific importance and constitute an inestimable precious resource for ichthyology and ecology studies. Active collaboration between fishermen and scientists was fundamental for this study and should be implemented in the future. In all cases, it would be useful to raise awareness in fishermen about the relevance of ecological studies to the preservation of fishery resources for the future, not only for ethical reasons, but also and especially to guarantee the continuity through time of commercial species populations. In this sense, fishermen could feel themselves to be a greater part of a large community of marine experts that encompasses fishermen, scientists, and anybody who interacts with the sea environment.

In conclusion, we believe that the ecology and food habits of Nemichthys scolopaceus are strongly linked to commercial fishing, both from the point of view of fishery resource management and for collaboration between fishermen and scientists, in particular to use bycatch species as scientific material, and, as highlighted, to aid knowledge of the sea habitat and ecological processes. This study, and potential future ones, may be useful to help raise awareness in fishermen and other people about the importance of knowledge about non-commercial and accidentally captured marine species.