A Multidisciplinary Approach to Identify Priority Areas for the Conservation of a Vulnerable Family of Fishes in Spanish Marine National Parks

Background: Syngnathid shes (Actinopterygii, Syngnathidae) are agship species strongly associated with seaweed and seagrass habitats. Seahorses and pipeshes are highly vulnerable to anthropogenic and environmental disturbances, but most species are currently Data Decient according to the IUCN (IUCN, 2019), requiring more biological and ecological research. This study provides the rst insights into syngnathid populations in the two marine Spanish National Parks (PNIA –Atlantic- and PNAC – Mediterranean). Fishes were collected periodically, marked, morphologically identied, analysed for size, weight, sex and sexual maturity, and sampled for stable isotope and genetic identication. Due the scarcity of previous information, habitat characteristics were also assessed in PNIA. Results: Syngnathid diversity and abundance were low, with two species identied in PNIA (Hippocampus guttulatus and Syngnathus acus) and four in PNAC (S. abaster, S. acus, S. typhle and Nerophis maculatus). Syngnathids from both National Parks (NP) differed isotopically, with much lower δ 15 N in PNAC than in PNIA. The dominant species were S. abaster in PNAC and S. acus in PNIA. Syngnathids preferred less exposed sites in macroalgal assemblages in PNIA and Cymodocea meadows in PNAC. The occurrence of very large specimens, the absence of small-medium sizes and the isotopic comparison with a nearby population suggest that the population of Syngnathus acus (the dominant syngnathid in PNIA) mainly comprised breeders that migrate seasonally. Novel 16S rDNA haplotypes and sequence variants were detected for H. guttulatus, N. maculatus, S. acus, and S. abaster. Our data suggest the presence of a cryptic Syngnathus species in PNAC. Conclusions: This is the rst multidisciplinary approach to the study of syngnathids in Spanish marine NPs. Habitat preferences and population characteristics in both NPs differed. Further studies are needed to assess potential species misidentications of genus Syngnathus in PNAC, and migratory events in PNIA.


Background
Syngnathidae is a singular sh family mostly inhabiting temperate and tropical sheltered, coastal marine waters (Foster and Vincent 2004;Kuiter 2009). Seahorses and pipe shes utilize rocky, muddy, sandy, and rubble bottom habitats, generally associated with macrophytes communities (Manning et al., 2019. Syngnathids are secondary consumers with specialized and opportunistic predatory strategies, ambushing small prey (mainly planktonic and nektonic crustaceans), showing a variety of diets, and foraging behaviours across genera and locations (Manning et al. 2019). Seaweeds and seagrass meadows promote the growth of most food sources and enhance the cryptic ability of syngnathids to avoid predators.
Syngnathids are valuable agship species for conservation programs that will simultaneously bene t other sh (Shokri et al. 2009). Many species are vulnerable and threatened by habitat loss (pollution, sedimentation and eutrophication) and disturbances through boating and shipping (IUCN 2019). More than half of syngnathid species (two seahorse and eleven pipe sh species) inhabiting Spanish coasts are currently classi ed as Data De cient, and further research is needed to understand their biology and ecology (IUCN 2019).
Confuse identi cations have been reported due to cryptic morphology and unclear diagnostic traits among species, stressed by historical reference labelling errors in particular cases (e.g. European genus Syngnathus) ( Studies on syngnathids in the Iberian Peninsula are scarce, and the present study is the rst approach for the global evaluation of syngnathid populations in Spanish coasts, particularly in marine National Parks (NP). Studies conducted in NPs would be highly valuable, considering their protection status and the supposed reduced impacts of most potential disturbances. Currently, there are two marine National Parks (NPs) in Spain, differing in their characteristics and biodiversity: Atlantic Islands National Park (PNIA) (Atlantic Ocean, NW Spain) and Cabrera Archipelago National Park (PNAC) (Balearic Islands, Mediterranean Sea). NPs are the best marine ecosystems to ensure species survival and success in biodiversity conservation. However, protection requires a deep knowledge and analysis of habitats, values and threats, particularly for exceptional species and populations.
Marine ecosystems in PNIA host complex habitats and numerous ecological niches due to the extraordinary rich biota inhabiting soft and rocky oors typical of protected, semi exposed and exposed environments. Rocky shores are covered by seaweeds, whereas the Western side is dominated by hard substrates covered by crusty, coralline and other turf-forming seaweeds (Piñeiro-Corbeira et al. 2020). That side is exposed to Atlantic open water and extreme sea currents and waves, mainly in winter. The Eastern side is less exposed due to its position facing the Ría de Vigo. That side is characterized by a high biodiversity and productivity, and therefore it is an area of special interest for shing. Such high productivity is promoted by important seasonal phytoplankton blooms (Álvarez-Salgado et al. 2000;Rodil et al. 2009), and secondary production (Valdés et al. 1990; Buttay et al. 2015), with high abundance in summer and seasonal changes in community structure. Copepods are largely predominant in winter, being accompanied in summer by other groups of fauna (Buttay et al. 2015). The aims of this study were threefold. First, to assess distribution and habitat use of syngnathids in PNIA and PNAC, each with highly distinctive environmental characteristics and vegetal assemblages. Second, to characterize syngnathid populations, which include the assessment of genetics identi cation and stable isotopes signals. Finally, the unavailability of historical data for syngnathids in the Iberian Peninsula prevents the assessment of population trends. Hence, the third aim of this study was the selection of speci c sites for further monitoring of distribution/abundance and temporal-seasonal patterns on important biological and ecological features (e.g. diet composition, animal migration). The results achieved would be valuable for the development of further conservation actions in both NPs.

Results
Habitat characterization in Cíes Archipelago (PNIA) Soft bottom sediments were mostly coarse sandy (569 µm), with 90% sand and a prevalence of a single mode (Additional le 2). Muddy sands, with > 20% mud (< 63 µm), were only located in the deepest (17.6 m to 21 m) and distal areas of TR5, in the immediate vicinity of the muddy bottoms characteristics of the central part of Ría de Vigo. The presence of two or three mode samples in TR2, TR4 and TR3 re ected a mixture of particle sizes, including bioclastic gravel (bivalves and gastropods shells) and maerl elements. Different sedimentary environments (wide variability of textural characteristics) were present along some transects (e.g. TR3). Syngnathids were mostly sighted in sheltered sectors, preferring habitats with medium sands, better sorted and lacking mud (Additional le 2).
Similarity of seaweed assemblages was analysed considering data of 55 species with medium-high abundance (Additional les 2, 3). Diversity (H') and species richness (S) were particularly low in TR1, TR2, TR7 and TR10, especially in spring (Additional le 2). Seaweed cover increased in summer, especially in TR8 (633.8%) and TR9 (861.0%), but it was noticeable low in TR10 (42% in spring; 107% in summer) (Additional le 2). PERMANOVA results showed signi cant differences in assemblage structures for transects (P < 0.05) and seasons (P < 0.01). Those differences are re ected in the two-dimensional PCOs plot ( Fig. 1). Spring (left) and summer (right) samples followed a gradient along axis 1 (20.4% of total variation). Abundance increased in summer for most species, especially for Treptacantha baccata, Padina pavonica, Corallina o cinalis or Codium tomentosum (strong negative correlation with PCO1; Spearman correlation > 0.65). Differences between transects were explained by axis 2 (18.1% of total variation), re ecting wave exposure. Transects TR1, TR8 and TR9 were clearly separated from the others, especially from TR10 and TR3. These results explained spatial differences between transects, with TR9, TR8 and TR7 as the most northern sites of Cíes Archipelago, and TR1 located in the west side of the southern island. The remaining transects (especially TR10) were located in areas less exposed to wave impact and current actions. Vectors overlay in PCO plot indicated that species such as T. baccata, P. pavonica or C. tomentosum were more abundant on less exposed areas, while Mesophyllum expansum, C. o cinalis, Plocamium cartilagineum and Kallymenia reniformis preferred more wave-exposed sites (Spearman correlation > 0.65).

Syngnathids in PNIA and PNAC
In PNIA, two species of syngnathids were identi ed morphologically and genetically: the long-snouted seahorse Hippocampus guttulatus Cuvier, 1829, and the greater pipe sh Syngnathus acus, Linnaeus 1758. A total of 28 specimens were sighted in PNIA from 4 to 15 m depth (mostly at < 8 m), with six transects providing at least one sh ( Table 1). None of the individuals marked in spring were recaptured in summer. All PCO showed a positive correlation of syngnathids with seaweed assemblages on transects TR3, TR4, TR5 and especially TR10 (Spearman correlation > 0.65) in summer (Additional le 2).
The highest abundance (0.06-0.13 syngnathids 100 m − 2 ) were recorded in mixed (sand-rock) or rocky substrates on transects TR3 and TR10 (32 and 43% of total specimens, respectively). Syngnathids were missing in the more exposed transects TR1, TR7, TR8 and TR9 (northern and southern areas with rocky substrate and coarse sand patches). TR1 was facing SW waves (prevalent component during storm winter conditions), while TR7, TR8 and TR9 were facing N waves (prevalent component during storm summer conditions). The most common species was S. acus (n = 24), which comprised 86% of total sh sighted.  A novel 16S rDNA haplotype of 521 pb (16S_NM01: MW080705) was detected for one pipe sh morphologically identi ed as N. maculatus, with 48 variable sites respect to a related reference species (N. ophidion; AF354994). For six specimens morphologically identi ed as S. abaster, two Cytb haplotypes were detected (Cytb_SAb01: MW080703 and Cytb_SAb02: MW080704) in four and two sh, respectively; showing 139 variable sites respect to the S. acus sequence (Additional le 2). These two Cytb_Sab haplotypes showed a higher sequence identity with S. typhle haplotype reference (JX228148; identities > 98%) than with S. abaster (JX228141; identities ≤ 95%) available at GenBank database. Phylogenetic analysis grouped the two Cytb_SAb haplotypes from PNAC in a monophyletic cluster, clearly differentiated from available GenBank sequences of S. abaster (Mwale et al. 2013) and other pipe sh species distributed in Mediterranean areas (S. acus, S. typhle and S. rostellatus) (Fig. 2).

Stable isotope signatures in syngnathids
In PNIA, H. guttulatus and S. acus (Table 3)  inter-seasonal differences only for δ 13 C, with spring values (-16.2 ± 0.3‰) lower than in summer (-15.6 ± 0.5‰) (MANOVA, P = 0.008), and values in males (-16.2 ± 0.2‰) lower than in females (-15.9 ± 0.6‰) (P = 0.021). Sex-maturity comparisons showed similar δ 13 C values for mature (-16.2 ± 0.3‰) and immature (15.9 ± 0.6‰) shes (P = 0.104) but lower δ 15 N signals in the later (11.1 ± 0.5‰ for mature, 10.4 ± 0.8‰ for immature) (P = 0.014). Table 3 PNIA -Mean (± sd) δ 13 C and δ 15  partially covered by large extensions of seagrass meadows (P. oceanica and C. nodosa), which is a typical cover enhancing the occurrence of syngnathids in some Mediterranean areas (Vizzini and Mazzola 2004). That is the case of S. typhle, a pipe sh that preferentially display an upright position in seagrasses with narrow leaves (e.g Zostera) (Vincent et al. 1995;Skóra 2001). Its absence in PNIA could rely on the lack of seagrass meadows, even though this species may adapt to different types of habitats (Tarnowska and Sapota 2007). Appropriate habitats for syngnathids may not be determined simply by the presence or absence of vegetation but also by the prevalence of seaweed communities that best enable them to remain inconspicuous to predators (Kendrick and Hyndes 2003). All pipe sh in PNAC were collected in C. nodosa and P. oceanica meadows, suggesting that macroalgal beds are less preferred than seagrass meadows.
In PNAC, the results showed unexpected low pipe sh abundance, which agrees with previous Due to the high dominance of S. acus in PNIA, the species deserves special consideration. The lengthweight relationship was similar to that in the western Black Sea (Yildiz et al. 2015 immature specimens in Cíes suggests that young shes prefer less exposed nearby sites, and/or that small juveniles are dispersed by currents to other areas. Sexual maturity in females and males was reached at 6.9 and 6.1 cm size, respectively. These sizes are smaller than reported in the Aegean Sea isotopic signals from that in Cíes (Fig. 3). Habitat and trophic web characteristics in both areas also differ considerably (Filgueira and Castro 2011). The former receives anthropogenic wastewater inputs and it is characterized by a complex trophic web, and a locally important microphytobenthos production available to primary consumers through resuspension. The cove is a community with a high diversity of organic matter sources but terrestrial particulate organic matter does not seem to contribute signi cantly to consumers' most plausible diets (

Directions for research and conservation
There is increasing public awareness of the challenges of marine biodiversity from habitat destruction, over-shing and development. Efforts are needed to protect and value marine biodiversity, especially species and communities that require relatively large areas of undisturbed habitat. NPs are areas set aside for the preservation of the natural environment to protect natural biodiversity along with its underlying ecological structure and supporting environmental processes, and to promote education and recreation (IUCN).
A lack of information on syngnathid populations has prevented conservation actions from being Those areas include rocky and sandy-gravel substrates, maerl beds as well as seaweed communities enhancing protection and habitat suitability for syngnathids. Rodas Bay is also interesting from a conservation point of view since it is also a preferential habitat for small Octopus vulgaris (Guerra et al. 2014). In PNAC, considering the benthic communities and the higher abundance of pipe sh compared to other locations, Es Burri Bay has the greatest interest value for further conservation actions and monitoring. Studies not based on long-term monitoring may lead to erroneous or incomplete assumptions. The study of temporal patterns of syngnathids diversity will provide valuable information on the assessment of species' sensitivity to habitat disturbances and climate warming (Faleiro et al. 2015), and on optimal conditions for captive breeding and further population reinforcement for the most endangered species if necessary.
-Some of the main questions arisen from the study is whether Cíes Archipelago should be considered a breeding sanctuary for S. acus. Understanding sh movement patterns and migrations from/to other nearby areas is another pivotal topic than needs addressing. For that, further isotopic and genetic information, and acoustic telemetric studies in specimens tagged with transmitters would provide valuable information to undertake further conservation actions (Villegas-Ríos 2013).
-The management of vessel transits to protect sensible areas against habitat loss (marine ora communities) and to mitigate anthropogenic sound is necessary (McKenna et al. 2017

Conclusions
This is the rst multidisciplinary study of syngnathid populations in Spanish coasts, speci cally in the two marine Spanish NPs. It will contribute to the knowledge of syngnathid populations, leading to more informed and e cient management of both NPs. Species diversity, abundance, habitat preference, and isotopic signatures differed in both NPs, depending on habitat characteristics. Syngnathids preferred sheltered macroalgal assemblages in PNIA and Cymodocea meadows in PNAC. Our results seem to indicate that PNIA is a breeding sanctuary for S. acus, which migrate seasonally. Genetic markers agreed with meristic characteristics, except for S. abaster in PNAC, suggesting the presence of cryptic Syngnathus species, and the need of further genetic and morphological sampling surveys to clarify the taxonomic status of the genus in PNAC and the consequences for species conservation. Preferential sites for future monitoring of syngnathid populations in both NPs, some actions to undertake for conservation purposes and further research priorities are proposed. Syngnathids, particularly seahorses, are agship species attracting the attention of citizens. E cient further actions will enhance public engagement with marine biodiversity, resulting also in social, economic and wellbeing pro ts.

Study areas
The study was carried out in (a) Cíes Archipelago (42°13'N, 8°54'W), in Atlantic Islands National Park (PNIA), located on the outer area of the Ría de Vigo (NW Iberian Peninsula) (Fig. 5
The study in PNIA was conducted in Cíes Archipelago, comprising three islands and various islets. Soft bottom substrates were found among rocky outcrops or in the edge of rocky reefs. To characterize sediments of the swept area in PNIA, the uppermost 2 cm of sediment were underwater manually collected using plastic pots along each transect and considering changes in bottom characteristics.
Through and crest zones were also sampled when bedforms were present. A total of 76 sediment samples was collected in spring (52) and summer (24), and conserved at 4 °C for further textural and compositional analysis.
PNAC includes a main island (Cabrera) and a group of four minor islands and several islets. Eleven subtidal sites (TR1 to TR11) including the main shallow benthic habitats present in PNAC were visited from 21st April to 1st December 2016 (17-26 °C) throughout the coast of Cabrera and Conillera islands (Fig. 6) for an overview of syngnathids distribution. We conducted 37 surveys using UVC (50 m length × The depth, position and habitat type (also substrate and seaweed assemblages in PNIA) were annotated for each sh captured. Flora and fauna nomenclature followed codes of Guiry and Guiry (2020) and WoRMS Editorial Board (2020). Swept areas were calculated according to Guerra et al. (2015), considering the effective sampling time, the net sampling distance, the distance between divers and the number of divers.

Fish collection
In UVC, syngnathids were hand-caught collected or manually extracted from the shing gear, introduced in numbered plastic bags and transferred to a support boat. In PNIA, once on land, the sh were morphologically identi ed, anesthetized with Ethyl 3-aminobenzoate methane sulfonate (MS-222; 0.1 g L − 1 ; Sigma-Aldrich Co., USA) and marked subcutaneously using visible implant uorescent elastomers (VIFE; Northwest Marine Technology Inc., USA) on the ventral surface of the trunk (pipe sh) or laterally (seahorses). All anaesthetized sh were weighted (W, g) and sized for standard length (SL, cm). In PNAC, the sh were morphologically identi ed on board, anesthetized, sized as reported above but not weighted because it was not possible to stabilize the balance in boat conditions. A fraction of the sh collected by shing in PNAC were sacri ced for sampling (stable isotopes and genetic analysis) due to their small size (with permission of NP authority).
Dorsal n samples were taken by n-clipping (Planas et al. 2008), transferred to screw-capped tubes containing 95% ethanol and conserved at 4 °C for further genetic and stable isotope analysis (SIA). The presence of previous marks (recapture events), sex, sexual status, meristics ( n rays, body rings) and body coloration were also annotated. The sexual status was recorded considering pregnancy in males and trunk shape (holding of hydrated eggs) in females. Species identi cation was evaluated genetically using DNA extracted from dorsal n samples available from PNIA and PNAC surveys. In PNIA, all shes from visual censuses were released at the capture site within 2-3 hours after sampling.
For SL measurement, the shes were placed on a plate including a measurement scale and photographed laterally (seahorses) or measured directly (pipe sh). Seahorse images were analysed in the laboratory to determine length using image-processing software (NIS Elements Nikon and ImageJ2). Seahorses were measured as head + trunk + tail length (curved measurement in seahorses) (Lourie 2003).

Sediment analysis
The analysis of sediments was only carried out in PNIA. For compositional analysis, the content of organic carbon and inorganic carbon (calcium carbonate content is equivalent to bioclastic component for this regional setting) was determined by a LECO CNS-2000 Macro Elemental Analyser at CACTI (University of Vigo). Those analyses were performed on the fractions < 2 mm, in order to avoid distortional results due to gravel components (> 2 mm, maerl, bivalve and gastropod shells).
For textural analysis, the bulk grain size distribution was performed by dry sieving. Previously to grain size analysis, the organic matter was removed using 30% H 2 O 2 for several days and salts were removed with further washings with distilled water. Afterwards, the samples were dried at 50 °C and dry sieved between 4 mm and 63 µm (sieve size intervals of 1/2 ø). The resulting grain size distribution was treated with the GRADISTAT program (Blott and Pye 2001). For statistical parameters (mean, selection, asymmetry and kurtosis or pointing of the grain size curve), the nomenclature of Folk and Ward (1957) classi cation was used.
DNA sequence analysis DNA was extracted from dorsal n tissue collected from the following morphologically identi ed specimens: i) twenty-two wild greater pipe sh (Syngnathus acus) and four long-snouted seahorses (Hippocampus guttulatus) from PNIA; and ii) six black-striped pipe sh (S. abaster) and one spotted pipe sh (Nerophis maculatus) from PNAC. Genomic DNA was isolated using NucleoSpin Tissue XS kit (Macherey-Nagel Inc., Germany) and for extremely small tissue samples further ampli ed using GenomiPhi V2 kit (Healthcare, USA).
Two mitochondrial markers (i.e. cytochrome b (Cytb) and 16S rDNA) were assayed for the molecular identi cation of sampled specimens from two divergent phylogenetic groups (Syngnathinae and

Stable isotopes analysis (SIA)
For δ 13 C and δ 15 N analysis in syngnathids, the samples were rinsed with distilled water, transferred to tin capsules, dried in oven at 60 °C for 24 h and weighted (± 1 µg). Due to the low lipid content in n samples conserved in ethanol (< 5% lipids, C/N < 3.56) (Post et al. 2007), further full defatting was not necessary (Valladares and Planas 2012). Samples were analysed at SAI (University of A Coruña) by continuous ow isotope ratio mass spectrometry using a FlashEA1112 elemental analyser (Thermo Finnigan, Italy) coupled to a Delta Plus mass spectrometer (FinniganMat, Germany) through a Con o II interface. Isotopic values are expressed as permil (‰) in conventional delta relative to VPDB (Vienna Pee Dee Belemnite) and Atmospheric Air. The precision (standard deviation) for SIA of the laboratory standard (acetanilide) was ± 0.15‰ (1-sigma, n = 10).

Geographic information
GIS was managed with ArcGIS v.10.5 software to represent the maps. Layers of bionomic maps for both NPs (OAPN, unpublished observations) were incorporated. Sampled sites/transects and syngnathid capture locations were recorded and added to a geodatabase. Biological information of the specimens (species, sex, size, weight and sexual stage) was joined to each register. Available abiotic information (topographic and bathymetric layers), as well as bionomic information, were also added to geodatabase.
Cartographic data were projected in UTM 29N/UTM 31N reference system (for PNIA and PNAC, respectively) using ETRS89 Datum.

Data analysis
All means are reported with standard deviation. The data were checked for normality and homogeneity of variances (Shapiro-Wilk and Levene's tests). Analyses of variance (ANOVA/MANOVA) were used to examine the effects of season, sex, reproductive status, length, weight and isotopic values in syngnathids. Tukey's HSD test adjusted for unequal sample sizes were performed for post hoc comparisons (Spjotvoll and Stoline, 1973). Statistical analyses were performed using R packages, with signi cance set at P = 0.05. Diversity, species richness and total number of species were estimated for seaweeds in PNIA. Differences between transects and seasons were analysed using PERMANOVA for each univariate variable. P-values were estimated with an asymptotic permutation distribution generated by the Monte Carlo method.
PERMANOVA was also used for seaweed assemblage comparisons across transects and seasons using Bray-Curtis pairwise similarities. Patterns in the structure of assemblages were visualized with principal author declares that all listed co-authors agreed to participate in the study and consent to being involved in the publication.
-Consent to publish: All listed co-authors consent to participate and publish this manuscript and understand that the text and any pictures or videos published in the article will be freely available on the internet and may be seen by the general public. The pictures, videos and text may also appear on other websites or in print, may be translated into other languages or used for commercial purposes. All authors have been offered the opportunity to read the manuscript.
-Availability of data and material: As we are working on a long-term project, the datasets used and analysed during the current study are available from the corresponding author on reasonable request.