Alien Species Threat across Marine Protected Areas of Turkey—An Updated Inventory

: This study presents the ﬁrst comprehensive assessment of alien species occurrences within the selected 11 Marine Protected Areas (MPAs) located on the Aegean and Levantine coasts of Turkey. The inventory includes a total of 289 species belonging to 15 phyla, in which lowest and highest diversities were observed in Saros Bay MPA (27 species, northern Aegean Sea) and Fethiye-Göcek Bay MPA (150 species, northwest Levantine Sea), respectively. Alien species distributions that were revealed in protected areas located in the southern Aegean and Levantine Seas were 56.9% similar (based on presence vs. absence data), while northern Aegean sites formed another distinct group. According to the breakdown of major phyla through the entire study areas, Mollusca had the highest alien diversity (22.1% of alien species), followed by Actinopterygii (19.0%), Arthropoda (15.2%) and Annelida (13.5%). Casual aliens were represented by very low proportions in each MPA, proving that most species were already established in the region, with a signiﬁcant proportion of invasive species. Regardless of the localities, the majority of the species originated from the Red Sea, whose primary pathway of introduction is the corridor, the Suez Canal. In the absence of effective management actions against bioinvasions, MPAs located along the Turkish coastline do not currently seem to provide any protection, revealing a large conservation gap to be ﬁlled.


Introduction
In the Mediterranean Sea, Marine Protected Areas (MPAs) are described as clearly defined marine geographical spaces (including subtidal, intertidal and supratidal ecosystems, together with coastal lakes/lagoons connected permanently or temporarily to the sea), which are recognized, dedicated and managed through legal or other effective means, to achieve the long-term conservation of nature with its associated ecosystem services and cultural values [1]. As clearly implied by this broad generic term, MPAs are the most effective conservation tool in the key management of marine ecosystem alterations around the world [2], by maintaining natural ecological processes, increasing ecosystem resilience, preserving genetic diversity, ensuring the sustainable utilization of species and ecosystems, restoring the biomass and structure of species assemblages, and providing socio-economic benefits [3][4][5]. These advantages, however, can only be noticeable in appropriately sited, strongly protected and effectively managed MPAs [3], which prominently protect natural habitats and species from multiple local human stressors, e.g., overexploitation of living resources and habitat destruction [6]. The recent assessments of the development of MPAs in the Mediterranean Sea during the last decade show progress towards the increased protection of marine and coastal areas; however, results are not encouraging and effectiveness of management measures are still a matter of concern [7]. Currently, 9.7% of the Mediterranean Sea is designated as MPAs, but only a small portion are associated with a properly implemented management plan and few countries have fulfilled the designated

Materials and Methods
Selection of the study sites was made according to their marine terrain coverage. Among the existing coastal and marine protected areas in Turkey, only those with over 40 km 2 of marine space were considered, including 7 MPAs in the Aegean Sea and 4 in the Levantine Sea (Table 1, Figure 1). Despite its huge area of nearly 11,000 km 2 , the Finike Seamounts special environment protection area (northwest Levant) was omitted from the analyses, since the area was characterized by unique habitats (i.e., mud volcanoes) hosting deep sea species and currently no data on the existence of any alien species were available.

Materials and Methods
Selection of the study sites was made according to their marine terrain coverage. Among the existing coastal and marine protected areas in Turkey, only those with over 40 km 2 of marine space were considered, including 7 MPAs in the Aegean Sea and 4 in the Levantine Sea (Table 1, Figure 1). Despite its huge area of nearly 11,000 km 2 , the Finike Seamounts special environment protection area (northwest Levant) was omitted from the analyses, since the area was characterized by unique habitats (i.e., mud volcanoes) hosting deep sea species and currently no data on the existence of any alien species were available.   Table 1).
In order to determine marine faunal hotspots in Turkey, species occurrence records (comprising all extant taxa) associated with accurate locality information and/or exact coordinates were plotted on a map where the Turkish coasts were divided into equivalent squares of 15 × 15 km 2 , which were then entered into an Excel file for further analysis (for full account see [29]). Initial entries were based on governmental marine biodiversity studies conducted at 8 MPAs [20][21][22][30][31][32][33], with a number of alien taxa as follows: Saros Bay  Table 1).
An updated version of this file (including floral elements) focusing solely on alien taxa was used in a recent comprehensive inventory [12], which formed the core data of the present study. For better presenting the distributional patterns of alien species, we also included previously unpublished observations (not new for the country or the region but overlooked occurrences in corresponding MPAs). The presence vs. absence matrix was used to reveal similarities within the MPAs, using cluster analysis with a group average sorting performed with the PRIMER 5.2 software [34].
The terminology of alien species followed [10], referring to species introduced by human activities, while species undergoing climate-shifted range expansions, without humanassisted spread, were not considered to be alien. Species that formed self-maintaining populations with at least two records in the area (three records for fish) spread over time and space, were classified as established species, while those having been recorded only once (no more than twice for fish) with no evidence of self-sustaining populations were classified as casual species [12]. Established aliens whose populations had proliferated and rapidly expanded their distributional range by overcoming biotic and abiotic barriers in the region were treated as invasive species [35]. Only primary pathways were considered in the classification of the pathways for the alien species introductions to Turkey (corridor via the Suez Canal, ships and aquaculture). Cryptogenic (species with no definite evidence of their native or introduced status) and questionable species were all left out of the inventory.

Results
The alien species inventory of Turkish MPAs included a total of 289 species belonging to 15 phyla, the majority of which had established successfully breeding populations throughout the study sites (61.6%) and a significant portion displayed an invasive character (29.4%) (Appendix A). Mollusca ranked first in terms of the number of alien species (64 sp.), followed by Chordata (55 sp.), Arthropoda (44 sp.) and Annelida (39 sp.). Percentage distribution of alien species phyla in each MPA is presented in Table 2. There were significant differences in local alien biodiversity, with a clear decreasing pattern in a clockwise direction from Levantine towards the northern Aegean Sea coasts. The cluster analysis showed that MPAs were clearly separated into two groups (northern Aegean Sea areas vs. southern Aegean Sea and Levant Sea combined) at 30.0% similarity, which split further at 46.0% and 56.9% similarities ( Figure 2). The numbers of taxa were typically higher in Levantine localities (ranging from 93 to 150 sp.), gradually decreasing to  It is striking that the number of casual aliens is quite low throughout the entire study area (none in Foça and Köyceğiz-Dalyan, ranging from 1.2% to 6.9% elsewhere), where the great majority of the species are either characterized by successfully breeding established populations or possess an invasive character ( Figure 3). The proportion of invasive taxa ranged from 37.0% (Göksu Delta) to 62.3% (Köyceğiz-Dalyan), displaying a largescale impact regardless of their occurrence localities. The two invasive fish, Lagocephalus sceleratus (Gmelin, 1789) and Siganus rivulatus (Forsskål, 1775) were present in each of the MPAs, while some other noxious species, such as Caulerpa cylindracea (Sonder, 1845) (Chlorophyta), Asparagopsis armata (Harvey, 1855) (Rhodophyta), Leodice antennata (Savigny, 1820) (Annelida), etc., were absent in just a few sites.
According to their origins, there was a very pronounced dominance of Red Sea originated species (201 sp., out of 289 sp.), while the contribution of rest of the categories was set at low levels ( Figure 4). This was an expected result, since Turkey was geographically located close to the Red Sea, which explained why corridors (Suez Canal) were the main vector of the species introductions ( Figure 5). Ship-transferred species were higher in proportion at the north Aegean MPAs (Saros Bay, Ayvalık Islands, Foça and Karaburun-Ildır, ranging between 28.1-48.1%), significantly reducing to levels of 8.7-18.0% in the rest of It is striking that the number of casual aliens is quite low throughout the entire study area (none in Foça and Köycegiz-Dalyan, ranging from 1.2% to 6.9% elsewhere), where the great majority of the species are either characterized by successfully breeding established populations or possess an invasive character ( Figure 3). The proportion of invasive taxa ranged from 37.0% (Göksu Delta) to 62.3% (Köycegiz-Dalyan), displaying a largescale impact regardless of their occurrence localities. The two invasive fish, Lagocephalus sceleratus (Gmelin, 1789) and Siganus rivulatus (Forsskål, 1775) were present in each of the MPAs, while some other noxious species, such as Caulerpa cylindracea (Sonder, 1845) (Chlorophyta), Asparagopsis armata (Harvey, 1855) (Rhodophyta), Leodice antennata (Savigny, 1820) (Annelida), etc., were absent in just a few sites.   According to their origins, there was a very pronounced dominance of Red Sea originated species (201 sp., out of 289 sp.), while the contribution of rest of the categories was set at low levels ( Figure 4). This was an expected result, since Turkey was geographically located close to the Red Sea, which explained why corridors (Suez Canal) were the main vector of the species introductions ( Figure 5). Ship-transferred species were higher in proportion at the north Aegean MPAs (Saros Bay, Ayvalık Islands, Foça and Karaburun-Ildır, ranging between 28.1-48.1%), significantly reducing to levels of 8.7-18.0% in the rest of the coastal areas. Saros Bay was the only locality that shipping-oriented introductions outnumbered Suez Canal introductions (13 sp. vs. 9 sp., respectively); the latter vector was dominant elsewhere. Aquaculture was the least impacting vector and only four such species were present in MPAs (for example, the Pacific Ocean originated invasive bivalve Ruditapes philippinarum (Adams & Reeve, 1850)).

Discussion
The present inventory revealed that (i) existing MPAs were exposed to different levels of alien biota impact, with a clear decreasing trend from the Levantine coasts to the northernmost regions of the Aegean Sea; (ii) the Suez Canal was the most important vector for species introductions, and was also responsible for the existence of the huge proportion of Red Sea-originated biota; and (iii) the high rates of established aliens and invasive species were obvious in every MPA. These results were in accordance with the general trends observed for the alien biota of the entire coast of Turkey during the last decade [12,35]. The low number of species determined in the northern Aegean protected areas should be carefully monitored, since they may be related to lower research efforts, in comparison to the northern Levantine coasts. The significant taxonomic similarity we found between the southern Aegean and Levantine MPAs was an issue to be taken seriously and may have indicated an ongoing biotic homogenization event, although concrete data for proving this phenomenon are currently lacking. It is a known fact that species invasions and extinctions lead to a decrease in β-diversity, by increasing the genetic, taxonomic or functional similarity of two or more locations over time [36].
When Mediterranean Sea coastal countries were taken into consideration, more alien taxa were recorded along the Turkish coastline than anywhere else. For example, the reported diversity was 452 sp. in Israel [37], 265 sp. in Italy [38], 214 sp. in Greece [39], 136 sp. in Tunisia [40] and 73 sp. in Libya [41]. Receiving 185 new alien species introductions just during the last decade, the immense impact of bioinvasions to the Turkish marine realm reflected the diversity estimates, now reaching to 539 species, 404 of which were established in the region [12]. Thus, it was not surprising that there was a high number of alien species (289 sp.) throughout Turkish MPAs, which we believe was merely an underestimation and could certainly be increased by further research. The relevant data are currently incomparable to any other regional datasets, due to the lack of country-based comprehensive alien species checklists, which focus on their presence in protected areas of the

Discussion
The present inventory revealed that (i) existing MPAs were exposed to different levels of alien biota impact, with a clear decreasing trend from the Levantine coasts to the northernmost regions of the Aegean Sea; (ii) the Suez Canal was the most important vector for species introductions, and was also responsible for the existence of the huge proportion of Red Sea-originated biota; and (iii) the high rates of established aliens and invasive species were obvious in every MPA. These results were in accordance with the general trends observed for the alien biota of the entire coast of Turkey during the last decade [12,35]. The low number of species determined in the northern Aegean protected areas should be carefully monitored, since they may be related to lower research efforts, in comparison to the northern Levantine coasts. The significant taxonomic similarity we found between the southern Aegean and Levantine MPAs was an issue to be taken seriously and may have indicated an ongoing biotic homogenization event, although concrete data for proving this phenomenon are currently lacking. It is a known fact that species invasions and extinctions lead to a decrease in β-diversity, by increasing the genetic, taxonomic or functional similarity of two or more locations over time [36].
When Mediterranean Sea coastal countries were taken into consideration, more alien taxa were recorded along the Turkish coastline than anywhere else. For example, the reported diversity was 452 sp. in Israel [37], 265 sp. in Italy [38], 214 sp. in Greece [39], 136 sp. in Tunisia [40] and 73 sp. in Libya [41]. Receiving 185 new alien species introductions just during the last decade, the immense impact of bioinvasions to the Turkish marine realm reflected the diversity estimates, now reaching to 539 species, 404 of which were established in the region [12]. Thus, it was not surprising that there was a high number of alien species (289 sp.) throughout Turkish MPAs, which we believe was merely an underestimation and could certainly be increased by further research. The relevant data are currently incomparable to any other regional datasets, due to the lack of country-based comprehensive alien species checklists, which focus on their presence in protected areas of the Mediterranean Sea. As previously outlined, Mediterranean MPAs face common challenges including a lack of baseline information and the inefficient reporting of biological invasions [10,18,42], constituting crucial data to draw robust conclusions in the effective management of protected sites [14,43].
No control of alien species is feasible that would not also harm other components of the biota once an invasion process is underway [10,44], thus the objective highlighted in the Convention on Biological Diversity (CBD) at art. 8(h), calling for contracting parties "as far as possible and as appropriate, (to) prevent the introduction of, control or eradicate those alien species which threaten ecosystems, habitats or species" can only be partially fulfilled, especially by the eastern Mediterranean countries where the existing invasion process is unique. By ratifying and signing the "International Convention for the Control and Management of Ship's Ballast Water and Sediments" in 2014, Turkey took an important step that could considerably decrease risks of ship-borne introductions, but the troublesome vector, the Suez Canal, remains wide open, unquestioningly destroying the strength of the proposed eradication measures. The first ever government-supported incentive notification on the bounty hunting of the noxious Lagocephalus sceleratus along Turkish coasts (for a price of EUR 0.5 per each captured individual) ended with little success, where only 46,000 individuals were eradicated out of the targeted 1 million fish [12]. Although L. sceleratus fishery was also conducted within the existing MPAs, the captured quantities in each site have not been announced yet. The above-mentioned notification has been updated recently (official gazette no. 31524, dated 27 June 2021), now encouraging the capture of all Tetraodontids inhabiting Turkey (native pufferfish were incomprehensibly included as well) from now until the end of 2023. There are also efforts to cull and create consumption demand to decrease the population trend of Pterois miles in the Kaş-Kekova region [45], though currently no official announcement has been made.
MPAs alone are unlikely to be sufficient in preventing biological invasions in the Mediterranean Sea, as evidenced from results of several recent studies [46][47][48][49]. On the contrary, the Red Sea invaders, which came from a highly competitive environment, find a suitable feeding and shelter ground in the Mediterranean MPAs, whereby they increase their population sizes enormously and utilize the areas as stepping stones in their distribution expansions [16]. Therefore, alien species' harvesting should be promoted in MPAs where they benefit from fishing bans and restrictions that apply within the MPAs. Considering the highly connected nature of the Mediterranean Sea, a basin-wide ecosystembased policy on bioinvasions is required [10]. The effective management of the Suez Canal is of utmost importance, and an issue which all Mediterranean countries have failed to put on their conservation agendas so far, highlighting the urgent need of international cooperation in the management of alien species [12].
In terms of their resilience to invasive species, an unmanaged MPA is no different from an unprotected coastal area, reflecting the current situation we are experiencing in Turkey. The drastic impact of invaders are therefore the expected and inevitable result of numerous "lacks", including a lack of legal background, lack of marine management plans (available only for Foça, Gökova Bay and Kaş-Kekova with no measures defined against the struggle with aliens), the lack of public/governmental awareness and understanding of the impacts of invasive species, the lack of trained staff devoted to monitor the impacts of alien taxa, the lack of funds to regularly carry out monitoring research (at least for selected invasive species), and the lack of fishery regulations in favor of apex predators, etc.