The target of the present study was to assess the mortality incidence on P. nobilis in local populations along the southeast coast of Italy (the Apulia peninsula) following the mass mortality event. At the same time, species distribution and densities were investigated in the Adriatic and the Ionian Sea, allowing us to build a picture of species populations before the MEE.
Spatial distribution
Concerning the data collected on P. nobilis distribution in the Apulia region before the MME, unfortunately, there is a lack of information at the regional scale and the reports found in the scientific literature are limited only to semi-enclosed systems such as the Taranto basins (Centoducati et al., 2007; Tiscar et al., 2019; Tursi et al., 2018) and the Aquatina lagoon (Marrocco et al., 2018; Pinna et al., 2018). No large-scale monitoring program on P. nobilis has been carried out previously along the Apulian coast, although these types of surveys are indispensable for the management of a protected species and must become mandatory for a critically endangered species such as the P. nobilis. To address this information gap, a data-gathering survey was undertaken to assess the current status of the P. nobilis population. In this study, thirty-one areas distributed along the entire Apulian region coast were explored with standardized visual-census surveys. Recently dead specimens with still intact shells allowed us to map P. nobilis distribution and densities in each area before the Mass Mortality Event.
Along the Ionian coast, P. nobilis was detected in all the areas studied, highlighting a continuous distribution of the species. This result indicated that the Ionian coastal region presents environmental conditions compatible with the species requirements throughout the entire coast extension. However, the distribution along the Adriatic coast was not continuous. Pinna nobilis occurrence was recorded in the areas surveyed in the south, from A7 to A17, but no traces were found along the northernmost areas except the Tremiti archipelago. These results indicate that the northernmost Adriatic coast of the region probably does not meet the environmental conditions suitable for hosting this species. The Gulf of Manfredonia may have represented the sole exception in the past. In this area, we collected multiple reports from fishermen indicating the presence of the species in a local Cymodocea nodosa meadow before the 1980s. Still, no shell traces were observed during the surveys. Therefore, we can assume that excessive fishing and anthropogenic activities in this area are likely to have caused the species to disappear many decades ago.
Mortality incidence
Data regarding the effects of the MME in Apulian populations is scarce. Panarese et al., in 2019 reported the advent of the disease in Mar Piccolo di Taranto but without describing the disease incidence. In this study, we recorded a mortality incidence of 100% in all habitat types, bathymetries and basins, demonstrating the seriousness of the situation along the entire Apulian coast. To achieve an exhaustive evaluation of the state of P. nobilis populations, different environments such as, inshore, offshore, and lagoon and marine protected areas were investigated.
While the availability of nutrients and the trophic conditions are assumed to be very different between offshore and inshore systems, the archipelago of Tremiti islands, located 13 miles from the coast, showed no differences in mortality incidence from sites along the coast, evidencing the same critical conditions in both inshore and offshore environments.
Many Mediterranean lagoon systems, including the Ebro Delta, Mar Menor Lagoon in Spain (Prado et al., 2021), the Rhone delta, Leucate and Thau in France (Foulquie et al., 2020; García-March et al., 2020; Katsanevakis et al., 2021), Venice, Grado-Marano and Faro in Italy (Curiel et al., 2021; Donato et al., 2021; Russo, 2017), Bizerte in Tunisia (Katsanevakis et al., 2021) are considered the last healthy shelters for P. nobilis populations in the central and eastern Mediterranean basin (Foulquie et al., 2020). These areas seem to offer a degree of resistance against the disease and are all characterized by high seasonal fluctuations of environmental parameters, such as temperature and salinity. It has been supposed that the effect of these fluctuations could make these environments less suitable for the spread of the disease and reduce the rate of transmission (Foulquie et al., 2020; Prado et al., 2021). For this purpose, three lagoon systems were investigated in the present study: Aquatina di Frigole, Varano, and Alimini. Aquatina di Frigole is the sole lagoon in the region where previous studies have reported the presence of P. nobilis (Marrocco et al., 2018). The results of the surveys highlighted a critical condition, showing that also the population of P. nobilis in the Aquatina di Frigole lagoon have totally collapsed. Considering that the lagoon refuges currently represent the main source of larval production for P. nobilis recruitment (Foulquie et al., 2020) the collapse of these populations confirms the severity of the situation for species conservation along the southeast coast of Italy.
Regarding the timeframe of the spread of the MME along the Apulian cost, the first report of the infection dates back to 2018 (Tiscar et al., 2019), in the Mar Piccolo di Taranto. Compared to the first MME event observed in the Spanish coast in 2016 (Darriba, 2017; Vázquez-Luis et al., 2017), over a period of two years, the disease has spread from the western to the eastern basin of the Mediterranean Sea. Our surveys, carried out in 2020, showed that 91% of the shells were still undamaged and with joined valves. Based on the state of conservation of the shells (Scarpa et al., 2021) it is possible to hypothesize that the death of the specimens was a recent phenomenon that had occurred in Apulia in the two years preceding our surveys, and most probably shud should be dated back to 2019.
Habitat association and trophic ecology
In our surveys, P. nobilis showed transverse distribution among habitat types occurring both in marine and lagoon systems, inside and outside seagrass meadows. Nevertheless, when the pen shell distribution was analysed on a spatial macro and mesoscale, the results showed an overlap with the distributional range of seagrass meadows. In fact, all the coastal areas where the P. nobilis was found were also characterized by the presence of seagrass meadows, either directly in the area under examination or in proximity. In fact, in the northern Adriatic coast of the region, where extended seagrass meadows are absent, no trace of P. nobilis was encountered, except in the Tremiti archipelago where both P.oceanica meadows and pen shells were found to be present. Focusing specifically on the occurrence data, it emerged that P. nobilis is associated with various seagrass species such as P. oceanica, C. nodosa and Zostera sp., which in turn are distributed in heterogeneous habitat including those of marines and lagoon-estuarines. From these results it can therefore be concluded that there is a macro and mesoscale association between P. nobilis and seagrass meadows sensu lato. This hypothesis is consistent with much of the data in literature reporting ubiquitous distribution of the species both in lagoon-estuarine ecosystems (Curiel et al., 2021; Donato et al., 2021; Foulquie et al., 2020; Katsanevakis et al., 2021; Prado et al., 2021; Tsatiris et al., 2018) and in marine ecosystems (Basso et al., 2015; Katsanevakis et al., 2021; Kersting et al., 2019; Šarić et al., 2020; Vázquez-Luis et al., 2017; Zotou et al., 2020).
However, taking a closer look at their micro-scale distribution, the pen shells in our surveys were commonly recorded also outside the seagrass meadows boundaries, at times up to one kilometer away. Hence, seagrass sheltering can potentially be ruled out as the sole explanatory factor for the distribution pattern of the species. The pattern emerging from our study led us to hypothesise that a trophic link with the seagrass detritus food-chain may explain both the macro-mesoscale association with seagrass species and the microscale cross-boundary distribution. In fact, seagrass detritus is highly refractory, since it is largely exported to the nearby areas where it can represent the major food source for other invertebrates (Boncagni et al., 2019a; Danovaro, 1996). This hypothesis is consistent with the stomach contents observations reported by Davenport and co-authors (2011) indicating detritus as the bulk component, accounting for 95% of the total ingested material.
One of the main factors underlying the distribution pattern in benthic invertebrates is indeed food availability (Palmer et al., 1996; Tregenza, 1995). According to the Ideal Free Distribution theory (IFD), in ecology the individuals in a population disperse to different resource patches within their environment, minimizing competition and maximizing fitness (Whitham, 1980). When the IFD assumptions are met, the number of individuals who aggregate in patches is proportional to the amount of food resource available in each one. Accordingly, the distribution of large, long life and sessile organisms such as P. nobilis would be expected to depict the species trophic supply, by analyzing the resources available in those patches.
Studies on the seagrass system energy flow have shown that seagrass debris must be fractionated before entering the food chain (Danovaro, 1996). In this way, plant material becomes fine particulates moving in the boundary layer over the sediment-water interface (Danovaro et al., 1998; Danovaro and Fabiano, 1997). These processes take time, and while the matter is transported, heterotrophic bacteria grow exponentially, turning it into a high quality and protein enriched food for consumers. Hence, bacteria adhering to seagrass detritus may play a key role in this benthic food chain and sediment-water interface consumers may incorporate more energy from associated microbes than from the detritus itself (Boncagni et al., 2019b; Danovaro et al., 1998; Rakaj et al., 2019, 2018). On the basis of these considerations, it is reasonable to conclude that the quantity, composition and origin of the suspended particles are regulated by a drift mechanism and that this mechanism may explain local densities of P nobilis as a response to sinking rates and resuspension effects. The assumption of the species’ ability to feed on seagrass detritus, together with the high biomasses reached (large size specimens and high density), lead us to suppose that P. nobilis played a key role in the processing of matter and in the energy pathway deriving from seagrass detritus in Mediterranean coastal areas. This makes the repercussions of the MME not only a problem of conservation, but also and above all, an ecological-functional issue.
We can, therefore, conclude that Mediterranean seagrass meadows not only constitute a habitat for P. nobilis, but most of all, a food source through refractory detritus generation which is transferred and transformed outside the meadows. Unfortunately, literature is lacking on this topic and further investigations are needed to define the trophic role and function of these filter feeders in the different seagrass meadows.
Local densities
The density values that emerged were significantly different among basins. In the Adriatic Sea, where all the coastal values were recorded, the densities were consistently lower than those reported in the Ionian Sea, except for the two southernmost areas. In the Adriatic basin, it was also possible to recognize a north-south trend when considering the densities of pen shells in the coastal areas. In the central area of the region, the values recorded were fewer than 0.07 individuals/100 m2(A7-A15). Instead, in the southernmost areas, density values of between 0.22 and 0.41 individuals/100 m2 were recorded (A16; A17). Although the values recorded along the southern coast of the region were much greater than those recorded in the central coast, they were far lower than those reported by Čižmek et al. (2020) in the Croatian coast (North Adriatic Sea). Similar values to ours within the same basin were reported by Celebicic et al. (2018) in Bosnian waters (0.12 individuals/100m2).
On the other hand, in the Ionian areas, the values recorded were consistently higher than 0.1 individuals/100m2. The highest density was recorded in the Gulf of Gallipoli (with a mean value of 3.94 individuals/100m2), whilst the mean density value for the entire basin was 0.88 individuals/100m2. The values recorded in the Mar Grande di Taranto, were higher than those reported by Centoducati et al. (2007) (0.1–0.7 ind/ha2). From interviews with fishermen it emerged that illegal trawling in this area has strongly impacted the natural populations of the Mar Grande di Taranto, and a partial reduction in recent years could explain the slight increase in density compared to the 2004 survey data (Centoducati et al. 2007).
Analyzing the data collected for Aquatina di Frigole - the only lagoon system in which the presence of P. nobilis specimens was detected- the density value recorded of 0.33 individuals/ 100m2 is far lower when compared with that reported in the Thau lagoon, Grado-Marano lagoon and Mljet lagoon (Curiel et al., 2021; Foulquie et al., 2020; Mihaljević et al., 2021), but is higher than the density values recorded in the Lake Faro (Sicily) in 2010 when density values of 0.18 individuals /100m2 were recorded. In this area, the density values decreased dramatically after the MME, dropping to 0.07 individuals /100m2 in 2020, although some live specimens are still present (Donato et al., 2021).
In interpreting our data, it should be considered that the surveys were carried out employing an extensive sampling protocol conceived to assess wide surface densities on coastal areas investigating across several habitat types. Therefore, literature density values focused only on local areas or habitat patchiness that were not randomly selected must be contextualized when compared with these data. In addition, given the scale of the prsented surveys, emphasis must be given to P. nobilis absence data of which the literature appears poor. Indeed, contrary to presence data, reliable absence data are difficult to obtain requiring much greater effort to rule out a rare occurrence (Gu & Swihart, 2004). The absence data obtained in this study derive from the merger of two different data types. The first one come from the local ecological knowledge arisng from the local fishermen intervew, which allowed us to confirm our data exluding pointlike occurrence in the same area. The other one dervies from a comprehensive view during the field surveys. Indeed the scuba diver overview was at least 10 times wider than 50 cm for side around the rope hence, the absence perception can be extended over a much larger investigated surface. Considering these two types of information together, we can assume the absence data exhaustive and full of information.