Laminaria rodriguezii, a Mediterranean deep-water kelp is in danger in the Adriatic Sea – part II

As Jabuka Pit is one of the most important trawling areas in the Adriatic (UNEP/CBD/EBSA/WS 2014), the disappearance of L. rodriguezii is likely a result of bottom trawling activities. Mechanisms of impact would be direct, by removal of algae as bycatch, and indirect by stirring up sediments (Jones 1992; Palanques et al. 2001). In the Jabuka Pit, a layer of dense bottom water can in some years persist over longer periods (Artegiani et al. 2001). Resuspended sediments, which circulate inside the enclosed water mass within the depression, would possibly decrease light intensity under the light compensation point of the algae. It was pointed out in other parts of the Mediterranean that frequent trawling might influence L. rodriguezii (UNEP/IUCN/GIS Posidonie 1990; Joher et al. 2012), but the disappearance of L. rodriguezii from Jabuka and Biševo is up to now the greatest reported decline of this species. The area of Palagruža is not an intensive trawling area like Jabuka (IOF, unpublished data) but still has important commercial fishing activity. Moreover, the locations where L. rodrigueziiwas recently reported by fishermen, observed and collected by ROV and grab (Fig. 1c), correspond to an area where trawling is not allowed (minimum 2 nautical miles from the Island) and only small scale fishing like trammel nets is performed. In addition, strong bottom currents in the Palagruža may allow fast removal of fine sediment dispersed in bottom water due to bottom trawling, and possibly also clean the surface of thalli from deposited sediments. The historical records of L. rodriguezii at Jabuka, at 150–260 m depth, were the deepest of that species in the entire Mediterranean. Such a depth is in the range of the deepest records of seaweeds worldwide (Littler et al. 1985; Markager and Sand-Jensen 1992). Sporadic findings of the alga in the Palagruža area during MEDITS surveys during the last 14 years were also in the depth range between 170 and 200 m. Collected specimens did not have hold-fasts, thus it is not possible to conclude with certainty whether they were attached or just dispersed fragments. Bottom currents of 50 cm s⁻¹ might be sufficient for proper fragment drifting. However, according to data provided from ADCP, bottom currents with E–SSE directions (Fig. S6) could not transport fragments from Palagruža Island to locations where L. rodiguezi was collected during MEDITS campaigns (Fig. 1b; 10–30 km in N–NE direction). Anyway, depths reported by earlier studies based on bottom trawling surveys must be considered with a certain caution. Without direct observation by ROV, exact data on light intensity on the bottom and ideally supporting, ecophysiological experiments, we cannot be sure that L. rodriguezii can develop at such extreme depth. The lower depth limits for kelp is assumed to be at 1 % of surface irradiance (Lüning and Dring 1979); however, this may not apply for L. rodriguezii. Secchi values measured in the last 15 years were around 20 m, with a maximum of 30 m. Such surface values of transparency only might be indicative and indeed reveal no information about the light intensity reaching the bottom. Giaccone (Giaccone 1967) noted how Secchi disks were visible up to 35 m near the island of Ustica (Italy) where L. rodriguezii populations occurred between 45 and 85 m depth. At 85 m, the seabed at Ustica was too steep and formed an environment with low luminosity that did not allow life of photosynthetic organisms. The locations where the alga was collected near Palagruža and historically at Jabuka are mostly flat, and therefore probably allow development of the alga at greater depth (no shading effect) even if the transparency is equal. Similar to the Palagruža site, costal detritic bottoms occur in Spain, where L. rodriguezii has been mainly reported between 70 and 80 m, with a maximum at 95 m (Joher et al. 2012; Sergi Joher Sais pers. comm.).

There is not much information on oceanographic conditions prevailing in the locations of other populations of L. rodriguezii in the Mediterranean. However, most reports suggest that favourable conditions include a hard coralligenous or rhodolithic substratum, dim light, highly transparent open sea water due to very low abundance of suspended particles, water temperatures permanently below 15 °C, and constant, mostly unidirectional weak to strong currents (Ercegović 1960; Giaccone 1967; UNEP/IUCN/GIS Posidonie 1990; Bellan-Santini et al. 2002). This study presents for the first time precise long-term measurements of temperature, currents, salinity and nutrients in the bottom seawater in the documented distribution range of L. rodriguezii. Seawater temperature is almost constantly around 14 °C with temperature extremes slightly above 15 °C and below 12 °C being a very rare occurrence over 60 years of seasonal measurements (Fig. S4). Current measurements from the ADCP station close to Palagruža showed that bottom currents were relatively strong and very stable. The strong bottom current in the area is the result of formation and sinking of cold and dense North Adriatic Dense Water (NAdDW) formed in the North Adriatic shelf and coastal area (Mihanović et al. 2013). One of the two major pathways of the NAdDW from the shallow northern Adriatic shelf and Jabuka Pit to the deep South Adriatic Pit is located to the north of Palagruža (Vilibić et al. 2004; Janeković et al. 2014). Predominant flow directions measured at the ADCP station were between east and southeast, corresponding to the northern pathway of the NAdDW outflow across the Palagruža Sill. The NAdDW formation event that occurred in January/February 2012 was exceptional and bottom density currents between March and June 2012 were particularly intense. Still, even after that period, weaker but persistent bottom currents, flowing mostly towards E–SE were observed in the area, till the end of measurements in December 2012 (Fig. S7). Both current and temperature measurements corroborate the global perception how persistent and stable currents and low temperatures play important roles in the formation of favourable condition for the development of L. rodriguezii. Therefore, NAdDW outflow with both strong currents and low and relatively stable temperatures is probably essential for the development of L. rodriguezii in the Central Adriatic. It thus appears that presence of L. rodriguezii in the Adriatic Sea coincides with oceanographic conditions that differ from the upwelling of cold waters at deep water kelp sites in tropical regions (Graham et al. 2007).

Total inorganic nitrogen, orthophosphate and orthosilicate concentrations in the bottom layer (Table 3) were generally higher than in the surface or middle layer of the water column (Vilibić et al. 2012), due to remineralization processes from the sediment or to advective mixing of water masses from the Mediterranean. Increased TIN concentrations are particularly expressed during the summer season (Barić et al. 2002), as a consequence of more intensive nitrogen flux from sediment that can be an additional source of nutrients facilitating algal growth. Laminaria rodriguezii has a disjunct distribution within the Mediterranean Sea. The population in the Adriatic Sea, although genetically similar according to ITS sequences (see below), does not seem to be connected to the populations in the Western Mediterranean, since there are no intermediate populations known. Long distances from the next populations and adverse water mass circulation probably inhibit transport of any reproductive phase. Natural reintroduction in the case of total regional extinction in the Adriatic Sea is therefore not possible. The IUCN Red List criteria are commonly used for evaluating the extinction risk of species (IUCN 2012). However, applying these to macroalgae is not simple and there are only a few seaweed species in the world on which they have been applied (Brodie et al. 2009; Phillips and Blackshaw 2011), basically due to lack of information on historical and recent distributions and the species’ biology and ecology. Except for the present comparison of historical and recent records, this also applies to L. rodriguezii in the Adriatic. There is still limited knowledge on its biology, population density and exact distribution as well as changes of the last two over exact periods of time, which are relevant criteria for IUCN evaluation. However, such data are clearly very difficult to obtain for such a relatively inaccessible species. In the light of the drastic decline of more than 85 % of the extent of occurrence in the Adriatic, which is estimated to be nowadays around 1200 km², and its disappearance in two of three historical distribution areas (Jabuka Pit and Biševo Island area), as well as still ongoing, strong bottom trawling activities (UNEP/CBD/EBSA/WS 2014), we propose that L. rodriguezii should be classified as “Endangered” under IUCN criteria B1ab(i,iii,iv), ver 3.1.in the Adriatic Sea.
A management plan should therefore be developed in order to enable maintenance of L. rodriguezii in the area of Palagruža Island, its last refuge within the Adriatic. Similar measures were proposed for an area with L. rodriguezii populations in the Menorca Channel (Barbera et al. 2012). The decline of canopy-forming species, especially those in shallow waters, is an expanding, worldwide trend, largely due to different human activities while only sporadic local actions are taken to remedy this situation (e.g. Thibaut et al. 2015; Yesson et al. 2015, and references therein). Further ROV expeditions should be organized with the objective of not destructively quantifying the abundance of L. rodriguezii and to establish detailed parameters for a further monitoring program. The greatest risk for the species appears to consist in continued bottom trawling. Temperature increases in Mediterranean deep water may represent another threat. The distribution of shallow water kelp species was reported to react rapidly to small changes in seawater temperatures, possibly due to temperature requirements for reproduction (Bartsch et al. 2013). On the other side, Yesson et al. (2015) showed that temperature alone can not account for long-term increases and decreases in abundance of the large brown seaweeds observed around the British Isles and suggested a combination of both physical and biological factors as drivers of these changes. Therefore, the possible future impact of seawater temperature increase on deep water L. rodriguezii must remain speculative, as there are no ecophysiological data on growth and reproduction of this species. So far, long-term measurements of sea bottom water in the Palagruža area have not shown any statistically significant temperature increase (Vilibić et al. 2012).

Sequence similarity in a highly variable part of ITS and the Rubisco spacer between our Croatian and Italian samples of L. rodriguezii suggested that there may not be much genetic differentiation between individuals from the two sites. The sequences of all markers confirmed that L. rodriguezii belongs to the Laminariaceae, which is at present (Guiry and Guiry 2015) understood as a clade formed by the type L. digitata and about 20 other species, including all Atlantic kelps with digitate blades, such as L. hyperborea, L. ochroleucaand L. pallida. According to the sequences studied, particularly ITS, L. rodriguezii is closely related to a clade formed by L. ochroleuca, L. pallida and L. abyssalis. Our sequences suggest that adaptation to deep water in L. rodriguezii and L. abyssalis possibly evolved independently and that the two are different species, which is born out by the fact that the adventive blades characteristic for L. rodriguezii are absent in L. abyssalis. However, these are preliminary data from three markers and few individuals. A more comprehensive phylogeographic hypothesis for L. rodriguezii would require sequences for several markers from replicate individuals from different populations of the two deep-water species as well as of all Atlantic kelp taxa. It is unknown if L. rodriguezii can cross with the members of the AOP clade; cross-fertility has been shown experimentally for L. pallida and L. abyssalis (Dieck and De Oliveira 1993). It is urgent to isolate gametophyte cultures of L. rodriguezii in order to study genetics and physiology of this impressive species.