assessment of the quality of coralligenous reefs: The method 34 ..... techâ methods that could be employed for monitoring activities on rocky benthic .... information about the orientation (compass reading) and the slope ...... working on coralligenous reefs, to develop an index universally valid and the necessity of testing.
University of Genoa DiSTAV - Department of Earth, Environment and Life Science European PhD in Environmental Science (Marine Science)
Integration between hi-tech and “low-tech” methods for the study of coastal rocky reef assemblages in the Ligurian Sea
PhD thesis of Giulia Gatti
Advisors: Prof. Paolo Povero Dr. Marina Locritani
March 2014
Table of contents 1. General Introduction ------------------------------------------------ 4 2. Bionomic transects and long-term changes in rocky benthic reefs. 11 2.1 Introduction ----------------------------------------------------- 11 2.2 Materials and methods ----------------------------------------- 13 2.2.1 Study area ------------------------------------------------ 13 2.2.2 Historical data and present field work ------------------ 14 2.2.3 Data management --------------------------------------- 14 2.3 Results----------------------------------------------------------- 15 2.4 Discussion ------------------------------------------------------- 25 Appendix 2.1 -------------------------------------------------------- 27 Appendix 2.2 -------------------------------------------------------- 28 3. A new Rapid Visual Assessment (RVA) approach for the characterisation and the assessment of the quality of coralligenous reefs: The method 34 3.1 Introduction ----------------------------------------------------- 34 3.2 Material and methods ------------------------------------------ 36 3.2.1 Study area and field activities --------------------------- 36 3.2.2 Data management --------------------------------------- 37 3.2.2.1 Characterisation ---------------------------------------- 37 3.2.2.2. Quality assessment ------------------------------------ 37 3.3 Results----------------------------------------------------------- 39 3.3.1 Characterisation ------------------------------------------ 39 3.3.2 Quality assessment -------------------------------------- 40 3.4 Discussion ------------------------------------------------------- 46 4. A new Rapid Visual Assessment (RVA) approach for the characterisation and the assessment of the quality of coralligenous reefs: Application and validation 49 4.1 Introduction ----------------------------------------------------- 49 4.2 Materials and methods ----------------------------------------- 51 4.2.1 Study area and field work ------------------------------- 51 4.2.2 Data management --------------------------------------- 53 4.2.2.1 Robustness to observer biases ------------------------ 53 4.2.2.2 Characterisation and quality assessment-------------- 54 4.3 Results----------------------------------------------------------- 54 4.3.1 Robustness to observer biases -------------------------- 54 4.3.2 Characterisation ------------------------------------------ 55 4.3.3 Quality assessment -------------------------------------- 55 4.4 Discussion ------------------------------------------------------- 59 4.4.1 The RVA protocol ---------------------------------------- 59 4.4.2 Robustness to observer biases -------------------------- 60 4.4.3 Characterisation and quality assessment --------------- 60 4.5 Conclusions ----------------------------------------------------- 62 5. Quantitative methods for the monitoring of infralittoral assemblages: Point Intercept Transects vs Quadrats ------------------------------------ 63 5.1 Introduction ----------------------------------------------------- 63 5.2 Material and Methods ------------------------------------------ 64
2
5.2.1 Study area and sampling designs ------------------------ 64 5.2.1.1 Bergeggi (B1): sampling design ------------------------ 65 5.2.1.2 Portofino promontory (P1): sampling design. --------- 66 5.2.1.3 Portofino promontory (P2): sampling design ---------- 67 5.2.1.4 Data management -------------------------------------- 67 5.3 Results ----------------------------------------------------------- 68 5.3.1 Bergeggi (B1) --------------------------------------------- 68 5.3.2 Portofino promontory (P1) ------------------------------- 69 5.3.3 Portofino promontory (P2) ------------------------------- 69 5.4 Discussion ------------------------------------------------------- 77 6. Integration between qualitative descriptions and photographic samples to follow the evolution of benthic assemblages ---------------------------------- 79 6.1 Introduction ----------------------------------------------------- 79 6.2 Material and methods ------------------------------------------- 81 6.2.1 Study area ------------------------------------------------- 81 6.2.2 Digital Elevation Model (DEM) and bathymetric map --- 83 6.2.3 Historical qualitative and quantitative information ------ 83 6.2.4 Data management and analysis -------------------------- 87 6.3 Results ----------------------------------------------------------- 88 6.3.1 Qualitative analysis of historical data -------------------- 89 6.3.2 Quantitative results --------------------------------------- 89 6.4 Discussion ------------------------------------------------------- 90 General Discussion --------------------------------------------------- 98 References ----------------------------------------------------------- 102
3
1 General Introduction
The interest toward marine life is documented since the Greek mythology, which mainly referred to the sea as a source of food or of monstrous creatures. It wasn’t until the Historia animalium of Aristotle, in the IV century BC, that specific references to marine organisms were recorded and a variety of vagile invertebrate and fishes were identified, giving him the title of father of the marine biology. Starting from the Renaissance, the exploration of the seas gave a strong impulse to the collection and to the study of marine organisms all over the world, but it was only a few centuries later that naturalists started to be interested in studying living organisms in their own environment. And at the end of the eighteenth century, the scientific diving was born (Riedl, 1980; Bianchi and Morri, 2000). However, several decades had been required for the SCUBA diving became common heritage among marine scientists. In fact, since it was considered as a “sporting activity”, not professional enough for a scientific institution, researchers had usually the habit of co-operate with sport divers for the field work (Bianchi and Morri, 2000). Two remarkable examples from the Ligurian Sea are represented by Tortonese (1958, 1961) and Rossi (1961, 1965a, 1964b), which undertook fundamental research on the benthic assemblages of rocky bottoms on the basis of descriptions, physical samples and/or underwater photographs provided by divers. The arrival of SCUBA diving, therefore, gave the impulse to the study of environments that were difficult to approach using indirect sampling techniques: the rocky reefs. In the Mediterranean Sea, coastal rocky bottoms (Box 1), although representing a minimal fraction of the total marine environment in terms of spatial extent, can claim a socio-economic and scientific interest totally comparable to soft substrates (Bianchi et al., 2004b). Besides their vertical extent, the variety of geomorphologic, hydrodynamic and exposure characteristics provides a number of different biotopes that consequently allow the coexistence of as many biocoenoses. Such an heterogeneity determines a number of ecological processes of general importance (e.g. competition, trophic cascades, habitat structure, etc.), a great biodiversity and the presence of organisms, like the sessile modular ones, that have no correspondence in any other environment (Bianchi et al., 2004b). At the same time, hard bottoms represent an important economic resource in coastal areas, for both fisheries, as they host an alieutic fauna of high commercial value, and for their role in nautical and diving tourism. Intense exploitation, pollution and coastal development strongly affected marine ecosystems (Halpern et al., 2008), inducing the necessity of implementing some management measures.
4
1. General Introduction In Italy, several national and European legislation measures were applied with the objective of the conservation of marine ecosystems, directly or indirectly considering rocky reefs. The national law in Defence of the Sea (L. 979/1982, Legge in Difesa del Mare,) and the framework low for the Natural Protected Areas (L. 394/1991, Legge quadro sulle Aree Naturali Protette) proposed some criteria for the identification of future marine protected areas (MPA), which, de facto, addressed the choice towards rocky shores. Unfortunately, the actual effectiveness of MPA, in Italy like in many other part of the world, is a problematic issue, as protected areas are often nothing more than “paper parks” (Mora et al., 2006), with little or no protection effects on the habitats that they host (Montefalcone et al., 2009). Then, the European Habitat Directive (92/43/EEC), citing “reefs” in general in the list of habitats of community interest (Annex I), lead to the inclusion of about the 50% of the Italian coasts in the network of Natura 2000 (Council European Communities, 1992); otherwise, in a context of heavy demographic, urban and industrial pressures (Benoit and Comeau, 2005), the management of such a wide area had to face with several problems, not to list those due to the overlapping among administrative authorities (Relini, 2009). A few years later, the Water Framework Directive (2000/60/EC) indirectly gave a transnational impulse for the assessment of the health state of marine habitats, encouraging, among others, the use of benthic flora and fauna as an indicator for the ecological status of marine waters. This resulted in a number of different indices based on soft bottom communities (Borja et al., 2000; Simboura and Zenetos, 2002; Rosenberg et al., 2004; Muxika et al. 2007) and seagrass meadows (Romero et al., 2007; Gobert et al., 2009; Lopez y Royo et al., 2009), but only a little contribution came from rocky reefs and it was limited to the upper infralittoral (Orfanidis et al., 2001; Ballesteros et al., 2007). Finally, the Marine Strategy Framework Directive (2008/56/EC) brought the attention back to the whole marine environment and introduced the concept of “seafloor integrity”, i.e. the natural functioning characteristics of the physical structure and of the associate benthic biota, as a descriptor of the Good Environmental Status (GES) (Rice et al., 2010). Independently from the final aim, any management measure requires monitoring programs, which should imply both the characterisation of biocoenoses and biotopes: over long-time scales (each 5-10 years) (the “inventory”), in order to highlight the general evolution of the assemblages; and a more frequent (each year at least) assessment of some biotic quantitative parameters (e.g. demographic structure, abundance of sensitive species, etc.) (the “control”), of the main physico-chemical features and of the health status of the habitats (Bianchi, 2002). Considering the conservational perspective, sampling activities should prefer non destructive approaches, such as the visual census or the photography. The high variability that characterise rocky reefs determined the development of several nondestructive methods for the investigation of qualitative and quantitative features of sessile benthic assemblages. A complete review of this topic is available in Bianchi et al. (2004b). Nevertheless, considering the vertical extent of rocky bottoms and the operational limitation imposed by SCUBA diving (Parravicini et al., 2009), the most of visual quantitative sampling methods has been addressed to shallower reefs, i.e. the infralittoral reef. The two main categories of visual census techniques are the point sampling in a well-defined reference area (quadrat) and the survey along a defined path (transect) (Bianchi et al., 2004b). The latter can be, in turn, divided, with respect to the coastline, in perpendicular (“depth”) transects, which
5
1. General Introduction
Box 1 – Subtidal rocky reefs: infralittoral and circalittoral planes. (Pérès and Picard, 1964) Infralittoral Vertical extent: from the limit of the low tide and the compensation depth, usually indicated by the lower limit of Posidonia oceanica meadows and/or of photophilic algae; in clear waters it can reach the maximum depth of 40 m. Biotic features: high algal diversity and biomass, and rich but less abundant invertebrate fauna. The typical, well developed, community shows a stratified structure, whose composition vary according to environmental factors. The first, basal, layer consists of calcareous algae and skeletal material of polychaetes, bryozoans and gastropods. Above this, there is an intermediate layer composed by small calcareous or soft algae. A third stratum consists of low arbustive, erect forms and the fourth and upper one is made of large photophilic brown and red algae (Ros et al., 1984). Circalittoral Vertical extent: from the compensation depth to the edge of the continental shelf. Biotic features: fixed and colonial animals dominate over algae, mainly represented by red calcareous algae and few sciaphilic green algae. Usually, the endemic biocostruction of coralligenous develops shaping a reef that represents one of the main biodiversity hotspots in the Mediterranean (Ballesteros, 2006). Reefs are characterised by a stratified structure: encrusting and small invertebrates and algae compose the basal layer, while the intermediate one consists of large bryozoans and sponges, in addition to some sciaphilic erect algae, polychaetes and ascidians; the upper layer is usually composed by gorgonians and large sponges. Finally, endobionts, burrowing and hiding microfauna generate an uderstratum, under or between the rests of the biomass (Ros et al., 1984).
provide the qualitative or semi-quantitative variability of the assemblages along the depth gradient, encompassing both infralittoral and circalittoral reefs (Bianchi et al., 1991), and parallel transects, which allow to assess the quali-quantitative composition of a specific assemblage (Loya, 1978). In temperate areas, the most frequent sampling units used for quantitative studies are 0.5-1 m2 quadrats and 10-20 m long transects. According to the size of the unit and to the sampling technique applied, the number of achievable replicates rapidly decreases with depth (Fig. 1.1) (Parravicini et al., 2010). This imply that for a monitoring activity of deeper reefs, i.e. circalittoral reefs, a hierarchical design based on the above mentioned methods is not likely. A second, but not less important, issue is that the assemblages that characterise circalittoral reefs are deeply different from the infralittoral, so appropriate sampling techniques should be developed.
6
1. General Introduction To date, standardised protocols internationally recognised for the study of circalittoral communities do not exist, although some recent essays have been done (Kipson et al., 2011; Deter et al., 2012b; Gatti et al., 2012). Nevertheless, the technological progress allows us to improve both the diving and the sampling techniques, going beyond the limits of underwater direct exploration. Starting from SCUBA diving, improved technologies such as mixed gas (NITROX, TRIMIX) and close circuits re-breathers allow the scientific diver reaching higher depths and lengthening the diving time, reducing the decompression stage and decreasing the gases narcosis and the risk of oxygen toxicity. Anyway, for the study of benthic assemblages, the use of such technologies is still limited in the Mediterranean Sea (Palma et al., 2011, Dether et al., 1012a, b), being more diffused in tropical areas of the USA and in the Caribbean, for the exploration of mesophotic reefs (e.g. Patterson and Relles, 2008; Sherman et al., 2009; Smith et al., 2010).
Figure 1.1 - Diver's safety implications of adopting different sampling procedures whit quadrats: frequency count (black symbols) or cover estimation (white symbols), with either 100×100 cm quadrats (circles) or 50×50 cm quadrats (diamonds). Grey line indicates the safety curve for no decompression dives (Parravicini et al., 2010).
Among the “hi-tech” sampling techniques, acoustic technologies, optical methods, image analysis and mosaicing, and robotic technologies, allow obtaining information on the distribution and composition of benthic assemblages (Zapata-Ramirez et al., 2013). Acoustic methods normally include Side-Scan Sonars (SSS) and Multibeam sonars (Gordini et al., 2012; Bonacorsi et al., 2012), eventually coupled with observations by scientific divers and/or remoteoperate vehicles (ROV). These methods allow mapping the morphology and the distribution of benthic habitats, and describing the spatial pattern of diversity in the area (Barberá et al., 2012). Despite the advantages of the acoustic technologies, these methods cannot resolve the threedimensional fine structure or discriminate benthic communities based on pigmentation, in addition to a very high implementation cost (Zapata-Ramirez et al., 2013). On the other hand, in 7
1. General Introduction situ optical methods (i.e. digital video and cameras) provide an important source of information on the biodiversity, even at small scales. These methods are more commonly used in research, to reduce the time spent underwater, provide high-resolution images and cover large areas (McKinnon et al., 2011; Van Rein et al., 2011, 2012); in addition, they generate permanent survey records and achieve greater sampling subjectivity, than traditional diver observations. The limits of photographic sampling are linked to the small sampling unit: when the scale at which ecological processes operate is wide, under-sampling can cause spatial information loss (Camilli, 2007). The use of dedicated image analysis softwares provides information on species abundance and diversity (Deter et al., 2012; Trygonis and Sini, 2012) and photo–video mosaicing techniques, overcoming the limitation of photographic sampling unit, allow enlarging the spatial scale of the survey (Lirman et al., 2007; Friedman et al., 2010; He et al., 2012; Prados et al., 2012). Therefore, it is possible to create images of the seabed while maintaining high image resolution, providing information in both seascape maps and small scale diversity (Reid et al., 2010; Van Rein et al., 2012). Two- and three-dimensional reconstruction of benthic assemblages integrated with precise georeferentiation may allow both mapping and monitoring activities (ZapataRamirez et al., 2013). Although observation skills and work abilities of scientific divers cannot be equalled, robotic technologies may provide a reliable alternative to survey deep rocky reefs. The choice among different technologies depends on the objective of the study, the characteristics of the environment and the amount of available resources. Usually, Remotely Operated Vehicles (ROVs) are preferred, because their use guarantees the teleoperation and a minimum of interaction with the explored environment; anyway, when the teleoperation is not crucial, Autonomous Underwater Vehicles (AUVs) can be used instead of ROVs (Zapata-Ramirez et al., 2012). A further improvement consists in enabling autonomous capabilities in ROVs (Sattar et al., 2008), introducing autopilot control systems that simplify piloting operations, leaving the researcher more concentrate over the purposes of his study (Toal et al., 2011). Coming back to the issue of monitoring rocky reefs, Table 1.1 summarises the hi-tech and “lowtech” methods that could be employed for monitoring activities on rocky benthic communities. Each technique own advantages and disadvantages, suggesting the necessity of integration among different approach. In the present PhD work, I explored limits and opportunities of several visual sampling methods, both qualitative and quantitative, for the study of infralittoral and circalittoral rocky reefs, through a series of case studies. The general aim was to analyse what kind of information could be provided by each technique, which one should be preferred according to the aim of the study, what is the result when using integrated methods and how “low-tech” visual methods can be improved using the new technologic approaches. The study is presented following a “gradient” of technological level, which was interpreted not only as the employment of hi-tech approaches, but also as the level of data accuracy provided by each method. In particular, starting from the lowest level, in Chapter 2, a visual qualitative and semiquantitative technique was used to analyse the evolution of benthic sessile assemblages over time. Data provided by bionomic descriptions (Tortonese, 1958, 1961) and depth transects
8
1. General Introduction (bionomic transects) were used to characterise infra- and circalittoral rocky reefs, in order to verify their attitude in detecting decadal changes in the composition of the assemblages. In Chapter 3 a new seascape approach for the characterisation and the quality evaluation of coralligenous reefs is presented. Based on a selection of environmental descriptors, the Rapid Visual Assessment (RVA) method was developed with the aim of providing information on the stability, the specific richness and the structure of coralligenous assemblages, through a rapid visual
underwater
work.
Collected
data
allow
the
geomorphologic
and
bionomic
characterisation of the reefs and the assessment of their health state, summarised in an index of quality. In Chapter 4, the response of the index was tested along a gradient of anthropic pressure, together with its sensitivity to the observer effect, with encouraging results. In Chapter 5, two quantitative visual sampling techniques, quadrats and Point Intercept Transects (PITs), were compared in detecting the pattern of composition of infralittoral rocky reefs in different areas and at different scales. According to the object and to the aim of the study, and depending on the scale at which environmental factors act, the proper method should be applied, in order to assure the reliability of the samples and of derived results. Chapter 6 presents an example of integration between descriptive data and photography in the analysis of the long-term evolution of coralligenous reefs. Quantitative data derived from the photographs provided some snapshot over the structure and the composition of the assemblages, giving solidity to a notable amount of qualitative information that allowed reconstructing a sixty year long biological history.
9
1. General Introduction
Table 1 – High-tech and low-tech methods for inventory and control monitoring , with respective pros and cons.
HI-TECH Inventory
Control
Acoustic methods (Multibeam, SSS) (a) Morphologic mapping and characterisation
Optical methods (photo, video) (b) Quali-quantitative characterisation, quality assessment
Optical methods (photo, video) (b) Bionomic characterisation Pros
Cons
Pros
Cons
Wide sampling area, detailed morphologies (a).
Cost, logistic, sea-truth necessary (a)
Reduced fieldwork, objectivity, permanent records, mosaicing techniques (b)
Long data treatment time (a-b)
b) Reduced fieldwork, objectivity, permanent records, mosaicing techniques
b) Loss of 3D structure, reduced size of sampling unit, visibility limited
Loss of 3D structure, reduced size of sampling unit, visibility limited (b)
LOW-TECH
Inventory
Control
Visual methods
Visual methods (PITs, quadrats)
Bionomic and morphologic characterisation, mapping
Quali-quantitative bionomic characterisation, quality assessment
Pros
Cons
Pros
Cons
Human eye, in situ identification
High sampling effort, lack of permanent records
Direct observation, seascape and small scale perception
High sampling effort, lack of permanent records
10
2 Bionomic transects and long-term changes in rocky benthic reefs.
The present work was done thanks to collaboration with C.N. Bianchi, M. Montefalcone, C. Morri (University of Genoa) and the Portofino Marine Protected Area.
2.1 Introduction Marine coastal ecosystems are changing, from both the functional and structural point of view (Gray, 1997; Petchey, 2000; Jackson et al., 2001; Thrush and Dayton, 2002; Micheli and Halpern, 2005). Such variations are due to the response of communities when facing the major disturbing factors, which have largely been recognised being human activities (e.g. fishery, coastal development, pollution) (Halpern et al., 2008) and climate change (Hughes et al., 2003). The most visible effects of the increasing water temperature are variation in the distribution of the species according to their thermic tolerance and mass mortalities of sensitive species (Harley et al., 2006). Human and climatic pressures may interact in a very complex way (Bianchi, 1997) favouring, for example, biological invasions of species in habitats yet affected by anthropic activities (Stachowicz et al., 2002; Bulleri and Benedetti-Cecchi, 2008; Montefalcone et al., 2010). Often, although the change is evident, evaluating its magnitude is difficult because proper reference conditions to which compare the actual situation are seldom available (Al-Abdulrazzak et al., 2012). This may lead to the so-called “sliding baseline syndrome” (Hobday, 2011), which imply that, in absence of any benchmark criteria, an already degraded environmental status might be accepted as reference (Knowlton and Jackson, 2008). The difficulty of delineating changes in marine ecosystems is even greater in the Mediterranean Sea, a semi-enclosed basin experiencing heavy human pressures (UNEP/GPA, 2006) and largely impacted by climate change (Bianchi and Morri, 2004). In these conditions, the availability of historical data and long-time series (Southward, 1991; Barry et al., 1995; Bianchi, 1997; Sagarin et al., 1999) assumes even more importance to measure ecosystem change. A way for setting reference conditions is the use of historical data (Borja et al., 2012). Long time series are precious to measure magnitude and rate of change (Sukhotin and Berger, 2013), and many ecologists have recognised the importance of historical knowledge for understanding and
11
2. Bionomic Transects and long-term changes managing ecosystems (Christensen et al., 1988). In absence of data series, revisiting sites that had been described in the past has proved being a successful alternative to assess stability or change (Barry et al., 1995; Hiscock, 2005; Schückel and Kröncke, 2013; Bianchi et al., 2014). In the northwestern Mediterranean Sea, the Portofino promontory (Genoa, Italy) can claim one of the longest historical series on marine coastal ecosystems. The first documented naturalistic researches date back to mid XIX century, consisting in a list of Ligurian ichthyic species (Pareto et al., 1846). Further research contribution over the geologic, biologic and the ecological characteristics of the reefs was given by pioneers such as A. Issel, R. Issel, A. Brian, R. Santucci, Ruffo, E. Tortonese and L. Rossi (Melegari, 1973) and led to the collection of an increasing amount of detailed data. Focusing on the ecology of benthic sessile rocky reef assemblages, the first study comparable to the “modern” ones date back to 1955-1956 (Tortonese, 1958, 1961). Since it referred to a period that preceded the arrival of mass tourists (Cigolini and Croce, 1997; Carbone, 2004), the increasing temperatures and the invasion by alien species, the assemblages described could be considered as the reference for Portofino area. In the seventies, the introduction of SCUBA diving for scientific exploration gave a new impulse to naturalistic, biologic and ecologic studies, allowing the direct observation by scientific operators (Bianchi and Morri, 2000). Among others, the bionomic transect (vertical or depth transect) (Bianchi et al., 1991; Bianchi et al., 2004a, b) was largely employed to study benthic sessile communities of the area of Portofino promontory, although the only published examples of its use refers to other zones (Balduzzi et al,. 1994; Lloret et al., 2006). Bionomic transects consist in following a route from the sea bed to the surface, along a metric line laid on the bottom, perpendicular to the coastline, and it is suitable for identifying biocoenoses on a physiognomic basis, therefore only conspicuous species, i.e. easily recognisable and identifiable underwater, non cryptic and physiognomically remarkable species (Hiscock, 1987), are recorded. The semi-quantitative abundance (or at least a descriptive indication) of species is referenced according to the depth of the seabed and the distance along the line, and coupled with information about the orientation (compass reading) and the slope (clinometer reading) of the substrate and any other useful data on seabed morphology (e.g. sedimentary and lithological features). All observations are referred to the line between two subsequent reference points, which can be identified in a salient topographic feature or bionomic discontinuity. Integrating all data together allow obtaining the vertical profile of the study site, where the distribution of benthic assemblages is superimposed on the geomorphologic features of the substrate. Such a sampling method allows obtaining data that are similar to the descriptions of the seabed given by Tortonese (1958, 1961), which were based on physical samples and information reported by non-scientific divers charged with doing field activities. Since I had access to a huge amount of bionomic data collected since the nineties, the purpose of the present study was to compare rocky benthic assemblages over time, on the basis of qualitative and semi-quantitative data. As recent researches has shown that, along the Ligurian Sea, marine rocky reefs have been undergoing rapid and considerable alteration in the last decades (Roghi et al., 2010; Parravicini et al., 2013), the final aim of the study was to verify whether non-quantitative approaches are useful to detect such a variation.
12
2. Bionomic Transects and long-term changes
2.2 Materials and methods 2.2.1 Study area The Portofino Promontory (Genoa, NW Mediterranean Sea) (Fig. 2.1) is an imposing rocky headland, mainly composed by holocoenic conglomerates, which is 6 km large and runs out toward the sea of about 5 km. The southern front shows high vertical or sub-vertical cliffs which frequently continue also underwater, while more gentle slopes characterise the eastern and the western sides. From an oceanographic point of view, the circulation of the coastal area is in the general northwestward direction, with only short periods of reversal associated to northern winds (Astraldi and Gasparini, 1986; Astraldi and Manzella, 1983). In correspondence of the Portofino Promontory, the presence of a narrow continental shelf produces a tunnel effect on the northwestward coastal current, which flows approximately along the isobaths, increasing the dynamics of the area. Furthermore, the proximity of the deep sea allows the interaction between coastal and open water masses (Doglioli et al., 2004). The direct contact with deep bottoms, the jagged profile and the highly heterogeneous characteristics (in terms of light, temperature, hydrodynamism, depth and feeding availability) of the Portofino Promontory, allows for the coexistence, in a limited area, of species belonging to very different habitats (Melegari, 1973) and to particularly interesting seascapes. Since the 1999, the Portofino Marine Protected Area was established around the promontory and was divided into three zones subjected to different levels of protection: the A zone (no entry – no take zone), limited to the Cala dell’Oro bay; the B zone (partial reserve zone), all along the southern front of the promontory; the C zone (buffer zone), along the eastern and western sides of the headland (Fig. 2.1).
Figure 2.1 – Study area. Dots represent all sites sampled between 1956 and 2012.
13
2. Bionomic Transects and long-term changes
2.2.2 Historical data and present field work The first semi-quantitative data were collected in 1956 and 1959, when Enrico Tortonese, director of the Natural History Museum of Genoa (Italy), organised a campaign aimed to describe the composition of the rocky reef macrobenthic communities of the southern side of Portofino Promontory. Sampling activities took place tanks to the diver Duilio Marcante, who directed underwater activities providing physical samples of the species and visual descriptions of benthic communities from 12 sites, between the surface and 40 m depth (Tortonese, 1958; 1961). During the 1970s, data from a single bionomic transect (hereaftre called BT) (see Introduction for details) performed along the eastern side of the promontory were provided by professor Carlo Nike Bianchi of the University of Genoa (personal communication). In 1991, 13 BTs were effectuated in 8 site along the southern and eastern sides of the promontory, during a campaign organized by ICRAM (Central Institute for Scientific and Technologic Research applied to the Sea) and ENEA (National committee for the research and the development of Nuclear Energy and Alternative Energies) to provide preliminary studies for the institution of the Portofino Marine Protected Area. Two years later, in 1993, during a wider study aimed to an environmental assessment of sewage pipes along the Ligurian coast, ENEA, again, carried out a campaign along the southern and eastern side of the promontory and provided 6 BTs in 4 sites. In 2008, the same study sites of 1993 were revisited by researchers of the University of Genoa with 12 BTs. In 2012, following the same sampling design of Tortonese (1958), the author collected data along 12 BTs in as many sites (campaign RiTo, see Appendix 2.1 for results); 2 additional sites were (5 BTs) placed along the eastern side of the promontory in order to obtain data from all the areas historically studied. Finally, in 2013, from a private initiative of professor Bianchi, 4 BTs were performed along the southern front of the promontory. All data refer to a depth range comprised between the surface and 40-45 m depth.
2.2.3 Data management Starting from descriptive data obtained by each observation, a first single data matrix (observation depth) species was compiled with simple semi-quantitative abundance scores (AS): 1 = scarce; 2 = abundant; 3 = very abundant (Bianchi et al., 2004b). In order to simplify the dataset for the analysis, depths were organised into four depth ranges: “Da” = 0-10 m; “Db” = 11-20 m; “Dc” = 21-30 m; “Dd” = 31-40 m. The global list of species, obtained pooling together all the observations, initially resulted in a big amount of species or higher taxa, a number of which were cryptic species recorded by Tortonese (1958; 1961) thanks to the laboratory analyses on physical samples. As all subsequent observation, based on bionomic transects, were focused only on conspicuous species, cryptic species were excluded from the analysis. Observations were then organised into three time intervals: “Time 1” (T1), comprehending all observations collected before the 1990s; “Time 2” (T2), with BTs referred to the 1991 and 1993 campaigns; “Time 3” (T3), including all BTs collected after the year 2000. Secondly, as the exact localisation of sampling sites, especially for T1, cannot be considered reliable for a site-level comparison over time, the study area was divided into four zones, referred to the current
14
2. Bionomic Transects and long-term changes zonation of the Portofino MPA (Fig. 2.1): “Az”; “Wz” and “Ez”, when sites were in correspondence of the Western or the Eastern portion of the B zone; “Cz”. The average value of AS calculated among observations determined the AS of the zone they belonged to. Therefore, a single data matrix (depth range zone time) species was organised for further analysis. The number of conspicuous species found in each or the three times was counted. In order to compare the abundance over time, for each species the average abundance score was computed for each depth range within the three periods. As further analysis (see Results) detected no relevant differences among zones, such factor was not considered. Therefore, variations of abundances over time for each depth range were compared using histograms. Results allowed distinguishing three main categories of species, as done by several work dealing with long-term change (Baskin, 1998; McKinney and Lockwood, 1999; Loya et al., 2001): 1) “winners”, i.e. species that increased their abundance over time; 2) “losers”, i.e. species that disappeared or reduced their abundance; 3) “commuters”, a new category introduced to represent those species that presented a variable trend of increase or decrease in their abundance. Results are described in the text, but, for reasons of space, only two examples of two representative species belonging to each of the three categories are graphically presented in Results. The overall composition of assemblages for each depth within time was characterised by summing up AS of all species belonging to the same phylum and results were compared, again, using histograms. The choice of considering the sum of AS instead of the simple number of species was addressed to “maintain memory” of their abundance. The overall multivariate change was explored through a Correspondence Analysis (CA), using the open access software PaSt. Centroids of zones and depth ranges, computed starting from axes values, were used in order to analyse the time trajectories for each depth range and of the different zones. Similarity percentage analysis (SIMPER) (Clarke and Gorley, 2006) was then used to identify the species major responsible for significant differences over time.
2.3 Results A total of 80 conspicuous species was found (Table 2.1): 67 species were detected in T1, 61 in T2 and 67 again in T3; 53 species were common to T1 and T2, 51 to T2 and T3, 44 to T1 and T3 and 43 to all periods. The overall composition showed, at Da (0-10 m), an assemblage dominated by Porifera followed by Cnidaria, Rhodophyta, Ochrophyta and Chlorophyta (with a little representation of Bryozoa) in T1, and the opposite situation in T2 and T3, becoming prevalent algal phyla; between the latter periods, a considerable higher abundance of Ochrophyta characterised T3 (Fig 2.2a). At Db (11-20 m), in T1 and T2, although the overall abundances reduced (except for Cnidaria), communities maintained nearly the same pattern of shallower waters; T3 showed a reduction of algae and a nearly equal repartition of abundance among all phyla (Fig 2.2b). At Dc and Dd, the assemblages are dominated by Porifera and Cnidaria, the first phylum being more abundant than the second only in T1; other phyla, with few differences, were equally represented in all
15
2. Bionomic Transects and long-term changes periods, with the exception of Bryozoa, more abundant in T1 at Dc, and Polychaeta, present only in T3 at Dd (Fig 2.2c, d). The analysis of abundance over time showed 9 “winners”, 21 “losers” and 5 “commuters”, while little or no changes were observed for the remaining species (Table 2.1). Among “winners”, 3 new species were found in T2 (Axinella cannabina, Laurencia spp., Pseudochlorodesmis furcellata) and 1 in T3 (Caulerpa racemosa) (Fig. 2.3c), while other 3 species, i.e. Eunicella verrucosa, Leptogorgia sarmentosa and Pentapora fascialis appeared at depths shallower than before and increases in depth (Fig 2.3c); Corallina elongata and Salmacina dysteri, were present since T1 and just increased their abundance. Among “losers”, 5 species were not found since T2 (Calyx nicaensis, Hippospongia communis, Osmundaria volubilis, Phyllophora crispa, Spongia agaricina) and only 1 in T3 (Alcyonium coralloides); Sargassum spp. disappeared in T3 and at Da since T2 (Fig. 2.3a); Spongia officinalis reduced at Dd and disappeared in shallower waters since T2; large hydroids decreased at Da in T3 and started reducing constantly since T2 in deeper water, until disappearing at Db (Fig. 2.3a); Cystoseira spp. decreased at Da and disappeared at Db in T2; Dyctiopteris polypodioides, in T3, was not found between Da and Db, reduced at Dc but increased at Dd. All remaining species showed only a little decreasing pattern. Among “commuters”, some species decreased or reduced in T2 (Eunicella cavolini, E. singularis and Tricleocarpa fragilis) (Fig. 2.3b), while Sphaerococcus coronopifolius showed its maximum abundance; Pinna nobilis was absent in T2 at Db and appeared at Dc and Dd in T3. CA performed on the original dataset showed three significant axes (Lebart test, p < 0.05), together explaining the 32% of the total variation (Fig. 2.4). On the plane formed by the two first axes, the ordination model plotted a triangular cloud for both species and sample-points. Sample-points showed a main clustering according to Time, more evident for T1, with groups distributed along the vertical axis; Zones and Depths will be analysed in detail later. Giving a look over the species, “losers” are mainly located in the upper-right portion of the cloud (A. coralloides, C. nicaensis, C. viridis, H. communis, large hydroids, M. sabatieri, M. truncata, O. volubilis, P. crispa, R. grimaldii, Sargassum spp., S. scrupea, S. agaricina and S. officinalis), with three exceptions (A. rigida, D. polypodioides and Cystoseira spp.). “Winners”, on the contrary, are distributed along the lower part of the plain and four groups can be distinguished: A. cannabina, E. verrucosa and L. sarmentosa at the extreme right; S. dysteri, P. fascialis and P. furcellata down left; C. racemosa alone and C. elongata and Laurencia sp. on the extreme left of the plot. Finally, “commuters” are located in the middle of the cloud of points creating two distinct groups, with P. nobilis, S. coronopifolius and T. fragilis on one side, and E. cavolini and E. singularis on their right. Position of sample-points and species in the CA diagram was used for axes interpretation (Fig. 2.4). The first axis showed mainly an expression of the difference among depths: species characteristic of high depths, like Eunicella verrucosa, Axinella cannabina, Leptogorgia sarmentosa, Aplysina cavernicola and Paramuricea clavata, were on the right of the diagram, while shallow water species, such as Laurencia spp., Corallina elongata and Jania rubens, on the left. The second axis seemed to depend mainly by the nature of the species, showing “winners” along the lower part of the plot, “loosers” in the upper part and “commuters” in the middle. The time trajectories of Zones showed, on the plain of the first two axes, a common direction for Az, Wz and Ez, while a different behaviour of Cz, which started from a different point and
16
2. Bionomic Transects and long-term changes experienced a consistent change of direction in T2 (Fig. 2.5a). Axes 1-3 pointed out that, actually, T2 represented the turning point for all Zones, which after a first moment, between T1 and T2, of variable evolution, totally changed direction; it should be also noticed that Zones behave differently, being all trajectories different, although pointing the same final direction (Fig. 2.5b). The time trajectories of Depths looked similar in the plane of axes 1-2 (Fig. 2.6a); at all Depths major changes occurred at time T2, as showed by the change of direction along the third axis in the plane of axes 1-3 (Fig. 2.6b). The main contributors to internal similarity of each Time greatly varied, indicating a change in the composition of the assemblages over time (SIMPER results, Appendix 2.1). Time 1 was characterised by invertebrates, i.e Eunicella cavolini, large hydroids and Reteporella grimaldii, and algae, like Flabellia petiolata and encrusting corallines. Differently, in Times 2 and 3 the assemblages were mainly represented by vegetal components, being more important potential canopy forming algae such as Dictyopteris polypodioides and Cystoseira sp. in T2 and other erected generalist species like Dictyota dichotoma, Padina pavonica and encrusting rhodophyta in T3. The invasive alga Caulerpa racemosa was recorded only in T3 and assumed a considerable role in delineating the composition of the assemblages, being abundant and widespread over all Depths and Zones.
17
2. Bionomic Transects and long-term changes Table 2.1 - List of the conspicuous species observed, ordered alphabetically according to their code (as used in Fig. 2.4). The category of abundance variation over time is also indicated. Codes
Species (Higher taxon)
Aae
Acetabularia acetabulum (Chlorophyta)
little or no change
Aar
Asparagopsis armata (Rhodophyta)
little or no change
Aau
Acanthella acuta (Porifera)
little or no change
Acn
Axinella cannabina (Porifera)
winner
Aco
Alcyonium coralloides (Cnidaria)
loser
Acv
Aplysina cavernicola (Porifera)
little or no change
Ada
Axinella damicornis (Porifera)
little or no change
Aor
Agelas oroides (Porifera)
little or no change
Apo
Axinella polypoides (Porifera)
loser
Ari
Amphiroa rigida (Rhodophyta)
loser
Ave
Axinella verrucosa (Porifera)
little or no change
Beu
Balanophyllia europaea (Cnidaria)
little or no change
Cbu
Codium bursa (Chlorophyta)
little or no change
Cca
Cladocora caespitosa (Cnidaria)
little or no change
Ccl
Clathrina clathrus (Porifera)
little or no change
Cco
Codium coralloides (Chlorophyta)
little or no change
Cel
Corallina elongata (Rhodophyta)
winner
Cin
Caryophyllia (Caryophyllia) inornata (Cnidaria)
little or no change
Cme
Cerianthus membranaceus (Cnidaria)
little or no change
Cni
Calyx nicaeensis (Porifera)
loser
Cnu
Chondrilla nucula (Porifera)
little or no change
Cra
Caulerpa racemosa (Chlorophyta)
winner
Cre
Chondrosia reniformis (Porifera)
little or no change
Cru
Corallium rubrum (Cnidaria)
loser
Csi
Colpomenia sinuosa (Ochrophyta)
little or no change
Cso
Cladostephus spongiosus (Ochrophyta)
little or no change
Csp
Cystoseira spp. (Ochrophyta)
loser
Cvi
Cliona viridis (Porifera)
loser
Dav
Dysidea avara (Porifera)
little or no change
Ddi
Dictyota dichotoma (Ochrophyta)
little or no change
Dpo
Dictyopteris polypodioides (Ochrophyta)
loser
Dve
Dasycladus vermicularis (Chlorophyta)
little or no change
Eca
Eunicella cavolini (Cnidaria)
commuter
Enco
encrusting corallines (Rhodophyta)
little or no change
Ensp
encrusting sponges (Porifera)
little or no change
Esi
Eunicella singularis (Cnidaria)
commuter
Eve
Eunicella verrucosa (Cnidaria)
winner
Fpe
Flabellia petiolata (Chlorophyta)
little or no change
Fve
Frondipora verrucosa (Bryozoa)
little or no change
Gre
Gloiocladia repens (Rhodophyta)
little or no change (continued)
18
2. Bionomic Transects and long-term changes (continued) Hcl
Hemimycale columella (Porifera)
little or no change
Hcm
Hippospongia communis (Porifera)
loser
Hpa
Halocynthia papillosa (Tunicata)
little or no change
Htu
Halimeda tuna (Chlorophyta)
little or no change
Hyd
large hydroids (Cnidaria)
loser
Isp
Ircinia spp. (Porifera)
little or no change
Jru
Jania rubens (Rhodophyta)
little or no change
Lpr
Leptopsammia pruvoti (Cnidaria)
little or no change
Lsa
Leptogorgia sarmentosa (Cnidaria)
winner
Lsp
Laurencia spp. (Rhodophyta)
winner
Msa
Microcosmus sabatieri (Tunicata)
loser
Mtr
Myriapora truncata (Bryozoa)
loser
Olo
Oscarella lobularis (Porifera)
little or no change
Ovo
Osmundaria volubilis (Rhodophyta)
loser
Pax
Parazoanthus axinellae (Cnidaria)
little or no change
Pca
Palmophyllum crassum (Chlorophyta)
little or no change
Pci
Phyllophora crispa (Rhodophyta)
loser
Pcl
Paramuricea clavata (Cnidaria)
little or no change
Pfa
Pentapora fascialis (Bryozoa)
winner
Pfi
Petrosia (Petrosia) ficiformis (Porifera)
little or no change
Pfu
Pseudochlorodesmis furcellata (Chlorophyta)
winner
Pno
Pinna nobilis (Mollusca)
commuter
Ppa
Padina pavonica (Ochrophyta)
little or no change
Psi
Pleraplysilla spinifera (Porifera)
little or no change
Psp
Peyssonnelia spp. (Rhodophyta)
little or no change
Rgr
Reteporella grimaldii (Bryozoa)
loser
Sag
Spongia (Spongia) agaricina (Porifera)
loser
Sce
Smittina cervicornis (Bryozoa)
loser
Sco
Sphaerococcus coronopifolius (Rhodophyta)
commuter
Sdy
Salmacina dysteri (Annelida)
winner
Sof
Spongia (Spongia) officinalis (Porifera)
loser
Ssa
Sabella spallanzanii (Annelida)
little or no change
Sso
Stypocaulon scoparium (Ochrophyta)
little or no change
Ssp
Sargassum spp. (Ochrophyta)
loser
Ssr
Scrupocellaria scrupea (Bryozoa)
loser
Ssv
Savalia savaglia (Cnidaria)
little or no change
Sve
Serpula vermicularis (Polychaeta)
little or no change
Tav
Turbicellepora avicularis (Bryozoa)
little or no change
Tfr
Tricleocarpa fragilis (Rhodophyta)
commuter
Zty
Zanardinia typus (Ochrophyta)
little or no change
19
Figure 2.2 – Overall composition and abundance (sum of Abundance Scores) of phyla for each depth range in each time: a) Da (0-10 m); b) Db (10-20 m); c) Dc (20-30 m); d) Dd (30-40 m). White bars: T1; grey: T2; black: T3.
2. Bionomic Transects and long-term changes
20
2. Bionomic Transects and long-term changes
Figure 2.3 – Mean (± SD) abundance score (AS) for selected examples: a) of species that decreased or disappeared over the three times (losers); b) of species whose abundance decreased only between T1 and T2, increasing again in T3 (commuters); c) of species that appeared or showed higher abundance in more recent times (winners). White bars indicate T1, grey bars T2 and black bars T3.
21
2. Bionomic Transects and long-term changes
Figure 2.4 – Correspondence Analysis ordination model on the plane formed by the first two axes extracted. Sample-points are represented by alphanumeric codes, where numbers represent Time, capital letters Zone and minuscule letters Depth. Time is also represented with colours: red=T1, blue=T2, green=T3. Species are identified by codes as in Table 1; in bolt are represented species belonging to the “winner”, “loser” or “commuter” category. “Winners” (below in the plot) and “losers” (above) are grouped to facilitate the interpretation.
22
2. Bionomic Transects and long-term changes a)
b)
Figure 2.5 – Correspondence Analysis ordination model on the plane formed by axes 1-2 (a) and 1-3 (b), obtained analysing the distribution of Zone-points over time; the coordinates of points were calculated as the mean (centroid) among sample-points coordinates for each Zone and Time. Time trajectories of individual Zones are also indicated. Zone-points are represented by alphanumeric codes, where numbers represent the Time and capital letters the Zone. Species are identified by codes as in Table 1; in bolt are represented species belonging to the “winner”, “loser” or “commuter” category.
23
2. Bionomic Transects and long-term changes a)
b)
Figure 2.6 – Correspondence Analysis ordination model on the plane formed by axes 1-2 (a) and 1-3 (b), obtained analysing the distribution of Depth-points over time; the coordinates of points were calculated as the mean (centroid) among sample-points coordinates, for each Depth range and Time. Time trajectories of individual Depths are also indicated. Depth-points are represented by alphanumeric codes, where numbers represent the Time and minuscule letters the Depth. Species are identified by codes as in Table 1; in bolt are represented species belonging to the “winner”, “loser” or “commuter” category.
24
2. Bionomic Transects and long-term changes
2.4 Discussion In this study the pattern shown by rocky reef communities in Time 1 (1950s-1970s) was compared with that the same communities exhibited in Time 2 (1990s) and Time 3 (2010s): results indicated significant change in species composition over time. The use of qualitative and semi-quantitative data, therefore, led to the same results obtained by previous studies based on quantitative data, in NW Mediterranean rocky reefs (Roghi et al., 2010; Parravicini et al., 2013). This is the first important outcome of the study, which enlarge the possibility of approaching the long term evolution of rocky reef assemblages also to non rigorous quantitative data. The analysis of the communities over three periods allowed identifying a consistent trend of variation between the half of the XX century and the first decade of the XXI. Such a long-term study permitted also to distinguish between two different patterns of change, having the nineties as turning point. A similar trend was already observed for circalittoral assemblages at Mesco Point (Ligurian sea) (Roghi et al., 2010), suggesting that it should not be considered a phenomenon of local concern. While long-term studies lacks, changes occurred during the last twenty years in the Mediterranean sea are well documented and indicate that both rocky and soft bottoms suffered changes in the specific composition, structure and trophic organisation of the assemblages (Morri and Bianchi, 2001; Roghi et al., 2010; Parravicini et al., 2013; Bianchi et al., 2014), The change in species composition experienced three different situations: in the 1950s the assemblages were characterised by invertebrates like Eunicella cavolini and large hydroids together with encrusting and erect algae, while in the 1990s the communities were dominated by algae, especially encrusting corallines and canopy-forming Fucales, such as Cystoseira spp., D. polypodioides and, although less abundant, Sargassum spp.; in 2010s those species almost disappeared and were substituted by erect generalist species like D. dichotoma and P. pavonica, and by the invasive alien green alga C. racemosa. A major impact of coastal development on reef communities has been shown to be the replacement of canopy-forming algae with algal turfs, with consequent habitat homogenization and loss of structural complexity (Thimbaud et al., 2005; Airoldi et al., 2008; Mangialajo et al., 2008). Therefore, the general trend seemed to lead to a change in the trophic organisation, with a strong decrease of heterotrophic species, and a reduction of three-dimensional structure, with the loss of higher species, all together leading to community homogenisation (Parravicini et al., 2013). Despite the creation of the Portofino MPA in 1999, the evolution of the assemblages did not show significant effects due to the different levels of protection. A (no entry–no take) and B (partial reserve) zones showed a similar pattern over time, unlike the C zone, which took a totally different direction. Probably, the main explication may be found in the geographic position of the zones rather than in an effect of protection measures. In fact, the C zone is located on the eastern side of the Portofino Promontory, while the others extend along the southern front. Different hydrodynamic conditions, proximity to anthropic settlement and high human frequentation may have been the major factors influencing the evolution of rocky reef assemblages. As very few differences were detected between the A and B zones, we cannot affirm an evidence of the effects of the protection (Montefalcone et al., 2009; Parravicini et al.,
25
2. Bionomic Transects and long-term changes 2013). Anyway, all zones are concord in indicating a change in the evolutionary trend occurred during the nineties. The pattern of variation observed along the depth gradient revealed, again, an abrupt change of direction in the evolution of benthic assemblages, during the nineties. Anthropic pressure and climate change combined their effects on the marine biota, leading to variation in the composition of communities (Wilkinson and Buddemeier, 1994; Bourcier, 1996). The first direct human action that affected benthic assemblages was the indiscriminate collection of commercial species (Peirano and Tunesi, 1989), which, for example, resulted in the disappearance, or strong reduction, of big massive sponges, such as S. agaricina, S. officinalis, and C. nicaensis. Human activities can also result in increasing water turbidity and fine sedimentation, which favour gorgonians like E. verrucosa and L. sarmentosa (Francour and Sartoretto, 1992), and determine the “movement” of more photophilic species toward shallower waters. Mass mortalities of gorgonians, sponges and other species occurred in the entire Mediterranean sea as a consequence of the thermal anomalies of 1999 and 2003, induced by the persistence of high water temperature and stratification (Cerrano et al., 2000; Pérez et al., 2000; Cupido et al., 2008; Coma et al., 2009; Garrabou et al., 2009; Lejeusne et al., 2010). A second major effect of climate change that took place in recent decades is the warming trend of marine water temperature in the northwestern Mediterranean sea (Bensoussan et al., 2009; Calvo et al., 2011), which favours the northward expansion of many thermophilic species, both native (e.g. P. furcellata) and alien (e.g. C. racemosa) (Bianchi and Morri, 1994; Parravicini et al., 2008). The northward range extension of species has been favoured also by modifications in the emphasis of water flow and in the pattern of water circulation forced by climate change (Astraldi et al., 1995; Bianchi, 2007a). So far, the leitmotif of the discussion is the abrupt variation in assemblage evolutionary trends occurred during the nineties, already documented in previous work (Morri and Bianchi, 2001; Roghi et al., 2010; Parravicini et al., 2013; Bianchi et al., 2014). A recent review of long-term records of Mediterranean ecological and hydro-climate variables (Conversi et al., 2010) found that all point to a synchronous change in the late 1980s, which determined such important variations in the ecological systems, that a regime shift can be considered (Montefalcone et al., 2011). Evidences of parallelisms with other European basins (North, Baltic and Black seas), showing ecological changes in the same period, revealed that we are not facing a basin level phenomenon, but it is likely linked as part of a larger scale, northern hemisphere change (Conversi et al., 2010). The complex interaction among several factors determined a consistent variation in the composition and structure of the rocky reef assemblages of Portofino promontory, over the last sixty years. Qualitative (semi-quantitative) data allowed detecting major changes in communities, and three points in the time permitted to identify two periods in the evolution of communities, underling that the long-term trend of variation is not linear, since new drivers of change appeared in the 1990s.
26
2. Bionomic Transects and long-term changes
Appendix 2.1 Results of SIMilarity PERcentage (SIMPER) analysis identifying taxa major contributing to differences between Times. Species
Time 1
Time 2
Time 3
Acanthella acuta
2,11
-
-
Axinella damicornis
3,58
1,20
2,04
Agelas oroides
4,46
-
3,31
Amphiroa rigida
1,39
-
-
-
-
1,87
Codium bursa
3,75
2,30
2,11
Caryophyllia inornata
3,01
-
-
-
-
7,52
Axinella verrucosa
Caulerpa racemosa Chondrosia reniformis
-
-
1,05
Cystoseira sp.
2,70
13,14
1,75
Dictyota dichotoma
1,83
6,23
7,40
Dictyopteris polypodioides
1,79
6,29
-
Eunicella cavolini
10,31
-
5,12
Encrusting corallines
6,12
14,18
10,29
Encrusting sponges
4,97
4,86
6,47
Eunicella singularis
3,83
-
1,37
Flabellia petiolata
6,46
7,66
6,77
Hippospongia communis
2,80
-
-
-
1,68
-
Halimeda tuna
2,32
3,80
1,50
Large hydroids
7,26
3,54
-
Ircinia sp.
1,22
1,54
1,41
-
1,13
1,22 3,57
Halocynthia papillosa
Jania rubens Leptopsammia pruvoti
1,91
-
Microcosmus sabatieri
2,03
1,62
-
Parazoanthus axinellae
1,17
5,53
5,07
Phyllophora crispa
1,25
-
-
Paramuricea clavata
1,68
2,51
-
Pentapora fascialis
-
2,42
2,20
Petrosia ficiformis
3,59
-
1,09
Pseudochlorodesmis furcellata
-
-
0,98
Padina pavonica
-
-
7,49
Reteporella grimaldii
5,23
2,63
-
Smittina cervicornis
1,11
-
-
-
2,11
-
Salmacina dysteri
-
1,43
1,07
Spongia officinalis
2,88
3,02
-
Stypocaulon scoparium
-
-
5,25
Tricleocarpa fragilis
-
-
2,54
Zanardinia typus
-
1,39
-
Sphaerococcus coronopifolius
27
2. Bionomic Transects and long-term changes
Appendix 2.2 Bionomic profiles from the RiTo campaign, showing the main geomorphologic and bionomic features of 11 sites along the Portofino promontory southern side. Alphanumeric codes indicate the main assemblages, according to the EuNIS habitats classification (Davies et al., 2004); dominant species are reported, with their respective codes (Table 2.1). Particularly significant species, e.g. alien species or sea fans, are graphically represented.
Substrate legend: Rock
Sand
Detritic whit sand
Detritic
Rocky blocks
Figure 2.6 – Station I, Grotta dei Gamberi
Figure 2.7 – Station II, Punta Targhetta (Punta Pidocchiaro)
28
2. Bionomic Transects and long-term changes
Figure 2.8 – Station III, Punta del Buco
Figure 2.9 – Station IV, Cala dell’Oro West
29
2. Bionomic Transects and long-term changes
Figure 2.10 – Station V, Cala dell’Oro
Figure 2.11 – Station VI, Punta Torretta
30
2. Bionomic Transects and long-term changes
Figure 2.13 – Station VII, Punta Torretta East
Figure 2.14 – Station VIII, San Fruttuoso
31
2. Bionomic Transects and long-term changes
Figure 2.15 – Station IX, Tetto del Dragone
Figure 2.16 – Station X, Punta Carega
32
2. Bionomic Transects and long-term changes
Figure 2.17 – Station XI, Chiesa di S. Giorgio
33
3 A new Rapid Visual Assessment (RVA) approach for the characterisation and the assessment of the quality of coralligenous reefs: The method
This work was done thanks to a collaboration with A. Rovere (Lamont-Doherty Earth Observatory, Columbia University, USA), V. Parravicini (Institut de Recherche pour le Développement, France), M. Montefalcone, C. Morri, G. Albertelli and C.N. Bianchi (University of Genoa).
3.1 Introduction Intense urbanization is one of the major drivers replacing natural ecosystems with humandominated landscapes, with obvious consequences on habitat structure, biodiversity and functioning (Sochat et al., 2006). The coastal zone is strongly urbanized today and about twothirds of the Mediterranean coastline is characterised by harbours and ports (UNEP/GPA, 2006). Coastal ecosystems are among the most threatened worldwide, but a full understanding of the effects of extensive coastal development on marine environment is still far from being reached (Bulleri, 2006; Connel and Glasby, 2006). The assessment of the status of coastal waters is required by the European Water Framework Directive (WFD, 2000/60/EEC) through the selection of appropriate Biological Quality Elements (BQEs) (European Community, 2000) and with the support of hydromorphology and physicochemical descriptors (Orlando-Bonaca et al., 2012). WFD attempts to achieve an ecosystem level assessment by evaluating separately selected ecosystem components. Similarly, the Marine Strategy Framework Directive (MSFD, 2008/56/EEC) underlines the necessity to assess the ecological status of marine habitats at ecosystem level rather than at species or chemical levels alone (Borja et al., 2008; Borja et al., 2010). Landscape or, more properly, seascape approaches (Montefalcone et al., 2010; Pittman et al., 2011) integrate various levels of information from species identification to habitat structure characterisation, as requested by the MSFD, and have great potential to enhance our understanding and management of coastal environments (Boström et al., 2011). One of the most important coastal habitats in the Mediterranean Sea is represented by the socalled ‘‘coralligenous’’ (Marion, 1883), an endemic underwater habitat (UNEP-MAP-RAC/SPA, 2008) shaped by bioconstructors and characterised by high species richness, biomass and
34
3. Rapid Visual Assessment: The method production. Its calcification rate, assessed around 103 g CaCO3 m-2 y-1, falls within the range of values calculated for tropical reefs (Bianchi, 2001). It develops on rocky reefs and biodetritic bottoms from about 20 m down to 120 m depth, in relatively constant conditions of temperature, currents and salinity. Coralligenous reefs result from the dynamic equilibrium between bioconstruction (mainly by encrusting red algae, with an accessory contribution by serpulid polychaetes, bryozoans and scleractinian corals), and destruction processes (by borers and physical abrasion) (Cerrano et al, 2001), which create morphologically complex substrates where highly diverse benthic assemblages develop (Laborel, 1961; Laubier, 1966; Hong, 1982; Ballesteros, 2006). Despite the occurrence of many species with high ecological value (some of which are also legally protected, e.g. Savalia savaglia, Spongia officinalis, etc.), coralligenous reefs were not listed among the priority habitats defined by the EU Habitat Directive (92/43/EEC). This implies that the most important Mediterranean bioconstruction still remained without formal protection and it was not included within the list of Sites of Community Interest (SCIs). Few years after the adoption of the Habitat Directive, coralligenous reefs were listed among the habitats needing rigorous protection by the Protocol for special protected areas (SPA/BIO) of the Barcelona Convention for the conservation of Mediterranean biodiversity (1995) (Cinelli et al., 2009). However, the concept of “rigorous protection” sounds somewhat vague and only recently, the “Action plan for the conservation of coralligenous and other calcareous concretions in the Mediterranean Sea” (UNEP-MAP-RAC/SPA, 2008) encouraged the conservation of the coralligenous by the establishment of marine protected areas and emphasized the need for standardised programs for its monitoring. Coralligenous is threatened by direct human activities, such as trawling and illegal exploitation of protected species, and is also vulnerable to the indirect effects of climate change (e.g. positive thermal peaks). To date, however, neither national legislation to protect coralligenous reefs nor rigorous scientifically-based management and monitoring programs have yet been proposed (Ballesteros, 2009). Due to its large bathymetric distribution and the consequent sampling constraints, coralligenous was subjected to limited spatio-temporal studies, so that its geographical distribution and health status remain poorly known at regional level. The operational restrictions imposed by SCUBA diving (Parravicini et al., 2010) reduce the amount of collected data during each dive and increase the sampling effort. To optimise the diving time, photo-quadrats (Parravicini et al., 2009), frequency counts (Parravicini et al., 2010) and point intercept transects (Gatti et al., 2010) have been proposed as efficient sampling techniques of benthic assemblages; however, most of them are not commonly employed on the coralligenous (but see Kipson et a., 2011). Since the coralligenous is characterised by high heterogeneity, extreme patchiness and coexistence of several biotic assemblages, a seascape approach seems to be the most reasonable solution for its characterisation. This study had three distinct aims: 1) to obtain a first characterisation of the coralligenous shoals of Vado Ligure (Italy, NW Mediterranean) employing a seascape approach, i.e. integrating the bionomic description of benthic assemblages (Cocito et al., 1991) with geomorphologic and mesologic (physical) characterisation of the shoals; 2) to assess preliminarily their quality; 3) to propose a Rapid Visual Assessment technique (RVA) for optimising underwater surveys, inspired by the one described by Bianchi et al. (2007) and mainly focused on seascape rather than on community aspects (Bianchi et al., 2010).
35
3. Rapid Visual Assessment: The method
3.2 Material and methods 3.2.1 Study area and field activities The study was carried out in April 2010 on the coralligenous shoals of Vado Ligure (Savona, Italy, NW Mediterranean), an area that is very close to an important commercial harbour (Fig. 3.1). Pre-existing information collected by a multi-beam provided the preliminary morphology and the exact position of all coralligenous shoals. Five shoals (each composed by various rocky outcrops) were then chosen for sea-truthing. One to four surveys were performed in each shoal, according to its extent and morphological heterogeneity: surveys 1.1, 1.2, 1.3 were conducted in shoal 1; surveys 2.1, 2.2, 2.3, 2.4 in shoal 2; surveys 3.1, 3.2, 3.3, 3.4 in shoal 3; surveys 4.1, 4.2, 4.3, 4.4 in shoal 4 and survey 5.1 in shoal 5. The geomorphologic characterisation of each shoal was obtained in situ by considering three main ‘‘morphotypes’’: 1) high rocky outcrops (HR); 2) low rocky outcrops (10 cm) vertical growth (e.g. Paramuricea clavata, Cystoseira zosteroides); 2) intermediate layer, constituted by organisms with moderate (1 cm to 10 cm) vertical growth (e.g. Reteporella grimaldii, Axinella damicornis); 3) basal layer, constituted by encrusting or with limited ( 25%, according Pérès and Picard (1964);
maximum height (MH): the maximum height of the tallest species was compared to the maximum height value available in literature (LMH) for that species. Score 1 was assigned when
MH < 0.5 LMH,
score
2
when
0.5 LHM ≤ MH ≤ 0.75 LMH,
score
3
when
MH > 0.75 LMH;
epibiosis-necrosis (EN): from the percentage of epibiosis and/or necrosis of organisms, score 1 was assigned when EN > 75%, score 2 when 10% ≤ EN ≤ 75%, score 3 when EN < 10%.
Intermediate layer:
specific richness (SR): preliminary investigations showed that, over an area of approximately 2m2, the maximum number of conspicuous species detected in two minutes was about 15. Then, score 1 was assigned when SR < 5, score 2 when 5 ≤ SR ≤ 10, score 3 when SR > 10;
seasonal-perennial species ratio (S/P): the persistence of coralligenous assemblages is strictly dependent on the maintenance of definite abiotic and biotic factors (Bellan-Santini et al., 2002). The dominance of seasonal life cycles may indicate opportunistic strategies, which typically occur under high disturbance regimes and unstable conditions; on the contrary, the dominance of long lived species may indicate environmental stability or good adaptation to predictably variable conditions. Therefore, the ratio between the number of seasonal and perennial species was calculated and score 1 was assigned when S/P > 0.5, score 2 when 0.5 ≤ S/P ≤ 0.2, score 3 when S/P < 0.2;
erect calcified bryozoans (ECB): according to Hong (1982; 1983), ECB have an important ecological role, since they are the most abundant bioconstructors among animals and their presence is an indicator of low human impact. Considering the number of species of erect bryozoans, score 1 was assigned when ECB = 1 species, score 2 when ECB = 2 to 4 species, score 3 when ECB > 4 species.
Basal layer:
NTDs cover: depending on their role in the bioconstruction, score 3 was assigned to ECR, because they are the main active producers of calcareous substrate; score 2.5 to NCEA and AN for their role in substrate protection; score 2 to TURF, which may protect the substrate but retains sediment; score 1 to SED, because its presence inhibits bioconstructors’ growth and may contribute to the abrasion of calcareous substrate. The formula (cover x
38
3. Rapid Visual Assessment: The method score)/100 was applied to percent cover of each NTD and resulting values were summed up to obtain the total quality score relative to NTDs cover in the basal layer of each survey;
thickness and consistency of calcareous layer: score 1 was assigned when penetration was null, meaning that the calcareous substrate was either absent or completely lithified (i.e. bioconstruction was not more active); score 2 was assigned when penetration was centimetric, suggesting the presence of an unconsolidated calcareous substrate that results from an active bioconstruction with little or no consolidation, undermined by the action of biotic and abiotic erosion; score 3 was assigned when penetration was millimetric, suggesting the presence of active bioconstruction resulting in a compact calcareous biogenic substrate;
borer marks: score 1 was assigned when borers were common, score 2 when they were occasional and score 3 when they were absent.
In order to get a total quality score for each layer (QL) in each survey, the following formula inspired by the one adopted by Bianchi (2007), was applied: QL = (XL x YL x ZL) x k(1-n) where XL, YL and ZL are the quality scores assigned to the three descriptors, k is the maximum value assumed by these scores (3 in this case), n is the number of descriptors considered. In our experience, adoption of an addictive model like arithmetic mean is inappropriate, because very different configurations of sub-scores (quality scores of descriptors, in our case) would give the same integrated value (QL). Therefore, we used this multiplicative formula that guarantees to obtain a QL score that reflects the configuration of each sub-score (Villa, 1994). According to the ecological status classification of the Water Framework Directive and its chromatic scheme, QL was divided into five classes of quality status: Bad (red) when 0 < QL ≤ 0.6; Poor (orange) when 0.6 < QL ≤ 1.2; Moderate (yellow) when 1.2 < QL ≤ 1.8; Good (green) when 1.8 < QL ≤ 2.4; High (blue) when 2.4 < QL ≤ 3. Finally, among QL values of each layer belonging to each habitat type, the maximum QL value (max QL) was assumed as the synthetic quality index of the habitat type.
3.3 Results 3.3.1 Characterisation The multi-beam survey showed a system of rocky shoals at Vado Ligure, scattered between 14 m and 40 m depth. Shoals 1, 2, and 4 were located between 20 m and 30 m depth, shoal 3 had rocky outcrops developing down to 30 m depth, whilst shoal 5 was shallower than 20 m (Fig. 3.5). Three habitat types were defined on the basis of the main species that dominated the upper layer (Fig. 3.2), all belonging to the same EuNIS habitat “Mediterranean coralligenous communities moderately exposed to hydrodynamic action” (code A4.26). The “A” habitat type (EuNIS code A4.26B, “Facies with Paramuricea clavata”) is dominated by the sea fan Paramuricea clavata, which was occasionally associated to other gorgonians (Eunicella verrucosa, Eunicella singularis), the antipatarian Savalia savaglia, sponges (Scalarispongia scalaris, Spongia officinalis, S. agaricina) and the polychaete Sabella spallanzanii. This habitat type was found only in the
39
3. Rapid Visual Assessment: The method shoal 3. The “B” habitat type (EuNIS code A4.261, “Association with Cystoseira zosteroides”) was characterised by the brown seaweed Cystoseira zosteroides together with tall colonies of Eudendrium racemosum. This type was widely distributed in the shoals of Vado Ligure (Fig. 3.3). The “B*” habitat type corresponds to the previous EuNIS habitat “Association with Cystoseira zosteroides” (code A4.261) but gorgonians (E. verrucosa, E. singularis, Leptogorgia sarmentosa) occurred together with Cystoseira zosteroides and Eudendrium racemosum. Also this habitat showed a wide distribution at Vado Ligure (Fig. 3.3). Geomorphologic and mesologic characteristics associated to each habitat type are summarised in Table 3.1.
Table 3.1 - Geomorphologic and mesologic characteristics associated with each habitat type, with their respective EuNIS code. Habitat type
Morfotype
Depth (m)
Slope (°)
Elevation (m)
EUNIS Code
A
HR , LR
31 – 34
15 – 30
1–3
A4.26B
B
HR
20.5 - 23
5 - 80
2.5 – 5
A4.261
B*
HR ,LD, LR
17 - 27
5 – 42.5
1-6
A4.261
3.3.2 Quality assessment Species of the upper layer had total cover values ranging from 2% to 60% (Fig. 3.3a). Maximum height (MH) of the tallest species varied from 50 cm of some Paramuricea clavata colonies to 9 cm of Cystoseira zosteroides (Fig. 3.3b). Comparing MH values with the maximum height values found in literature (LMH), quality scores have been defined as shown in Table 3.2. Epibiosis and necrosis usually exceeded 75%, with the exception of surveys 1.1, 1.2, 4.1 where comparatively lower values (EN < 10%) were detected. Specific richness (SR) of the intermediate layer varied between 4 – 14 species (Fig. 3.3c and Table 3.3). The number of erect bryozoans ranged between 1 – 5. Seasonal and perennial species ratio (S/P) was always lower than 0.25 and, in the 50% of surveys, conspicuous seasonal species were absent. In the basal layer, sediment always showed the highest cover values, in some cases exceeding 80% (Fig. 3.3d). As a consequence, quality scores obtained from the cover descriptor were always lower than 2 (see Table 3.4). Penetration of calcareous substrate was null in the 50% of the surveys; only in a single case (survey 2.4) it was higher than 2 cm. Borer marks were always absent, except for the occasional occurrence of Cliona viridis papillae in survey 1.3.
40
3. Rapid Visual Assessment: The method
Figure 3.2 - Box and whiskers plot representing species composition and classes of abundance of the upper layers in the three habitat types: A = Facies with Paramuricea clavata (EuNIS code A4.26B); B = Association with Cystoseira zosteroides (EuNIS code A4.261); B* = ditto together with gorgonians.
41
Figure 3.3 - Percentages of cover of the upper layer’s species (a); maximum height of the upper layer’s species (b); number of species detected in the intermediate layer, with the dark portion representing erect bryozoans (c); percentages of cover of the five NTDs in the basal layer (d). SED = sediment, TURF = turf-forming algae, ANIM = encrusting animals, NCEA = non-calcified encrusting algae, ECR = encrusting calcified rodhophyta.
3. Rapid Visual Assessment: The method
42
3. Rapid Visual Assessment: The method the ecological status classification of the WFD (Fig. 3.5), the basal layer showed Bad status in seven surveys (more than half being found in the shoal 4), Poor in five (mostly found in shoal 3), Moderate in three and Good in only one survey (in shoal 2); the intermediate layer had Bad status in one survey (shoal 2), Poor in five, Moderate in seven and Good in three surveys; the upper layer showed seven surveys with Bad status (mostly found in shoals 2 and 4), two with Poor and Good, five with Moderate status (more than half in shoal 3). The ecological status of each layer in the three habitat types, considering the synthetic quality index (max QL), resulted as follow (see Fig. 3.4):
type A: basal and upper layers in Moderate status (max QL = 1.4 and 1.3, respectively), intermediate layer in Good status (max QL = 2);
type B: basal and intermediate layers in Good status (max QL = 2), upper layer in Moderate status (max QL = 1.3);
type B*: basal layer in Moderate status (max QL = 1.6), intermediate and upper layers in Good status (max QL = 2).
Table 3.2 - Quality scores associated with the classes of height of the upper layer species. Cystoseira Paramurice Scor zosteroide a e s clavata 1
< 15 cm
< 40 cm
2
15 - 20 cm
40 - 60 cm
3
> 20 cm
> 60 cm
Eunicella singularis
Eunicella verrucosa
Leptogorgia sarmentosa
< 20 cm
< 20 cm
< 30 cm
20 - 30 cm
20 - 30 cm
30 - 50 cm
> 30 cm
> 30 cm
> 50 cm
Figure 3.4 - Quality scores per layer (QL) of each survey grouped in the three habitat types. A = Facies with Paramuricea clavata; B = Association with Cystoseira zosteroides; B* = Association with Cystoseira zosteroides and gorgonians. UL = upper layer; IL = intermediate layer; BL = basal layer. Classes of the ecological status classification of the Water Framework Directive are also reported (from Bad to Good status).
43
3. Rapid Visual Assessment: The method Table 3.3 - List of species of the intermediate layer. Algae
Anthozoa
Codium bursa
Aiptasia mutabilis
Dictyopteris polypodioides
Balanophyllia europaea
Dictyota dichotoma
Cerianthus membranaceus
Flabellia petiolata
Cladocora caespitose
Halimeda tuna
Condylactis aurantiaca
Porifera
Parazoanthus axinellae Phyllangia americana mouchezii
Acanthella acuta Agelas oroides
Serpuloidea
Axinella damicornis
Protula tubularia
Axinella verrucosa
Salmacina dysteri
Chondrosia reniformis
Bryozoa
Clathrina clathrus
Bugula fulva
Dysidea avara
Bugula plumosa
Dysidea sp.
Myriapora truncata
Haliclona cratera
Pentapora fascialis
Hemimycale columella
Reteporella grimaldii
Ircinia variabilis
Rhynchozoon sp.
Oscarella lobularis
Schizoporella errata
Petrosia ficiformis
Smittina cervicornis
Pleraplysilla spinifera
Tunicata
Hydroida
Clavelina lepadiformis
Aglaophaenia sp.
Halocynthia papillosa
Garveia grisea
44
3. Rapid Visual Assessment: The method
Table 3.4 - Quality scores assigned to layers descriptors in each survey. 1.1
1.2
Cover Ul
2
2
2
2
1
1
2
3
3
2
3
2
2
1
2
2
Maximum height
2
3
3
2
1
1
3
2
2
2
2
3
2
2
2
2
Epibiosis-Necrosis
3
3
2
1
1
1
1
2
2
2
2
3
1
1
1
1
Species richness
2
3
3
2
2
2
1
2
2
2
2
2
2
1
2
1
Seasonal/Perennial
3
2
3
3
2
3
3
3
3
3
3
3
3
3
3
3
Bryozoans
2
1
2
2
1
1
2
3
2
2
2
3
2
2
2
3
1,6
1,9
1,8
1,9
1,6
2,0
2,0
1,2
1,4
1,2
1,2
1,9
1,7
1,8
1,6
2,1
Penetrometry
3
1
3
1
1
3
2
3
3
2
2
1
1
1
1
1
Borers
3
3
2
3
3
3
3
3
3
3
3
3
3
3
3
3
Cover Bl
1.3 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 4.1 4.2 4.3 4.4 5.1
Figure 3.5 - Quality scores per layer in each survey represented according to the chromatic scheme of the WFD. Red = Bad status; orange = Poor status; yellow = Moderate status; green = Good status; blue = High status.
45
3. Rapid Visual Assessment: The method
3.4 Discussion To date, no general consensus has been achieved in the definition of coralligenous, being defined as eco-ethological crossroad (Laubier, 1966), biocoenosis (Hong, 1982), polybiocoenotic entity (Picard, 1985), assemblage (UNEP/IUCN/GIS Posidonie, 1990), community (Garrabou et al, 1998), community puzzle (Ballesteros, 2006) and seascape (Giaccone, 2008; UNEP-MAPRAC/SPA, 2008). We embraced the latter definition of coralligenous as a seascape, because its heterogeneity reflects exactly the mosaic of habitat patches forming a landscape (Dunning et al., 1992). A seascape approach describes relationships between ecological processes and environmental patchiness and between spatial configuration of habitats and typology of territorial elements (Farina, 2006). The choice to combine geomorphology, mesology and bionomy, as proposed earlier by Cocito et al. (1991), resulted in an effective way to characterise and evaluate the quality of coralligenous assemblages. This approach is consistent with the indication of the Marine Strategy Framework Directive (MSFD) to integrate abiotic and biotic features for assessing the ecological status of marine habitats. Although preliminary, this first description of the coralligenous shoals of Vado Ligure was obtained using some of the indicators proposed by the MSFD for seafloor integrity assessment (Rice et al., 2012), such as the status of the biogenic substrate (thickness and consistency of calcareous concretion in our case), the presence of sensitive species, the benthic community condition and functionality (e.g. species richness), the number of individuals over a specified size. The employment of a Rapid Visual Assessment technique (RVA) solved most of the constraints linked with the wide bathymetric distribution of coralligenous: it optimised underwater work allowing the direct collection of a sufficient amount of data with a congruent diving effort. The RVA turned out to be much effective when joined with a detailed cartography of the seafloor: the preliminary map based on the multi-beam surveys was indispensable to localise exactly each rocky outcrop and to better finalize field activities. A first characterisation of the coralligenous shoals of Vado Ligure was achieved through the recognition of habitat types as defined by the EuNIS classification on the basis of the dominant species of the upper layer (Bianchi and Morri, 2001; Davies et al., 2004). High rocky outcrops below 30m depth showed the A habitat type, i.e. the facies with Paramuricea clavata. This habitat was always associated with the highest values of sediment cover in the basal layer. Landslide deposits and rocky outcrops between 17m and 30m depth had the B habitat type, i.e. the association with Cystoseira zosteroides. At Vado Ligure, Cystoseira zosteroides was often found together with gorgonians, thus calling for the necessity to define this variant of the habitat B as a new habitat typology, namely B*. In both B and B* habitats the basal layer was characterised by a higher percentage of encrusting calcified rhodophyta (ECR) than the habitat A: G. Giaccone observed a positive relationship between active bioconstruction in the basal layer and the occurrence of Cystoseira zosteroides, instead of dense gorgonian canopies, in the upper layer (personal communication). Although sediment cover may be strongly affected by short-term hydrodynamic condition, the high cover of sediment in the basal layer and the concomitant occurrence of species indicative of turbidity, such as Eunicella verrucosa and Leptogorgia sarmentosa (Carpine and Grasshoff, 1975), suggest that high sedimentation rate is possibly the main stressor affecting the
46
3. Rapid Visual Assessment: The method coralligenous of Vado Ligure. Our results are consistent with the situation observed in other areas of the Ligurian Sea, e.g. the “Cinque Terre” marine protected area (La Spezia, Italy), where high turbidity and sedimentation rate were associated with the same species of gorgonians (Roghi et al., 2010 and references therein). High sedimentation was also locally associated with low cover by encrusting algae in the basal layer and with the absence of Cystoseira zosteroides from the upper layer Balata et al., 2007 and references therein; Piazzi and Balata, 2011). This confirms a negative influence of sedimentation on the structure of coralligenous assemblages (Irving and Connel, 2002). Although both the basal and the upper layers exhibited a clear response to water turbidity and sediment deposition, no evidences of such impacts had been shown by the intermediate layer. The different dynamics observed in the three layers of the coralligenous (Bianchi et al., 2007; Ponti et al., 2011b) justify our choice to keep them separated, both operationally during field activities and analytically during assessment of their quality. QL scores of the three layers were different in most of the cases and the resulting quality states were often discordant among layers (Table 3.5). A “global” quality index computed from the QL values of the three layers taken together would be, therefore, inappropriate to depict the overall status of the coralligenous seascape. Working on distinct layers would also be useful in response to specific objectives of management and conservation: when the major interest is the aesthetic value, for instance, efforts should be focused on evaluating the nature and the three-dimensional structure of the upper layer; when the focus is biodiversity, the intermediate layer would be an opportune proxy; finally, if the goal is the maintenance of bioconstruction, the quality of the basal layer would be the most informative. As we did in our study, the synthetic quality index (max QL) grouped together the values of surveys belonging to each habitat type per layer. Alternatively, a synthetic quality index could be computed either for each shoal (grouping QL values of each survey on this shoal) or for the whole system of shoals in the area. Three different methods can be adopted for computing a synthetic quality index starting from a number of quality indices, i.e. summing values, averaging values and considering the maximum value, each with its pros and cons (Table 3.6). The max QL index we used in this study was computed from the maximum value of QL indices belonging to each habitat type: with this method, the potentiality of coralligenous assemblages is evidenced, thus providing indications for undertaking protection measures, as required by the MSFD (Borja et al., 2012).
Table 3.5 - Concordances and discordances of the total quality score (QL) between layers.
CONCORDANCE
DISCORDANCE
All layers
4
12
Upper x Intermediate
7
9
Upper x Basal
8
8
Intermediate x Basal
6
10
47
3. Rapid Visual Assessment: The method Table 3.6 - The three main methods that can be used to obtain the synthetic quality index of the habitat type from QL scores of each layer belonging to each habitat type, with their respective pros and cons.
Method
Pros
Sum of the values of surveys (sum QL)
Cons
It takes into account the surplus
value
of
ecodiversity Maximum
value
of
surveys (max QL)
surveys (mean QL)
number
of
surveys should be balanced
Surveys with high values and potentiality of layers
It does not consider ecodiversity
are evidenced Many surveys with low
Mean of the values of
The
Averaging low values
value give a low value
with
the
produces
synthetic
index
quality
high
values
mediocrity
Although the quality of the coralligenous of Vado Ligure was found Bad or Poor in most cases, the max QL ranged from Moderate to Good, thus suggesting that these assemblages may have the potential to recover if chronic stresses such as sedimentation are properly addressed by conservation measures. When the main aim is to assess the overall quality of habitat types, the use of the mean QL would be more appropriate. Selection of the method for computing the synthetic quality index should be guided by the objective of the study. The approach we proposed has been developed and applied for the first time to characterise the status of the coralligenous shoals in the specific area of Vado Ligure. The RVA technique combined with the seascape approach results were promising, although some improvements are necessary and applications in other areas are needed in order to make it repeatable, comparable and to reduce the subjectivity of underwater operators.
Acknowledgements This work was partially done within the frame of the research agreement between DipTeRis (currently DiSTAV) and CIMA foundation. I am grateful to the Savona Port Authority (Savona, Italy) for conceding baseline multi-beam data, collected in 2006 by Drafinsub Survey, partially reproduced in Fig. 3.5. R. Ivaldi and the Istituto Geografico della Marina (Genoa, Italy) are thanked for raw bathymetric data of Fig. 3.1 (Prot. SRE 1653) and F. Siccardi for scientific cooperation. Finally, we would like to acknowledge A. Balduzzi (University of Genoa), F. Mastrototaro (University of Bari), A. Occhipinti Ambrogi (University of Pavia) and R. Pronzato (University of Genoa) for advice on some species.
48
4 A new Rapid Visual Assessment (RVA) approach for the characterisation and the assessment of the quality of coralligenous reefs: Application and validation
This study has been made possible thanks to the collaboration with the Environment and Resources Laboratory of the French Research Institute for the Exploitation of the Sea (Ifremer) – Mediterranean Centre, in particular with Dr. S. Sartoretto (Ifremer).
4.1 Introduction Coralligenous reefs are an endemic Mediterranean bioconstruction and represent the second pole of species diversity in the Mediterranean Sea, after the Posidonia oceanica meadows (Boudouresque, 2004). Nevertheless, due to their complex structure (Pérès and Picard, 1964; Ros et al., 1985), diverse fauna (Laubier, 1966) and the restrict number of studies dealing with their biodiversity, they probably host more species than other Mediterranean habitat (Ballesteros, 2006). Estimates on the number of invertebrates (Laubier, 1966; Ros et al., 1984; Ballesteros, 2006) and macroalgae (Boudouresque, 1973) exist, while only one study was addressed to assess the number of fishes (Harmelin, 1990), so it is necessary to refer to the available literature regarding the biology of Mediterranean fishes (e.g., Whitehead et al. 1984, 1986; Corbera et al., 1996; Mayol et al., 2000). A first global and still conservative estimate of the total number of coralligenous species is given by Ballesteros (2006) and accounts 1666 species. Despite their evident importance, the most significant European Directives in the field of environmental protection never refer directly to coralligenous reefs. Nevertheless, the Habitat Directive (HD, 92/43/EEC) cites “reefs” in general in the list of habitats types of community interest (Annex I), so including automatically coralligenous reefs in the network of Natura 2000 sites (Council European Communities, 1992). On the contrary, the Water Framework Directive (WFD, 2000/60/EC) mentions the assessment of composition and abundance of the benthic flora and fauna only as a means for the monitoring of the ecological status of marine waters, and this probably addressed the researchers to some “easiest-to-study” and most known benthic habitats, like soft bottoms (Borja et al., 2000; Simboura and Zenetos, 2002; Rosenberg et al., 2004; Muxika et al. 2007), upper infralittoral rocky shores (Orfanidis et al., 2001; Ballesteros et al., 2007) or seagrass meadows (Romero et al., 2007; Gobert et al., 2009; Lopez y Royo et al., 2009).
49
4. Rapid Visual Assessment: Application and validation Finally, introducing the concept of “seafloor integrity” as a descriptor of the status of marine environment, the Marine Strategy Framework Directive (MSFD, 2008/56/EC) encouraged the study of all benthic habitats and internationally aroused the interests on the subject of coralligenous reefs. This allowed pinpointing the gaps concerning the knowledge about the geographical and bathymetrical distribution, the taxonomy, the functioning and the dynamics of coralligenous communities (UNEP-MAP-RAC/SPA, 2008), thus justifying the difficulties in finding some means to measure their quality state. Usually, to assess the Ecological Status (ES) of a habitat it is necessary to follow three steps (Borja et al., 2012): i) the definition of a reference condition or, where not possible, some environmental targets for the good status of the habitat; ii) the computation of an indicator for the reference condition for the considered habitat, in order to obtain an Ecological Quality Ratio (EQR) (sensu WFD); iii) the conversion of the EQR into an ES. Concerning coralligenous reefs, the paucity of knowledge coupled with the diffused problem of the sliding (or shifting) baselines (Al-Abdulrazzak et al., 2012), does not allow to define general reference conditions or to set targets using “traditional” methods (see Borja et al., 2012). A possible solution could be to provide a large-scale current baseline to which refer to for future evaluation of habitat state, as did Sala et al. (2012) for nearshore rocky reefs. Paucity of knowledge and operational restrictions imposed by scuba diving (Parravicini et al., 2010) when working at depths at which coralligenous reefs develop, limited the number of studies concerning the assessment of their health status. The Ecological Status of Coralligenous Assemblages index (ESCA index) (Cecchi and Piazzi, 2010) and the Coralligenous Assemblage Index (CAI) (Deter et al., 2012) both adopted the state of coralligenous assemblages as an indicator of ES of coastal waters, according to the WFD, and they are based on data collected by photographic sampling and further analysis of images. The Rapid Visual Assessment (RVA) method proposed by the author (Gatti et al., 2012), on the contrary, is aimed to assess the quality of coralligenous reefs as an indicator of seafloor integrity, as defined by MSFD, and is based on a seascape approach allowed by direct SCUBA diving observations. The main critics moved to the RVA protocol are that the variability between observers can easily affect the visual estimations of percent cover (Meese and Tomich, 1992) and the correct identification of species (Thompson and Mapstone, 1997); such a discrepancy may be due to variability in observer experience, as demonstrated in measuring habitat variables in terrestrial habitats by Block et al. (1987), however Meese and Tomich (1992) did not found evidence of an experience effect. Therefore, the first objective of the present paper was to test the robustness of the RVA to the observer bias. Through the application of the method at different biogeografical sector, at larger scale than that of the site where it was developed and in sites affected by different levels of human pressures, the second objective of the work was to verify the effectiveness and the sensitivity of the RVA. Finally, despite a global quality score can be inappropriate to depict the overall status of coralligenous seascape (Gatti et al., 2012), usually a single number that resume the state of the habitat is what is requested in order to comply with management issues. Therefore, the third objective of the present paper was to propose the most appropriate formula to give a global quality score for each site which better reflects the health state of coralligenous reefs.
50
4. Rapid Visual Assessment: Application and validation
4.2 Materials and methods 4.2.1 Study area and field work The study area is located along French Mediterranean coasts, around the cities of Toulon, La Ciotat and Marseille (Provence-Alpes-Côte-d’Azur – PACA region) (Fig. 4.1). The area is characterised by high population density, which determines the presence, along an 80 km long coast, of four stations of sewage treatment and discharge, for a total capacity of 2.3 millions of equivalent inhabitants. During the summer, this number increases because of the touristic presence, concurrently with the intensification of all the marine activities (e.g. SCUBA diving, yachting, etc.) usually practised during the whole year. The local fishery completes the picture of an area where the coastal marine environment is strongly affected by the human presence. The study area is characterised by high rocky shores mainly composed by limestone in the zone of Marseille, conglomerates at La Ciotat and phylades (siliceous rocks) near Toulon. Between the surface and the continental shelf, the area is dominated by the Liguro-Provencal current, which normally flows westwards. Nevertheless, coastal circulation is also constrained by two dominant winds: north-western (upwelling favourable) and south-eastern (downwelling favourable) winds (Pairaud et al., 2011). Locally (e.g. around Marseilles), the upwelling induce a decrease of temperature reaching more than 5°C (Millot, 1990), which influence the dynamic of benthic communities associated to hard bottoms, like coralligenous reefs. Sampling stations (Fig. 4.1) were divided into three groups corresponding to levels of pressure they are subjected to, according to an expert judgement:
High pressure, mainly due to sewage outfalls: Ile Plane Nord (8), Figuerolle (12), Sêche des Pêcheurs West (17), Large Oursinière (20);
Moderate pressure, mainly due to scuba diving, fishery, sediment resuspension or few effects of sewage discharge: Méjean (1), Large Niolon (2), Fromages (6), Imperial du Milieu (9), Ile Plane South (10), Bec de l’Aigle West (13), Pierre du Levant (16), Formigue (21);
Low pressure: Tiboulen (3), Ile du Planier (4), Cap Caveau (5), Moyade (7), Sêche des Pêcheurs Est (18), Morgiou (11), Bec de l’Aigle Est (14), Les Rosiers (15), Les Deux Frères (19).
Sampling activities took place between Mars and December 2013. In each site, three replicates at constant depth were collected by a single SCUBA diving operator (hereafter called Op1), for a total of 54 replicates. In Table 4.1 is summarised the kind of data collected to obtain the geomorphology and the mesologic features of the site and the bionomic characterisation of coralligenous reefs. For details about the RVA protocol see Gatti et al. (2012). In order to assess the sensitivity of RVA to observer biases, a second operator (hereafter called Op2) applied the protocol (with the exception of the six random photographs for the intermediate layer) on the same surfaces in independent dives. Finally, four 60 x 40 cm photographs (hereafter called Ph) for each RVA replicate were taken on the same surfaces observed by SCUBA diving operators, in order to compare the visual percent cover estimation of encrusting species with the more “objective” percent cover obtained by the digital analysis of images.
51
4. Rapid Visual Assessment: Application and validation In addition, in order to verify if the experience of operators could influence the visual estimation of NTG’s percent cover, in two sites data were collected by four operators: Op1 and Op2, which have a good experience in visual estimations and identification of benthic species, and Op3 and Op4, which have absolutely no experience. A rapid pre-sampling briefing gave them the essential information for the identification of the NTG.
Figure 4.1 – Study area and sampling stations. The white arrows indicate the position of water treatment stations discharges; from west to east: Cortiou (Marseille), Figuerolle (La Ciotat), Cap Cicié (Toulon west), Pont de la Clue (Toulon east).
52
4. Rapid Visual Assessment: Application and validation Table 4.1 – Data collected for each replicate. In the bionomic characterisation, the height of species considered part of each layer is reminded.
Geomorphologic characterisation Morphotypes
Shoal/outcrop/cliff/landslide/deposits/detritic Mesologic characterisation
Physical features
Depth, slope, exposure, elevation from the bottom Bionomic characterisation Percent cover of non-taxonomic groups (NTG): encrusting calcified rhodophyta
Basal layer < 1 cm height
(ECR), non-calcified encrusting algae (NCEA), encrusting animals (EA), turf-forming algae (TURF), sediment (SED) Semi-quantitative abundance of boring species Thickness and consistency of calcareous layer
Intermediate List of species layer 1 to 10 cm 6 random photographs without frame height Upper layer > 10 cm height
Visual estimation of percent cover of each species Maximum height of each species Percentage of necrosis (also if covered by epibiosis)
4.2.2 Data management 4.2.2.1 Robustness to observer biases Photographs 40 cm x 60 cm were analysed using the software Vision 1.0 (Rende et al., 2009), which allowed obtaining the percent cover of ECR, NCEA, AN, TURF and SED through a freehand outline technique. Then, one-way analyses of variance (ANOVA) were used to test: i) the differences among the percent cover of the former groups visually estimated underwater by Op1 and Op2 and those obtained by the analysis of the photographs (Ph); ii) differences among the visual estimation of Op1, Op2, Op3, and Op4. The homogeneity of variances was verified by Cochran’s test and, whenever necessary, percentage cover data were transformed to arcsin√(x/100). Whenever transformations did not produce homogeneous variances, ANOVA was used after setting α = 0.01 in order to compensate for the increased likelihood of Type I error (Underwood, 1997). Post hoc comparisons were made by the Tukey’s Honestly Significance Difference (Tukey’s HSD) test. The differences between Op1 and Op2 where tested by the comparison of row data collected for each layer, with the exception of the cover of NTG (already tested with the former analysis of variance). In each case, Cochran’s test was made in order to verify the homogeneity of variances and, whenever necessary, cover data were transformed to arcsin√(x/100); then one way ANOVAs were performed to test the differences between operators.
53
4. Rapid Visual Assessment: Application and validation 4.2.2.2 Characterisation and quality assessment From the quali-quantitative composition of the assemblages, the dominant species allowed for the definition of the habitat types of sites. In Gatti et al. (2012), habitat types were identified according to the European Nature Information System (EuNIS) (Davies et al., 2004) codification. Here the codification approach was maintained where possible, but in case of facies or associations that were not present in EuNIS list, the habitat type was identified by the name of dominant species. Geomorphologic, abiotic and bionomic features were summarised in a table in order to obtain a comprehensive characterisation of each site. The quality of coralligenous reefs was assessed for each layer individually on the basis of the different descriptors. A total of nine descriptors, three for each layer, were used and were converted to quality scores ranging from 1 to 3. As the protocol was improved compared to what described in Gatti et al. (2012), in Table 4.2 are summarised the new criteria for the assignation of quality scores to each descriptor. Briefly, differences with the previous protocol consist in: considering TURF and SED as a unique NTG for the assignation of quality scores in the basal layer; reducing the upper limit for the assignation of the maximum quality score (3) for the number of species and of erect bryozoans in the intermediate layer; substituting the descriptor “Seasonal-perennial species ratio (S/P)” with “Sensitivity of bryozoans (SB)” in the intermediate layer. In order to get the quality score for each layer (QL), the formula [1] formerly described in Gatti et al. (2012) and inspired by the one adopted by Bianchi (2007) was applied: QL = (XL YL ZL) k(1-n) [1]. Then, to obtain a global quality score (QG) for each site starting from the QL of the three layers, different formulas were tested in order to find the best one: formula [1], sum, maximum value, and arithmetical, geometric, harmonic and quadratic means. On the basis of results, the harmonic mean [2] was then used to compute the global quality score of coralligenous reefs: QG = n / (1/QLB + 1/QLI + 1/ QLU [2] where n is the number of layers and QLB, QLI and QLU are respectively the quality scores of Basal, Intermediate and Upper layer. Finally, in order to obtain a classification of quality, for both QL and QG, only three classes were considered: Bad (red) when 0 < QL or QG ≤ 1, Moderate (yellow) when 1 < QL or QG ≤ 2, Good (green) when 2 < QL or QG ≤ 3.
4.3 Results 4.3.1 Robustness to observer biases Analyses of variance performed on percent cover of NTG showed not significant differences among the visual estimations of Op1, Op2 and Ph for ECR (F = 0.204, d.f. = 2, p = 0.816), NCEA (F = 0.409, d.f. = 2, p = 0.666) and AN (F = 0.049, d.f. = 2, p = 0.953); a significant difference was detected only for TURF/SED (F = 3.663, d.f. = 2, p < 0.05) and Tukey’s HSD test revealed that the percent cover was lower for Op2 compared to Op1.
54
4. Rapid Visual Assessment: Application and validation A significant difference between data collected by Op1 and Op2 was found in the number of species of the intermediate layer (F = 11.48; d.f. = 1; p < 0.01) and Tukey HSD test revealed that Op2 found a lower number of species than Op1. For the other descriptors the analysis of variance showed not significant differences between operators (Table 4.3). The comparison among Op1, Op2, Op3 and Op4 showed not significant differences in the percent cover estimation of all NTG (ECR: F = 2.552, d.f. = 3, p = 0.0844; NCEA: F = 0.592, d.f. = 3, p = 0.628; AN: F = 0.805, d.f. = 3, p = 0.506; TURF/SED: F = 0.279, d.f. = 3, p = 0.84).
4.3.2 Characterisation The geomorphologic and mesologic characterisation shows that the most diffused morphotypes are cliffs and platforms with variable slope. In most cases they are characterised by the EuNIS habitat “Mediterranean coralligenous communities moderately exposed to hydrodynamic action” (code A4.26), and in particular by the facies with Paramuricea clavata (code A4.26B) or the facies with Eunicella cavolini (code A4.269), which in some cases coexist. However, two sites, one belonging to the high pressured (site 8) and one to the moderate pressured (site 2), are characterised by a so-called “facies of impoverishment”, which shows the vestiges of ancient populations of gorgonians affected by some pressures. The morphotype outcrop, less represented among our sampling sites, is generally characterised by associations of sciafilic (Flabellia petiolata, Halimeda tuna) or photophilic green algae (Codium bursa, Codium coralloides) and only in one case by a habitat type codified in the EuNIS list, the association with F. petiolata and Peyssonnelia squamaria (code A3.23J), which belongs to the habitat “Mediterranean communities of infralittoral algae moderately exposed to wave action“ (code A3.23). The only exception is represented by site 15, where the vertical southern wall of the shoal is dominated by a facies with P. clavata (code A4.26B). A global vision of all sites is summarised in Table 4.4.
4.3.3 Quality assessment Layer’s Quality Scores (QL) show a high variability both within and among sites. In sites characterised by high levels of pressure, the basal layer shows bad quality (QLB ≤ 1) in all sites excepte for site 20 (QLB = 1.1), the intermediate layer has scores lower than 1 with the exception of site 8 (Q LI = 1.4) and the upper layer always shows QLU ≤ 1. In sites subjected to a moderate pressure, the basal layer shows bad quality in one case (site 6, QLB = 0.8), while all the other sites exhibit a moderate quality (1.1> QLB ≤2); the intermediate layer indicate a moderate quality in three out of the eight sites (1, 10 and 16) and a good quality score (QLI > 2) in the remaining ones; the upper layer show very variable situations: low quality in site 2, moderate in sites 1, 9, 13 and 21, and good quality in sites 6 and 10. Finally, concerning low pressured sites, the basal layer shows bad quality in site 14 (QLB = 0.7), good in site 4, 5 and 7 and moderate in all the other sites; intermediate and upper layer show good quality in all sites except sites 14 (QLI = 1.7, QLU = 1.7) and 15 (QLI = 0.6, QLU = 0.7). Global Quality Scores show bad quality (QG ≤ 1) for sites 8, 12, 15, 17 and 20 all subjected to high pressure with the exception of site 15, which was classified among the low impacted ones,
55
4. Rapid Visual Assessment: Application and validation according to the expert judgement. A moderate quality (1.1> QG ≤2) of coralligenous reefs is detected for sites 1, 2, 4, 6, 9, 10, 13, 14 and 16, all subjected to a medium pressure whit the exception of site 14, which was considered subjected to low pressure conditions. Finally, good quality scores (QG > 2) are assessed for sites 3, 4, 5, 7, 11, 18 and 19, all subjected to a low pressure.
Table 4.2 – Criteria for the assignation of quality scores to each descriptor, for each layer. BASAL LAYER 1 -> TURF/SED Percent cover of NTD (PC)
2 -> NCEA/AN 3 -> ECR
Thickness and consistency of calcareous layer (CL)
1 -> null penetration 2 -> penetration > 1 cm 3 -> penetration up to 1 cm 1 -> common
Borer marks (BM)
2 -> occasional 3 -> absent INTERMEDIATE LAYER 1 -> SR < 5
Specific Richness (SR)
2 -> 5 ≤ SR ≤ 8 3 -> SR > 8 1 -> ECB ≤ 1
Erect Calcified Bryozoans (ECB)
2 -> 1 < ECB ≤ 3 3 -> ECB > 3 1 -> M. truncata
Sensitivity of bryozoans (SB)
2 -> P. fascialis, A. calveti 3 -> S. cervicornis, R. grimaldii UPPER LAYER 1 -> cover < 5%
Total cover of species (TC)
2 -> 5% ≤ cover ≤ 25% 3 -> cover > 25%
Maximum height (MH)
1 -> MH < 0.3 LMH
LMH = Literature max height, the maximum
2 -> 0.3 LMH ≤ MH ≤ 0.6 LMH
height find in literature for each species
3 -> MH > 0.6 MH 1 -> N > 75%
Necrosis (N)
2 -> 10% ≤ N ≤ 75% 3 -> N < 10%
56
4. Rapid Visual Assessment: Application and validation Table 4.3 – Results of ANOVA testing the differences in the row data collected by Op1 and Op2 for each descriptor. ns = not significant; ** p
