ACTA GEOGRAPHICA GEOGRAFSKI ZBORNIK SLOVENICA 2020 60 1 ACTA GEOGRAPHICA SLOVENICA GEOGRAFSKI ZBORNIK 60-1 • 2020 Contents Mojca POKLAR Comparison of the sonar recording method and the aerial photography methodfor mapping seagrass meadows 7 Vanja PAVLUKOVIĆ, Uglješa STANKOV, Daniela ARSENOVIĆ Social impacts of music festivals: A comparative study of Sziget (Hungary) and Exit (Serbia) 21 Péter János KISS, Csaba TÖLGYESI, Imola BÓNI, László ERDŐS, András VOJTKÓ,István Elek MAÁK, Zoltán BÁTORI The effects of intensive logging on the capacity of karst dolines to provide potential microrefugia for cool-adapted plants 37 Radu SĂGEATĂ Commercial services and urban space reconversion in Romania (1990–2017) 49 Kristina IVANČIČ, Jernej JEŽ, Blaž MILANIČ, Špela KUMELJ, Andrej ŠMUC Application of a mass movement susceptibility model in the heterogeneous Miocene clastic successions of the Slovenj Gradec Basin, northeast Slovenia 1 Andrej GOSAR Measurements of tectonic micro-displacements within the Idrija fault zone in the Učjavalley (W Slovenia) 79 Piotr RAŹNIAK, Sławomir DOROCKI, Anna WINIARCZYK-RAŹNIAK Economic resilienceofthe command andcontrolfunctionof citiesin Centraland EasternEurope 95 Mateja FERK, Rok CIGLIČ, Blaž KOMAC, Dénes LÓCZY Management of small retention ponds and their impact on flood hazard prevention in the Slovenske Gorice Hills 107 Gregor KOVAČIČ Sediment production in flysch badlands: A case study from Slovenian Istria 127 Vesna LUKIĆ, Aleksandar TOMAŠEVIĆ Immigrant integration regimes in Europe: Incorporating the Western Balkan countries 143 Mitja DURNIK Community development: LocalImmigrationPartnershipsin Canadaand implications forSlovenia 155 ISSN 1581-6613 9 771581 661010 COMPARISON OF THE SONAR RECORDING METHOD AND THE AERIAL PHOTOGRAPHY METHOD FOR MAPPING SEAGRASS MEADOWS Mojca Poklar Underwater image of a meadow in Semedela Bay. DOI: https://doi.org/10.3986/AGS.5161 UDC: 911.2:582.533.1(497.4Semedelski zaliv) 681.883:582.533.1(497.4Semedelski zaliv) 528.7:582.533.1(497.4Semedelski zaliv) COBISS: 1.01 Mojca Poklar1 Comparison of the sonar recording method and the aerial photography method for mapping seagrass meadows ABSTRACT: This article presents a new perspective on the study of the spatial distribution of seagrass meadows,which–duetotheirsensitivitytocoastalhydrodynamics,sedimenttransport,changesinnutri­entcontent,anddisruptionsduetohumaninterventionintheirenvironment–areagoodindirectindicator of the properties of seawater. Monitoring their extent and characteristics is essential for determining the properties of seawater, but this requires developing a precise methodology that involves acquiring data on the occurrence of seagrass meadows and mapping them. The base data for the survey presented are sonar recording and aerial photography data, which were utilized to create a seabed classification using geographicinformationsystems(GIS).Thisprovidedinformationontheextentandcharacteristicsofthe seagrass meadows. Spatial analysis offers a new look at the coastal belt and reveals some new features. KEYWORDS:geography,SemedelaBay,seagrassmeadows,multibeamsonardata,aerialphotography,GIS, line transect method, coastal area Primerjava metode sonarskega snemanja in metode zračne fotografije za namen kartiranja morskih travnikov POVZETEK: Prispevek prikazujenov pogled na preučevanje prostorske porazdelitve morskih travnikov, ki so zaradi njihove občutljivosti na obalno hidrodinamiko, transport sedimentov, spremembe vsebnos­tihranilinmotnjezaradičlovekovegaposeganjavnjihovookolje,doberposrednipokazateljlastnostimorske vode.Spremljanjenjihovegaobsegainlastnostijenamrečbistvenopriugotavljanjulastnostimorskevode, zahteva pa natančno izdelano metodologijo, ki vključuje pridobivanje podatkov o razširjenosti morskih travnikov in njihovo kartiranje. Izhodišče za izvedeno raziskavo so bili podatki sonarskega snemanja in zračnefotografije,nakaterihsmozuporabogeografskoinformacijskihsistemovizvedlitipizacijomorskega dna, kjer je bila posebna pozornost posvečena morskim travnikom. S tem smo dobili podatke o obsegu inlastnostimorskihtravnikov.Prostorskeanalizesoomogočilenovpoglednaobalnipasinrazkrilenekatere nove značilnosti. KLJUČNE BESEDE: geografija, Semedelski zaliv, morski travniki, podatki večsnopnega sonarja, zračna fotografija, GIS, metoda linijskih presekov, obalno območje The article was submitted for publication on July 5th, 2017. Uredništvo je prejelo prispevek 5. julija 2017. 1 University of Primorska, Faculty of Humanities, Koper, Slovenia mojca.poklar@fhs.upr.si 8 1 Introduction Seagrass meadows are one of the most important marine ecosystems in the world in terms of the goods they produce and ecosystem services they provide (Telesca et al. 2015). Due to their characteristics and sensitivitytocoastalhydrodynamics,sedimenttransport,changesinnutrientcontent,anddisruptiondue to human intervention in their environment, seagrasses are important species in determining the quali­tyofacoastalecosystem(Krause-Jensenetal.2004;Ralphetal.2007;McMahonetal.2013;Peterlin2013; Vacchi et al. 2014). Despite their importance, they are constantly threatened by numerous human activ­ities that eventually lead to their degradation and rapid loss (Duarte 2002), estimated at a rate of 110km2/year since 1980 (Waycott et al. 2009). Consequently, seagrass meadows are regularly included in monitoringprograms,bothfortheirprotectionandfortheirvalueasabioindicator(McMahonetal.2013). To understand the dynamic nature of seagrass meadows and to predict their response to future environ-mentalchanges(Unsworthetal.2014),itisnecessarytosynopticallymonitorthechangesinthecomposition of a meadow, its spatial distribution or cover, andits biomass. Therefore, developing an effective method­ology for monitoring meadows is a very topical issue (Comas Gonzales 2015). Thisstudyfocusesonmappingseagrassmeadowsandthusexaminingtheirspatialdistributionasone of the parameters of monitoring seagrass meadows (Hossain et al. 2014). Because the changes in spatial distribution occur on small (<1km2) and large (>100km2) spatial scales, traditional field surveys (diver observations, sampling using rakes or scrapers, and other methods) are often inconvenient for mapping large areas (McKenzie 2003; Hossain et al. 2014). With the development of geographic information sys­tems (GIS) and technical improvements in remote sensing techniques (Robbins 1997), indirect methods have become more popular. Due to the ability of remote sensing to detect changes in the spatial distrib­ution of seagrass meadows on larger spatial and time scales, it is among the most important tools in the managementofseagrassmeadowsbecauseoftimeefficiency,speedofuse,largecoverage,andreproducibility of observations (Hossain et al. 2014). On a global scale, mapping seagrass meadows using remote sensing techniques is already well known (Hossain et al. 2014), whereas in Slovenia mapping seagrass meadows is still carried out using only field surveys (Turk et al. 2002; Lipej et al. 2007). Despite the fact that there have been some individual attempts to map the seabed using remote sensing techniques (Berden Zrimec, Poklar and Moškon 2015; Berden Zrimec et al. 2015; Moškon et al. 2015), the traditional approach is still predominant. In order to reduce theconstraintsimposedbythisapproach,thisstudycomparedthesonarrecordingmethodandaerialpho­tography method, and it verified this with the already established line transect method for determining the spatial distribution of seagrass meadows. The aim of this research was to evaluate the selected meth­odsbasedontheobtaineddataqualityandtodeterminetheirsuitabilityandopportunitiesforuseinfurther research.Ofparticularinterestwastheaccuracyofbothmethods,especiallyaerialphotography,forwhich it was assumed that the significant water turbidity typical for Slovenian waters and for a large part of the northern Adriatic Sea would be a limiting factor. 2 Methods 2.1 Research area TheresearchareacoveredSemedelaBayasthesoutheasternmostpartofKoperBaybetweenŽusternaand theold town of Koper. This is a shallow bay with an average depth of 6m (Harpha sea 2013), and, despite its strong anthropogenic transformation, its coastline has the characteristics of a depositional coast. Due to its erodible flysch hinterland (Zorn 2009), the Badaševica River carries sediments that are deposited in the sea (Malačič 1994; Orožen Adamič 2002). The area is a uniquehabitatbecause it differs from the cen­tralpartofKoperBayinitsnaturalcharacteristics.Themixingofseawaterandfreshwatervariesconsiderably over the course of the year (Poklar 2016). Due to this variability, the area is suitable for researching the impact of changing water properties on seagrass coverage. Two types of seagrass are found in Semedela Bay: little Neptune grass (Cymodocea nodosa) and common eelgrass (Zostera marina; Lipej et al. 2006). 2.2 Definition of a seagrass meadow Because the perimeter of a seagrass meadow, which is the basis of determining its entire area, cannot be absolutely determined, problems may arise in defining it. In measuring phenomena that are not directly measurable, the need for an operational definition arises. This ensures that the understanding of phenomena and the data collection method are unified and repeatable (Adanza 1995). Therefore, for the purpose of this survey, the operational definition of a seagrass meadow and thereby the minimum mapping unit of 0.01 ha were defined. Even though very sparse seagrass may indicate that seagrass appears in a certain area, such areas were excluded from the operational definition of the seagrass meadow. There are several rea­sons for this: very sparse seagrasses have very little ecological value and also visually do not correspond to the idea of a meadow. In addition, monitoring very sparse seagrass and tracking its changes is very dif­ficult (Virnstein et al. 2000). 2.3 Mapping seagrass meadows using the sonar recording method Sonar data, which are essential for this survey, were obtained from bathymetric measurements with a Reson SeaBat 8125 multibeam echosounder. Measurements were conducted within seven working days (August 28th, September 26th, October 15th, 17th, 23rd, and 25th, and November 5th, 2013) in the morning in clear to cloudy weather with precipitation with winds from 0.0m/s (smooth sea level) up to 5.9m/s (small waves, peaks already breaking; Internet 1; Internet 2). The measurements provided a georeferenced point cloud, which was manually examined in order to avoid incorrect data that occasionally arise due to disturbances in measurements. From processed and systematically organized data, a bathymetric model with a reso­lution of 0.5×0.5m was created, which served as a basis for mapping seagrass meadows. Based on this mapping, a spatial seabed slope analysis was made. Seagrass meadows are higher than the seabed and it was Legend depths (m) 0–1 1–2 2–3 3–4 4–5 5–6 6–7 0 100 200 m Scale: Content by: Mojca Poklar Map by: Mojca Poklar Source: DOF 2014, TK 50 © 2013, Harpha Sea, Koper ´ Figure 1: Depths of the Semedela Bay research area (Source: Podatki snemanja morskega dna z večsnopnim sonarjem 2013). expectedthattheslopesatthetransitionsbetweensiltandmeadowwouldbequitehigh.Theresultinglayer was examined in detail. Because the area of seagrass meadow occurrence was previously recognized from an orthophoto (Digitalni ortofoto 2012), it was known in advance where they could be expected. In these areas, an attempt was made to identify key patterns or edges of seagrass meadows. A vector layer of sea-grass meadows was acquired from the seabed slope raster by exporting all contours of slopes greater than 40°, which was completed and verified with raw sonar data at the end. 2.4 Mapping seagrass meadows using the aerial photography method Aerial photography was used to obtain aerial photos, which were used to digitize seagrass meadows. This wascarriedoutwithaprofessionalcamerawithautomatictriggeringintermsofaircraftheightandveloc­ity, providing 60% overlap of the photos in the forward direction of the flight. The exact location of the aerial photos was ensured by monitoring the position and orientation of the camera on the aircraft using a GNSS receiver and a gyroscope. Aerial photography was carried out in one working day (September 6th, 2013), in the morning during clear weather. Prior to the digitization of seagrass meadows, pre-processing of aerial photos was carried out, which included geometric and lighting corrections. Aerial photos were then merged into a unique photo of the entire research area, which was orthorectified and georeferenced, and its contrast was improved. Dataonthespatialdistributionandthereforeedgesofseagrassmeadowswereobtainedthroughasuper-visedimageclassificationoftheRGBlayersoftheaerialphoto,incombinationwithitsvisualinterpretation. Intheprocessofasupervisedimageclassification,trainingsampleswerefirstcreated;theseareareaswith aknown type of seabed, on which the spectral signature of the seabed type was calculated. Training sam­ples were marked interactively using the training sample drawing tools and were determined by manual limitation. Twelve training samples were determined for various seabed types and in various situations (shadows,seagrassmeadowdensity,etc.).Fortheclassification,themaximumlikelihoodclassificationmethod was used because it is the most accurate, although it is a very demanding computing process (Oštir 2006). The quality of the classification was improvedby visual interpretation of the entire photograph, for which the edges of seagrass meadows were manually corrected by evaluating the basic elements of visual photo interpretation(tone,shape,size,pattern,texture,shadows,etc.);thisisthemostsubjectivepartofthemethod. 2.5 Verification of both methods by comparison with the line transect method and the final map of seagrass meadows Because the sonar recording and aerial photography methods are indirect remote sensing methods, after their implementation they always require ground truth observations toverify the results already obtained (Komatsu et al. 2003). They are helpful in interpreting the distinctive characteristics of seagrasses from sonar data or aerial photos, where they also serve as a reference point for verifying the interpretation of photos; for example, tocheckthatnomacroalgaeorshellsweremisidentifiedasseagrassmeadows(Krause-Jensenetal.2004). Accordingly, to verify the spatial distribution of seagrass meadows obtained by sonar data or aerial photography,andtoevaluatetheaccuracyofselectedmethods,afieldsurveywascarriedout,whichinvolved seabed recordings with an underwater camera. Underwater recordings were made directly from a vessel on predetermined line transects and meadow centroids (Figure 2). Because the line transects were plot­ted by a computer, the precise geographical position and the angles of the recordings were verified with a GNSS receiver and a gyroscope simultaneously with the recordings. The recorded videos of line tran­sects were then processed and converted into underwater photos or raster data. The raster data obtained represented the reference state for the verification of sonar and aerial pho­tographydata.Thefirstphasecomparedthemappededgesofseagrassmeadowsandmeasureddeviations fromthereferencestate.Alongfivelinetransects,forty-sixcontrolpointswererandomlyselected,onwhich the seabed type was determined (the analysis was limited to two types: silt and seagrass meadows) and then compared with sonar and aerial photography data. A comparison also included four points that rep-resentedthecentroidsoftheseagrassmeadows.Basedonthecomparison,theaccuracyofseagrassmeadows mapped with each method was assessed using a confusion matrix (Mumby and Green 2000). Based on the evaluation of the accuracy of two remote sensing methods, the polygon layer of seagrass meadowswasestablished.Wheretheedgesofseagrassmeadowsweredetectedbybothmethods,theedge with the greater accuracy was considered. Where the edges were detected by only one method, the avail­able ones were considered. When checking raster data from the field survey, it did not occur that an edge was not be detected by any method. The result was then mapped using the tools for analysis and spatial display of measured data. 3 Results The multibeam sonar data and the seabed slope analysis showed that seagrass meadows’ edges are clear­lyvisibleinmostcasesbecausetheslopeatthetransitionbetweenthesiltandthemeadowcanrangefrom 0°to80°.Largeranddenserseagrassmeadowsarewellvisible(Figure3),whereastheareaswherethemead­ows are sparse are not. Such areas are difficult to separate from the silt, and so accurate mapping requires a review of raw data (the distance between points was about 10cm) or the combination of sonar data with data from another method. The seagrass meadow edges in Semedela Bay mapped from sonar data are shown in Figure 4. Not all the edges are connected. The western edge, which lies in the eastern part of the bay, was not completely visible because of a gradual transition between seagrass and silt. In this area, the seagrass is sparse and lower in growth, making it difficult to determine its edge. The same applies to seagrass meadows around themouthoftheBadaševicaRiver.Thesonarrecordingmethodmadeitpossibletodrawtheseagrassmead­ow edges with a total length of 5,310.10m. From aerial photography data (i.e., classified aerial photos), it was determined that the edges between theseagrassmeadowsandsiltweremostlyvisible.However,therewereareaswherethephotodidnotmake Figure 2: Selected line transects and sampling points (centroids) for verifying seagrass meadow occurrence. it possible to recognize whether there is a meadow or not. This is especially true for deeper areas, where the lower edge of the meadow is often difficult to determine due to the smaller proportion of light that can penetrate to the seabed or to the meadow. Problems also occurred in areas where the meadow edges were less visible due to reflection of light from the sea surface. The problem was solved by changing the direction of the flight. For the research area, it turned out that the reflection of sunlight from the sea sur­faceislessvisibleintheaerialphotos,whichweretakenbyflyinginanorth–southdirection.Nevertheless, it was not possible to completely solve these problems, and so in the previously described areas the mead­ow edges were difficult to determine. Figure 5 shows edges of seagrass meadows mapped using the aerial photography method and with a total length of 5,727.30m. Anoverviewoftheunderwaterphotosofthefieldsurveyshowedthepresenceofbothtypesofseabed as predicted with the sonar recording method and aerial photography method (silt and seagrass mead­ows). In addition, other species were also found in underwater photos; specifically, various macroalgae that appeared closer to the mouth of the Badaševica River and the noble pen shell or fan mussel (Pinna nobilis), found on the outer part of the seagrass meadow along the harbor at Koper. Figure 3: Example of a seagrass meadow on a seabed slope raster. Figure 4: Seagrass meadow edges in Semedela Bay obtained using the sonar recording method. Figure 5: Seagrass meadow edges in Semedela Bay obtained using the aerial photography method. 14 Acomparisonoftheseagrassmeadowedgesinunderwaterphotoswithmeadowedgesobtainedusing thesonarrecordingmethodshoweddeviationsofupto1m,whereasatthemeadowedgesobtainedthrough aerial photography deviations of up to 3m occurred. Considering the position errors – which were esti­mated between 0.2 and 0.3m (sonar data), between 0 and 1m (aerial photography), and between 0.2 and 0.3m(linetransectdata)–deviationsoccurforvariousreasons.Inthecaseofsonardata,deviationsoccur in areas of gradual transition between seagrass and silt, whereas at the sharp edges of seagrass meadows the contours are completely coincident (centimeter-level accuracy). Major deviations in aerial photogra­phydatacanbeattributedtoerrorsingeoreferencingofaerialphotos,aswellasthepoorvisibilityofseagrass meadow edges from the aerial photo in areas of greater depth and in areas where light reflected from the sea surface during the shooting. To assess the classification accuracy of selected methods (Lillesand and Kiefer 1994), two confusion matriceswereproduced,comparingthepredicteddataofthesonarrecordingoraerialphotographymethod with ground truth (reference) data of the field survey. Figure 6: Example of verifying the sonar recording and aerial photography method by using the line transect method, recorded with an underwater camera on line transect T4. Table 1: Confusion matrix of the a) sonar recording method and b) aerial photography method. a) Reference data Seagrass Silt User accuracy Sonar recording method Seagrass Silt 30 1 3 26 90.9% 96.3% Total number of sampling points Producer accuracy 31 96.8% 29 89.7% Overall accuracy=93.3% b) Reference data Seagrass Silt User accuracy Aerial photography method Seagrass Silt 24 7 13 16 64.9% 69.6% Total number of sampling points Producer accuracy 31 77.4% 29 55.2% Overall accuracy=66.6% Figure7:SeagrassmeadowedgesinSemedelaBaybasedonthemappingmethodused,andthefinalmapofthemeadowsinSemedelaBayinautumn2013. 16 Table1showsthatsamplingpointslabelingseagrassweremorecorrectlyclassifiedthanthosethatlabeled silt. The difference is small (7.1%) for the sonar recording method, whereas for the aerial photography method it is considerably larger and amounts to 22.2%. In contrast, user accuracy, which serves as a guide to the results’ reliability as a prediction tool, shows that in both the sonar recording and aerial photogra­phy methods the silt class is more correctly classified (96.3% by sonar recording and 69.6% by aerial photography).Nevertheless,themostnoticeableinformationinTable1isthedifferencebetweentheover­all accuracy of the methods by which the seagrass meadows were detected. For the sonar recording the overall accuracy was 93.3%, and for the aerial photography it was 66.6%. Based on the accuracy of both methods, a spatial data layer of seagrass meadows was created using complete sonar data (because in this case it is more accurate than aerial photography data) supplement­ed with aerial photography data. Based on the mapping method, a map of seagrass meadows edges was created (Figure 7). Most of the meadow edges (52%) were plotted using the sonar recording method, and theaerialphotographymethodwasusefulfor24%oftheplottededges.Someedgesweredetectedbyboth methods, which is shown as an independent category in Figure 7 (24%). Figure 7, which also shows the spatial distribution of seagrass meadows in Semedela Bay, shows that seagrass meadows are distributed along the coast and in the inner part of the bay. There are two major seagrass meadows as well asa number of minor ones, constituting »islands,« separated from major mead­ows. Seagrass meadows were not detected at depths exceeding 5m, where light conditions do not allow the growth of seagrass, and directly along the coast, especially at the mouth of the Badaševica River. In the past, the Badaševica deposited contaminated and nutrient-rich water in the bay, which contributed to the extremely depleted vegetation at its mouth. Because both Cymodocea nodosa and Zostera marina are sensitive to elevated levels of nutrients in the water column (Lipej et al. 2006; Orfanidis et al. 2007), this could be the main reason for the lower coverage of the seabed with seagrass in the area. Lower coverage of the seabed with seagrass directly along the Semedela promenade in the eastern part of the bay can be attributed to the renovation of the promenade in 2010. The renovation works also consisted of deepen­ing the seabed, which led to physical damage to the seabed and associated vegetation. 4 Discussion Measurements of the spatial distribution of seagrass meadows with the methods presented for Semedela Bayprovidedsomekeyfindingsregardingtheircharacteristics.Thefirstrelatestothetimeframeformak­ingthemeasurements.Theaerialphotographymethodisfasterincomparisontothesonarrecordingmethod because photographing the entire research area was carried out in one day, whereas sonar measurements lasted several days. Another characteristic investigated was the spatial and temporal dependence of the method. It was determined that the sonar recording method, in contrast to the aerial photography method, is a spatial-lyandtime-independentmethodbecausethedatacapturewithamultibeamsonarisindependentofwater transparencyandsunlight,andwithaccurateGNSSandINSreceiversitispossibletoperformqualitymea­surements in the undulating sea. In contrast, the use of the aerial photography method in the Slovenian seais limiteddue tohigh water turbidity. It turned out that the greatest problems arise inthebays (the sea currents are not so strong, the influence of waves is greater than in the open sea, and siltation is promi­nent), where the largest share of seagrass meadows is located. In addition to increased water turbidity, the problem also lies in the refraction and reflection of light on sea surface, which makes it necessary to cap­ture photos at the best time of the day and under the best environmental conditions. The greatest weakness of the aerial photography method is certainly its subjectivity in determining thedistributionofseagrassmeadows.Inordertodetermineseagrassmeadowsfromaerialphotos,animage classification was made, partially also with a manual capture of the edges of seagrass meadows, where it was necessary tovisually evaluate the basic elements of photointerpretation (such as tone, color, contrast, texture, shadows, etc.), which each individual can recognize differently. Because the multibeam sonar spreads beams at ± 60° steering angles (Fridl, Kolega and Žerjal 2008), the method is useful for flat and for more morphologically diverse seabeds, and in addition it is also pos­sibletomeasuretheheightofseagrassanditsbiomassabovetheseabed.However,ifoneisonlyinterested in information on the occurrence of a meadow in a certain area or if the required precision of the mapped meadows is low, the aerial photography method is more appropriate from a user perspective. This is espe­ciallytruewhenanalyzingalreadyexistingaerialorsatelliteimages(inthiscase,onemusttakeintoaccount thelowerresolutionandthusthelowerqualityofsuchimages),which,incontrasttothemethodsdescribed, are more accessible. Inadditiontothesecharacteristicsofbothmethods,theiraccuracywasofprimaryinterest.Considering positionerrors–whichwereestimatedbetween0.2and0.3m(sonardata),0and1m(aerialphotography), and0.2and0.3m(linetransectdata)–theoverallaccuracyofthesonardatawas93.3%,whereastheover-all accuracy of the aerial photos was only 63.3%. Considering that accuracy of classification over 90% is good,andthatover80%issatisfactory(Oštir2006),thesonarrecordingmethodwasgoodinthisresearch case, whereas the aerial photography method did not yield the most accurate results. In this study, sonar recording is a more reliable method of data acquisition on the spatial distribution of seagrass meadows. Ofcourse, this does not apply to less turbid waters and thus to more accurate visibility of seagrass mead-ows,wheretheaerialphotographymethodcanachievethesameaccuracyasthesonarrecordingmethod. In the case at hand, this was noticeable in determining the edges of seagrass meadows in the area from the mouth of the Badaševica River to the inner part of the bay. In that area, seagrass meadows were poor­lyvisibleonsonardataduetotheaforementionedgradualtransitionbetweensparseseagrassandsilt,whereas they were clearly visible in aerial photos due to shallow water (0 to 2m in depth) and thus increased light penetration through the water column to the seabed. 5 Conclusion The purpose of this study was to show that the use of modern remote sensing technologies and GIS tech­niques makes it possible to obtain new and more useful results in mapping seagrass meadows, yielding betterknowledgeoftheecologicalstatusoftheseaandabetterunderstandingoftheprocessesinitscoastal area.Comparisonofthesonarrecordingandaerialphotographymethodswiththealreadyestablishedline transect method for determining the spatial distribution of seagrass meadows in part of the Slovenian sea showed that both methods allow efficient mapping of seagrass meadows, but they differ significantly in certain characteristics. The sonar recording method proved to be more accurate, more objective, and, in contrasttotheaerialphotographymethod,spatiallyandtemporallyindependent,whichisaconsequence ofthehigherwaterturbiditytypicaloftheSloveniansea.Intermsofaffordability,bothmethodsareexpen­sive because high-precision data require high-priced equipment. Iftherequiredprecisionofthemappedseagrassmeadowsisloworoneisonlyinterestedintheoccur­renceofseagrassmeadowsinacertainarea,theaerialphotographymethodismoreappropriatefromauser perspective because already existing aerial or satellite images are much easier to access. However,thechoiceofmethodologyprimarilydependsonthepurposeofresearch/mapping,andthen ontheenvironmentalconditions,wherethewatertransparencyorturbidity,bathymetryandmorphology ofthebay,weather conditions,andavailableresources mustbetaken intoaccount. Itshouldalso benoted thattheremotesensingmethodsusedalonearenotenough.Regardlessofthechoiceofthemappingmethod, after applying it ground truth observations are always required in order to verify the results obtained and to explain the characteristics of seagrass recognized by remote sensing images or sonar measurements. 6 References Adanza, E. G. 1995: Research methods, principles and applications. Manila. BerdenZrimec,M.,Poklar,M.,Moškon,S.2015:Underwaterhabitatmappingusingremotesensingtech­niques. Proceedings GEOSUB – Underwater geology. Trieste. Berden Zrimec, M., Poklar, M., Mozetič, D., Vranac, D. 2015: Seabed habitat mapping with underwater photogrammetry. Abstract book, 52nd European Biology Symposium. Piran. Comas Gonzales, R. 2015: Mapping seagrass meadows, using low altitude aerial images. Ph.D. thesis, Universidade de Lisboa, Faculdade de Ciencias. Lisboa. Internet: http://repositorio.ul.pt/bitstream/ 10451/23040/1/ulfc117350_tm_Robert_Comas_Gonzalez.pdf. Digitalniortofotoposnetki,območjeMestneobčineKoper.LjubljanaGeodetskaupravaRepublikeSlovenije. Ljubljana, 2012. Digitalni ortofoto posnetki, območje Mestne občine Koper. Geodetska uprava Republike Slovenije. Ljubljana, 2014. Duarte,C.M.2002:Thefutureofseagrassmeadows.EnvironmentalConservation29-2.DOI:https://doi.org/ 10.1017/S0376892902000127Fridl, J., Kolega, N., Žerjal, A. 2008: Pomen digitalnega batimetričnega modela za trajnostni razvoj morja. Geodetski vestnik 52-4. Hidrografija 2014: Internet: http://www.hidrografija.si/p1/4-1-5.php (26.11.2014). Hossain, M. S., Bujang, J. S., Zakaria, M. H., Hashim, M. 2014: The application of remote sensing to sea-grassecosystems:anoverviewandfutureresearchprospects.InternationalJournalofRemoteSensing 36-1. DOI: https://doi.org/10.1080/01431161.2014.990649 Komatsu, T.,Mikami, A.,Sultana, S., Ishida,K., Hiraishi,T., Tatsukawa,K. 2003:Hydro-acoustic methods as a practical tool for cartography of seagrass beds. Otsuchi marine science 28. Krause-Jensen, D., Queresma, A. L., Cunha, A. H., Greve, T. M. 2004: European seagrasses: an introduc­tion to monitoring and management. Internet: http://www.seagrasses.org/handbook/european_ seagrasses_high.pdf (5.7.2017). Lillesand, T. M., Kiefer, R. W. 1994: Remote sensing and image interpretation. New York. Lipej, L., Dobrajc, Ž., Forte, J. Mavrič, B., Orlando Bonaca, M., Šiško, M. 2007: Kartiranje habitatnih tipov inpopisvrstnamorskihzavarovanihobmočjihNSDebelirtič,NRStrunjaninNSRtMadona.Nacionalni inštitut za biologijo, Morska biološka postaja. Piran. Lipej, L., Turk, R., Makovec, T. 2006: Ogrožene vrste in habitatni tipi v slovenskem morju. Ljubljana. Malačič, V. 1994: Razvojni projekt občine Koper 2020: varstvo morja in priobalnega pasu. Raziskovalno poročilo. Nacionalni inštitut za biologijo, Morska biološka postaja. Piran. McKenzie, L. 2003: Guidelines for the rapid assessment and mapping of tropical seagrass habitats. Cairns. McMahon, K., Collier, C. J., Lavery, P. S. 2013: Identifying robust bioindicators of light stress in seagrasses: A meta-analysis. Ecological Indicators 30. DOI: https://doi.org/10.1016/j.ecolind.2013.01.030Moškon, S., Žibert, J., Kavšek, B. 2015: Kartiranje morskih travnikov s podatkih mnogosnopnega sonarja. Geografski vestnik 87-1. DOI: https://doi.org/10.3986/GV87106 Mumby, P. J., Green, E. P. 2000: Field survey: building the link between image and reality. Remote Sensing Handbook for Tropical Coastal Management. Paris. OpazovaniinmerjenimeteorološkipodatkipoSloveniji2014.Internet:http://meteo.arso.gov.si/met/sl/archive (18.11.2014). Orfanidis, S., Papathanasiou V., Gounaris, S. 2007: Body size descriptor of Cymodocea nodosa indicates anthropogenicstressincoastalecosystem.TransitionalWatersBulletin2.DOI:https://doi.org/10.1285/i1825229Xv1n2p1 OroženAdamič,M.2002:GeomorfološkeznačilnostiTržaškegazalivainobrobja.Geografijainnjeneaplika­tivne možnosti. Ljubljana. Oštir, K. 2006: Daljinsko zaznavanje. Ljubljana. Peterlin, M. 2013: Načrt upravljanja morskega okolja. Začetna presoja morskih voda v pristojnosti Republike Slovenije. Bistvene lastnosti in značilnosti morskih voda. Ljubljana. Podatki snemanja morskega dna z večsnopnim sonarjem, ožji morski pas. Harpha sea d.o.o. Koper, 2013. Poklar,M.2016:Metodologijageografskegaraziskovanjastikasladkovodnegainmorskegaokoljanaprimeru Semedelskegazaliva.Ph.D.thesis,FakultetazahumanističneštudijeUniverzenaPrimorskem.Koper. Prostorsko-podatkovni sloj meje obale morja. Agencija Republike Slovenije za okolje. Ljubljana, 2010. Ralph, P. J., Durako, M. J., Enriquez, S., Collier, C. J., Doblin, M. A. 2007: Impact of light limitation on seagrasses.JournalofExperimentalMarineBiologyandEcology350,1-2.DOI:https://doi.org/10.1016/ j.jembe.2007.06.017 Robbins, B. D. 1997: Quantifying temporal change in seagrass areal coverage: The use of GIS and low res­olutionaerialphotography.AquaticBotany58-3-4.DOI:https://doi.org/10.1016/S0304-3770(97)00039-9 Telesca, L., Belluscio, A., Criscoli, A., Ardizzone, G., Apostolaki, E. T., Fraschetti, S., Gristina, M., Knittweis, L., Martin, C. S., Pergent. G., Alagna, A., Badalamenti, F., Garofalo, G., Gerakaris, V., Pace, M. L., Pergent-Martini, C., Salomidi, M. 2015: Seagrass meadows (Posidonia oceanica) distribution and trajectories of change. Scientific Reports 5. DOI: https://doi.org/10.1038/srep12505 Turk, R., Orlando Bonaca, M., Makovec, T., Vukovič, A., Lipej, L. 2002: A topographical survey of habitat types in the area characterized by seagrass meadow of Posidonia oceanica in the southern part of the Gulf of Trieste (Northern Adriatic). Annales, Series historia naturalis 12-2. Unsworth, R. K. F., Keulen, M., Coles, R. G. 2014: Seagrass meadows in a globally changing environment. Marine Pollution Bulletin 83-2. DOI: https://doi.org/10.1016/j.marpolbul.2014.02.026 Vacchi, M., Montefalcone, M., Schiaf.no, C. F., Parravicini, V., Bianchi, C. N., Morri, C., Ferrari,M. 2014: TowardsapredictivemodeltoassessthenaturalpositionofthePosidoniaoceanicaseagrassmeadows upper limit. Marine Pollution Bulletin 83-2. DOI: https://doi.org/10.1016/j.marpolbul.2013.09.038 Virnstein, R., Avery, W. M., Johansson, R. O. 2000: On defining the »edge« of a seagrass bed. Reports 145. Waycott, M., Duarte, C. M., Carruthers, T. J. B., Orth, R. J., Dennison, W. C., Olyarnik, S., Calladine, A., Fourqurean, J. W., Heck, K. L., Hughes, A. R., Kendrick, G. A., Kenworthy, W. J., Short, F. T., Williams, S. L. 2009: Accelerating loss of seagrasses across the globe threatens coastal ecosystem Proceedings of the National Academy of Sciences 106-30. DOI: https://doi.org/10.1073/pnas.0905620106 Zorn,M. 2009: ErosionprocessesinSlovene Istria – part1:Soilerosion. ActageographicaSlovenica 49-1. DOI: https://doi.org/10.3986/AGS49102