Marine-turtle survey at Akrotiri-Episkopi,

Marine-turtle survey at Akrotiri-Episkopi, 2010-11 Version 1.3 January 2012 Pantelis Charilaou Alexia Perdiou SBAA Environment Department, Episkopi

2 Acknowledgments The survey was carried out with contribution, assistance and advice at various levels from the following: SBA Police SBA Customs Wing Commander Paul Jones Cyprus Wildlife Society Cyprus Fisheries and Marine Research Department Cyprus Environment Department Episkopi Turtlewatch Akrotiri Turtlewatch Medasset JNCC Defence Estates Akrotiri Community Council Episkopi Community Council Akrotiri Environmental Education and Information Centre

3 Executive summary Numbers of dead turtles washing up on SBA beaches at Episkopi-Akrotiri cause a conservation concern for the two priority species of Loggerhead and Green turtle. These numbers are disproportionately higher than the nesting interest of the two species at Episkopi-Akrotiri (hosting less than 5% of the nesting interest on island). Aerial and boat surveys and fishermen interviews have been carried out to complement death and nesting data, with a view to confirm the turtle interest and the connection of deaths with fishing. The results estimate a population of several hundred turtles within the study area and suggest more interest to the east, west and further away from the shore than the study area. The composition of this population is estimated to be similar to that of the dead turtles outlined below - with a juvenile majority - but more information is required for confirmation purposes. It is suggested that the interest is lower during the period November-January than for the rest of the year. Most of the interest was in waters deeper than 10 metres, but more information would be useful in terms of the mating and nesting movements. More turtle presence occurred in some areas than others, with more observations in the north-western part of Episkopi Bay and less south of Akrotiri Peninsula. Death data have systematically been recorded only during the last three years with an average of more than 50 dead turtles washing up at Episkopi-Akrotiri per year. The composition of the deaths has been consistent over the last three years, with percentages more than: 50% Green juveniles; 30% Loggerhead adult; 10% Loggerhead juveniles; and 5% Green adult. The majority of deaths are due to fishing by-catch in set nets. An estimated 60-70% of turtles caught in nets drown and a significant proportion of the remaining 30-40%, which are retrieved alive are deliberately killed. The conflict between turtles and fishing covers areas from Cape Aspro to Lady s Mile and is likely to be higher than the dead turtles found at Episkopi-Akrotiri, but, equally, some of the dead turtles washing up at Episkopi-Akrotiri may be coming from other more distant areas. The majority of turtles are likely to be caught in water depths bigger than 10 meters, where most of the turtle interest occurs. Also, the nesting activity does not seem to increase the death rate of nesting females during the nesting compared to the nonnesting season. Most of the deaths for all ages of both species occur during April- September in comparison with October March, the difference appearing to be due to a combination of reduced fishing intensity and turtle presence during the winter. In order to assess whether the death numbers found at Episkopi-Akrotiri are significant, more information is required to put these numbers in context to the population affected and the death rates and concerns in the wider region. This should conclude which areas are hotspots of turtle interest and fishing conflict, with a view to inform management decisions. The recommended increase of minimum fishing depth is not likely to decrease the death rate. More research is recommended, focused on wider aerial surveys, mapping of marine habitats and possibly tagging/satellite tracking, where appropriate. Once hotspot areas of turtle presence and fishing conflict have been confirmed, appropriate measures should be considered to address mortality concerns, including fishing regulation, designations and eco-tourism. In the meantime, enforcement of fishing regulations, awareness raising and promotion of sustainable fishing should be a priority.

4 1. INTRODUCTION 1.1 BACKGROUND The need for the survey originated from concerns over the large number of dead/injured marine turtles found on SBA beaches in Akrotiri/Episkopi bay during 2008-2009. As explained in the proposal attached as Annex 1, the Sovereign Base Areas Administration (SBAA) recognised the increased death figures but was unable to draw reliable conclusions due to a lack of systematic information on the turtle interest in the area and its conflict with fishing and other activities. The results of the survey could also inform the management process for the marine turtle interest in the area. 1.1.1 STATUS OF MARINE TURTLES IN CYPRUS The Loggerhead (Caretta caretta) and the Green turtle (Chelonia mydas) are protected as priority species under the Protection and Management of Nature and Wildlife Law of the Republic of Cyprus which transposes Council Directive 92/43/EC on the conservation of natural habitats and wild fauna and flora. The two species are also protected by the Bern Convention on the protection of European wildlife and natural habitats and by the Bonn Convention on the conservation of migratory species of wild animals. The nesting habitat at the west coast (Lara-Toxeftra) was declared as a turtle reserve under the Fisheries Law. It includes a coastline of 10 Km and the sea area up to the 20 metre isobath. An 11 Km coastline and the adjacent sea area up to the 50 metre isobath at Chrysochou Bay was proposed as a Natura 2000 site, which in 2008 was accepted by the European Commission as a Site of Community Interest. The average number of turtle nests at Chrysochou Bay and West Coast during the last decades has been around 300 Loggerhead and 50 Green and during the last few years around 600 and 50 respectively. On Morphou Bay, Famagusta Bay and Kyrenia Coast the average nests during 1993-2007 were 236 Loggerhead and 104 Green. Marine turtles in the SBAs are afforded the same level of protection as in the Republic of Cyprus. The SBA authorities have enacted the Protection and Management of Nature and Wildlife Ordinance in 2007 which broadly mirrors the Republican Protection and Management of Nature and Wildlife Law. The European LIFE project on the NATURA 2000 Network undertaken by the Republic of Cyprus identified the turtle nesting interest at Akrotiri. The current intention of the SBA authorities is to designate qualifying nesting sites as Special Areas of Conservation under the Protection and Management of Nature and Wildlife Ordinance. These sites are also within the scope of the Environmental Management Plan underway for Akrotiri Peninsula. The candidate sites in question enjoy the same protection as if they were formally designated in accordance with the SBAA Administrative Secretary s 2007 Policy statement on designated and candidate SPAs and SACs. The Bern Convention was extended by the UK to the SBAs in 2001 and the Bonn convention in 1985.

5 The Western SBAs (Akrotiri Episkopi Paramali - Evdhimou) have been supporting an average of 22 nests per season of Loggerhead turtles and 4 of Green, since systematic recording started in 1994. 1.1.2 DEAD TURTLES AT AKROTIRI-EPISKOPI Records of dead turtles washing up at beaches at Akrotiri and Episkopi since 2003 are as follows: Year Dead turtles 2011 up to early November 46 2010 61 2009 50 2008 32 2007 16 2006 6 2005 2 2004 4 2003 10 Due to the high number of deaths recorded in 2008, the SBAA decided to undertake a post-mortem project in an effort to confirm the cause of death and the level of decomposition of turtles. In 2009 autopsies were carried out by a veterinarian on 14 turtles. These were not chosen on a random basis but on a presumption that they were more fresh than others. The vet s results include probable cause of death for ten turtles, entanglement in fishing nets and four killed deliberately. Three turtles were deemed to be fresh, two dead for 2-3 days, two dead for 3-4 days, one as having started to decompose and five to be decomposed. In 2009 the number of dead turtles increased to fifty, which raised concerns. In response, the SBAA prepared an action plan in 2010 (attached at Annex 3) in an effort to address the problem. This survey was part of this action plan. At the end of the 2010 turtle-nesting season (September 2010) and while the SBAA survey was in progress a complaint by Episkopi Turtlewatch, Medasset and Terra Cypria was launched against the SBAA for inadequate turtle protection measures at sea. The complainants claimed that the root of the problem was the depth permitted by the SBAA Fisheries Regulations for the casting of fishing nets, currently at 5 metres between June and September. They suggested that a change to the limit from 5 to 10 metres (which was the limit permitted before 2005) would solve the problem. The SBAA considered that the problem was more multifaceted and not confined within SBA waters. 1.2 OBJECTIVES 1.2.1 Confirm the marine turtle interest within Akrotiri-Episkopi Bay. 1.2.2 Investigate the connection of dead turtles washing up on Akrotiri-Episkopi beaches with fishing activities in the area.

6 1.2.3 Provide background for future association of turtle presence with marine habitats in the area. 2. METHODS 2.1. SURVEY METHODS Originally, three types of surveys were proposed: aerial; boat; and fishermen interviews. Stakeholder consultation produced useful comments and advice on the proposed surveys by Turtlewatch Episkopi, Medasset, Defence Estates and Cyprus Wildlife Society. 2.1.1. Aerial survey The aerial survey was based on observations of turtles from small aircraft based at RAF Station Akrotiri. The aircraft followed zigzag transects along the coast between Cape Gata and the SBA boundary at Evdhimou, extending 4-5 Km from the shore up to the 50-100 metre isobaths, to cover the fishing areas. A typical example of a transect is shown in map 1 (the relatively more straight part represents the return trip). Map 1: GPS track of survey 6-11-10 The transect spacing was designed to avoid double/multiple counts, taking account of aircraft and turtle speeds. Under agreement with the Republic of Cyprus, two surveys were extended approximately 4 Km beyond the SBA boundary to Cape Aspro in order to confirm information from fishermen and others for turtle interest in that area. The aircraft flied at a mean altitude of about 170 metres, which was the minimum for safety reasons, and an average airspeed of 150 km/hour. On every flight there was at least one trained observer who was recording turtle observations using a GPS. Observations were also

7 noted in a simple table in relation to the distance of each turtle from the plane in three categories: near represented approximately the range between 50 and 100 meters; medium between 100 and 200; and far between 200 and 250. All observations were carried out from the same window next to the pilot on the starboard side. Each survey sheet included general notes on the conditions of wind, sea state and visibility. Most of the aerial surveys were carried out in the morning hours, as this period of the day provided the best observation conditions, avoiding the deteriorating sea state in the afternoon. However, some surveys were inevitably carried out in the afternoon due to the availability of aircraft and pilot. The pilot for the surveys was Wing Commander Paul Jones and the observers, Alexia Perdiou and Pantelis Charilaou. Some flights carried other persons, besides the pilot and observer(s), who attempted to take photographs and obtain more details using binoculars. However, small space in the aircraft, the aircraft screen and the speed of the plane restricted such possibilities. Photo 1 shows the best photos taken during the survey, with quite obvious quality limitations. Extra persons also had a chance to collect information not used in the survey such as for fishing boats, pollution etc. The spotting efficiency of the observer was estimated by a second observer sitting at the rear, starboard window, in three of the surveys. Some turtle sightings other than the observer s, as well as observations during the return trip were recorded. However, these were not systematic and have, therefore, been excluded from analysis. Photo 1: Photos taken during aerial survey 30-6-10 by Thomas Hadjikyriakou 2.1.2. Boat survey The original plan for the boat surveys was to follow transects across isobaths between 5-20 meters at 10 knots, along the Episkopi-Akrotiri coastline with two observers, one on either side of the boat with camera and binoculars who would collect data about the presence of turtles, habitat and behaviour, weather and environmental conditions and fishing/sporting activities. However, it became evident after the first attempt that the value of such surveys was limited, compared to the aerial survey due to: a. the extensive time and effort needed in the boat survey; b. the influence of the boat (noise, visual etc) on the presence of turtles; and c. the limited information obtained from turtle observations, as these mainly included surfacing activity with little opportunity to observe details.

8 It was, therefore, decided to replace the boat surveys with snorkelling/diving ones. The available resources and the higher efficiency of snorkelling compared to diving led to a decision to concentrate on snorkelling for this survey. The snorkelling surveys were based on using a boat to reach hotspots of turtle presence as indicated by the first aerial surveys and initial information from other sources. The boat was anchored at each such location and one or more observers snorkelled away from the boat in different directions for around 30 minutes. The boat and observers stayed longer in one location when the turtle interest was high. Observers tried to keep a horizontal position in the water and avoid diving/vertical positions as per expert advice. The location of the anchored boat was registered via GPS. There was no systematic approach to the selection of the locations, other than an effort to cover different hotspot areas on a trial basis to gain as much information as possible relating to the above objectives. After the first surveys it became apparent that underwater cameras were necessary for photo-identification purposes. Therefore an underwater camera was procured and used. After coming back to the boat from each location, surveyors took notes on the numbers, species, habitat, behaviour and location in the water column of the observations. Photographs and videos were downloaded from the camera for species and numbers confirmation. The boat surveys were also used as an opportunity to gain a better understanding of the fishing and sporting activities at sea within the bay. Boats were approached for quick discussions with fishermen and others on turtle conservation issues. Such discussions focused on numbers, dates, locations, and other circumstances of turtle sightings. The boat and crew used for the surveys was mainly provided by SBA Marine Police and Customs, whereas private boats were used for two of the surveys. The main observer for the boat surveys was Pantelis Charilaou of the SBAA Environment Department who is a qualified scuba and free diver. 2.1.3. Fishermen interviews The interviews were based on the questionnaire proposed at Annex 1. However, in order to minimise fishermen reservations, the interviews took a form of friendly discussion rather than a structured approach. The interviews were carried out at the fishing ports of Akrotiri and Zapalo and notes were taken at a later stage by the interviewer. In addition to these interviews much effort was spent on obtaining information from divers, snorkelers, amateur fishermen and others (referred to as other sources in this report) in relation to the sightings and other experiences with marine turtles within SBA waters and elsewhere in Cyprus. The interviews were carried out by Pantelis Charilaou with valuable assistance from Varnavas Michael of the SBAA Environment Department who is also an amateur fisherman from Akrotiri Village. 2.1.4. Dead turtles Data for dead turtles washing up on SBA beaches at Akrotiri and Episkopi have been collected on a systematic basis only during the last three years. Dead turtles are reported to the SBAA Environment Department and the Police who also attend dead turtle scenes. A protocol with the procedure for reporting dead turtles has been established with a standard report form (attached as Annex 2) that must be completed for each turtle. This process has been designed to facilitate the timely reporting of turtles and avoid double counting or disposing of turtles before the necessary information has been collected. Members of Episkopi and Akrotiri Turtlewatch

9 and the SBAA Environmental Wardens record deaths. SBA Police and the SBAA Environment Department consolidate the records. Each dead turtle is sprayed with red paint in order to avoid double counts. 2.2. DATA ANALYSIS 2.2.1. Aerial surveys 2.2.1.1. Only systematic data from the aerial surveys were used in the analysis: only from the main observer, along the transect within the SBAs. 2.2.1.2. Observations recorded on Garmin GPS were downloaded using Garmin Map Source software and transferred to ArcGIS. A nautical map of the area was used as a base map on GIS, where isobaths of 10, 20, 30 and 50 meters were digitised. For the purposes of the fishing depth analysis, observations close to the 10 meter isobath were corrected based on the orientation of the aircraft and the distance of each observation from the plane. 2.2.1.3. Two surveys have been excluded from analysis due to the fact that visibility was low on the particular dates. Also the first survey on 18 June 2010 was excluded due to the fact that the observations were not systematic. 2.2.1.4. The transect width used in the analysis for extrapolation purposes was 250 meters. The average transect length was 90 Km and the total area 100 square Km. Therefore, the extrapolation factor calculates to 4.4. 2.2.1.5. Observer sighting efficiency was estimated at 80%, using observations from a second observer looking out from the rear, starboard window during three of the surveys. 2.2.2. Boat surveys 2.2.2.1. Species, habitat and distance from the surface of each observation were based on the photos and videos in combination with the notes taken by the observer. Only records where the species could be confirmed and double counting eliminated were included in the analysis. Only the approximate size of turtles could be confirmed and not age. 2.2.3. Fishermen interviews 2.2.3.1. Data from the interviews were analysed by applying an inevitable degree of judgment, due to: the qualitative nature of data collected; the flexible structure of the interviews; and the sensitivities involved with information such as the number of turtles caught, the fishing grounds etc. 2.2.4. Dead turtles

10 2.2.4.1. The age estimation of dead turtles was based on size of curved carapace. More than 60 and 70 cm were deemed to be adult for Loggerhead and Green respectively. 2.2.4.2. Where appropriate, figures were extrapolated in the analysis to accommodate cases where information was missing. 2.3. CONSTRAINTS 2.3.1. Aerial survey The aerial survey was subject to the availability of plane, pilot and good weather conditions. As a result, the survey could not be planned on a regular basis and some periods such as the mating season were not covered adequately. The loops followed by the plane could not be spaced evenly. As a result, some loops were more open than others and in rare occasions overlapping and double counting could not be avoided completely. Parts of the aerial survey were subject to glare, which could have been minimised by using polarised glasses. The aerial surveys could not cover certain times such as late afternoon and night. Due to the speed of the plane, inevitably there were small time delays (1-2 seconds) between sightings and GPS recordings. No good photographs of turtles could be taken or details recorded from the plane. 2.3.2. Boat survey The boat survey was subject to the availability of a boat and crew. As a result, only a limited number of surveys were carried out. The boat survey was affected by weather conditions, especially the sea state and water turbidity which affected the snorkelling visibility. Much of the data collected could not be used in the analysis due to the low quality of the photographs taken. Without good photos, the observations were subject to double counting, as turtles often came round the observer several times. Snorkelling caused disturbance when mating turtles were approached. Recording the exact location of snorkelling observations was subject to inaccuracy. This could be overcome by having a GPS in the safety buoy of the observer, the clock of which could be synchronised with the camera clock. Diving for habitat and turtle information at depths more than 15 meters was difficult and safety of the observer was an issue. Teams of trained free-divers are required for such surveys. Observer presence affected the turtle behaviour. A variety of reactions were observed, with some turtles fleeing, some circling around the observer and rarely even attracted to the observer for a short period. Mating activity was particularly sensitive to disturbance.

11 2.3.3. Fishermen interviews The information collected in the fishermen interviews was subject to various sensitivities and a certain degree of judgement had to be applied in its analysis. Amateur fishermen were not interviewed systematically, as most of them fish on weekends and holidays. It was not possible to interview trawler owners, but according to the Fisheries Department who monitor trawlers, the possibilities for turtle by-catch are remote, due to the slow trawling speed. Even if turtles get caught the chances of drowning are low due to the early retrieval of nets. Also, due to the nature of trawling nets, the probabilities for entanglement are low. 2.3.4. Dead turtle data The information on dead turtles was collected by a number of people, some of which had little training. As a result, certain information is missing from the records, some may be inaccurate and the whole system lacks surveyor consistency. The searching effort for dead turtles is lower in the winter as there are fewer people on the beaches. Some remote and/or inaccessible beaches may be completely unchecked. 3. RESULTS 3.1. Aerial survey 3.1.1. Graph 1 below shows the sightings of all surveys. These include only systematic observations used in the analysis, ie. along the transect within the SBAs from the main observer and excluding surveys with limited visibility.

12 80 70 30/6/2010 64 AERIAL TURTLE OBSERVATIONS 18/8/2011 75 NUMBER OF TURTLES 60 50 40 30 29/8/2010 45 5/9/2010 55 10/10/2010 34 26/9/2010 31 11/5/2011 50 4/9/2011 48 20 10 0 1/6/2010 16/6/2010 1/7/2010 28/7/2010 17 16/9/2010 14 9/1/2011 18 14/11/2010 22/12/2010 13 12 6/11/2010 3 16/7/2010 31/7/2010 15/8/2010 30/8/2010 14/9/2010 29/9/2010 14/10/2010 29/10/2010 13/11/2010 28/11/2010 13/12/2010 28/12/2010 12/1/2011 27/1/2011 SURVEY DATE 11/2/2011 26/2/2011 13/3/2011 28/3/2011 12/4/2011 Graph 1: Number of turtles observed in the aerial surveys 30/6/10 4/9/11 27/4/2011 12/5/2011 27/5/2011 11/6/2011 26/6/2011 11/7/2011 26/7/2011 10/8/2011 25/8/2011 9/9/2011 24/9/2011 The average number of turtles observed is 34.2, which, extrapolated over the whole area comes to 150. Applying a correction for observer efficiency, this number becomes 188. The graph also includes a trend-line, indicating a reduction of numbers during the period November 10 - January 11 in relation to the periods of June 10 October 10 and May 11 September 11. There is a statistically significant difference between the medians of the two samples at a 95% confidence level (U=2, P<0.05, Mann-Whitney U-test) The surveys on 16/9/10 and 6/11/10 were carried out in the afternoon as opposed to the rest which were carried out in the morning. However, the sampling units in this case (2, 12) are too small with overlapping observations to confirm whether the differences are statistically significant. 3.1.2. The locations of the observations for each of the 14 surveys included in the analysis are shown on maps 2-15. After applying the correction described at 2.2.1.2 above, 50 observations turned out to be in water depths of less than 10 meters and 420 in more than 10 meters. There is a statistically significant difference between the medians of the two samples at a 99% confidence level (U=15.5, P<0.01, Mann-Whitney U-test).

13 Map 2: Aerial survey 30-6-11 Map 3: Aerial survey 28-7-10

14 Map 4: Aerial survey 29-8-10 Map 5: Aerial survey 5-9-10

15 Map 6: Aerial survey 16-9-10 Map 7: Aerial survey 26-9-10

16 Map 8: Aerial survey 10-10-10 Map 9: Aerial survey 6-11-10

17 Map 10: Aerial survey 14-11-10 Map 11: Aerial survey 22-12-10

18 Map 12: Aerial survey 9-1-11 Map 13: Aerial survey 11-5-11

19 Map 14: Aerial survey 18-8-11 Map 15: Aerial survey 4-9-11 3.1.3. Map 16 shows the locations of all the 14 surveys referred to in 3.1.2. This overall picture indicates more turtle presence in some areas than others, with more observations in the north-western part of Episkopi Bay and less south of the peninsula.

20 Map 16: Aerial surveys 30-6-10 to 4-9-11 3.1.4. Maps 17-18 show the extended observations to Cape Aspro, indicating that the turtle interest continues beyond the confines of the bay. Map 17: Aerial survey 22-12-10 including extension west

21 Map 18: Aerial survey 9-1-11 including extension west 3.2. Boat survey 3.2.1. Twenty observations in total have been made where the species could be identified and double counting avoided through photo-identification. Out of these, 14 were green and 6 loggerhead. 3.2.2. Four (all Green) out of the 20 observations were at the sea floor, three (Loggerhead) were mating on the surface (one female, two males) and the remaining 13 (10 Green, 3 Loggerhead) were seen within 5 metres from the surface. 3.2.3. The habitat in 13 out of the 17 non-mating observations was established, with 2 (Green turtles) at Posidonia beds, and 11 at Cymodocea habitat. 3.3. Fishermen interviews 3.3.1. Professional fishermen based at Akrotiri fishing shelter and Zapalo mostly fish in the areas marked dark blue on map 19, mostly using nets.

22 Map 19: Fishing grounds at Episkopi-Akrotiri 3.3.2. The estimated average number of turtles caught in fishing nets per year per professional fisherman based at Akrotiri/Zapalo is 2-3. This yields a total of 30 to 45 turtles caught by these fishermen within the fishing grounds at Akrotiri- Episkopi, which extend beyond the SBA boundary up to roughly Cape Aspro. A rough estimation shows that approximately 30-40% of these turtles are retrieved alive and the remaining dead. Eight turtles were released alive from nets last year by Akrotiri fishermen. 3.3.3. More turtles are caught at relatively deeper waters than shallow. 3.3.4. The hotspot for turtle by-catch at Akrotiri is the area around Cape Zevgari. 3.3.5. There is no hotspot of by-catch at Episkopi, as this takes place everywhere within the local fishing grounds. 3.3.6. Fishermen think that turtle presence in the area has been increasing in the last few years as a result of turtle nest protection. This is in agreement with information obtained from other sources. 3.3.7. Fishermen think that amateur fishing has been continuously increasing in the last few years. 3.3.8. Some fishermen suggested that in the past some dead turtles caught in fishing nets were not thrown back to the sea, but were retrieved in order to keep the carapace as a souvenir. They indicated that this has been less common in recent years due to the fact that this is now prohibited by the law. 3.4. Dead turtles

23 3.4.1. Table 1 shows the total number of dead turtles found on SBA beaches at Episkopi and Akrotiri since 2003. Year Dead turtles 2011 up to early November 46 2010 61 2009 50 2008 32 2007 16 2006 6 2005 2 2004 4 2003 10 Table 1: Dead turtles found on SBA beaches at Episkopi and Akrotiri 2003-2011 3.4.2. Table 2 shows the number of dead turtles found on the south-westerly and easterly (wind/current) exposed coasts for 2009-2011. Year South and western Coast Eastern Coast Total 2011 (-early November) 41 5 46 2010 53 8 61 2009 45 5 50 Table 2: Dead turtles at Episkopi Akrotiri 2009-2011 by coast exposure to wind/current 3.4.3. Table 3 shows the numbers and percentages of dead turtles per species for 2009, 2010 and 2011 (excluding deaths with missing species information), along with relative percentages. Graph 2 shows the percentages per species for this period. Deaths 2009 2010 2011 Loggerhead 20 26 20 Green 29 34 24 Loggerhead % 41 % 43 % 45 % Green % 59 % 57 % 55 % Table 3: Turtle deaths per species 2009-2011

24 Percentage of dead turtles per species for 2009-2011 Percentage 80 60 40 20 Green Loggerhead 0 2009 2010 2011 Year Graph 2: Turtle death percentages per species 3.4.4. Table 4 shows the numbers and percentages of dead turtles per species and age for 2009-2011 (excluding deaths with missing information). Graph 3 shows the trend in the percentages per species and age for this period. Deaths 2009 2010 2011 Green Juvenile 27 32 20 Adult Loggerhead 17 17 11 Average Green Adult 2 2 4 Loggerhead Juvenile 3 4 8 Green Juvenile % 55 % 58 % 46 % 53 % Adult Loggerhead % 35 % 31 % 26 % 30.7 % Green Adult % 4 % 4 % 9 % 5.7 % Loggerhead Juvenile % 6 % 7 % 19 % 10.7 % Table 4: Turtle deaths per species and age 2009-2011 Percentage of deaths per species and age for 2009-2011 Percentage 80 60 40 20 0 2009 2010 2011 Year Green Juvenile Adult Loggerhead Green Adult Loggerhead Juvenile

25 Graph 3: Turtle death percentages per species/age 2009-2011 3.4.5. Graph 4 shows dead turtles at Akrotiri and Episkopi on a monthly basis for years 2009, 2010 and 2011, along with trend-lines for each year. Total deaths 2009-2011 Deaths 14 12 10 8 6 4 2 0-2 -4 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 2010 2011 Poly. (2009) Poly. (2011) Poly. (2010) Graph 4: Total turtle deaths per month 2009-2011 3.4.6. Graph 5 shows Green juvenile deaths on a monthly basis for the three last years, along with trend-lines for each year. Green Juvenile deaths AKI/EPI 2009-2011 10 8 Deaths 6 4 2 0-2 1 2 3 4 5 6 7 8 9 10 11 12 2009 2010 2011 Poly. (2009) Poly. (2010) Poly. (2011) -4 Month Graph 5: Green juvenile deaths per month 2009-2011

26 3.4.7. Graph 6 shows adult Loggerhead deaths on a monthly basis for the last three years, along with trend-lines for each year. Graphs 7 and 8 show juvenile Loggerhead and adult Green deaths respectively on a monthly basis for 2009-2011. Adult Loggerhead deathsaki/epi 2009-2011 6 5 Deaths 4 3 2 1 0-1 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 2010 2011 Poly. (2009) Poly. (2010) Poly. (2011) Graph 6: Loggerhead adult deaths per month 2009-2011 Juvenile Loggerhead Deaths AKI/EPI 2009-2011 4 Deaths 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 2010 2011 Graph 7: Juvenile loggerhead deaths per month 2009-2011

27 Adult Green Deaths AKI/EPI 2009-2011 Deaths 2,5 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 2010 2011 Graph 8: Adult green deaths per month 2009-2011 3.4.8. Table 5 lists the dead adult turtles per species Year Total Green Total Log. Missing Information Total 2009 2 17 0 19 2010 2 17 1 20 2011 4 11 0 15 Table 5: Dead adult turtles per species 2009-2011 3.4.9. Table 6 shows the dead adult turtles per species and gender for the years 2009-2011 Year Green Female Green Male Green Missing information Total Green Log Female Log Male Log Missing information Total Log 2009 0 0 2 2 11 4 2 17 2010 1 1 0 2 11 1 5 17 2011 1 3 0 4 3 2 6 11 Table 6: Dead adult turtles per species/gender 2009-2011 3.5. Information from other sources Incidental information collected during the boat survey, at the Akrotiri Environmental Education and Information Centre, as well as in other occasions, from divers, snorkelers, amateur fishermen and others who use Akrotiri-Episkopi and other marine areas in Cyprus suggest the following: 3.5.1. Sightings of marine turtles have been increasing during the last few years in various areas in Cyprus.

28 3.5.2. Sightings in the region, besides the study area include Lady s Mile and the area west of Evdhimou up to Cape Aspro. 3.6. Turtle nests The available data on the number of nests at Akrotiri and Episkopi are listed in table 7. A plus sign after a number is used where information for one of the two areas is missing. Year Logger Episkopi Green Episkopi Total Episkopi Logger Akrotiri Green Akrotiri Total Akrotiri 1994 - - - 15 5 20 20+ 1995 - - - 19 6 25 25+ 1996 - - - 20 1 21 21+ 1997 - - - 16 0 16 16+ 1998 10 2 12 - - - 12+ 1999 3 0 3 13 6 19 22 2000 - - - 11 7 18 18+ 2001 2 0 2 23 3 26 28 2002 1 0 1 23 0 23 24 2003 3 0 3 5 1 6 9 2004 6 0 6 17 2 19 25 2005 2 0 2 10 0 10 12 2006 13 0 13 24 3 27 40 2007 12 0 12 33 0 33 45 2008 4 0 4 20 5 25 29 2009 18 0 18 49 5 54 72 2010 13 2 15 40 12 52 67 2011 6 2 8 29 9 38 46 Table 7: Turtle nests per species at Akrotiri and Episkopi 1994-2011 Total Akrotiri Episkopi Graph 9 compares the Loggerhead nest numbers at Episkopi-Akrotiri with Chrysochou- West Coast between 1999 and 2011. The ratio fluctuates between 7.5 and 28, with an average of around 15. A similar comparison for green turtles would be less meaningful, due to the very small numbers of green nests at Akrotiri-Episkopi. The Loggerhead nesting at Akrotiri-Episkopi represents less than 5% of the island total in Cyprus and Green nesting an even lower percentage.

29 Ratio of number of Loggerhead nests Chrysochou-West Coast/Episkopi-Akrotiri Ratio of number of Log nests Chry-Wcoast/Epi-Akr 30 25 20 15 10 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Year (1999-2011) Graph 9: Ratio of Chrysochou-West Coast/Episkopi-Akrotiri numbers of Loggerhead nests 4. DISCUSSION 4.1. Turtle population The estimated number of turtles, within the survey area, at or near the surface, based on the average aerial observations is 188. In order to estimate the total population in the survey area, it would be necessary to have information on average times spent by turtles at or near the surface, where they could be observed. The literature suggests that the surfacing behaviour of sea turtles is quite variable by species, season, time of day and region. The number of observations presented in 3.2.2 are too small and the boat surveys not systematic enough to serve this purpose, but at least they provide an indication that a substantial part of the population would occur below depths visible in the aerial survey. A rough estimation, based on the average figure of 188 above and the extrapolation factors used in literature, would be that the population within the survey area could be several hundred turtles. During most parts of the aerial surveys turtles were observed up to the maximum distance of the aircraft from the coast. This is an indication that turtle interest most probably extends beyond the average 4.5 Km distance from the coast covered in the survey. It is also evident that the turtle interest within the bay extends beyond the survey area to the west towards Cape Aspro. Dead turtles found at Lady s Mile when prevailing winds are from the east suggest that turtle interest occurs within Limassol Bay as well. The turtle interest east and west of the survey area is also confirmed by incidental sightings of turtles from other sources. Another unknown factor for census purposes is the movement of turtles in and out of Episkopi Bay. 4.2. Temporal variation of turtle interest According to the results at 3.1.1 the turtle observations decreased in late-autumn/winter in comparison to spring/summer. Since the observation effort and efficiency remained relatively constant, this reduction could mean that some turtles moved out of the study area and/or spent on average less time at or near the surface. In order to enhance its statistical validity, the aerial study could be continued for a few years and include snorkelling and/or diving surveys preferably on the same day and time of aerial surveys.

30 Also satellite tracking, where appropriate, could provide useful information on the movements of turtles across different territories and the water column. 4.3. Spatial variation of turtle interest The results of turtle-location observations show significantly more interest beyond the 10 meter isobath than before it. This conclusion is subject to two limitations of the aerial surveys: the fact that the surveys did not cover adequately the mating season, when turtles could be using shallower waters; and the fact that the surveys did not cover late afternoon early morning hours (whilst certain fishing activity is allowed), when there is a possibility of turtle movements in shallower waters, particularly nesting females. Further surveys during the mating season and satellite tracking could help resolve these uncertainties. The death data for the last three years can only provide an indication whether adult turtles have a higher death rate during the mating/nesting season than the rest of the year. To address some of the confounding involved, it is useful to compare the difference in death rate between the mating/nesting and the rest of the year of adult turtles with the same difference for juveniles. Table 8 shows the relevant figures. Adult deaths/month Juvenile deaths/month Ratio Year April- Sept. Ratio April- Sept. - Ratio Adult ratio/juv. August March August March Ratio 2009 2.2 1.8 1.2 1.9 2.4 0.8 1.5 2010 1.8 1.4 1.3 4.8 1.7 2.8 0.5 2011 2.0 1.0 2.0 4.0 1.6 2.5 0.8 Table 8: Comparison between adult and juvenile turtles of death rate changes between the mating/nesting season and the rest of the year The ratio of adult /juvenile death rate difference between the mating/nesting season and the rest of the year shows that for 2009 the death rate difference was higher for adults, but for 2010 and 2011 it was higher for juveniles. This indicates that there is no safe conclusion that adult deaths increase more than juvenile deaths during the mating/nesting season in comparison with the rest of the year. This said, it should be recognised there may still be compounding factors involved such as higher congregation of adult turtles in the region during the mating/nesting season, by-catch sensitivity differences between adults and juveniles etc. 4.4. Species and ages of turtle interest Information on species and ages can be obtained from the dead turtle records and to a lesser degree from the boat survey. It can be seen from table 3 and graph 2 that during the last three years the numbers of green deaths have been slightly higher (57%) than the loggerhead (43%). If it is assumed that there is no death rate bias between the two species, it can be inferred that the two species have almost equal populations in the wider area, where deaths are likely to be occurring. The ratio of Green to Loggerhead observations in the boat surveys has been higher (70%/30%), but the boat observations have a weak statistical basis in general and a weak correlation with the deaths, due to the small number of boat surveys and confirmed observations, the seasonal bias of the boat surveys (May-August) and the fact that the boat surveys covered specific hotspot locations within the bay whereas the deaths were likely to be coming from a wider area. In term of ages, most deaths (53%) have been Green juvenile, followed by Loggerhead adult (30.7%), Loggerhead juvenile (10.7%) and Green adult (5.7%). Again, assuming

31 no species/age bias in the deaths, a similar representation can be assumed to apply in the turtle population in the wider area. Information can also be drawn from the nesting interest. The average number of nests at Akrotiri and Episkopi during the last 3 years has been 52 Loggerhead and 10 Green. Assuming equal number of clutches per season for the two species, the ratio of Loggerhead to Green nesting females is 5.2 to 1. In order to be able to use the nest numbers to estimate the female nesting population, it is necessary to know the average number of clutches for the particular period in the particular area. What can be safely confirmed by the turtle population numbers discussed at 4.1, the nest numbers at 3.6, and the death figures (to a certain degree only, as deaths relate to areas wider than the study area as discussed at 4.8) at 3.4.4, is that the total population within the study area is much higher than the breeding/nesting interest at Akrotiri- Episkopi. 4.5. Habitat association The limited information on habitat mapping suggests association of most of the turtle interest with Cymodocea nodosa, but much more work is required on habitat association. 4.6. Death numbers Numbers of dead turtles found in a particular area depend on three factors: number of turtle deaths in the wider sea area; collection basin properties of the area in relation to the deaths; and searching effort/efficiency for dead turtles. Systematic surveying for dead turtles along the whole coastline has only been taking place during the last three years (2009-2011), with a relatively constant trend of over 50 dead turtles per year. During this 3-year period the dead turtle finding effort was at its highest and 2008 was associated with relatively more effort than previous years, as it coincided with the promulgation and enforcement of the Protection and Management of Nature and Wildlife Ordinance. The number of dead turtles recorded in 2008 was 32 in comparison to an average of 8 between 2003 and 2007. In addition to the reduced effort, there was no central collection of data during the period 2003-2007. An example of the underestimation of figures is the result of a search in the data kept at Area Office Akrotiri on payments made for the burial of dead animals. This revealed a few records for dead turtles during 2003-2007, which appear to be missing from the dead turtle records (these numbers are not included in the data of this report due to the lack of details). Another example of the influence of searching effort is Dhekelia with two small beaches in the Eastern SBA, where no dead turtles had been reported until a proper turtle watch program was established in 2009. After this, a total of 39 deaths were reported in the last three years. Therefore, the only systematic set of dead turtle data, at least for trendanalysis purposes, is for 2009-2011. It should be noted that a number of dead turtles may also be washing up at isolated, cliffy parts of the coast - such as the southern cliffs at Akrotiri - and are never recorded. In addition, leisure and other activities on the beaches are much more intensive during the months of April October, maximising the chances of finding and reporting dead turtles. Unfound carcasses in general could be disappearing after a period as they would be subject to scavenging such as by dogs and foxes. 4.7. Death causes The results of the post-mortem study and the fishermen interviews support that the majority of turtles washing up dead on the SBA beaches at Episkopi-Akrotiri are a result of fishing by-catch in set nets, professional and amateur. Most turtles caught in nets drown because of the long time nets stay in the water. The percentage of live retrievals,

32 estimated through the fishermen surveys, is 30-40 %. Despite its statistical weakness - non randomly selected sample - it can be seen from the post mortem survey that 29% of the examination involved deliberate killing, presumably after live retrieval from nets. The data collected from dead turtles cannot be used to estimate accurately the percentage of deliberate kills, due to the weaknesses outlined in 2.3.4. Despite the weaknesses involved and the possibility of some injuries created by accidental boat strikes, the data from the post mortem survey and the dead turtles confirm that a significant proportion of live retrievals are deliberately killed. The release of 8 live turtles from nets last year at Akrotiri, appears to be a result of raised awareness and a longer term potential for improvement. 4.8. Death locations Most of the dead turtles (92%) during the last three years were found on the coastline with south and western exposure to wind/currents. Only 8% were found on the east coast of the Peninsula. This does not conclude that there are more deaths in Episkopi Bay (west bay) than in Limassol Bay (west bay), but is rather consistent with the prevailing winds in the area, which are from the west and south-west during 85-90% of the year. There is a clear conflict between fishing and turtle presence as shown in map 20 overlaying all aerial observations included in the analysis onto the fishing grounds. The fishing grounds in Limassol Bay have not been covered in the aerial survey, but the information from the fishermen surveys, information from other sources and the dead turtles on the eastern coast suggest that the conflict extends to those fishing grounds as well. Equally, the dead turtles on the eastern coast could be coming from other fishing grounds further to the east towards Limassol, or - in rarer occasions - even coming round/back from other areas when there are changes in the winds/currents. The conflict clearly extends to the west of the aerial survey area, as confirmed by the two aerial surveys extended west, the fishermen survey information and the information from other sources. Map 20: Fishing grounds and aerial turtle observations

33 The numbers suggested in the fishermen surveys show that the majority of dead turtles (30-45 turtles minus the live releases) are a result of fishing net by-catch relating to the professional fishermen based at Akrotiri and Zapalo (15 boats). These estimations should add the impact of the professional fishermen who come on a relatively regular basis from their base in Limassol (2-3 boats) and the amateur fishermen who mainly launch boats from Lady s Mile, Akrotiri fishing shelter and Evdhimou; this impact has not been estimated in this survey, but it is expected to be much lower, due to the lower level of the fishing activity involved. The fishing grounds covered by all professional and amateur fishermen are shown on map 20 and cover areas up to Cape Aspro to the west and Limassol Bay to the east. Some dead findings may relate to more distant areas towards Paphos in the west and Limassol in the east, but the numbers involved would be relatively smaller. An indication of the time a carcass stays in the water before it washes up - as carcasses are detected relatively quickly - is their decomposition state. The percentage of decomposed carcasses in the last three years is roughly 30% (15 turtles), but a small proportion may have, in rare occasions, died nearby, drifted away and turned back when winds/currents changed direction. Another indication of deaths at distant areas is the fact that decomposed dead turtles wash up on SBA beaches after storms, when fishing activities had been suspended for a few days. Dead turtles wash up together with other marine debris including carcasses of sheep, goats and cows presumably coming from animal cargo ships. In terms of the depth of water at the death locations, the results of the aerial survey suggest that the chances of deaths are less close/before the 10 metre isobath than beyond it. This is subject to the weaknesses of the aerial survey, namely the lack of adequate data during the mating season and during late afternoon early morning, which includes the nesting activity. A discussion of the possibilities for mating and nesting turtles being caught in relatively shallow waters has been included in paragraph 4.3. Further analysis can be carried out in relation to adult female deaths. Table 9 shows the adult female deaths per month for 2009-2011. The adult female deaths per month for the nesting season are 1.25, 1.5 and 0.5 compared to 0.75, 0.88 and 0.25 during the non-nesting season for 2009, 2010 and 2011 respectively. The ratios between the deaths per nesting month and non-nesting month for the three years are 1.7, 1.7 and 2 respectively. Adult female deaths May June July Aug. Jan. Feb. March April Sept. Oct. Nov. Dec. 2009 3 1 1 1 2 1 2 2010 1 2 2 1 1 1 3 2 2011 1 1 2 Table 9: Adult female deaths per month for 2009-2011 Table 10 shows the deaths other than adult females per month for 2009-2011. The other than adult females deaths per month for the nesting season are 4.25, 5.25 and 7.5 compared to 2.75, 3.125 and 1.83 during the non-nesting season for 2009, 2010 and 2011 respectively. The rations between the deaths per nesting month and non-nesting month for the three years are 1.5, 1.7 and 4.1 respectively. Deaths other than adult females

34 May June July Aug. Jan. Feb. March April Sept. Oct. Nov. Dec. 2009 1 6 5 5 1 1 2 2 9 2 4 1 2010 6 5 6 6 2 1 5 5 2 10 2011 5 11 7 7 0 0 2 0 8 1 Table 10: Deaths other than adult females per month for 2009-2011 A comparison of the ratios for adult female deaths (1.7, 1.7, 2) with the ratios of other than adult female deaths (1.5, 1.7, 4.1) shows that the monthly adult female death rate increase between the nesting and the non-nesting period has been lower in 2011 (2 compared to 4.1), equal to in 2010 (1.7) and slightly higher in 2009 (1.7 compared to 1.5) the equivalent increase for the non adult females. This shows, subject to confounding and statistical factors, that adult female deaths do not increase more than the rest of the deaths during the nesting season in comparison with the rest of the year. The information provided by fishermen appears to be generally in line with all the above, suggesting that turtles are caught in relatively deeper waters, with no hotspot at Episkopi and a hotspot around Cape Zevgari for Akrotiri. These are also generally consistent with the trends for turtle observations and the fishing grounds shown in map 20. 4.9. Death times-period There is no information on the death times of the day. Most of the total deaths for the period 2009-2011 were in the months April-September and this applies to all species and ages as shown in table 11 and charts 10-14. This may be attributed to two factors, namely the decreased fishing activities and the probable reduction in turtle interest in the area during autumn-winter. Green juvenile Green adult Loggerhead juvenile Loggerhead Adult Total Apr- Sept Oct- Mar Apr- Sept Oct- Mar Apr- Sept Oct- Mar Apr- Sept Oct- Mar Apr- Sept Oct- Mar 2009 19 8 1 1 2 1 11 6 33 16 2010 26 6 2 0 2 2 10 7 40 15 2011 20 0 3 1 6 2 9 2 39 4 Total 65 (82%) 14 (18%) 6 (75%) 2 (25%) 10 (67%) 5 (33%) 30 (67%) 15 (33%) 112 (76%) 35 (24%) Table 11: Turtle deaths per species, age and period for 2009-2011

35 Total deahs 2009-2011 24% Apr-Sept Oct-Mar 76% Graph 10: Total deaths per season 2009-2011 Green juvenile deaths 2009-2011 18% Apr-Sept Oct-Mar 82% Graph11: Green juvenile deaths per season 2009-2011

36 Green adult deaths 2009-2011 25% Apr-Sept Oct-Mar 75% Graph 12: Green adult deaths per season 2009-2011 Loggerhead juvenile deaths 2009-2011 33% Apr-Sept Oct-Mar 67% Graph 13: Loggerhead juvenile deaths per season 2009-2011

37 Loggerhead adult deaths 2009-2011 33% Apr-Sept Oct-Mar 67% Graph 14: Loggerhead adult deaths per season 2009-2011 4.10. Composition and significance of deaths Dead turtle figures during 2009-2011 are quite consistent in terms of species, with Greens slightly more than Loggerheads (57% / 43%). In relation to ages, the numbers are also fairly consistent over the 3-year period with roughly one out of two deaths being Green juvenile, one out of three Loggerhead adult, one out of twenty Green adult and one out of ten Loggerhead juvenile. As discussed at paragraph 4.4, assuming no species/age bias in deaths, these percentages would reflect the population structure of turtles in the area where the deaths are taking place. A bias in the death data could have been introduced by the fact that smaller size turtles are more difficult to spot and are affected more by scavenging. In absolute numbers of deaths, there are on average 26 Green juvenile, 15 Loggerhead adult, 3 Green adult and 5 Loggerhead juvenile per year. All above numbers do not include turtles where information was missing. In order to assess the impact of these deaths on the turtle conservation interest it is necessary to know the size and composition of the population affected by the deaths and take account of the relative importance of the species and ages of deaths. In addition, this impact should be placed in context with the wider problem of marine turtle by-catch in the whole of Cyprus and probably at a wider scale. Ideally, it would be useful if it was possible to know which of the dead turtles are nesting females, as their importance would be the highest. This information could be obtained under a tagging programme. Under the circumstances where this information is missing, it may be appropriate to assign the highest importance to the adult deaths. During the last three years there have been 8 adult Green and 45 adult Loggerhead deaths. Extrapolation to include the cases where the species information was missing (1 turtle) results to 8 adult Green and 46 adult Loggerhead. Out of these deaths, the extrapolated figures to include the ones where the gender information was missing (17 turtles), yields 3 Green female, 5 Green male, 35 Loggerhead female and 10 Loggerhead male. The extrapolation is made under an assumption that the set of missing information is not biased in relation to the gender. The averages per year are 1 Green female, 2 Green male, 12 Loggerhead female and 3 Loggerhead male. The question is whether this death rate has a significant impact on the affected population. However, such an assessment is subject to the uncertainty about the size and territory of the population affected by the deaths, as this

38 population does not appear to be restricted to the confines of Episkopi Bay but to be extending to the east in Limassol Bay, the west towards Paphos and possibly other more distant areas. Although on the face of it and in absolute numbers, the average annual death rates of 18 adult turtles - especially the 13 females, and roughly 30 juveniles appear substantial and cause concern, these numbers should be put in context with bycatch rates/numbers and populations in the wider region. At the same time it would be useful to keep in mind that IUCN has a wider concern on the impact of fishing by-catch on turtles, with 150,000 estimated by-catches and 50,000 deaths per year in the Mediterranean. Estimation and monitoring of populations can also be informed by data collected at the wider rookery level. An important issue that needs to be confirmed is the extent to which Episkopi Bay (or the wider area including Limassol Bay and areas to the west) is a hotspot for turtles in relation to the rest of Cyprus. This will form the basis for further action to address the issue. Specific recommendations are included in paragraph 5 and 6. 5. Recommendations 5.1. The suggested change of fishing depth from 5 to 10 meters during the period June- September, which covers - among other interest protected in the law - the turtle nesting season, is not likely to decrease the general mortality rate, as discussed at paragraph 4.8. The only possibility for improvement would be in respect of the mating and nesting interest; nesting is definitely and mating possibly related to shallow waters (literature for the rest of Cyprus suggests 1 Km off the shore). However, the analysis at 4.8 suggests that the increased death rate of adult females during the nesting season is not higher than the general increase during that period. The impact of fishing depths on the mating interest requires further information. 5.2. Satellite tracking on adult nesting turtles could answer questions as to their movements, migration patterns and concentration in areas which are also fishing grounds. Results can direct whether additional protection measures are necessary. 5.3. Tagging of nesting turtles would be useful in determining the extent to which they are affected by net entanglement and subsequent death. However further investigation is required as to the disturbance tagging could cause to the relatively small nesting population within the SBAs as it could require the tagging of most nesting females in order to reach any meaningful conclusions. 5.4. Aerial, nesting and death surveys, following the same methodology, could continue in order to reach conclusions with more statistical validity. Furthermore, the aerial surveys did not establish species and age. As this is important information alternative options (such as flying lower with the screen open, parasailing with a camera, helicopter flying etc) which will allow for the taking of photos for identification purposes should be investigated. 5.5. A mapping of marine habitat and its subsequent association to turtle presence and fishing activities is necessary and could assist in determining further conservation measures. At the moment such a survey is underway by the Republic of Cyprus Fisheries and Marine Research Department. This survey includes the Akrotiri SBAs but only up to Cape Gata. An SBA survey will need to be undertaken to map habitats in the rest of the bay, which can also inform the designation process for marine areas.

39 5.6. Eco tourism as an economic alternative to fishing should be promoted to the local communities. The benefits of ecotourism focusing on the turtle interest should be presented with successful examples from other regions and awareness should be raised as to the direct benefits to the communities. Possibilities for EU or other funding for such initiatives should be explored and presented to the communities. 5.7. More effort should be put into raising awareness on turtle conservation among fishermen and the local communities in general. As the bay is overfished, benefits to local fishermen from a more sustainable fishing which could be combined with ecotourism initiatives (as suggested in 5.6 above) should be presented. Such efforts should be encouraged by the SBAA, for example in the form of aiding local fishermen to meet some of their needs. 5.8. It is widely accepted that enforcement of fishing regulations in the sea is very difficult. Nevertheless, the efforts to educate fishermen on the provisions of the legislation and their obligations should continue. These should focus on explaining the importance of casting set nets only during the permitted times and the damage they can cause to turtles. Fishermen should be reminded of the penalties they will be liable to when they do not comply with the restrictions in relation to the casting of nets, but at the same time ways to promote live releases from nets should be explored as an interim measure to wider management decisions. 5.9. The aerial surveys have missed the mating season. Even though mortality of adults during this period is not higher than for the rest of the year, surveys should be carried out to confirm presence and movements of turtles during the months of April and May. The possibility of carrying out some land surveys with binoculars from cliffs at Episkopi and Akrotiri should also be investigated. This would be possible provided mating takes place at very shallow waters. Boat surveys and snorkelling should be avoided as they could disturb mating turtles. 5.10. IUCN reports on distribution, threats and conservation priorities for turtles in the Mediterranean have not included work undertaken in the SBAs so far. Results on nesting interest, deaths, conservation practices and research work done in the SBAs should be reported to the IUCN so that it is included in the wider context of turtle conservation work in Cyprus and the Mediterranean. 5.11. A wider aerial survey, in cooperation with the Fisheries and Marine Research Department, should be carried out to establish the Cyprus context of turtle interest in the sea of Episkopi-Akrotiri Bay and the level and significance of by-catch deaths. In addition to areas in the Republic of Cyprus, this should include distances further out of the average 4.5 Km covered in the current survey, as well as areas off Lady s Mile.

40 6. Proposed action plan s/n Action Time Cost (excluding staff cost) in Euros Prio rity 1 Joint SBAA/RoC aerial survey in wider April October 5,000 1 area 2012 2 Marine habitat mapping within Episkopi April October 30,000 1 Bay 3 Consideration of management options after completion of aerial survey and habitat mapping, including sustainable fishing - fishing regulation, designation, eco-tourism etc 4 Continuing awareness and enforcement efforts 5 Exploration of tagging and satellite tracking options 6 Communication to IUCN of turtle information within the SBAs 2012 December 2012-1 Continuing 2000 1 2012? 2 2012 2 References: Cesale, P. and Margaritoulis, D. (Eds.), 2010, Sea turtles in the Mediterranean; Distribution, threats and conservation priorities, Gland, Switzerland Conover, W.J., 1999, Practical nonparametric statistics. 3 rd ed. New York: John Wiley & Sons, Inc Demetropoulos, A. and Hadjichristophorou, M., 1995, Manual on Marine Turtle Conservation in the Mediterranean, Republic of Cyprus, UNEP(MAP)SPA/IUCN/CWS/Fisheries Department, MANRE (Cyprus) Dytham, C., 2011, Choosing and using statistics, a biologists guide, 3 rd ed. New York: Wiley-Blackwell Eckert, K., L., K.A. Bjorndal, F.A. Abreu-Grobois, and M. Donnelly (Editors), 1999, Research and Management Techniques for the Conservation of Sea Turtles, IUCN/SSC Marine Turtle Specialist Group Publication No. 4 Gilman, E. and Bianchi, G., 2009, Guidelines to reduce sea turtle mortality in fishing operations, Food and Agricultural Organisation of the United Nations, Rome

41 Annexes Annex 1 SBA-HQ-10-6-1 UNDERSTANDING THE SEA TURTLE ENVIRONMENT STRATEGY References: A: SBA Administration s Marine Turtle Conservation in the SBAs: Proposal for Nesting Season 2010 B: Fishermen questionnaire prepared 7 th April 2010 C: Map dated 7 th April 2010 (map base: Admiralty Map Cape Aspro to Cape Pyla) BACKGROUND Numbers of dead/injured turtles 1. The need for a survey originates from concerns over the large number of dead/injured marine turtles found on SBA beaches in Akrotiri/Episkopi bay during 2009. The number of dead/injured turtles found in 2009 was 50. In 2008, 2007, 2006, 2005, 2004 and 2003 there was 32, 16, 6, 2, 4 and 10 respectively. The SBA Administration (SBAA) has recognised the increase in deaths but has so far been unable to draw reliable conclusions about the trends due to a lack of systematic information. In February 2010 the SBAA undertook to conduct a survey of sea turtles in the SBAs in order to better understand the turtle environment. This forms part of the SBAA s wider strategy on marine turtle conservation as outlined at reference A. Species, ages and periods 2. Analysis of the 2009 data has shown that the majority of deaths were juvenile Green (which according to existing information in the Republic of Cyprus may not be associated with the nesting population) and adult Loggerhead (mostly female). In particular, out of the 50 cases 25 were juvenile Green and 17 adult Loggerhead. The same tendency appears to apply in dead turtles reported in other areas in Cyprus. 3. The majority of the dead juvenile Greens (23 out of 27) were found during the period between June and November, whereas the adult Loggerheads were spread out during the rest of the year. This data is shown in the graphs below.

42 Total deaths 2009 Deaths 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 Deaths 8 7 6 5 4 3 2 1 0 Green juv deaths Akr/Epi 2009 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009 Adult Loggerhead deaths 6 5 4 Deaths 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Month 2009

43 Cause of death/injury 4. Post mortem examinations carried out on 14 relatively fresh carcasses in 2009 has indicated that the probable cause of death in the majority of cases has not been natural. In most cases there are strong indications of injuries and/or drowning, which would link the deaths to fishing or sea sporting activities. The by-catch impact of fishing on marine turtles is confirmed in international literature and in Cyprus it could be associated to fishing nets, long lines and trawlers. The international literature 1 also proposes mitigation that can be applied on fishing activities to reduce the turtle by-catch. The sporting element could include jet-skis, pleasure boats etc. Locations of deaths/injuries 5. The locations of the deaths/injuries need to be investigated further as Episkopi Bay seems to be a catchment for a wider area, due to its topography, the sea currents and the prevailing winds. THE SURVEY 6. The survey objectives are: To confirm the turtle presence within Episkopi-Akrotiri in numbers, species, ages, gender (where possible), timings and locations; To confirm the methods, intensity, locations and timings of fishing and sporting activities and the connection they may have with the dead turtles washing up on SBA beaches. 7. The survey will comprise of the following elements, each examined in turn: Questioning local sources; An aerial survey; A boat survey; A diving survey. Local sources 8. One of the best primary sources of information is the local fishermen. The questionnaire at reference B will be used by SBAA Environment Department to seek information from local fishermen. The questions are uncomplicated in order to facilitate rapport with the fishermen and minimise reluctance and suspicion. The interviews will be carried out at the fishing shelters for the convenience of the fishermen on a casual discussion basis. The same questions will be used by SBAA Environment Department to find out information from other local sources such as the SBA Marine Police Unit, SBA Customs and Turtlewatch Groups. Boat survey 1 Guidelines to reduce sea turtle mortality in fishing operations, UN FAO, Rome, 2009. Research and Management Techniques of the Conservation of Marine Turtles, IUCN/SSC Marine Turtle specialist Group, 1999.

44 9. The boat survey will follow a methodology similar to the one carried out at Polis by the Republic of Cyprus. The boat survey will collect data on: The presence of sea turtles; Habitat and turtle behaviour; Weather and environmental conditions; Fishing and sporting activity along the transects. 10. It will follow isobath transects between 5 and 20 meters (usual fishing-net depths) as shown on the map at reference C. A boat equipped with GPS and an echo-sounding depth meter will be used. There will be two observers on either side of the boat with camera and binoculars.. 11. The boat survey will be carried out once a week between May and November (both inclusive) in order to cover the periods with most deaths. Weather restrictions (wind, currents and visibility) may only allow certain days depending on the season. Therefore, the survey days will be flexible and sometimes arranged at short notice based on the weather forecast. The days that are more convenient for the survey will more than likely be the days that are more suitable for fishing: this will help collect data about the fishing activity. The boat speed will be approximately 10 knots. 12. The survey will be carried out as early in the morning as possible to avoid the deteriorating sea state in the afternoon. Aerial survey 13. The aerial surveys will comprise two trained observers aboard an aircraft/ helicopter using GPS and binoculars. Altitudes will be 150 meters or lower and airspeed not more than 100-150 km/hr, depending on the aircraft used. Sightings will be recorded along with species, location, distance from transect, time and environmental conditions. 14. The transects shown at reference C have been spaced to avoid double/multiple counts and to take account of other practical parameters such as flying conditions. The transects are parallel to each other and perpendicular to the shoreline and the isobaths up to the 100 meter depth. 15. Timing of the survey will depend on the availability of aircraft, weather and operational restrictions. It is preliminarily considered useful to carry out the survey as early in the morning as possible to avoid the deteriorating sea state in the afternoon (which may be influencing observation ability and turtle behaviour), but other hours of the day will be tried as well. Diving surveys 16. The location and timing of the diving surveys will be decided in due course, based on the preliminary information collected in the interviews and from other sources (turtlewatch groups, sailing clubs, customs, marine police etc) as well as the initial boat and aerial surveys. Current/wind movements

45 17. It would be useful to the SBAA to find out more about the drifting behaviour of objects in the wider area and the possibility of dead turtles coming from other areas into Episkopi/Akrotiri bay. This will require co-operation with the Republic of Cyprus. Information will be obtained from the search and rescue organisations as well as CYCOFOS. If necessary, experiments with drifting beacons or other objects will also be carried out. At the same time, the dead turtle monitoring system will be standardised in order to collect systematic information, which will be useful to the surveys and the wider assessment for the turtle deaths. 20 May 2010 Pantelis Charilaou Environment and Conservation Officer HQ SBAA Reference B Survey to collect information about the possible relationship of fishing activities in the Bay of Akrotiri Episkopi within WSBA and the deaths of marine turtles and possible measures to reduce the deaths Name of fisherman, base of boat, status (professional/amateur). How many turtles have you seen in the sea, when, where? How many turtles (sizes-species) have you caught in the last year, where and how (nets, lines etc)? Which areas do you use to cast nets (locations, depths, sizes/length of nets)? How long have you been using this area for fishing? Which areas do you use for other fishing methods (e.g. long lines) locations, depths? How often do you fish within the bay and in other areas? How many other boats do you estimate use the bay, which areas do they use and where are they based? Has the fishing activity increased in the last few years? Did the turtles die, if they did how were they disposed of? How did you handle them if they survived. Do you know how to handle them? Do you know that they are protected? Do you know of any other turtles being caught by other fishermen, when, where, how were they disposed of? What can be done to reduce the number of turtle deaths. What are your views on the relationship of turtles with fishing?

46 Annex 2 DEAD TURTLE REPORT Name of person who reported dead turtle, time first seen and whether found in water or on the shore: Observer s name, organisation, telephone number: Date and time: Location (please also note on map): Species (check one) Loggerhead: Green: Sex (check one) Undetermined: Female: Male: Tag or other distinguishing mark (check one) Yes: No: Level of decomposition (check one) Fresh: Partly decomposed (dead for 2-3 days): Decomposed: Measurements: Curved Carapace Length: Curved carapace Width: Final disposal (check) Left on beach where found: painted? Yes: No: Buried: on beach: off beach: Pulled up on beach: painted? Yes: No: Left floating, not recovered: painted? Yes: No: Disposed by: Photo taken and attached to this report (check one) Yes: No: Describe wounds/abnormalities: two wounds on carapace Please e-mail or send report to:

Alexia Perdiou Environment Department HQ SBAA BFPO 53 E-mail: Cyprusnet: SBAA-EPI SBA ADM AECO, External: Alexia.Perdiou192@mod.uk 47

48 Annex 3 SBA-HQ-16-2-1 Sovereign Base Area Administration MARINE TURTLE CONSERVATION IN THE SOVEREIGN BASE AREAS PROPOSAL FOR NESTING SEASON 2010 Introduction 1. During 2009, 49 dead turtles were found on Western Sovereign Base Area (WSBA) beaches: the highest number recorded in the last five years. To address this, the SBA Administration (SBAA) has drawn up this action plan. The RoC Fisheries Department, Cyprus Wildlife Society, MEDASSET, Terra Cypria, and local Turtlewatch groups have been involved in the production of this proposal. 2. The SBAA plans to tackle this issue by: consolidating and improving SBA fishing legislation; better understanding the turtle environment; designating a Special Area for Conservation (SAC) at Akrotiri; an education and awareness campaign; promoting conservation best practice; targeted policing of the foreshore. SBA legislation 3. The SBAA has considered changing the minimum depth in which fishermen may lay their nets from 5 to 10 metres but there was insufficient evidence to warrant deviation from the legislation in place in the Republic of Cyprus. The SBAA will therefore conduct a detailed survey of the turtle population, as detailed at paragraph 8, and will reconsider this proposal should the evidence support it. 4. SBA fisheries legislation will be consolidated and simplified and updated to include greater enforcement powers and recognition of the different Fishing licences issued by the Republic. This will be completed before the start of the nesting season. 5. The revised SBA Foreshore Policy will be enforced in the summer. This will reiterate SBAA s commitment to protecting the environment and will provide enhanced protection to turtle nesting areas. For example, organised sports will not be allowed,