Environmental Complaint No.	Date of Complaint Received	Description of Environmental Complaints
COM-2017-122	1823 Integrated Call Centre received a complaint lodged by a member of the public on 30 September 2017. ET received complaint details on 3 October 2017.	Cleanliness problem at Tung Fai Road

Description of Activities	Site Area
Stockpiling	WA7
Dismantling/trimming of temporary 40mm stone platform for construction of seawall	Portion X
Construction of seawall	Portion X
Loading and unloading of filling materials	Portion X
Backfilling at Scenic Hill Tunnel (Cut & Cover Tunnel)	Portion X
Excavation for HKBCF to Airport Tunnel & construction of tunnel box structure	Portion X
Works for diversion	Airport Road
Utilities detection	Airport Road/ Airport Express Line/ East Coast Road
Establishment of site access	Airport Road/ Airport Express Line/ East Coast Road
Mined tunnel lining / box jacking transition zone rebar fixing underneath Airport Road and Airport Express Line	Airport Road and Airport Express Line
Construction of Tunnel Box Structure	Shaft 3 Extension North Shaft
Excavation and lateral support works & Construction of Tunnel Box Structure for HKBCF to Airport Tunnel West (Cut & Cover Tunnel)	Airport Road
Excavation and lateral support works & construction of tunnel box structure for HKBCF to Airport Tunnel East (Cut & Cover Tunnel)	Portion X
Sub-structure, superstructure and finishing works for Highway Operation and Maintenance Area Building	Portion X
Superstructure and Finishing works for Scenic Hill Tunnel West Portal Ventilation Building	West Portal

Environmental Monitoring	Description	Monitoring Station	Frequencies	Remarks
Air Quality	1-hr TSP	AMS 5 & AMS 6	At least 3 times every 6 days	While the highest dust impact was expected.
24-hr TSP	At least once every 6 days	--
Noise	L_eq_(30mins)_, L₁₀_(30mins)and L₉₀_(30mins)	NMS 5	At least once per week	Daytime on normal weekdays (0700-1900 hrs).
Water Quality	· Depth · Temperature · Salinity · Dissolved Oxygen (DO) · Suspended Solids (SS) · DO Saturation · Turbidity · pH	· Impact Stations: IS5, IS(Mf)6, IS7, IS8, IS(Mf)9 & IS10, · Control/Far Field Stations: CS2 & CS(Mf)5, · Sensitive Receiver Stations: SR3, SR4, SR5, SR10A & SR10B	Three times per week during mid-ebb and mid-flood tides (within ± 1.75 hour of the predicted time)	3 (1 m below water surface, mid-depth and 1 m above sea bed, except where the water depth is less than 6 m, in which case the mid-depth station may be omitted. Should the water depth be less than 3 m, only the mid-depth station will be monitored).
Dolphin	Line-transect Methods	Northeast Lantau survey area and Northwest Lantau survey area	Twice per month	--
Mudflat	Horseshoe crabs, seagrass beds, intertidal soft shore communities, sedimentation rates and water quality	San Tau and Tung Chung Bay	Once every 3 months	--

Parameter (unit)	Water Depth	Action Level	Limit Level
Dissolved Oxygen (mg/L)	Surface and Middle	5.0	4.2 except 5 for Fish Culture Zone
Bottom	4.7	3.6
Turbidity (NTU)	Depth average	27.5 or 120% of upstream control station’s turbidity at the same tide of the same day; The action level has been amended to “27.5 *and* 120% of upstream control station’s turbidity at the same tide of the same day” since 25 March 2013.	47.0 or 130% of turbidity at the upstream control station at the same tide of same day; The limit level has been amended to “47.0 *and* 130% of turbidity at the upstream control station at the same tide of same day” since 25 March 2013.
Suspended Solid (SS) (mg/L)	Depth average	23.5 or 120% of upstream control station’s SS at the same tide of the same day; The action level has been amended to “23.5 *and* 120% of upstream control station’s SS at the same tide of the same day” since 25 March 2013.	34.4 or 130% of SS at the upstream control station at the same tide of same day and 10mg/L for Water Services Department Seawater Intakes; The limit level has been amended to “34.4 *and* 130% of SS at the upstream control station at the same tide of same day and 10mg/L for Water Services Department Seawater Intakes” since 25 March 2013

3 Environmental Monitoring and Audit

3.1 Implementation of Environmental Measures

3.1.1 In response to the site audit findings, the Contractor have rectified all observations identified in environmental site inspections undertaken during the reporting period. Details of site audit findings and the corrective actions during the reporting period are presented in Appendix F.

3.1.2 A summary of the Implementation Schedule of Environmental Mitigation Measures (EMIS) is presented in Appendix E.

3.1.3 Regular marine travel route for marine vessels were implemented properly in accordance to the submitted plan and relevant records were kept properly.

3.1.4 Dolphin Watching Plan was implemented during the reporting period. No dolphins inside the silt curtain were observed. The relevant records were kept properly.

3.2 Air Quality Monitoring Results

3.2.1 The monitoring results for 1-hour TSP and 24-hour TSP are summarized in Tables 3.1 and 3.2 respectively. Detailed impact air quality monitoring results and relevant graphical plots are presented in Appendix G.

Table 3.1 Summary of 1-hour TSP Monitoring Results Obtained During the Reporting Period

Reporting Period	Monitoring Station	Average (mg/m³)	Range (mg/m³)	Action Level (mg/m³)	Limit Level (mg/m³)
September 2017	AMS5	53	5 – 302	352	500
September 2017	AMS6	40	10 – 151	360
October 2017	AMS5	36	12 – 61	352
October 2017	AMS6	41	25 – 69	360
November 2017	AMS5	47	22 – 72	352
November 2017	AMS6	59	23 – 109	360

Table 3.2 Summary of 24-hour TSP Monitoring Results Obtained During the Reporting Period

Reporting Period	Monitoring Station	Average (mg/m³)	Range (mg/m³)	Action Level (mg/m³)	Limit Level (mg/m³)
September 2017	AMS5	40	22 – 71	164	260
September 2017	AMS6	48	31 – 72	173
October 2017	AMS5	74	48 – 100	164
October 2017	AMS6	110	73 – 163	173
November 2017	AMS5	72	30 – 110	164
November 2017	AMS6	105	71 – 136	173

3.2.2 No Action Level and Limit Level exceedances of 1-hr TSP and 24-hr TSP were recorded at AMS5 and AMS6 during the reporting period.

3.2.3 Record of notification of environmental quality limit exceedances are provided in Appendix M.

3.3 Noise Monitoring Results

3.3.1 The monitoring results for construction noise are summarized in Table 3.3 and the monitoring results and relevant graphical plots for this reporting period are provided in Appendix H.

Table 3.3 Summary of Construction Noise Monitoring Results Obtained During the Reporting Period

Reporting period	Monitoring Station	Average L_{eq (30 mins)}, dB(A)*	Range of L_{eq (30 mins)}, dB(A)*	Action Level	Limit Level L_{eq (30 mins)}, dB(A)
September 2017	NMS5	58	56 – 70	When one documented complaint is received	75
October 2017		59	58 – 60
November 2017		59	58 – 59

*A correction factor of +3dB(A) from free field to facade measurement was included.

3.3.2 No Action and Limit Level exceedances for noise were recorded during daytime on normal weekdays of the reporting period.

3.3.3 Major noise sources during the noise monitoring included construction activities of the Contract and nearby traffic noise and insect noise.

3.4 Water Quality Monitoring Results

3.4.1 Impact water quality monitoring was conducted at all designated monitoring stations during the reporting period. Impact water quality monitoring results and relevant graphical plots are provided in Appendix I.

3.4.2 No Action Level and Limit Level exceedances for turbidity level and dissolved oxygen level were recorded during the reporting period. There were 17 Action Level exceedances and 3 Limit Level exceedances of suspended solids level during the reporting period. The exceedances of suspended solids level recorded during reporting period were considered to be attributed to other external factors such as sea condition, rather than the contract works. The exceedances were considered as non-contract related. Record of “Notification of Environmental Quality Limit Exceedances” is provided in Appendix M.

3.4.3 Water quality impact sources during the water quality monitoring were the construction activities of the Contract, nearby construction activities by other parties and nearby operating vessels by other parties.

3.5 Dolphin Monitoring Results

Data Analysis

3.5.1 Distribution Analysis – The line-transect survey data was integrated with the Geographic Information System (GIS) in order to visualize and interpret different spatial and temporal patterns of dolphin distribution using sighting positions. Location data of dolphin groups were plotted on map layers of Hong Kong using a desktop GIS (ArcView^©3.1) to examine their distribution patterns in details. The dataset was also stratified into different subsets to examine distribution patterns of dolphin groups with different categories of group sizes, young calves and activities.

3.5.2 Encounter rate analysis – Encounter rates of Chinese white dolphins (number of on-effort sightings per 100 km of survey effort, and total number of dolphins sighted on-effort per 100 km of survey effort) were calculated in NEL and NWL survey areas in relation to the amount of survey effort conducted during each month of monitoring survey. Dolphin encounter rates were calculated in two ways for comparisons with the HZMB baseline monitoring results as well as to AFCD long-term marine mammal monitoring results.

3.5.3 Firstly, for the comparison with the HZMB baseline monitoring results, the encounter rates were calculated using primary survey effort alone, and only data collected under Beaufort 3 or below condition would be used for encounter rate analysis. The average encounter rate of sightings (STG) and average encounter rate of dolphins (ANI) were deduced based on the encounter rates from six events during the present quarter (i.e. six sets of line-transect surveys in North Lantau), which was also compared with the one deduced from the six events during the baseline period (i.e. six sets of line-transect surveys in North Lantau).

3.5.4 Secondly, the encounter rates were calculated using both primary and secondary survey effort collected under Beaufort 3 or below condition as in AFCD long-term monitoring study. The encounter rate of sightings and dolphins were deduced by dividing the total number of on-effort sightings (STG) and total number of dolphins (ANI) by the amount of survey effort for the present quarterly period.

3.5.5 Quantitative grid analysis on habitat use – To conduct quantitative grid analysis of habitat use, positions of on-effort sightings of Chinese White Dolphins collected during the quarterly impact phase monitoring period were plotted onto 1-km² grids among NWL and NEL survey areas on GIS. Sighting densities (number of on-effort sightings per km²) and dolphin densities (total number of dolphins from on-effort sightings per km²) were then calculated for each 1 km by 1 km grid with the aid of GIS. Sighting density grids and dolphin density grids were then further normalized with the amount of survey effort conducted within each grid. The total amount of survey effort spent on each grid was calculated by examining the survey coverage on each line-transect survey to determine how many times the grid was surveyed during the study period. For example, when the survey boat traversed through a specific grid 50 times, 50 units of survey effort were counted for that grid. With the amount of survey effort calculated for each grid, the sighting density and dolphin density of each grid were then normalized (i.e. divided by the unit of survey effort).

3.5.6 The newly-derived unit for sighting density was termed SPSE, representing the number of on-effort sightings per 100 units of survey effort. In addition, the derived unit for actual dolphin density was termed DPSE, representing the number of dolphins per 100 units of survey effort. Among the 1-km² grids that were partially covered by land, the percentage of sea area was calculated using GIS tools, and their SPSE and DPSE values were adjusted accordingly. The following formulae were used to estimate SPSE and DPSE in each 1-km² grid within the study area:

SPSE = ((S / E) x 100) / SA%

DPSE = ((D / E) x 100) / SA%

where S = total number of on-effort sightings

D = total number of dolphins from on-effort sightings

E = total number of units of survey effort

SA% = percentage of sea area

3.5.7 Behavioural analysis – When dolphins were sighted during vessel surveys, their behaviour was observed. Different activities were categorized (i.e. feeding, milling/resting, traveling, socializing) and recorded on sighting datasheets. This data was then input into a separate database with sighting information, which can be used to determine the distribution of behavioural data with a desktop GIS. Distribution of sightings of dolphins engaged in different activities and behaviours would then be plotted on GIS and carefully examined to identify important areas for different activities of the dolphins.

3.5.8 Ranging pattern analysis – Location data of individual dolphins that occurred during the 3-month baseline monitoring period were obtained from the dolphin sighting database and photo-identification catalogue. To deduce home ranges for individual dolphins using the fixed kernel methods, the program Animal Movement Analyst Extension, was loaded as an extension with ArcView^© 3.1 along with another extension Spatial Analyst 2.0. Using the fixed kernel method, the program calculated kernel density estimates based on all sighting positions, and provided an active interface to display kernel density plots. The kernel estimator then calculated and displayed the overall ranging area at 95% UD level.

Summary of Survey Effort and Dolphin Sightings

3.5.9 During the period of September to November 2017, six sets of systematic line-transect vessel surveys were conducted to cover all transect lines in NWL and NEL survey areas twice per month.

3.5.10 From these surveys, a total of 802.12 km of survey effort was collected, with 96.0% of the total survey effort being conducted under favourable weather conditions (i.e. Beaufort Sea State 3 or below with good visibility). Among the two areas, 297.00 km and 505.12 km of survey effort were conducted in NEL and NWL survey areas respectively.

3.5.11 The total survey effort conducted on primary lines 578.16 km, while the effort on secondary lines was 223.96 km. Survey effort conducted on both primary and secondary lines were considered as on-effort survey data. A summary table of the survey effort is shown in Annex I of Appendix J.

3.5.12 During the six sets of monitoring surveys in September to November 2017, 13 groups of 50 Chinese White Dolphins were sighted, with the summary table of the dolphin sightings is shown in Annex II of Appendix J. All dolphin sightings were made during on-effort search, while 12 of the 13 on-effort dolphin sightings were made on primary lines. In addition, all dolphin groups were sighted in NWL, and no dolphin was sighted at all in NEL. In fact, since August 2014, only two sightings of two lone dolphins were made respectively in NEL during HKLR03 monitoring surveys.

Distribution

3.5.13 Distribution of dolphin sightings made during monitoring surveys in September to November 2017 is shown in Figure 1 of Appendix J. Almost all sightings were made at the northwest portion of the North Lantau region, mainly to the east of Lung Kwu Chau and at the mouth of Deep Bay near Black Point (Figure 1 of Appendix J). One dolphin group was also sighted at the southwestern end of NWL survey area, or near the HKLR09 alignment. As consistently recorded in the previous monitoring quarters, the dolphins were completely absent from the central and eastern portions of North Lantau waters (Figure 1 of Appendix J).

3.5.14 All dolphin sightings were located far away from the HKLR03 and HKBCF reclamation sites as well as along the alignment and Tuen Mun-Chek Lap Kok Link (TMCLKL) (Figure 1 of Appendix J). However, one sighting was made near the alignment of HKLR09 as mentioned above.

3.5.15 Sighting distribution of dolphins during the present impact phase monitoring period (September to November 2017) was drastically different from the one during the baseline monitoring period (Figure 1 of Appendix J). In the present quarter, dolphins have disappeared from the NEL region, which was in stark contrast to their frequent occurrence around the Brothers Islands, near Shum Shui Kok and in the vicinity of HKBCF reclamation site during the baseline period (Figure 1 of Appendix J). The nearly complete abandonment of NEL region by the dolphins has been consistently recorded in the past 18 quarters of HKLR03 monitoring, which has resulted in zero to extremely low dolphin encounter rates in this area.

3.5.16 In NWL survey area, dolphin occurrence was also significantly different between the baseline and impact phase periods. During the present impact monitoring period, dolphins were infrequently sighted here, and mainly at the northwestern end of the area, which was in stark contrast with their frequent occurrences throughout the area during the baseline period (Figure 1 of Appendix J).

3.5.17 Another comparison in dolphin distribution was made between the five quarterly periods of autumn months in 2013-17 (Figure 2 of Appendix J). Among the five autumn periods, dolphins were still sighted regularly in NWL waters in 2013 and 2014, but their usage there was progressively reduced in the three subsequent autumn periods, with the only occurrences mostly concentrated at the northwestern portion of the survey area (Figure 2 of Appendix J).

Encounter Rate

3.5.18 During the present three-month study period, the encounter rates of Chinese White Dolphins deduced from the survey effort and on-effort sighting data from the primary transect lines under favourable conditions (Beaufort 3 or below) for each set of the surveys in NEL and NWL are shown in Table 3.4. The average encounter rates deduced from the six sets of surveys were also compared with the ones deduced from the baseline monitoring period (September – November 2011) (Table 3.5).

3.5.19 To facilitate the comparison with the AFCD long-term monitoring results, the encounter rates were also calculated for the present quarter using both primary and secondary survey effort. The encounter rates of sightings (STG) and dolphins (ANI) in NWL were 2.5 sightings and 9.9 dolphins per 100 km of survey effort respectively, while the encounter rates of sightings (STG) and dolphins (ANI) in NEL were both nil for this quarter.

Table 3.4 Dolphin Encounter Rates (Sightings Per 100 km of Survey Effort) During Reporting Period (September to November 2017)

Survey Area	Dolphin Monitoring	Encounter rate (STG) (no. of on-effort dolphin sightings per 100 km of survey effort)	Encounter rate (ANI) (no. of dolphins from all on-effort sightings per 100 km of survey effort)
Survey Area	Dolphin Monitoring	Primary Lines Only	Primary Lines Only
Northeast Lantau	Set 1 (15 & 18 Sep 2017)	0.00	0.00
	Set 2 (22 & 29 Sep 2017)	0.00	0.00
	Set 3 (4 & 9 Oct 2017)	0.00	0.00
	Set 4 (18 & 26 Oct 2017)	0.00	0.00
	Set 5 (1 & 8 Nov 2017)	0.00	0.00
	Set 6 (17 & 24 Nov 2017)	0.00	0.00
Northwest Lantau	Set 1 (15 & 18 Sep 2017)	0.00	0.00
	Set 2 (22 & 29 Sep 2017)	3.63	16.34
	Set 3 (4 & 9 Oct 2017)	1.86	9.30
	Set 4 (18 & 26 Oct 2017)	4.89	4.89
	Set 5 (1 & 8 Nov 2017)	4.99	26.60
	Set 6 (17 & 24 Nov 2017)	3.33	5.00

Table 3.5 Comparison of average dolphin encounter rates from impact monitoring period (September to November 2017) and baseline monitoring period (September – November 2011)

Survey Area	Encounter rate (STG) (no. of on-effort dolphin sightings per 100 km of survey effort)		Encounter rate (ANI) (no. of dolphins from all on-effort sightings per 100 km of survey effort)
Survey Area	Reporting Period	Baseline Monitoring Period	Reporting Period	Baseline Monitoring Period
Northeast Lantau	0.0	6.00 ± 5.05	0.0	22.19 ± 26.81
Northwest Lantau	3.12 ± 1.91	9.85 ± 5.85	10.35 ± 9.66	44.66 ± 29.85

Notes:
1) The encounter rates deduced from the baseline monitoring period have been recalculated based only on the survey effort and on-effort sighting data made along the primary transect lines under favourable conditions.

2) ± denotes the standard deviation of the average encounter rates.

3.5.20 In NEL, the average dolphin encounter rates (both STG and ANI) in the present three-month impact monitoring period were both zero with no on-effort sighting being made, and such extremely low occurrence of dolphins in NEL have been consistently recorded in the past 18 quarters of HKLR03 monitoring (Table 3.6). This is a serious concern as the dolphin occurrence in NEL in the past few years (0.0-1.0 for ER(STG) and 0.0-3.9 for ER(ANI)) have remained exceptionally low when compared to the baseline period (Table 3.6). Dolphins have been virtually absent from NEL waters since January 2014, with only three groups of six dolphins sighted there since then despite consistent and intensive survey effort being conducted in this survey area.

Table 3.6 Comparison of Average Dolphin Encounter Rates in Northeast Lantau Survey Area from All Quarters of Impact Monitoring Period and Baseline Monitoring Period (Sep – Nov 2011)

Monitoring Period	Encounter rate (STG) (no. of on-effort dolphin sightings per 100 km of survey effort)	Encounter rate (ANI) (no. of dolphins from all on-effort sightings per 100 km of survey effort)
September-November 2011 (Baseline)	6.00 ± 5.05*	22.19 ± 26.81*
December 2012-February 2013 (Impact)	3.14 ± 3.21	6.33 ± 8.64
March-May 2013 (Impact)	0.42 ± 1.03	0.42 ± 1.03
June-August 2013 (Impact)	0.88 ± 1.36	3.91 ± 8.36
September-November 2013 (Impact)	1.01 ± 1.59*	3.77 ± 6.49*
December 2013-February 2014 (Impact)	0.45 ± 1.10	1.34 ± 3.29
March-May 2014 (Impact)	0.00	0.00
June-August 2014 (Impact)	0.42 ± 1.04	1.69 ± 4.15
September-November 2014 (Impact)	0.00*	0.00*
December 2014-February 2015 (Impact)	0.00	0.00
March-May 2015 (Impact)	0.00	0.00
June-August 2015 (Impact)	0.44 ± 1.08	0.44 ± 1.08
September-November 2015 (Impact)	0.00*	0.00*
December 2015-February 2016 (Impact)	0.00	0.00
March-May 2016 (Impact)	0.00	0.00
June-August 2016 (Impact)	0.00	0.00
September-November 2016 (Impact)	0.00*	0.00*
December 2016-February 2017 (Impact)	0.00	0.00
March-May 2017 (Impact)	0.00	0.00
June-August 2017 (Impact)	0.00	0.00
September-November 2017 (Impact)	0.00*	0.00*

Notes:
1) The encounter rates deduced from the baseline monitoring period have been recalculated based only on survey effort and on-effort sighting data made along the primary transect lines under favourable conditions.

2) ± denotes the standard deviation of the average encounter rates.

3) The encounter rates in autumnr months were in blue and marked with asterisk.

3.5.21 On the other hand, the average dolphin encounter rates (STG and ANI) in NWL during the present impact phase monitoring period (reductions of 68.3% and 76.8% respectively) were only small fractions of the ones recorded during the three-month baseline period, indicating a dramatic decline in dolphin usage of this survey area as well during the present impact phase period (Table 3.7).

3.5.22 Dolphin encounter rates in NWL during autumn 2017 was similar to the previous autumn period in 2016, but was much lower than the ones in the autumn periods of 2013, 2014 and 2015 (Table 3.7). Such temporal trend should be closely monitored in the upcoming monitoring quarters whether the dolphin occurrence would continue to increase as the construction activities of HZMB works have been mostly completed in coming months.

Table 3.7 Comparison of Average Dolphin Encounter Rates in Northwest Lantau Survey Area from All Quarters of Impact Monitoring Period and Baseline Monitoring Period (Sep – Nov 2011)

Monitoring Period	Encounter rate (STG) (no. of on-effort dolphin sightings per 100 km of survey effort)	Encounter rate (ANI) (no. of dolphins from all on-effort sightings per 100 km of survey effort)
September-November 2011 (Baseline)	9.85 ± 5.85*	44.66 ± 29.85*
December 2012-February 2013 (Impact)	8.36 ± 5.03	35.90 ± 23.10
March-May 2013 (Impact)	7.75 ± 3.96	24.23 ± 18.05
June-August 2013 (Impact)	6.56 ± 3.68	27.00 ± 18.71
September-November 2013 (Impact)	8.04 ± 1.10*	32.48 ± 26.51*
December 2013-February 2014 (Impact)	8.21 ± 2.21	32.58 ± 11.21
March-May 2014 (Impact)	6.51 ± 3.34	19.14 ± 7.19
June-August 2014 (Impact)	4.74 ± 3.84	17.52 ± 15.12
September-November 2014 (Impact)	5.10 ± 4.40*	20.52 ± 15.10*
December 2014-February 2015 (Impact)	2.91 ± 2.69	11.27 ± 15.19
March-May 2015 (Impact)	0.47 ± 0.73	2.36 ± 4.07
June-August 2015 (Impact)	2.53 ± 3.20	9.21 ± 11.57
September-November 2015 (Impact)	3.94 ± 1.57*	21.05 ± 17.19*
December 2015-February 2016 (Impact)	2.64 ± 1.52	10.98 ± 3.81
March-May 2016 (Impact)	0.98 ± 1.10	4.78 ± 6.85
June-August 2016 (Impact)	1.72 ± 2.17	7.48 ± 10.98
September-November 2016 (Impact)	2.86 ± 1.98*	10.89 ± 10.98*
December 2016-February 2017 (Impact)	3.80 ± 3.79	14.52 ± 17.21
March-May 2017 (Impact)	0.93 ± 1.03	5.25 ± 9.53
June-August 2017 (Impact)	2.20 ± 2.88	6.58 ± 8.12
September-November 2017 (Impact)	3.12 ± 1.91*	10.35 ± 9.66*

Notes:

1) The encounter rates deduced from the baseline monitoring period have been recalculated based only on survey effort and on-effort sighting data made along the primary transect lines under favourable conditions.

2) ± denotes the standard deviation of the average encounter rates.

3) The encounter rates in autumn months were in blue and marked with asterisk.

3.5.23 As discussed in Hung (2017), the dramatic decline in dolphin usage of NEL waters in the past few years (including the declines in abundance, encounter rate and habitat use in NEL, as well as shifts of individual core areas and ranges away from NEL waters) was possibly related to the HZMB construction works that were commenced since 2012. Apparently such noticeable decline has already extended to NWL waters progressively in the past few years with no sign of recovery, even though the HZMB-related construction activities have well past the peak.

3.5.24 A two-way ANOVA with repeated measures and unequal sample size was conducted to examine whether there were any significant differences in the average encounter rates between the baseline and impact monitoring periods. The two variables that were examined included the two periods (baseline and impact phases) and two locations (NEL and NWL).

3.5.25 For the comparison between the baseline period and the present quarter (20^th quarter of the impact phase being assessed), the p-values for the differences in average dolphin encounter rates of STG and ANI were 0.0057 and 0.0278 respectively. If the alpha value is set at 0.05, significant differences were detected between the baseline and present quarters in both the average dolphin encounter rates of STG and ANI.

3.5.26 For the comparison between the baseline period and the cumulative quarters in impact phase (i.e. the first 20 quarters of the impact phase being assessed), the p-values for the differences in average dolphin encounter rates of STG and ANI were 0.000000 and 0.000000 respectively. Even if the alpha value is set at 0.00001, significant differences were still detected in both the average dolphin encounter rates of STG and ANI (i.e. between the two periods and the locations).

3.5.27 As indicated in both dolphin distribution patterns and encounter rates, dolphin usage has been significantly reduced in both NEL and NWL survey areas during the present quarterly period, and such low occurrence of dolphins has also been consistently documented in previous quarters of the past few years.

3.5.28 The dramatic decline in dolphin usage of North Lantau region raises serious concern, as the timing of the decline in dolphin usage in North Lantau waters coincided well with the construction schedule of the HZMB-related projects (Hung 2017). Apparently there was no sign of recovery of dolphin usage even though almost all marine works associated with the HZMB construction have been completed.

Group Size

3.5.29 Group size of Chinese White Dolphins ranged from one to 12 individuals per group in North Lantau region during September to November 2017. The average dolphin group sizes from these three months were compared with the ones deduced from the baseline period in September to November 2011, as shown in Table 3.8.

Table 3.8 Comparison of Average Dolphin Group Sizes between Reporting Period (Sep – Nov 2017) and Baseline Monitoring Period (Sep – Nov 2011)

Survey Area	Average Dolphin Group Size
Survey Area	Reporting Period	Baseline Monitoring Period
Overall	3.85 ± 3.39 (n = 13)	3.72 ± 3.13 (n = 66)
Northeast Lantau	---	3.18 ± 2.16 (n = 17)
Northwest Lantau	3.85 ± 3.39 (n = 12)	3.92 ± 3.40 (n = 49)

Note:

1) ± denotes the standard deviation of the average group size.

3.5.30 The average dolphin group size in NWL waters during September to November 2017 was only slightly higher than the one recorded during the three-month baseline period, but it should also be noted that the sample size of 13 dolphin groups in the present quarter was very small when compared to the 66 groups sighted during the baseline period(Table 3.8).

3.5.31 Notably, 8 of these 14 dolphin groups were composed of 1-3 individuals only while there were four medium-sized groups with 5-8 dolphins per group, and one large group of 12 dolphins (Annex II of Appendix J).

3.5.32 Distribution of the larger dolphin groups with five individuals or more per group during the present quarter is shown in Figure 3 of Appendix J, with comparison to the one in baseline period. The medium-sized group with 5-8 dolphins were scattered at the northwestern portion of the NWL survey area with no particular concentration, while the one large group of 12 dolphins was sighted at the mouth of Deep Bay (Figure 3 of Appendix J). Such distribution pattern was very different from the baseline period, when the larger dolphin groups were frequently sighted and evenly distributed in NWL waters, and a few were also sighted in NEL waters (Figure 3 of Appendix J).

Habitat Use

3.5.33 From September to November 2017, four of the five grids with moderately high to high dolphin densities were located to the north of Lung Kwu Chau, while one grid to the east of Sha Chau also recorded moderately high dolphin density (Figures 4a and 4b of Appendix J). All grids near HKLR03/HKBCF reclamation sites as well as TMCLKL alignment did not record any presence of dolphins at all during on-effort search in the present quarterly period (Figures 4a and 4b of Appendix J).

3.5.34 However, it should be emphasized that the amount of survey effort collected in each grid during the three-month period was fairly low (6-12 units of survey effort for most grids), and therefore the habitat use pattern derived from the three-month dataset should be treated with caution. A more complete picture of dolphin habitat use pattern should be examined when more survey effort for each grid will be collected throughout the impact phase monitoring programme.

3.5.35 When compared with the habitat use patterns during the baseline period, dolphin usage in NEL and NWL has drastically diminished in both areas during the present impact monitoring period (Figure 5 of Appendix J). During the baseline period, many grids between Siu Mo To and Shum Shui Kok in NEL recorded moderately high to high dolphin densities, which was in stark contrast to the complete absence of dolphins there during the present impact phase period (Figure 5 of Appendix J).

3.5.36 The density patterns were also very different in NWL between the baseline and impact phase monitoring periods, with high dolphin usage throughout the area, especially around Sha Chau, near Black Point, to the west of the airport, as well as between Pillar Point and airport platform during the baseline period. In contrast, only several grids with moderately high to high dolphin densities were located near Lung Kwu Chau and Sha Chau during the present impact phase period(Figure 5 of Appendix J).

Mother-calf Pairs

3.5.37 During the present quarterly period, no young calf was sighted at all among the 13 groups of dolphins.

Activities and Associations with Fishing Boats

3.5.38 One of the thirteen dolphin groups were engaged in feeding activity, while another two groups were engaged in socializing activity. However, none of them was engaged in traveling or milling/resting activity during the three-month study period.

3.5.39 The percentages of sightings associated with feeding activities (7.7%) was lower than the one recorded during the baseline period (11.6%), while the one for socializing activities (15.4%) was much higher than the ones recorded during the baseline period (5.4% respectively). However, it should be noted the sample sizes on total numbers of dolphin sightings were very different between the two periods.

3.5.40 Distribution of dolphins engaged in various activities during the present three-month period and baseline period is shown in (Figure 6 of Appendix J). The one dolphin group engaged in feeding activity was sighted at the southeast corner of Lung Kwu Chau, while the two dolphin groups engaged in socializing activities were both located to the north of Lung Kwu Chau (Figure 6 of Appendix J).

3.5.41 When compared to the baseline period, distribution of various dolphin activities during the present impact phase monitoring period was drastically different with a much more restricted area of occurrences (Figure 6 of Appendix J).

3.5.42 Notably, one group of a single dolphin was found to be associated with an operating purse-seiner adjacent to Lung Kwu Chau within the marine park during the present impact phase period.

Summary Photo-identification works

3.5.43 From September to November 2017, over 2,500 digital photographs of Chinese White Dolphins were taken during the impact phase monitoring surveys for the photo-identification work.

3.5.44 In total, 23 individuals sighted 42 times altogether were identified (see summary table in Annex III of Appendix J and photographs of identified individuals in Annex IV of Appendix J). All of these re-sightings were made in NWL. (i.e. CH34, NL33, NL46, NL49, NL320, NL322, NL328 and WL05) were re-sighted twice, while four other individuals (i.e. NL136, NL182, NL202 and NL286) were re-sighted 3-4 times during the three-month period (Annex III of Appendix J).

3.5.45 Notably, ten of these 23 individuals (i.e. CH34, NL12, NL49, NL104, NL136, NL182, NL202, NL320, NL321 and WL05) were also sighted in Northwest Lantau during the HKBCF monitoring surveys under the same three-month period. Moreover, six individuals (i.e. CH34, NL12, NL49, NL182, NL210 and WL05) were also sighted in West Lantau waters during the HKLR09 monitoring surveys from September to November 2017, showing their extensive individual movements across different survey areas.

Individual range use

3.5.46 Ranging patterns of the 23 individuals identified during the three-month study period were determined by fixed kernel method, and are shown in Annex V of Appendix J.

3.5.47 All identified dolphins sighted in the present quarter were utilizing NWL waters only, but have completely avoided NEL waters where many of them have utilized as their core areas in the past (Annex V of Appendix J). This is in contrary to the extensive movements between NEL and NWL survey areas observed in the earlier impact monitoring quarters as well as the baseline period.

3.5.48 On the other hand, three individuals (i.e. NL12, NL182 and NL210) consistently utilized North Lantau waters in the past have extended their range use to WL during the present quarter.

3.5.49 In the upcoming quarters, individual range use and movements should be continuously monitored to examine whether there has been any consistent shifts of individual home ranges from North Lantau to West or Southwest Lantau, as such shift could possibly be related to the HZMB-related construction works (see Hung 2017).

Action Level / Limit Level Exceedance

3.5.50 There was one Limit Level exceedance of dolphin monitoring for the quarterly monitoring data (between September 2017 – November 2017). According to the contractor’s information, the marine activities undertaken for HKLR03 during the quarter of September 2017 – November 2017 included seawall construction, box culvert construction, road and drainage construction and road and drainage works.

3.5.51 There is no evidence showing the current LL non-compliance directly related to the construction works of HKLR03 (where the amounts of working vessels for HKLR03 have been decreasing), although the generally increased amount of vessel traffic in NEL during the impact phase has been partly contributed by HKLR03 works since October 2012. It should also be noted that reclamation work under HKLR03 (adjoining the Airport Island) situates in waters which has rarely been used by dolphins in the past, and the working vessels under HKLR03 have been travelling from source to destination in accordance with the Marine Travel Route to minimize impacts on Chinese White Dolphin (CWD). In addition, the contractor will implement proactive mitigation measures such as avoiding anchoring at Marine Department’s designated anchorage site – Sham Shui Kok Anchorage (near Brothers Island) as far as practicable.

3.5.52 According to Monitoring of Chinese White Dolphins in Southwest Lantau Waters – Fourth Quarterly Report (December 2015 to February 2016) which is available on ENPO’s website, with their primary ranges centered in North and West Lantau waters, some individuals showed apparent range shifts or extensions to Southwest Lantau waters in 2015-16. For example, three individual dolphins (NL120, WL46 and WL221) indicated obvious shifts in their range use from NWL to West Lantau (WL) and Southwest Lantau (SWL) waters. Moreover, many individuals (e.g. NL212, NL260, WL200, SL55, WL232, WL237 and WL265) have extended their ranges from WL waters to SWL waters. It remains to be seen whether some of these individuals have permanently shifted their ranges away from their primary ranges in North Lantau, or begin to spend more times in SWL waters as part of their ranges.

3.5.53 ENPO updated that the Hong Kong-Zhuhai-Macao Bridge Authority (HZMBA) for the Mainland section of Hong Kong-Zhuhai-Macao Bridge (HZMB) has commenced an interim survey on fisheries resources and CWD in the Mainland waters. ENPO presented the preliminary findings of the HZMBA interim survey on CWD sighting and photo-identification works which provide solid evidence that some CWD that were previously more often sighted in HK waters have expanded their ranges into the Mainland waters, and some with reduced usage in HK waters. These preliminary data were mentioned in Monitoring of Chinese White Dolphins in Southwest Lantau Waters – Fourth Quarterly Report (December 2015 to February 2016) which is available on ENPO’s website.

3.5.54 A two-way ANOVA with repeated measures and unequal sample size was conducted to examine whether there were any significant differences in the average encounter rates between the baseline and impact monitoring periods. The two variables that were examined included the two periods (baseline and impact phases) and two locations (NEL and NWL).

3.5.55 For the comparison between the baseline period and the present quarter (20^th quarter of the impact phase being assessed), the p-values for the differences in average dolphin encounter rates of STG and ANI were 0.0057 and 0.0278 respectively. If the alpha value is set at 0.05, significant differences were detected between the baseline and present quarters in both the average dolphin encounter rates of STG and ANI.

3.5.56 For the comparison between the baseline period and the cumulative quarters in impact phase (i.e. first 20 quarters of the impact phase being assessed), the p-values for the differences in average dolphin encounter rates of STG and ANI were 0.000000 and 0.000000 respectively. Even if the alpha value is set at 0.00001, significant differences were still detected in both the average dolphin encounter rates of STG and ANI (i.e. between the two periods and the locations).

3.5.57 The AFCD monitoring data during September 2017 to November 2017 has been reviewed by the dolphin specialist. During the same quarter, no dolphin was sighted from 74.65 km of survey effort on primary lines in NEL, while four groups of 18 dolphins were sighted from 133.52 km of survey effort on primary lines in NWL. This review has confirmed that the low occurrence of dolphins reported by the HKLR03 monitoring surveys in autumn 2017 in NEL and NWL survey area is accurate.

3.5.58 All dolphin protective measures are fully and properly implemented in accordance with the EM&A Manual. According to the Marine Travel Route Plan, the travelling speed of vessels must not exceed 5 knots when crossing the edge of the proposed marine park. The Contractor will continue to provide training for skippers to ensure that their working vessels travel from source to destination to minimize impacts on Chinese White Dolphin and avoid anchoring at Marine Department’s designated anchorage site - Sham Shui Kok Anchorage (near Brothers Island) as far as practicable. Also, it is recommended to complete the marine works of the Contract as soon as possible so as to reduce the overall duration of impacts and allow the dolphins population to recover as early as possible.

3.5.59 All dolphin protective measures are fully and properly implemented in accordance with the EM&A Manual. According to the Marine Travel Route Plan, the travelling speed of vessels must not exceed 5 knots when crossing the edge of the proposed marine park. The Contractor will continue to provide training for skippers to ensure that their working vessels travel from source to destination to minimize impacts on Chinese White Dolphin and avoid anchoring at Marine Department’s designated anchorage site - Sham Shui Kok Anchorage (near Brothers Island) as far as practicable. Also, it is recommended to complete the marine works of the Contract as soon as possible so as to reduce the overall duration of impacts and allow the dolphins population to recover as early as possible.

3.5.60 A meeting was held on 7 March 2018 with attendance of representative of ENPO, Resident Site Staff (RSS), Environmental Team (ET) and dolphin specialist for Contract Nos. HY/2013/01, HY/2011/03, HY/2011/09, HY/2012/07, HY/2012/08. The discussion/ recommendation as raised in the meeting which might be relevant to HKLR03 Contract are summarized below.

3.5.61 It was concluded that the HZMB works is one of the contributing factors affecting the dolphins. It was also concluded the contribution of impacts due to the HZMB works as a whole (or individual marine contracts) cannot be quantified nor separate from the other stress factors.

3.5.62 The dolphin specialists of the projects confirmed that the CWD sighting around the North of Sha Chau and Lung Kwu Chau Marine Park (SCLKCMP) has significantly decreased, and it was apparently related to the re-routing of high speed ferry (HSF) from Skypier.

3.5.63 It was reminded that the ETs shall keep reviewing the implementation status of the dolphin related mitigation measures and remind the contractor to ensure the relevant measures were fully implemented.

3.5.64 It was recommended that the marine works of HZMB projects should be completed as soon as possible so as to reduce the overall duration of impacts and allow the dolphins population to recover as early as possible.

3.5.65 It was also recommended that the marine works footprint (e.g., reduce the size of peripheral silt curtain) and vessels for the marine works should be reduced as much as possible, and vessels idling / mooring in other part of the North Lantau shall be avoided whenever possible.

3.5.66 It was suggested that the protection measures (e.g., speed limit control) for the proposed Brothers Marine Park (BMP) shall be brought forward as soon as possible before its establishment so as to provide a better habitat for dolphin recovery. It was noted that under the Regular Marine Travel Route Plan, the contractors have committed to reduce the vessel speed in BMP. HyD updated that the draft map of the proposed BMP was gazetted in February 2016. ENPO updated that the BMP was approved by the Chief Executive in the Executive Council in August 2016. The ETs were reminded to update the BMP boundary in the RMTR Plan. The BMP was designated on 30 December 2016. It was reminded that trespassing the BMP is not allowed under the RMTR Plan.

3.5.67 There was a discussion on exploring possible further mitigation measures, for example, controlling the underwater noise. It was noted that the EIA reports for the projects suggested several mitigation measures, all of which have been implemented.

3.6 Mudflat Monitoring Results

Sedimentation Rate Monitoring

3.6.1 The baseline sedimentation rate monitoring was in September 2012 and impact sedimentation rate monitoring was undertaken on 16 September 2017. The mudflat surface levels at the four established monitoring stations and the corresponding XYZ HK1980 GRID coordinates are presented in Table 3.9 and Table 3.10.

Table 3.9 Measured Mudflat Surface Level Results

	Baseline Monitoring (September 2012)			Impact Monitoring (September 2017)
Monitoring Station	Easting (m)	Northing (m)	Surface Level (mPD)	Easting (m)	Northing (m)	Surface Level (mPD)
S1	810291.160	816678.727	0.950	810291.088	816678.776	1.088
S2	810958.272	815831.531	0.864	810958.265	815831.586	0.973
S3	810716.585	815953.308	1.341	810716.480	815953.308	1.463
S4	811221.433	816151.381	0.931	811221.423	816151.385	1.085

Table 3.10 Comparison of Measurement

	Comparison of measurement			Remarks and Recommendation
Monitoring Station	Easting (m)	Northing (m)	Surface Level (mPD)	Remarks and Recommendation
S1	-0.072	0.049	0.138	Level continuously increased
S2	-0.007	0.055	0.109	Level continuously increased
S3	-0.105	0.000	0.122	Level continuously increased
S4	-0.01	0.004	0.154	Level continuously increased

3.6.2 This measurement result was generally and relatively higher than the baseline measurement at S1, S2, S3 and S4. The mudflat level is continuously increased.

Water Quality Monitoring

3.6.3 The mudflat monitoring covered water quality monitoring data. Reference was made to the water quality monitoring data of the representative water quality monitoring station (i.e. SR3) as in the EM&A Manual. The water quality monitoring location (SR3) is shown in Figure 2.1.

3.6.4 Impact water quality monitoring in San Tau (monitoring station SR3) was conducted in September 2017. The monitoring parameters included dissolved oxygen (DO), turbidity and suspended solids (SS).

3.6.5 The Impact monitoring result for SR3 were extracted and summarised below:

Table 3.11 Impact Water Quality Monitoring Results (Depth Average)

Date	Mid Ebb Tide			Mid Flood Tide
Date	DO (mg/L)	Turbidity (NTU)	SS (mg/L)	DO (mg/L)	Turbidity (NTU)	SS (mg/L)
1-Sep-17	6.6	7.4	4.7	6.9	6.0	7.0
4-Sep-17	Remark 1	Remark 1	Remark 1	6.0	8.7	9.7
6-Sep-17	5.5	11.4	9.1	5.5	10.3	9.9
8-Sep-17	6.1	11.4	7.7	5.6	5.7	8.0
11-Sep-17	5.7	6.8	12.1	5.4	5.4	11.3
13-Sep-17	5.2	6.6	6.2	5.8	11.5	38.7
15-Sep-17	6.2	5.4	3.5	6.7	7.8	4.9
18-Sep-17	6.1	3.6	6.4	8.3	6.5	10.1
20-Sep-17	5.9	12.5	19.1	5.9	5.6	11.6
22-Sep-17	11.3	9.3	12.1	5.3	7.5	11.9
25-Sep-17	5.7	5.0	7.8	5.9	5.8	7.7
27-Sep-17	6.6	5.9	8.2	6.0	2.1	5.7
29-Sep-17	5.9	4.3	8.0	8.0	5.5	9.2
Average	6.4	7.5	8.7	6.3	6.8	11.2
Remark: 1) All water quality monitoring was conducted as scheduled in the reporting month except water quality monitoring during mid-ebb tide on 4 September 2017 which was cancelled due to adverse weather condition (Strong Wind Signal No. 3).

Mudflat Ecology Monitoring

Sampling Zone

3.6.6 In order to collect baseline information of mudflats in the study site, the study site was divided into three sampling zones (labeled as TC1, TC2, TC3) in Tung Chung Bay and one zone in San Tau (labeled as ST) (Figure 2.1 of Appendix O). The horizontal shoreline of sampling zones TC1, TC2, TC3 and ST were about 250 m, 300 m, 300 m and 250 m respectively (Figure 2.2 of Appendix O). Survey of horseshoe crabs, seagrass beds and intertidal communities were conducted in every sampling zone. The present survey was conducted in September 2017 (totally 5 sampling days between 2^nd and 17^th September 2017).

3.6.7 Since the field survey of Jun. 2016, increasing number of trashes and even big trashes (Figure 2.3 of Appendix O) were found in every sampling zone. It raised a concern about the solid waste dumping and current-driven waste issues in Tung Chung Wan. Respective measures (e.g. manual clean-up) should be implemented by responsible units.

Horseshoe Crabs

3.6.8 Active search method was conducted for horseshoe crab monitoring by two experienced surveyors in every sampling zone. During the search period, any accessible and potential area would be investigated for any horseshoe crab individuals within 2-3 hours of low tide period (tidal level below 1.2 m above Chart Datum (C.D.)). Once a horseshoe crab individual was found, the species was identified referencing to Li (2008). The prosomal width, inhabiting substratum and respective GPS coordinate were recorded. A photographic record was taken for future investigation. Any grouping behavior of individuals, if found, was recorded. The horseshoe crab surveys were conducted on 2nd (for TC2), 3^rd (for TC1) and 6^th (for TC3 and ST) September 2017. The weather was generally hot on all field days without rainfall.

3.6.9 In Jun. 2017, a big horseshoe crab was tangled by a trash gill net in ST mudflat (Figure 2.3 of Appendix O). It was released to sea once after photo recording. The horseshoe crab of such size should be inhabitating sub-tidal environment while it forages on intertidal shore occasionally during high tide period. If it is tangled by the trash net for few days, it may die due to starvation or overheat during low tide period. These trash gill nets are definitely ‘fatal trap’ for the horseshoe crabs and other marine life. Manual clean-up should be implemented as soon as possible by responsible units.

Seagrass Beds

3.6.10 Active search method was conducted for seagrass bed monitoring by two experienced surveyors in every sampling zone. During the search period, any accessible and potential area would be investigated for any seagrass beds within 2-3 hours of low tide period. Once seagrass bed was found, the species, estimated area, estimated coverage percentage and respective GPS coordinates were recorded. The seagrass beds surveys were conducted on 2^nd (for TC2), 3^rd (for TC1) and 6^th (for TC3 and ST) September 2017. The weather was generally hot on all field days without rainfall.

Intertidal Soft Shore Communities

3.6.11 The intertidal soft shore community surveys were conducted in low tide period on 2^nd (for TC2), 3^rd (for TC1), 16^th (for TC3) and 17^th (for ST) September 2017. In every sampling zone, three 100m horizontal transect lines were laid at high tidal level (H: 2.0 m above C.D.), mid tidal level (M: 1.5 m above C.D.) and low tidal level (L: 1.0 m above C.D.). Along every horizontal transect line, ten random quadrats (0.5 m x 0.5 m) were placed.

3.6.12 Inside a quadrat, any visible epifauna were collected and were in-situ identified to the lowest practical taxonomical resolution. Whenever possible a hand core sample (10 cm internal diameter ´ 20 cm depth) of sediments was collected in the quadrat. The core sample was gently washed through a sieve of mesh size 2.0 mm in-situ. Any visible infauna were collected and identified. Finally the top 5 cm surface sediments was dug for visible infauna in the quadrat regardless of hand core sample was taken.

3.6.13 All collected fauna were released after recording except some tiny individuals that are too small to be identified on site. These tiny individuals were taken to laboratory for identification under dissecting microscope.

3.6.14 The taxonomic classification was conducted in accordance to the following references: Polychaetes: Fauchald (1977), Yang and Sun (1988); Arthropods: Dai and Yang (1991), Dong (1991); Mollusks: Chan and Caley (2003), Qi (2004).

Data Analysis

3.6.15 Data collected from direct search and core sampling was pooled in every quadrat for data analysis. Shannon-Weaver Diversity Index (H’) and Pielou’s Species Evenness (J) were calculated for every quadrat using the formulae below,

H’= -Σ ( Ni / N ) ln ( Ni / N ) (Shannon and Weaver, 1963)

J = H’ / ln S, (Pielou, 1966)

where S is the total number of species in the sample, N is the total number of individuals, and Ni is the number of individuals of the i^th species.

Mudflat Ecology Monitoring Results and Conclusion

Horseshoe Crabs

3.6.16 In the present survey, two species of horseshoe crab Carcinoscorpius rotundicauda (total 130 ind.) and Tachypleus tridentatus (total 77 ind.) were recorded. For one sight record, grouping of 2-18 individuals was observed at same locations with similar substratum (fine sand or soft mud, slightly submerged). Photo records were shown in Figure 3.1 of Appendix O while the complete survey records were listed in Annex II of Appendix O.

3.6.17 Table 3.1 of Appendix O summarizes the survey results of horseshoe crab in the present survey. For Carcinoscorpius rotundicauda, moderate number of individuals (18 ind.) were found in TC1 that search record was at low-moderate level (4.5 ind. hr^-1 person^-1). The average body size was 30.91 mm (prosomal width ranged 19.99-47.55 mm) in TC1. There was one individual found in TC2 only (prosomal width 37.85 mm) resulting in very low search record (0.3 ind. hr^-1 person^-1). More individuals were found in TC3 (72 ind.) and ST (39 ind.) resulting in relatively higher search records (6.5-12.0 ind. hr^-1 person^-1). Smaller individuals were found in TC3 that the average body size was 35.02 mm (prosomal width ranged 13.30-80.27 mm). The average body size was 50.18 mm (prosomal width ranged 19.87-78.00 mm) in ST.

3.6.18 For Tachypleus tridentatus, there was one individual found in TC1 only (prosomal width 38.34 mm) resulting in very low search record (0.3 ind. hr-1 per son-1). No individual was found in TC2. Similarly, more individuals were found in TC3 (48 ind.) and ST (28 ind.) respectively. In TC3, the search record was relatively higher (8.0 ind. hr-1 person-1) while the average body size was 47.52 mm (prosomal width ranged 34.76-88.93 mm). In ST, the search record was 4.7 ind. hr^-1 person^-1 while the average body size was 47.21 mm (prosomal width ranged 29.61-64.36 mm).

3.6.19 In the previous survey of Mar. 2015, there was one important finding that a mating pair of Carcinoscorpius rotundicauda was found in ST (prosomal width: male 155.1 mm, female 138.2 mm) (Figure 3.2 of Appendix O). It indicated the importance of ST as a breeding ground of horseshoe crab. In another survey of Jun. 2017, mating pairs of Carcinoscorpius rotundicauda were also found in TC2 (prosomal width: male 175.27 mm, female 143.51 mm) and TC3 (prosomal width: male 182.08 mm, female 145.63 mm) (Figure 3.2 of Appendix O). It indicated that breeding of horseshoe crab could occur along the coast of Tung Chung Wan rather than ST only, as long as suitable substratum was available. The mating pairs were found nearly burrowing in soft mud at low tidal level (0.5-1.0 m above C.D.). The smaller male was holding the opisthosoma (abdomen carapace) of larger female from behind.

3.6.20 In the previous surveys (Jun. 2016, Jun. 2017) and present survey (Sep. 2017), there were occasional records of large individuals of Carcinoscorpius rotundicauda (prosomal width ranged 117.37- 178.67 mm, either single or in pair) in ST (Fig. 3.3). Based on their sizes, it indicated that individuals of prosomal width larger than 100 mm would progress its nursery stage from intertidal habitat to sub-tidal habitat of Tung Chung Wan. These large individuals might move onto intertidal shore occasionally during high tide for foraging and breeding. Because they should be inhabiting sub-tidal habitat most of the time. Their records were excluded from the data analysis to avoid mixing up with juvenile population living on intertidal habitat.

3.6.21 No marked individual of horseshoe crab was recorded in the present survey. Some marked individuals were found in the previous surveys of Sep. 2013, Mar. 2014 and Sep. 2014. All of them were released through a conservation programme in charged by Prof. Paul Shin (Department of Biology and Chemistry, The City University of Hong Kong (CityU)). It was a re-introduction trial of artificial bred horseshoe crab juvenile at selected sites. So that the horseshoe crab population might be restored in the natural habitat. Through a personal conversation with Prof. Shin, about 100 individuals were released in the sampling zone ST on 20 June 2013. All of them were marked with color tape and internal chip detected by specific chip sensor. There should be second round of release between June and September 2014 since new marked individuals were found in the survey of Sep. 2014.

3.6.22 The artificial bred individuals, if found, would be excluded from the results of present monitoring programme in order to reflect the changes of natural population. However, the mark on their prosoma might have been detached during moulting after a certain period of release. The artificially released individuals were no longer distinguishable from the natural population without the specific chip sensor. The survey data collected would possibly cover both natural population and artificially bred individuals.

Population difference among the sampling zones

3.6.23 Figures 3.4 and 3.5 of Appendix O show the changes of number of individuals, mean prosomal width and search record of horseshoe crabs Carcinoscorpius rotundicauda and Tachypleus tridentatus respectively in every sampling zone throughout the monitoring period.

3.6.24 For TC3 and ST, medium to high search records (i.e. number of individuals) of both species were always found in wet season (Jun. and Sep.). The search record of ST was higher from Sep. 2012 to Jun. 2014 while it was replaced by TC3 from Sep. 2014 to Jun. 2015. The search records were similar between two sampling zones from Sep. 2015 to Jun. 2016. In Sep. 2016, the search record of Carcinoscorpius rotundicauda in ST was much higher than TC3. From Mar. to Jun. 2017, the search records of both species were similar again between two sampling zones. It showed a natural variation of horseshoe crab population in these two zones due to weather condition and tidal effect. No obvious difference of horseshoe crab population was noted between TC3 and ST. In Sep. 2017, the search records of both horseshoe crab species decreased except the Carcinoscorpius rotundicauda in TC3. The survey results were different from previous findings that there were usually higher search records in Sep.. One possible reason was that the serial cyclone hit decreased horseshoe crab activity (totally 4 cyclone records between Jun. and Sep. 2017, to be discussed in 'Seagrass survey' section).

3.6.25 For TC1, the search record was at low to medium level throughout the monitoring period. The change of Carcinoscorpius rotundicauda was relatively more variable than that of Tachypleus tridentatus. Relatively, the search record was very low in TC2 (2 ind. in Sep. 2013; 1 ind. in Mar., Jun., Sep. 2014, Mar. and Jun. 2015; 4 ind. in Sep. 2015; 6 ind. in Jun. 2016; 1 ind. in Sep. 2016, Mar., Jun. and Sep. 2017).

3.6.26 About the body size, larger individuals of Carcinoscorpius rotundicauda were usually found in ST and TC1 relative to those in TC3. For Tachypleus tridentatus, larger individuals were usually found in ST followed by TC3 and TC1. Throughout the monitoring period, it was obvious that TC3 and ST (western shore of Tung Chung Wan) was an important nursery ground for horseshoe crab especially newly hatched individuals due to larger area of suitable substratum (fine sand or soft mud) and less human disturbance (far from urban district). Relatively, other sampling zones were not a suitable nursery ground especially TC2. Possible factors were less area of suitable substratum (especially TC1) and higher human disturbance (TC1 and TC2: close to urban district and easily accessible). In TC2, large daily salinity fluctuation was a possible factor either since it was flushed by two rivers under tidal inundation. The individuals inhabiting TC1 and TC2 were confined in small foraging area due to limited area of suitable substrata. Although a mating pair of Carcinoscorpius rotundicauda was found in TC2, the hatching rate and survival rate of newly hatched individuals were believed very low.

Seasonal variation of horseshoe crab population

3.6.27 Throughout the monitoring period, the search record of horseshoe crab declined obviously during dry season especially December (Figures 3.3 and 3.4 of Appendix O). In Dec. 2012, 4 individuals of Carcinoscorpius rotundicauda and 12 individuals of Tachypleus tridentatus were found only. In Dec. 2013, no individual of horseshoe crab was found. In Dec. 2014, 2 individuals of Carcinoscorpius rotundicauda and 8 individuals of Tachypleus tridentatus were found only. In Dec. 2015, 2 individuals of Carcinoscorpius rotundicauda, 6 individuals of Tachypleus tridentatus and one newly hatched, unidentified individual were found only. The horseshoe crabs were inactive and burrowed in the sediments during cold weather (<15 ºC). Similar results of low search record in dry season were reported in a previous territory-wide survey of horseshoe crab. For example, the search records in Tung Chung Wan were 0.17 ind. hr-1 person-1 and 0.00 ind. hr-1 person-1 in wet season and dry season respectively (details see Li, 2008). Relatively the serach records were much higher in Dec. 2016. There were totally 70 individuals of Carcinoscorpius rotundicauda and 24 individuals of Tachypleus tridentatus in TC3 and ST. Because the survey was arranged in early December while the weather was warm with sunlight (~22 ºC during dawn according to Hong Kong Observatory database, Chek Lap Kok station on 5 Dec). In contrast, there was no search record in TC1 and TC2 because the survey was conducted in mid December with colder and cloudy weather (~20 ºC during dawn on 19 Dec). The horseshoe crab activity would decrease gradually with the colder climate.

3.6.28 From Sep. 2012 to Dec. 2013, Carcinoscorpius rotundicauda was a less common species relative to Tachypleus tridentatus. Only 4 individuals were ever recorded in ST in Dec. 2012. This species had ever been believed of very low density in ST hence the encounter rate was very low. Since Mar. 2014, it was found in all sampling zones with higher abundance in ST. Based on its average size (mean prosomal width 39.28-49.81 mm), it indicated that breeding and spawning of this species had occurred about 3 years ago along the coastline of Tung Chun Wan. However, these individuals were still small while their walking trails were inconspicuous. Hence there was no search record in previous sampling months. Since Mar. 2014, more individuals were recorded due to larger size and higher activity (i.e. more conspicuous walking trail).

3.6.29 For Tachypleus tridentatus, sharp increase of number of individuals was recorded in ST during the wet season of 2013 (from Mar. to Sep.). According to a personal conversation with Prof. Shin (CityU), his monitoring team had recorded similar increase of horseshoe crab population during wet season. It was believed that the suitable ambient temperature increased its conspicuousness. However similar pattern was not recorded in the following wet seasons. The number of individuals increased in Mar. and Jun. 2014 followed by a rapid decline in Sep. 2014. Then the number of individuals fluctuated slightly in TC3 and ST until Mar. 2017. Apart from natural mortality, migration from nursery soft shore to subtidal habitat was another possible cause. Since the mean prosomal width of Tachypleus tridentatus continued to grow and reached about 50 mm since Mar. 2014. Then it varied slightly between 35-65 mm from Sep. 2014 to Mar. 2017. Most of the individuals might have reached a suitable size (e.g. prosomal width 50-60 mm) strong enough to forage in sub-tidal habitat. In Jun. 2017, the number of individuals increased sharply again in TC3 and ST. Although mating pair of Tachypleus tridentatus was not found in previous surveys, there should be new round of spawning in the wet season of 2016. The individuals might have grown to a more conspicuous size in 2017 accounting for higher search record.

3.6.30 Recently, Carcinoscorpius rotundicauda was a more common horseshoe crab species in Tung Chung Wan. It was recorded in the four sampling zones while the majority of population located in TC3 and ST. Due to potential breeding last year, Tachypleus tridentatus became common again and distributed in TC3 and ST only. Since TC3 and ST were regarded as important nursery ground for both horseshoe crab species, box plots of prosomal width of two horseshoe crab species were constructed to investigate the changes of population in details.

Box plot of horseshoe crab populations in TC3

3.6.31 Figure 3.6 of Appendix O shows the changes of prosomal width of Carcinoscorpius rotundicauda and Tachypleus tridentatus in TC3. As mentioned above, Carcinoscorpius rotundicauda was rarely found between Sep. 2012 and Dec. 2013 hence the data were lacking. In Mar 2014, the major size (50% of individual records between upper (top of red box) and lower quartile (bottom of blue box)) ranged 40-60 mm while only few individuals were found. From Mar. 2014 to Jun. 2017, the median prosomal width (middle line of whole box) and major size (whole box) decreased after Mar. of every year. It was due to more small individuals found. It indicated new rounds of spawning. Also there were slight increasing trends of body size from Jun. to Mar. of next year since 2015. It indicated a stable growth of individuals. Focused on larger juveniles (upper whisker), the size range was quite variable (prosomal width 60-90 mm) along the sampling months. Juveniles reaching this size might gradually migrate to sub-tidal habitats.

3.6.32 For Tachypleus tridentatus, the major size ranged 20-50 mm while the number of individuals fluctuated from Sep. 2012 to Jun. 2014. Then a slight but consistent growing trend was observed from Sep. 2014 to Jun. 2015. The prosomal width increased from 25-35 mm to 35-65 mm. As mentioned, the large individuals might have reached a suitable size for migrating from the nursery soft shore to subtidal habitat. It accounted for the declined population in TC3. From Mar. to Sep. 2016, slight increasing trend of major size was noticed again. From Dec. 2016 to Jun. 2017, similar increasing trend of major size was noted with much higher number of individuals. It reflected new round of spawning. In Sep. 2017 (present survey), the major size decreased while the trend was different from previous two years. Such decline might be the cause of serial cyclone hit between Jun. and Sep. 2017 (to be discussed in the 'Seagrass survey' section). Across the whole monitoring period, the larger juveniles (upper whisker) reached 60-80 mm in prosomal width while it could reach 90 mm in present survey. Juveniles reaching this size might gradually migrate to sub-tidal habitats.

Box plot of horseshoe crab populations in ST

3.6.33 Figure 3.7 of Appendix O shows the changes of prosomal width of Carcinoscorpius rotundicauda and Tachypleus tridentatus in ST. As mentioned above, Carcinoscorpius rotundicauda was rarely found between Sep. 2012 and Dec. 2013 hence the data were lacking. From Mar. 2014 to Sep. 2017, the size of major population decreased and more small individuals (i.e. lower whisker) were recorded after Jun. of every year. It indicated new round of spawning. Also there were similar increasing trends of body size from Sep. to Jun. of next year between 2014 and 2017. It indicated a stable growth of individuals. Across the whole monitoring period, the larger juveniles (i.e. upper whisker) usually ranged 60-80 mm in prosomal width except one individual (prosomal width 107.04 mm) found in Mar. 2017. It reflected juveniles reaching this size would gradually migrate to sub-tidal habitats.

3.6.34 For Tachypleus tridentatus, a consistent growing trend was observed for the major population from Dec. 2012 to Dec. 2014 regardless of change of search record. The prosomal width increased from 15-30 mm to 60-70 mm. As mentioned, the large juveniles might have reached a suitable size for migrating from the nursery soft shore to subtidal habitat. From Mar. to Sep. 2015, the size of major population decreased slightly to a prosomal width 40-60 mm. At the same time, the number of individuals decreased gradually. It further indicated some of large juveniles might have migrated to sub-tidal habitat, leaving the smaller individuals on shore. There was an overall growth trend. In Dec. 2015, two big individuals (prosomal width 89.27 mm and 98.89 mm) were recorded only while it could not represent the major population. In Mar. 2016, the number of individual was very few in ST that no boxplot could be produced. In Jun. 2016, the prosomal width of major population ranged 50-70 mm. But it dropped clearly to 30-40 mm in Sep. 2016 followed by an increase to 40-50 mm in Dec. 2016, 40-70 mm in Mar. 2017 and 50-60mm in Jun. 2017. Based on overall higher number of small individuals from Jun. 2016 to Sep. 2017 (present survey), it indicated new round of spawning. Throughout the monitoring period, the larger juveniles ranged 60-80 mm in prosomal width. Juveniles reaching this size would gradually migrate to sub-tidal habitats.

3.6.35 As a summary for horseshoe crab populations in TC3 and ST, there were spawning of Carcinoscorpius rotundicauda from 2014 to 2016 while the spawning time should be in spring. There were consistent, increasing trends of population size in these two sampling zones. For Tachypleus tridentatus, small individuals were rarely found in both zones from 2014 to 2015. It was believed no occurrence of successful spawning. The existing individuals (that recorded since 2012) grew to a mature size and migrated to sub-tidal habitat. Hence the number of individuals decreased gradually. In 2016, new round of spawning was recorded in ST while increasing number of individuals and body size was noticed.

Impact of the HKLR project

3.6.36 It was the 20^th survey of the EM&A programme during the construction period. Based on the results, impact of the HKLR project could not be detected on horseshoe crabs. The population change was mainly determined by seasonal variation while new rounds of spawning were observed for both species. In case, abnormal phenomenon (e.g. very few numbers of horseshoe crab individuals in wet season, large number of dead individuals on the shore) is found, it would be reported as soon as possible.

Seagrass Beds

3.6.37 In the present survey, no seagrass bed was recorded in Tung Chung Wan. Extensive area of mudflat, where used to be covered by seagrass beds, re-exposed along TC3 and ST (Figure 3.8 of Appendix O). In the previous survey of Jun. 2017, two species of seagrass Halophila ovalis and Zostera japonica were recorded in TC3 and ST (Figure 3.9 of Appendix O). There was still extensive seagrass area (~17046.5 m²) of Halophila ovalis along the mudflat between TC3 and ST at 0.5-2.0 m above C.D.. Another seagrass species Zostera japonica, which was much lower in vegetation area (~105.4 m²), was co-existing with few patches of Halophila ovalis nearby the mangrove strand. The disappearance of seagrass beds would be discussed in later paragraphs.

3.6.38 According to the previous results, majority of seagrass bed was confined in ST, the temporal change of both seagrass species were investigated in details:

Temporal variation of seagrass beds

3.6.39 Figure 3.10 of Appendix O shows the changes of estimated total area of seagrass beds in ST along the sampling months. For Zostera japonica, it was not recorded in the 1^st and 2^nd surveys of monitoring programme. Seasonal recruitment of few, small patches (total seagrass area: 10 m²) was found in Mar. 2013 that grew within the large patch of seagrass Halophila ovalis. Then the patch size increased and merged gradually with the warmer climate from Mar. to Jun. 2013 (15 m²). However the patch size decreased and remained similar from Sep. 2013 (4 m²) to Mar. 2014 (3 m²). In Jun. 2014, the patch size increased obviously again (41 m²) with warmer climate followed by a decrease between Sep. 2014 (2 m²) and Dec. 2014 (5 m²). From Mar. to Jun. 2015, the patch size increased sharply again (90 m2). It might be due to the disappearance of the originally dominant seagrass Halophila ovalis resulting in less competition for substratum and nutrients. From Sep.2015 to Jun.2016, it was found coexisting with seagrass Halophila ovalis with steady increasing patch size (from 44 m² to 115 m²) and variable coverage. In Sep. 2016, the patch size decreased again to (38 m²) followed by an increase to a horizontal strand (105.4 m²) in Jun. 2017 (present survey). And it was no longer co-existing with Halophila ovalis. Between Sep. 2014 and Jun. 2017, an increasing trend was noticed from Sep. to Jun. of next year followed by a rapid decline in Sep. of next year. It was possibly the causes of heat stress, typhoon and stronger grazing pressure during wet season. In the present survey, no seagrass patch of Zostera japonica was found. Such disappearance matched the findings of previous monitoring period.

3.6.40 For Halophila ovalis, it was recorded as 3-4 medium to large patches (area 18.9-251.7 m²; vegetation coverage 50-80%) beside the mangrove vegetation at tidal level 2 m above C.D. in Sep. 2012 (first survey). The total seagrass bed area grew steadily from 332.3 m² in Sep. 2012 to 727.4 m² in Dec. 2013. Flowers were observed in the largest patch during its flowering period. In Mar. 2014, 31 small to medium patches were newly recorded (variable area 1-72 m² per patch, vegetation coverage 40-80% per patch) in lower tidal zone between 1.0 and 1.5 m above C.D. The total seagrass area increased further to 1350 m². In Jun. 2014, these small and medium patches grew and extended to each other. These patches were no longer distinguishable and were covering a significant mudflat area of ST. It was generally grouped into 4 large patches (1116 – 2443 m²) of seagrass beds characterized of patchy distribution, variable vegetable coverage (40-80%) and smaller leaves. The total seagrass bed area increased sharply to 7629 m². In Sep. 2014, the total seagrass area declined sharply to 1111 m2. There were only 3-4 small to large patches (6-253 m²) at high tidal level and 1 patch at low tidal level (786 m²). Typhoon or strong water current was a possible cause (Fong, 1998). In Sep. 2014, there were two tropical cyclone records in Hong Kong (7^th-8^th Sep.: no cyclone name, maximum signal number 1; 14^th-17^th Sep.: Kalmaegi, maximum signal number 8SE) before the seagrass survey dated 21st Sep. 2014. The strong water current caused by the cyclone, Kalmaegi especially, might have given damage to the seagrass beds. In addition, natural heat stress and grazing force were other possible causes reducing seagrass beds area. Besides, very small patches of Halophila ovalis could be found in other mud flat area in addition to the recorded patches. But it was hardly distinguished due to very low coverage (10-20%) and small leaves.

3.6.41 In Dec. 2014, all the seagrass patches of Halophila ovalis disappeared in ST. Figure 3.10 of Appendix O shows the difference of the original seagrass beds area nearby the mangrove vegetation at high tidal level between Jun. 2014 and Dec. 2014. Such rapid loss would not be seasonal phenomenon because the seagrass beds at higher tidal level (2.0 m above C.D.) were present and normal in December 2012 and 2013. According to Fong (1998), similar incident had occurred in ST in the past. The original seagrass area had declined significantly during the commencement of the construction and reclamation works for the international airport at Chek Lap Kok in 1992. The seagrass almost disappeared in 1995 and recovered gradually after the completion of reclamation works. Moreover, incident of rapid loss of seagrass area was also recorded in another intertidal mudflat in Lai Chi Wo in 1998 with unknown reason. Hence Halophila ovalis was regarded as a short-lived and r-strategy seagrass that could colonize areas in short period but disappears quickly under unfavourable conditions (Fong, 1998).

Unfavourable conditions to seagrass Halophila ovalis

3.6.42 Typhoon or strong water current was suggested as one unfavourable condition to Halophila ovalis (Fong, 1998). As mentioned above, there were two tropical cyclone records in Hong Kong in Sep. 2014. The strong water current caused by the cyclones might have given damage to the seagrass beds.

3.6.43 Prolonged light deprivation due to turbid water would be another unfavouable condition. Previous studies reported that Halophila ovalis had little tolerance to light deprivation. During experimental darkness, seagrass biomass declined rapidly after 3-6 days and seagrass died completely after 30 days. The rapid death might be due to shortage of available carbohydrate under limited photosynthesis or accumulation of phytotoxic end products of anaerobic respiration (details see Longstaff et al., 1999). Hence the seagrass bed of this species was susceptible to temporary light deprivation events such as flooding river runoff (Longstaff and Dennison, 1999).

3.6.44 In order to investigate any deterioration of water quality (e.g. more turbid) in ST, the water quality measurement results at two closest monitoring stations SR3 and IS5 of the EM&A programme were obtained from the water quality monitoring team. Based on the results from June to December 2014, the overall water quality was in normal fluctuation except there was one exceedance of suspended solids (SS) at both stations in September. On 10^th Sep., 2014, the SS concentrations measured during mid-ebb tide at stations SR3 (27.5 mg/L) and IS5 (34.5 mg/L) exceeded the Action Level (≤23.5 mg/L and 120% of upstream control station’s reading) and Limit Level (≤34.4 mg/L and 130% of upstream control station’s reading) respectively. The turbidity readings at SR3 and IS5 reached 24.8-25.3 NTU and 22.3-22.5 NTU respectively. The temporary turbid water should not be caused by the runoff from upstream rivers. Because there was no rain or slight rain from 1st to 10th Sep. 2014 (daily total rainfall at the Hong Kong International Airport: 0-2.1 mm; extracted from the climatological data of Hong Kong Observatory). The effect of upstream runoff on water quality should be neglectable in that period. Moreover the exceedance of water quality was considered unlikely to be related to the contract works of HKLR according to the ‘Notifications of Environmental Quality Limits Exceedances’ provided by the respective environmental team. The respective construction of seawall and stone column works, which possibly caused turbid water, were carried out within silt curtain as recommended in the EIA report. Moreover there was no leakage of turbid water, abnormity or malpractice recorded during water sampling. In general, the exceedance of suspended solids concentration was considered to be attributed to other external factors, rather than the contract works.

3.6.45 Based on the weather condition and water quality results in ST, the co-occurrence of cyclone hit and turbid waters in Sep. 2014 might have combined the adverse effects on Halophila ovalis that leaded to disappearance of this short-lived and r-strategy seagrass species. Fortunately Halophila ovalis was a fast-growing species (Vermaat et al., 1995). Previous studies showed that the seagrass bed could be recovered to the original sizes in 2 months through vegetative propagation after experimental clearance (Supanwanid, 1996). Moreover it was reported to recover rapidly in less than 20 days after dugong herbivory (Nakaoka and Aioi, 1999). As mentioned, the disappeared seagrass in ST in 1995 could recover gradually after the completion of reclamation works for international airport (Fong, 1998). The seagrass beds of Halophila ovalis might recolonize the mudflat of ST through seed reproduction as long as there was no unfavourable condition in the coming months.

Recolonization of seagrass beds

3.6.46 Figure 3.10 of Appendix O shows the recolonization of seagrass bed area in ST from Dec. 2014 to Jun. 2017. From Mar. to Jun. 2015, 2-3 small patches of Halophila ovalis were newly found coinhabiting with another seagrass species Zostera japonica. But its total patch area was still very low relative to the previous records. The recolonization rate was low while cold weather and insufficient sunlight were possible factors between Dec. 2014 and Mar. 2015. Moreover, it would need to compete with seagrass Zostera japonica for substratum and nutrient. Since Zostera japonica had extended and had covered the original seagrass bed of Halophila ovalis at certain degree. From Jun. 2015 to Mar. 2016, the total seagrass area of Halophila ovalis had increased rapidly from 6.8 m² to 230.63 m². It had recolonized its original patch locations and covered Zostera japonica. In Jun. 2016, the total seagrass area increased sharply to 4707.3 m². Similar to the previous records of Mar to Jun. 2014, the original patch area increased further to a horizontally long strand. Another large seagrass beds colonized the lower tidal zone (1.0-1.5 m above C.D.). In Sep. 2016, this patch extended much and covered significant soft mud area of ST, resulting in sharp increase of total area (24245 m²). It indicated the second extensive colonization of this r-strategy seagrass. In Dec. 2016, this extensive seagrass patch decreased in size and had separated into few, undistinguishable patches. Moreover, the horizontal strand nearby the mangrove vegetation decreased in size (Figure 3.10 of Appendix O). The total seagrass bed decreased to 12550 m². From Mar. to Jun. 2017, the seagrass bed area remained generally stable (12438-17046.5 m²) but the vegetation coverage fluctuated (20-50% in Mar. 2017 to 80-100% in Jun. 2017).

Re-disappearance of seagrass bed

3.6.47 In present survey, the whole seagrass bed of Halophila ovalis disappeared again along the shore of TC3 and ST (Figure 3.10 of Appendix O). It was similar to the case between Sep. and Dec. 2014. As mentioned, strong water current (e.g. cyclone) or deteriorated water quality (e.g. high turbidity) were the possible causes.

3.6.48 Between the survey periods of Jun. and Sep. 2017, there were four tropical cyclone records in Hong Kong (Merbok in 12-13^th, Jun.; Roke in 23^rd, Jul.; Hato in 22-23^rd, Aug.; Pakhar in 26-27^th, Aug.) (online database of Hong Kong Observatory). All of them reaches signal 8 or above especially Hato (highest signal 10).

3.6.49 According to the water quality monitoring results (Jul. to Aug. 2017) of the two closest monitoring stations SR3 and I5 of the respective EM&A programme, the overall water quality was in normal fluctuation. There was one exceedance of suspended solids (SS) at SR3 on 12 Jul. 2017. The SS concentration reached 24.7 mg/L during mid-ebb tide. It exceeded the Action Level (≤23.5 mg/L) but was far below the Limit Level ((≤34.4 mg/L). Since such exceedance was slight and temporary, its effect to seagrass bed should be minimal.

3.6.50 Overall, the disappearance of seagrass beds in ST was believed the cause of serial cyclone hit in Jul and Aug. 2017. Based on previous findings, the seagrass beds of both species were expected to recolonize the mudflat as long as the vicinal water quality was normal. The recolonization would be a gradual process lasting for about 1.5 years.

Impact of the HKLR project

3.6.51 It was the 20^th survey of the EM&A programme during the construction period. According to the results of present survey, the disappearance of seagrass beds was believed the cause of serial cyclone hits rather than impact of HKLR project. Based on previous findings, the seagrass beds were expected to recolonize the mudflat gradually in the future, as long as the vicinal water quality remained normal.

Intertidal Soft Shore Communities

3.6.52 Table 3.2 and Figure 3.12 of Appendix O show the types of substratum along the horizontal transect at every tidal level in all sampling zones. The relative distribution of different substrata was estimated by categorizing the substratum types (Gravels & Boulders / Sands / Soft mud) of the ten random quadrats along the horizontal transect. The distribution of substratum types varied among tidal levels and sampling zones:

· In TC1, high percentage of ‘Gravels and Boulders’ (70%) was recorded at high tidal level followed by ‘Sands’ (30%). Even distribution of ‘Gravels and Boulders’ (40%) and ‘Sands’ (40%) were recorded at mid tidal level. Relatively, the substratum types of low tidal level were different while higher percentages of ‘Sands’ (60%) and ‘Soft mud’ (30%) were recorded.

· In TC2, the substartum types were recorded evenly at high and mid tidal levels ('Soft mud' 40-50%, 'Gravels and Boulders' 30%, 'Sands' 20-30%,). At low tidal level, the major substratum type was 'Soft mud' (80%) followed by 'Gravels and Boulders' (20%).

· In TC3, high percentages of ‘Sands’ (70-90%) were recorded at high and mid tidal levels followed by ‘Soft mud’ (10-30%). At low tidal level, the major substratum type was ‘Gravels and Boulders’ (80%).

· In ST, ‘Gravels and Boulders’ was the main substratum (100%) at high and mid tidal levels. At low tidal level, the substartum types were mainly ‘Soft mud’ (60%) and 'Sands' (40%).

3.6.53 There was neither consistent vertical nor horizontal zonation pattern of substratum type in all sampling zones. Such heterogeneous variation should be caused by different hydrology (e.g. wave in different direction and intensity) received by the four sampling zones.

3.6.54 Table 3.3 of Appendix O lists the total abundance, density and number of taxon of every phylum in this survey. A total of 12099 individuals were recorded. Mollusca was clearly the most abundant phylum (total abundance 11160 ind., density 372 ind. m^-2, relative abundance 92.2%). The second to fourth abundant phya were Arthropoda (803 ind., 27 ind. m^-2, 6.6%), Annelida (72 ind., 2 ind. m^-2, 0.6%) and Sipuncula (37 ind., 1 ind. m^-2, 0.3%) respectively. Relatively other phyla were very low in abundances (density £1 ind. m^-2, relative abundance £0.2%). Moreover, the most diverse phylum was Mollusca (37 taxa) followed by Arthropoda (14 taxa) and Annelida (6 taxa). There was 1-2 taxa recorded only for other phyla. The taxonomic resolution and complete list of collected specimens are shown in Annexes IV and V of Appendix O respectively.

3.6.55 Table 3.4 of Appendix O shows the number of individual, relative abundance and density of each phylum in every sampling zone. The total abundance (1838-4026 ind.) varied among the four sampling zones while the phyla distributions were similar. In general, Mollusca was the most dominant phylum (no. of individuals: 1720-3638 ind.; relative abundance 89.0-95.2%; density 229-485 ind. m^-2). Other phyla were much lower in number of individuals. Arthropoda was the second abundant phylum (107-402 ind.; 3.7-10.0%; 14-54 ind. m^-2). Annelida was the third abundant phylum in TC2 and TC3 (26-33 ind.; 0.7-1.4%; 3-4 ind. m^-2) and fourth abundant in TC1 (12 ind.; 0.3%; 2 ind. m^-2). Sipuncula was relatively common in TC1 and TC3 (12-16 ind.; 0.3-0.4%; 2 ind. m^-2). Relatively other phyla were low in abundance in all sampling zones (≤ 0.2%).

Dominant species in every sampling zone

3.6.56 Table 3.5 of Appendix O lists the abundant species (relative abundance >10%) in every sampling zone. In the present survey, most of the listed abundant species were of low to moderate densities (50-250 ind. m^-2). Few listed species of high or very high density (> 250 ind. m^-2) were regarded as dominant species. Other listed species of lower density (< 50 ind. m^-2) were regared as common species.

3.6.57 In TC1, the major substratum type was ‘Gravels and Boulders’ at high tidal level. There was dominant gastropod Batillaria multiformis (337 ind. m^-2, relative abundance 63%) followed by gastropods Cerithidea djadjariensis (84 ind. m^-2, 16%) and Cerithidea cingulata (62 ind. m^-2, 11%) at low densities. At mid tidal level (substratum types ‘Gravels and Boulders’ and ‘Sands’), there were few abundant speices at low to moderate densities including gastropods Cerithidea djadjariensis (109 ind. m^-2, 23%), Monodonta labio (81 ind. m^-2, 17%), Batillaria multiformis (59 ind. m^-2, 13%), Cerithidea cingulata (58 ind. m^-2, 12%), as well as rock oyster Saccostrea cucullata (106 ind. m^-2, 22%, attached on boulders). At low tidal level, few intertidal fauna were recorded in the major substratum type ‘Sands’. However abundant rock oyster Saccostrea cucullata (176 ind. m^-2, 29%) and barnacle Balanus amphitrite (115 ind. m^-2, 19 %) were attaching on the boulders.

3.6.58 In TC2, gastropod Cerithidea djadjariensis (172 ind. m^-2, 40%) was abundant at moderate-high density at high tidal level (major substratum types: ‘Soft mud’ and 'Sands') followed by gastropod Cerithidea cingulata (89 ind. m^-2, 21%) and rock oyster Saccostrea cucullata (56 ind. m^-2, 13%, attached on boulders). At mid and low tidal levels (major substratum types: ‘Soft mud’ and 'Gravels and Boulders'), gastropods Cerithidea djadjariensis (32-83 ind. m^-2, 15-25%), Batillaria zonalis (42-63 ind. m^-2, 19-20%) and rock oyster Saccostrea cucullata (61-62 ind. m^-2, 18-29%) were found common at low-moderate densities generally.

3.6.59 In TC3, the major substratum type was ‘Sands’ at both high and mid tidal levels. Gastropod Cerithidea djadjariensis was the dominant species of moderate to high density (256-365 ind. m^-2, 51-58%) followed by another gastropod Cerithidea cingulata (129-163 ind. m^-2, 26%). Besides gastropod Batillaria multiformis (73 ind. m^-2, 12%) was relatively abundant at high tidal level. At low tidal level (major substratum: ‘Gravels and Boulders’), rock oyster Saccostrea cucullata (163 ind. m^-2, 42%) and gastropod Monodonta labio (126 ind. m^-2, 32%) were abundant at moderate densities.

3.6.60 In ST, there was no clearly abundant species at all tidal levels. Gastropod Batillaria multiformis (56 ind. m^-2, 24%), Monodonta labio (46 ind. m^-2, 19%) and rock oyster Saccostrea cucullata (35 ind. m^-2, 15%) were common species at high tidal level (major substratum: ‘Gravels and Boulders’). At mid tidal level (major substratum: ‘Gravels and Boulders’), rock oyster Saccostrea cucullata (117 ind. m^-2, 33%) was abundant at moderate density followed by low-density gastropods Monodonta labio (63 ind. m^-2, 18 %), Lunella coronata (46 ind. m^-2, 13%) and Cerithidea djadjariensis (41 ind. m^-2, 12%). At low tidal level (major substratum types: ‘Sands’ and ‘Soft mud’), there were common taxa only including rock oyster Saccostrea cucullata (28 ind. m^-2, 20%, attached on boulders), barnacle Balanus amphitrite (25 ind. m^-2, 18%, attached on boulders), gastropods Cerithidea djadjariensis (24 ind. m^-2, 17%) and Batillaria zonalis (14 ind. m^-2, 10%).

3.6.61 In general, there was no consistent zonation pattern of species distribution across all sampling zones and tidal levels. The species distribution should be determined by the type of substratum primarily. In general, gastropods Cerithidea djadjariensis (total number of individuals: 2993 ind., relative abundance 24.7%), Cerithidea cingulata (1548 ind., 12.8%), Batillaria multiformis (1443 ind., 11.9%) and Batillaria zonalis (509 ind., 4.2%) were the most commonly occurring species on sandy and soft mud substrata. Rock oyster Saccostrea cucullata (2101 ind., 17.4%), gastropods Monodonta labio (1116 ind., 9.2%), Lunella coronata (323 ind., 2.7%), barnacle Balanus amphitrite (438 ind., 3.6%) were commonly occurring species inhabiting gravel and boulders substratum.

Biodiversity and abundance of soft shore communities

3.6.62 Table 3.7 of Appendix O shows the mean values of species number, density, biodiversity index H’ and species evenness J of soft shore communities at every tidal level and in every sampling zone. As mentioned above, the differences among sampling zones and tidal levels were determined by the major type of substratum primarily.

3.6.63 Among the sampling zones, there was no obvious difference of mean species number, H' and J across all tidal levels. The mean species numbers of TC1, TC2 and ST (9-10 spp. 0.25 m^-2) were slightly higher than TC3 (7 spp. 0.25 m^-2). The mean densities of TC1 and TC3 (509-537 ind. m^-2) were higher than TC2 and ST (245-322 ind. m^-2). Since TC3 was higher in mean density and was highly dominant by few species, the mean H’ (1.1) was relatively lower than other three sampling zones (1.4-1.6). Overall the mean J was similar among the four sampling zones (0.6-0.8).

3.6.64 Across the tidal levels, there was no consistent difference of the mean species number, H' and J in all sampling zones. For the mean density, there were generally decreasing trends in TC2, TC3 and ST from high to low tidal level.

3.6.65 Figures 3.13 to 3.16 of Appendix O show the temporal changes of mean species number, mean density, H’ and J at every tidal level and in every sampling zone along the sampling months. In general, all the biological parameters fluctuated seasonally throughout the monitoring period. Lower mean species number and density were recorded in dry season (Dec.) but the mean H' and J fluctuated within a stable range.

3.6.66 Focusing on the changes of mean density in ST, there were steady decreasing trends regardless of tidal levels since the beginning of monitoring period. It might be an unfavourable change that reflected environmental stresses. The mean densities increased again from Dec. 2016 to Jun. 2017 reflecting a recovery process. But it decreased again in Sep. 2017 (present survey) while the heat stress and serial cyclone hit of this wet season were believed the causes. Because similar decreases of density were noted in other sampling zones either.

Impact of the HKLR project

3.6.67 It was the 20^th survey of the EM&A programme during the construction period. Based on the results, impacts of the HKLR project were not detected on intertidal soft shore community. In case of other abnormal phenomena (e.g. rapid or consistent decline of fauna densities and species number) are observed, it would be reported as soon as possible.

3.7 Solid and Liquid Waste Management Status

3.7.1 The Contractor registered with EPD as a Chemical Waste Producer on 12 July 2012 for the Contract. Sufficient numbers of receptacles were available for general refuse collection and sorting.

3.7.2 The summary of waste flow table is detailed in Appendix K.

3.7.3 The Contractor was reminded that chemical waste containers should be properly treated and stored temporarily in designated chemical waste storage area on site in accordance with the Code of Practice on the Packaging, Labelling and Storage of Chemical Wastes.

Monitoring Activity		Monitoring Date
Monitoring Activity		September 2017	October 2017	November 2017
Air Quality	1-hr TSP	1, 7, 13, 19, 25 and 29	3, 9, 13, 19, 25 and 31	6, 10, 16, 22 and 28
Air Quality	24-hr TSP	6, 12, 18, 22, 28 and 30	6, 12, 18, 24 and 30	3, 9, 15, 22 and 27 for AMS5; 3, 9, 15, 21 and 27 for AMS6
Noise		7, 13, 19 and 25	3, 9, 19, 25 and 31	6, 16, 22 and 28
Water		1, 4, 6, 8, 11, 13, 15, 18, 20, 22, 25, 27 and 29	2, 4, 6, 9, 11, 13, 16, 18, 20, 23, 25, 27 and 30	1, 3, 6, 8, 10, 13, 15, 17, 20, 22, 24, 27 and 29
Chinese White Dolphin		15, 18, 22 and 29	4, 9, 18 and 26	1, 8, 17 and 24
Mudflat Monitoring (Ecology)		2, 3, 6, 16 and 17	--	--
Mudflat Monitoring (Sedimentation rate)		16	--	--
Site Inspection		6, 13, 20 and 29	4, 11, 18, 25 and 31	8, 15, 22 and 28

	North Lantau Social Cluster
	NEL	NWL
Action Level	STG < 70% of baseline & ANI < 70% of baseline	STG < 70% of baseline & ANI < 70% of baseline
Limit Level	STG < 40% of baseline & ANI < 40% of baseline

	North Lantau Social Cluster
	NEL	NWL
Action Level	STG < 4.2 & ANI < 15.5	STG < 6.9 & ANI < 31.3
Limit Level	(STG < 2.4 & ANI < 8.9) and (STG < 3.9 & ANI < 17.9)

Environmental Monitoring	Parameters	Action Level (AL)	Limit Level (LL)
Air Quality	1-hr TSP	0	0
Air Quality	24-hr TSP	0	0
Noise	L_{eq (30 min)}	0	0
Water Quality	Suspended solids level (SS)	17	3
	Turbidity level	0	0
	Dissolved oxygen level (DO)	0	0
Dolphin Monitoring	Quarterly Analysis (Sep 2017 to Nov 2017)	0	1

1 Introduction

1.1.5 This is the twenty-first Quarterly Environmental Monitoring and Audit (EM&A) report for the Contract which summarizes the monitoring results and audit findings of the EM&A programme during the reporting period from 1 September to 30 November 2017.

1.2.1 The project organization structure and lines of communication with respect to the on-site environmental management structure with the key personnel contact names and numbers are shown in Appendix A.

1.3 Construction Programme

1.3.1 A copy of the Contractor’s construction programme is provided in Appendix B.

1.4 Construction Works Undertaken During the Reporting Period

1.4.1 A summary of the construction activities undertaken during this reporting period is shown in Table 1.1. The Works areas of the Contract are showed in Appendix C.

2 EM&A Requirement

2.1 Summary of EM&A Requirements

2.1.1 The EM&A programme requires environmental monitoring of air quality, noise, water quality, dolphin monitoring and mudflat monitoring as specified in the approved EM&A Manual.

2.1.2 A summary of Impact EM&A requirements is presented in Table 2.1. The locations of air quality, noise and water quality monitoring stations are shown as in Figure 2.1. The transect line layout in Northwest and Northeast Lantau Survey Areas is presented in Figure 2.2.

2.2.1 Table 2.2 presents the Action and Limit Levels for the 1-hour TSP, 24-hour TSP and noise level.

2.2.2 The Action and Limit Levels for water quality monitoring are given as in Table 2.3.

2.2.3 The Action and Limit Levels for dolphin monitoring are shown in Tables 2.4 and 2.5.

2.3 Event Action Plans

2.3.1 The Event Actions Plans for air quality, noise, water quality and dolphin monitoring are annexed in Appendix D.

2.4 Mitigation Measures

2.4.1 Environmental mitigation measures for the contract were recommended in the approved EIA Report. Appendix E lists the recommended mitigation measures and the implementation status.

3 Environmental Monitoring and Audit

3.1 Implementation of Environmental Measures

3.1.1 In response to the site audit findings, the Contractor have rectified all observations identified in environmental site inspections undertaken during the reporting period. Details of site audit findings and the corrective actions during the reporting period are presented in Appendix F.

3.1.2 A summary of the Implementation Schedule of Environmental Mitigation Measures (EMIS) is presented in Appendix E.

3.1.3 Regular marine travel route for marine vessels were implemented properly in accordance to the submitted plan and relevant records were kept properly.

3.1.4 Dolphin Watching Plan was implemented during the reporting period. No dolphins inside the silt curtain were observed. The relevant records were kept properly.

3.2.1 The monitoring results for 1-hour TSP and 24-hour TSP are summarized in Tables 3.1 and 3.2 respectively. Detailed impact air quality monitoring results and relevant graphical plots are presented in Appendix G.

3.2.2 No Action Level and Limit Level exceedances of 1-hr TSP and 24-hr TSP were recorded at AMS5 and AMS6 during the reporting period.

3.2.3 Record of notification of environmental quality limit exceedances are provided in Appendix M.

3.3 Noise Monitoring Results

3.3.1 The monitoring results for construction noise are summarized in Table 3.3 and the monitoring results and relevant graphical plots for this reporting period are provided in Appendix H.

3.3.2 No Action and Limit Level exceedances for noise were recorded during daytime on normal weekdays of the reporting period.

3.3.3 Major noise sources during the noise monitoring included construction activities of the Contract and nearby traffic noise and insect noise.

3.4.1 Impact water quality monitoring was conducted at all designated monitoring stations during the reporting period. Impact water quality monitoring results and relevant graphical plots are provided in Appendix I.

3.4.3 Water quality impact sources during the water quality monitoring were the construction activities of the Contract, nearby construction activities by other parties and nearby operating vessels by other parties.

Data Analysis

3.5.9 During the period of September to November 2017, six sets of systematic line-transect vessel surveys were conducted to cover all transect lines in NWL and NEL survey areas twice per month.

3.5.11 The total survey effort conducted on primary lines 578.16 km, while the effort on secondary lines was 223.96 km. Survey effort conducted on both primary and secondary lines were considered as on-effort survey data. A summary table of the survey effort is shown in Annex I of Appendix J.

Distribution

3.5.14 All dolphin sightings were located far away from the HKLR03 and HKBCF reclamation sites as well as along the alignment and Tuen Mun-Chek Lap Kok Link (TMCLKL) (Figure 1 of Appendix J). However, one sighting was made near the alignment of HKLR09 as mentioned above.

Encounter Rate

Group Size

3.5.31 Notably, 8 of these 14 dolphin groups were composed of 1-3 individuals only while there were four medium-sized groups with 5-8 dolphins per group, and one large group of 12 dolphins (Annex II of Appendix J).

Habitat Use

Mother-calf Pairs

3.5.37 During the present quarterly period, no young calf was sighted at all among the 13 groups of dolphins.

Activities and Associations with Fishing Boats

3.5.38 One of the thirteen dolphin groups were engaged in feeding activity, while another two groups were engaged in socializing activity. However, none of them was engaged in traveling or milling/resting activity during the three-month study period.

3.5.41 When compared to the baseline period, distribution of various dolphin activities during the present impact phase monitoring period was drastically different with a much more restricted area of occurrences (Figure 6 of Appendix J).

3.5.42 Notably, one group of a single dolphin was found to be associated with an operating purse-seiner adjacent to Lung Kwu Chau within the marine park during the present impact phase period.

Summary Photo-identification works

3.5.43 From September to November 2017, over 2,500 digital photographs of Chinese White Dolphins were taken during the impact phase monitoring surveys for the photo-identification work.

Individual range use

3.5.46 Ranging patterns of the 23 individuals identified during the three-month study period were determined by fixed kernel method, and are shown in Annex V of Appendix J.

3.5.48 On the other hand, three individuals (i.e. NL12, NL182 and NL210) consistently utilized North Lantau waters in the past have extended their range use to WL during the present quarter.

Action Level / Limit Level Exceedance

3.5.61 It was concluded that the HZMB works is one of the contributing factors affecting the dolphins. It was also concluded the contribution of impacts due to the HZMB works as a whole (or individual marine contracts) cannot be quantified nor separate from the other stress factors.

3.5.62 The dolphin specialists of the projects confirmed that the CWD sighting around the North of Sha Chau and Lung Kwu Chau Marine Park (SCLKCMP) has significantly decreased, and it was apparently related to the re-routing of high speed ferry (HSF) from Skypier.

3.5.63 It was reminded that the ETs shall keep reviewing the implementation status of the dolphin related mitigation measures and remind the contractor to ensure the relevant measures were fully implemented.

3.5.64 It was recommended that the marine works of HZMB projects should be completed as soon as possible so as to reduce the overall duration of impacts and allow the dolphins population to recover as early as possible.

3.5.65 It was also recommended that the marine works footprint (e.g., reduce the size of peripheral silt curtain) and vessels for the marine works should be reduced as much as possible, and vessels idling / mooring in other part of the North Lantau shall be avoided whenever possible.

3.5.67 There was a discussion on exploring possible further mitigation measures, for example, controlling the underwater noise. It was noted that the EIA reports for the projects suggested several mitigation measures, all of which have been implemented.

3.6 Mudflat Monitoring Results

Sedimentation Rate Monitoring

3.6.2 This measurement result was generally and relatively higher than the baseline measurement at S1, S2, S3 and S4. The mudflat level is continuously increased.

Water Quality Monitoring

3.6.3 The mudflat monitoring covered water quality monitoring data. Reference was made to the water quality monitoring data of the representative water quality monitoring station (i.e. SR3) as in the EM&A Manual. The water quality monitoring location (SR3) is shown in Figure 2.1.

3.6.4 Impact water quality monitoring in San Tau (monitoring station SR3) was conducted in September 2017. The monitoring parameters included dissolved oxygen (DO), turbidity and suspended solids (SS).

3.6.5 The Impact monitoring result for SR3 were extracted and summarised below:

Sampling Zone

Horseshoe Crabs

Seagrass Beds

Intertidal Soft Shore Communities

3.6.13 All collected fauna were released after recording except some tiny individuals that are too small to be identified on site. These tiny individuals were taken to laboratory for identification under dissecting microscope.

3.6.14 The taxonomic classification was conducted in accordance to the following references: Polychaetes: Fauchald (1977), Yang and Sun (1988); Arthropods: Dai and Yang (1991), Dong (1991); Mollusks: Chan and Caley (2003), Qi (2004).

Data Analysis

3.6.15 Data collected from direct search and core sampling was pooled in every quadrat for data analysis. Shannon-Weaver Diversity Index (H’) and Pielou’s Species Evenness (J) were calculated for every quadrat using the formulae below,

Mudflat Ecology Monitoring Results and Conclusion