Utilization of Marine Geospatial Data for Determining Optimal FAD Locations in Improving the Living Standards of the North Gorontalo Coastal Community

Abstract

The use of fishing aids called Fish Aggregating Devices (FADs) has become a polemic in the context of realizing sustainable fisheries management. On the one hand, the use of FADs can increase fishermen’s catches. However, on the other hand, the use of FADs without proper management will result in unsustainable fish availability due to overfishing. The FADs used are also not always correlated with increased catches. At certain times, FADs used by fishermen can be lost because they are carried away by the current, which is clearly very detrimental to both fishermen and the environment. Utilization of geospatial information is one of the efforts to connect existing policies regarding the use of FADs, safety aspects, and optimizing the number of fishermen’s catches. The main focus of this study is the utilization of geospatial information such as the modeling of ocean currents, analysis of potential fishing zones, analysis of sea depth positions, and analysis of prohibited zones for the placement of FADs. The model and policy analysis results are used to determine recommendations for the optimal placement of FADs in the northern sea of the province of Gorontalo. In this study, an analysis of the influence of FADs on the economy of the community, especially the coastal community of North Gorontalo Regency, was also carried out. Recommended areas for FAD placement are divided into class 1 (low recommendation), class 2 (medium recommendation), and class 3 (high recommendation). The results of the recommendation for FAD placement areas resulted in 5 FADs being in the level 1 area, 27 FADs being in the level 2 area, and 6 FADs being in the level 3 area. Installed FADs cannot be moved to a different location because the installation is permanent. The findings of this study will help those who install new FADs to decide on a secure site for their installation. By paying attention to the shape of FADs, fishing gear, and the conditions of placement of existing FADs, it can be analyzed that the main catch targets of fishermen are pelagic (pelagic) fish, which is evidenced by the catches of fishermen in North Gorontalo Regency, which are dominated by large pelagic fish and small pelagic fish with a total catch of 21,535,604 kg. in 2021. Thus, it can be inferred that FADs have a significant role in supporting the economy that revolves around fishers, especially in North Gorontalo Regency, and with the stock of pelagic fish, which is still relatively high, it is possible to support the economy of the people of North Gorontalo Regency in the long term.

1. Introduction

The Sulawesi Sea is part of Fisheries Management Area (WPP) 716, which is listed in the Minister of Marine Affairs and Fisheries Regulation 18/PERMEN-KP/2014. In addition to the Sulawesi Sea, the geographical area of WPP 716 covers the north of Halmahera Island and borders the territory of three friendly countries, namely Malaysia, the Philippines, and Palau [1]. Geographically, the position of FMA 716 is depicted in Figure 1.

WPP 716 has abundant fishery potential, so it has the potential for catching economic fish and fishing for fish on a small scale as well as on an industrial scale. The fish found in WPP 716 include small pelagic (pelagic) fish, large pelagic (pelagic) fish, demersal (demergere) fish, shrimp, and other crustaceans [1]. The estimated state and utilization of fishery resources in the areas included within Fishery Management Areas (FMA) 716 is explained in

Table 1. The estimated state and utilization of fishery resources in the areas included within Fishery Management Areas (FMA) 716.

No	Fish Type	Estimated Potential (Ton)	Utilization Level	Stock Status
1	Demersal (demergere)	34.650	0.49	High
2	Reef Fish	54.194	1.11	Low
3	Small Pelagic (pelagic)	222.946	0.49	High
4	Squid	1.103	1.40	Low
5	Large Pelagic (pelagic)	154.329	0.74	Medium
6	Penaeid Shrimp	8.465	0.75	Medium
7	Lobster	685.000	1.02	Low
8	Crab	1.969	0.94	Medium
9	Rajungan	424.000	1.09	Low

Source: [1] Adapted from Nasution et.al (2019).

below.

In the waters of the Sulawesi Sea, there are a number of tools commonly referred to by the local community as FADs. Using these FADs, the fish is aggregated into a concentrated area, so that it will be easier for the fishers to harvest the fish[NO_PRINTED_FORM] [2]. This fishing tool offers an alternative solution to overcome the location uncertainty for fishing in both shallow and deep waters. Based on research on the effectiveness of installing FADs in Tomini Bay, it was found that the use of FADs was not efficient at increasing fishing yields [3].

Fish Aggregating Devices (FADs), or rumpon, the term that is used in the local fishing communities in Gorontalo, refers to multiple types of devices that are aimed at binding and attracting fish into a relatively concentrated area. FADs are aimed at increasing and enhancing fishery operations at sea and are regulated by the Decree of the Minister of Maritime and Fishery Affairs of Indonesia No. 26/2014 regarding FADs.

These conventional surface FADs have a straightforward design and structure and are typically constructed of organic materials such as [4]:

• A raft-shaped float constructed from bamboo.
• For fisherman from Java and Madura, the anchor rope is constructed of palm fiber, although rattan is also an option (for Sulawesi fishermen). Nowadays, synthetic rope is commonly used.
• The lure (attractor) makes use of thatch, palm leaves, and fronds of coconut leaves.
• Stone ballast with iron or wooden anchors that is strung together.

The shape of a typical FAD is illustrated in the following Figure 2.

The FADs deployed by fishers in Gorontalo typically adhere to the following criteria:

• The type of lines that are used are of Polyethilene (PE) type with sizes 21, 22, or 23 (diameter in mm) [5].
• The length of the line is adjusted to the depth of the water column in which the FADs are deployed (In the Sulawesi Sea, between 2281 and 4195 m), so as to give them tolerance and resistance to oceanic currents. The top and bottom 100 m of the line are given an extra layer of protection consisting of rubber hoses to protect the lines from friction with the water.
• Buoys are improvised by combining several unused metal drums so that they float.
• Anchors are improvised by submerging unused metal drums that have been filled with concrete.
• The fish are attracted using the leaves of Nypa fruticans (mangrove palm/nipah palm).

The usage of FADs presents several issues from the standpoint of policymakers. The function of FADs increases the possibility of increasing fishery yields, which in turn leads to the unchecked growth of FAD installations. Because in practice many FADs are not registered, their management is ineffective owing to differences between the data and the actual numbers in the field. Unsustainable fish stocks are a result of the possibility of overfishing due to the largely ungoverned number of FADs. The potential for overfishing as a result of a large number of FADs causes unsustainable fish stocks. FADs have been regulated in PERMEN-KP Number 26/2014 concerning FADs, related to the number of FADs a fisherman can own, which is limited to a maximum of three, attractors being made from easily biodegradable materials, and the distance between FADs being 10 nautical miles. In fact, in the field, only the attractors are filled. The rules in place are largely disobeyed due to the lack of understanding of fishermen about the negative impacts of loosely regulated FADs, lack of supervision from the corresponding authorities, and the lack of enforcement of penalties for offenders [6].

For the fishermen themselves, having a large number of FADs does not guarantee large fish catches. Especially when the ″knife current″ phenomenon occurs, FADs belonging to fishermen can simply disappear into the current, causing huge losses [7]. The geographic location of the Celebes Sea coincides with the pattern of oceanic currents moving from the Northern Pacific Ocean to the Indian Ocean, internationally recognized as the Trans-Indonesian Current (Arus Laut Indonesia or Arlindo for short, in Indonesian). This current brings turbulence, sinking, upwelling, and downwelling to the waters along the coast of Northern Gorontalo. In the Celebes Sea, this current is immune to the effects caused by monsoonal wind patterns. The scheme of current circulations in the Celebes Sea is illustrated in Figure 3 below:

The Trans-Indonesian Current (Arlindo), which crosses Indonesian waters with a transport volume of approximately 11 × 106 m³ s⁻¹, is a sea current that travels through Indonesia from the Pacific Ocean to the Indian Ocean [9,10]. The Trans-Indonesian Current flow, an inter-basin current system connecting the Pacific and Indian oceans, demonstrates that the sea level on the Pacific side is almost always higher than the sea level on the Indian side [11,12]. The pattern of the Trans-Indonesian Current in the Celebes Sea is best described with four main characteristics, such as:

• The main axis of the Trans-Indonesian Current from the site of the inflow to the outflow, which is indicated by the strong current vector heading west. The depth of the maximum magnitude of the current is at an estimated 150 to 250 m below the sea surface.
• Recirculation of the current’s vortex in a clockwise manner in the northern edge of the Trans-Indonesian Current.
• A vortex that rotates in a counterclockwise manner [8]

The greater than average magnitude of turbulence in the Celebes Sea, which is indicated by the large value of vertical eddy diffusion because of a strong interaction between the moving masses of water and the seabed [13].

One of the risks of employing FADs is the formation of the knife current phenomena, which is caused by the Trans-Indonesian Current phenomenon and can alter the character of the currents that occur in the Sulawesi Sea, particularly in the research region. Referring to these conditions, the existing regulations have not answered fishermen’s needs, so many violations occur. Therefore, the problem of FAD placement needs to be investigated further so that its use is optimal.

2. Materials and Methods

2.1. Method

The method used in this study is the inductive method, which combines and uses data acquired directly on-site to be analyzed and used as a basis to formulate solutions to the problem [14]. Referring to the data that are used in this study, the methods that function to break down the problem are specifically the spatial and ex-ante analysis.

Referring to the Decree of the Minister of Maritime and Fishery Affairs of Indonesia No. 26/2014 regarding FADs, it is stated that they must be located at least 10 nautical miles from each other so as not to disrupt the navigation channels and shipping routes used by vessels out on the sea. This approach can be developed by adding a parameter that is used in an effort to improve fishing yields. Other probable parameters are adding locations or sites that may inflict damage or carry the FADs away, compliance with the Zoning Masterplan for Coastal Areas and Small Islands (RZWP3K) of Gorontalo Province, or possible overlapping with undersea infrastructures such as telecommunication wiring or pipelines.

If the FADs happen to be situated in a restricted area, they must be relocated by also adhering to the 10 nautical mile safe distance regulation. The parameters of objects that must be avoided in establishing and deploying a FAD will be mapped throughout the Celebes Sea to provide clearer information. Fishermen, as the possible users of this information, will also obtain more in-depth guidance regarding improving fishing yields by deploying FADs, while also complying with the regulations that are in place. Following the processing of data, an overlay is used to determine the classification of regions where it is safe to place FADs.

2.2. Data

The data used in this study consist of spatial data and other supporting data that will be elaborated further in this section as follows.

2.2.1. Locations of Existing FADs in Gorontalo Province Northern Coastline

The position of the existing FADs was identified and detected from an on-site survey carried out by The Marine and Fisheries Department of Gorontalo Province in 2021. The spatial distribution as well as the coordinates of the FADs are summarized in

Table 2. The spatial distribution and the coordinates of the FADs.

No	Name	Latitude	Longitude	Depth	No	Name	Latitude	Longitude	Depth
1	AWAL	1° 41′ 41.3″	122° 48′ 40.6″	3941	20	UMI	1° 23′ 5.8″	122° 39′ 39.5″	2959
2	DUA	1° 24′ 13.4″	122° 23′ 55.1″	2960	21	UMI	1° 30′ 25.5″	122° 48′ 8.6″	3689
3	BANG	1° 28′ 8.1″	122° 8′ 26.6″	3991	22	46MANT	1° 21′ 57.7″	122° 30′ 45.1″	3199
4	BRAVO	1° 35′ 28.9″	122° 40′ 30.6″	3594	23	73GTAT	1° 41′ 40.6″	122° 57′ 26.2″	4162
5	MARKAS	1° 33′ 43.5″	122° 34′ 1.3″	3480	24	B3YSNT	1° 46′ 34.4″	122° 56′ 55.7″	4187
6	NANGGALA54	1° 25′ 37.2″	122° 19′ 8.8″	2837	25	49MANT	1° 19′ 26.4″	122° 33′ 0.1″	3048
7	AWAL	1° 25′ 13.2″	122° 22′ 49.1″	2898	26	02NIKT	1° 27′ 54.2″	122° 27′ 32.6″	3088
8	SASY	1° 40′ 5.5″	123° 3′ 18.7″	°4082	27	33SLBT	1° 30′ 3.4″	122° 29′ 6.6″	3127
9	MNDALA	1° 21′ 45.5″	122° 27′ 29.3″	3023	28	06CEMT	1° 34′ 52.7″	123° 7′ 59.8″	4195
10	MNDALA	1° 19′ 55.7″	122° 24′ 0.1″	2891	29	02MOPT	1° 21′ 11.3″	123° 13′ 25.7″	2281
11	DHEO	1° 25′ 17.5″	122° 22′ 16.1″	2893	30	ARHIMT	1° 20′ 10.8″	123° 10′ 23.6″	1888
12	86	1° 27′ 7.2″	122° 8′ 25.7″	2326	31	77RAIT	1° 18′ 29.9″	122° 34′ 59″	2989
13	ALVA	1° 25′ 40.1″	122° 12′ 4.6″	2512	32	JRAR45	1° 13′ 7.2″	122° 27′ 38.6″	1744
14	UMI	1° 30′ 25.5″	122° 48′ 8.6″	3689	33	JKAR49	1° 15′ 48.3″	122° 18′ 37.8″	2387
15	1PTRABRT	1° 18′ 53.2″	122° 44′ 35.7″	2208	34	JKRS93	1° 16′ 48.3″	122° 13′ 14.9″	2617
16	2A	1° 16′ 49.2″	122° 33′ 57.2″	2963	35	2MPORT	1° 16′ 51.6″	122° 13′ 16.6″	2608
17	85BOBRA	1° 27′ 34.0″	122° 37′ 0.8″	3456	36	2MPORT	1° 17′ 6.6″	122° 12′ 30.3″	2632
18	UMI	1° 29′ 15.7″	122° 39′ 25.9″	3605	37	NPAGOB	1° 3′ 22.9″	123° 25′ 4.9″	208
19	UMI	1° 22′ 48.2″	122° 41′ 54.5″	2833	38	BTDULK	1° 1′ 9.9″	122° 35′ 49.8″	14

Source: UPT PPI Gentuma Port, North Gorontalo.

.

2.2.2. Current Modeling of Gorontalo Province Northern Coastline

The current modeling in this study uses the two-dimensional hydrodynamic approach as one of the parameters in determining the optimal and safe locations for FAD deployment. The sampling approach used in this study, which is quantitative, is used to choose the sampling locations. The study of the two-dimensional hydrodynamic approach is carried out using the MIKE 21 flow model with a flexible mesh domain. The model is simulated in all model domains to determine the direction and velocity of oceanic currents in the waters of Northern Gorontalo. The simulation is set to a lapsed run time of 1 year with a time step of 1 h. The hydrodynamic models will be constructed using MIKE 21 with the Flow Model FM Module to identify the patterns of current movement based on the data obtained [15]. The input data used during the modeling process are summarized in

Table 3. Flow modeling data.

No	Data	Source	Remarks
1	Bathymetric data	GEBCO	30 arc seconds resolution
2	Tidal data	Prediction of tide elevation MIKE 21 Toolbox	Per 1 h for 1 year observation period
3	Wind data	Wind velocity reanalysis- ECMWF	Per 3 h for 1 year observation period

.

The equation of the model used in this study refers to the DHI module with consideration of the momentum generated by the mass of the water body, as formulated in the following equation [16].

Continuity equation:

$$\frac{\partial \zeta }{\partial t}+\frac{\partial p}{\partial x}+\frac{\partial q}{\partial y}=\frac{\partial d}{\partial t}$$

(1)

The momentum of the mass of water at the X-axis:

$$\frac{\partial p}{\partial t}+\frac{\partial }{\partial x}\left(\frac{p^2}{h}\right)+\frac{\partial }{\partial y}\left(\frac{pq}{h}\right)+gh\frac{\partial \zeta }{\partial t}+\frac{gp\sqrt{p^2+q^2}}{C^2 h^2}-\frac{1}{{\rho }_w}\left[\frac{\partial }{\partial x}\left(h{\tau }_{xx}\right)+\frac{\partial }{\partial y}\left(h{\tau }_{xy}\right)\right]-\mathsf{\Omega }p-{\displaystyle \int } V{V}_x + \frac{h}{{\rho }_w} \frac{\partial }{\partial x}\left(P_a\right) = 0$$

(2)

The momentum of the mass of water at the Y-axis:

$$\frac{\partial p}{\partial t}+\frac{\partial }{\partial y}\left(\frac{q^2}{h}\right)+\frac{\partial }{\partial x}\left(\frac{pq}{h}\right)+gh\frac{\partial \zeta }{\partial y}+\frac{gp\sqrt{p^2+q^2}}{C^2 h^2}-\frac{1}{{\rho }_w}\left[\frac{\partial }{\partial y}\left(h{\tau }_{yy}\right)+\frac{\partial }{\partial x}\left(h{\tau }_{xy}\right)\right]-\mathsf{\Omega }p-{\displaystyle \int } V{V}_y + \frac{h}{{\rho }_w} \frac{\partial }{\partial y}\left(P_a\right)= 0$$

(3)

Given:

h(x,y,t)	=	Depth (ζ-d,m)
d(x,y,t)	=	Variation of depth towards time (m)
ζ(x,y,t)	=	Height of the water surface (m)
p,q(x,y,t)	=	Flux directions at the x and y axis (m3/s/m) = (uh,vh); (u,v) at average depth and velocity alongside the x and y axis
C(x,y)	=	Chezy resistance coefficient (m1/2/s)
g	=	Gravitational pull (m2/s)
f(V)	=	Wind friction factor
V,Vx, Vy(x,y,t)	=	Components of velocity at the x and y axis (m/s)
Ω(x,y)	=	Latitude-dependent Coriolis factor (s−1)
Pa	=	Atmospheric pressure (kg/m/s2)
ρw	=	Density of water (kg/m3)
x,y	=	Coordinates in space (m)
t	=	Time (s)
τxx, τxy, τyy	=	Effective shifting tension component

The results of the current modeling process are then visualized by averaging the current velocity throughout a year, visualized, and compared with the results and findings of the validation survey carried out directly on-site to determine how the model differs from real-world conditions. This test is carried out with the objective of knowing the errors and accuracy of the generated model [17]. The equation used in this stage is described as follows:

$$RMS difference=\sqrt{\frac{({\sum }_{i=0}^n\left(x_i-\overline{x}\right)}{N}}$$

(4)

Remarks:

$x$	=	Observed value
$\overline{x}$	=	Model-generated value
$N$	=	Number of data points

2.2.3. Determining the Distribution of “Knife Currents”

The spatial distribution of the knife currents is carried out using the Peak Over Threshold (POT) method using Microsoft Excel software. The input data used in this stage are the current velocity generated from the modeling process. The POT method takes all values that are above a certain threshold. Values above the set threshold are considered to be extreme points. An easier method would be the 10% quantile method. Even though this method is more practical, the threshold determination outputs are accurate. The 10% quantile method based on previous studies by Chavez and Embrechts (2002) [18] are as follows:

• Arranging data points of observation from the biggest to the smallest.
• Counting how many data points exceed the threshold with the formula n = 10% × N. N is the total number of observed data points. Data that is in the order of 1 to extreme data.
• Determining the value of the threshold (u) with the formula of u = n + 1. Therefore, the data that is at the (n + 1) sequence is the threshold value (u).

The obtained threshold value is the minimum velocity of the knife current. This study uses that value to determine when and where knife currents appear or occur.

2.2.4. Zona Potensi Perikanan Indonesia (ZPPI) or Indonesian Fishery Potential Zones

The Indonesian Fishery Potential Zones (ZPPI) are a group of areas in Indonesian territorial waters that are highly likely to contain agglomerations of fish, and are a potential optimal spot for fishery-related activities. The potential spots are detected through three parameters, namely thermal front, chlorophyll-A distribution, and the presence of upwelling. Thermal front and upwelling can be detected using remote sensing satellite imagery that provides information regarding sea surface temperature [19].

The distribution of chlorophyll in the sea is obtained from Aqua-MODIS satellite imagery with a spatial resolution of 1 km. The imagery is then processed using the standard MODIS OC-3 algorithm and is verified using direct survey carried out on-site [20]. These activities are performed by the Indonesian Aeronautical and Space Agency (Lembaga Antariksa dan Penerbangan Nasional/LAPAN). The existence of chlorophyll in a certain part of the sea is one of the input parameters to determine Indonesian Fishery Potential Zones (ZPPI)which are issued monthly by LAPAN to assist fishers around the nation in improving their fishing yields. Figure 4 below is a ZPPI map of seas in Northern Gorontalo for May 2021.

The month of May 2021 proved to be the month with the highest number of knife current occurrences based on the data processing carried out in this study. The method of data processing used in this study is the POT method, as elaborated further in Section 2.2.3, which inputs the current velocity data. The threshold is set at the 10% quantile. The current data is processed on a monthly basis. The results show that May is the month with the highest number of knife current occurrences, with a total of 24. For this reason, the month of May 2021 is designated as the sample month in analyzing the effects of the knife current on the FADs deployed in Northern Gorontalo.

2.2.5. Zoning Masterplan for Coastal Areas and Small Islands (RZWP3K) Map

The placement of FADs must take into account the function of the water bodies that are stated in the Zoning Masterplan for Coastal Zones and Small Islands. According to Article 1, Section 19 of The Gorontalo Provincial Law No. 4/2018 concerning RZWP-3-K, the Zoning Masterplan for Coastal Areas and Small Islands, hereinafter abbreviated as RZWP-3-K, is a plan that defines the purpose and intended long-term use cases for bodies of water. The regulation also sets out structures and spatial patterns within the area by clearly stating activities that may and may not be carried out as well as activities that can only be carried out after obtaining official permits from government officials in the said area. The RZWP-3-K map is shown in Figure 5.

3. Results

3.1. Analysis of Current FAD Distribution

The current spatial distribution of FADs was obtained from the authorities at Gentuma Port, North Gorontalo Regency, which was last updated from a 2021 survey. Figure 6 below is the distribution of FADs in North Gorontalo in 2021.

The following table summarizes the spatial distribution of existing FADs situated in the waters of Northern Gorontalo and is sorted based on their depth: refer to Table 2.

The variation in size and depth of the multiple FADs indicates that they have different preferred fishing targets. On the northern coast of Gorontalo, most FADs are categorized as deep-water FADs, which are situated at a range of 1200 to 4000 m below the sea surface in order to attract large pelagic (pelagic) fish such as tuna (Thunnus sp.) and cakalang (Katsuwonus pelamis). These types of fish normally dwell in depths of more than 60 m below the sea surface [22].

3.2. Results of Current Modeling

The current modeling in the waters in Northern Gorontalo is carried out by averaging the current velocity for an observation period of 1 year. The results are visualized in Figure 7.

The modeling results are then validated by comparing them with the results from the on-site survey. The on-site survey measures the oceanic currents using a current meter in two sample locations to represent the modeling process and at two different observation epochs, with 1 h as the minimum interval between the two. The data acquisition is carried out throughout a single day using one instrument, which is the current meter. This limits the number of sample sites where data are collected. Therefore, a few points are designated as sample points to represent the greater scope of interest of this study. The data are acquired in points of varying depths and latitudes. The sample points for current profiling are visualized in the map in Figure 8.

The results of the on-site validation survey are described in

Table 4. Data acquired from on-site Validation.

	Point	Time	Coordinates		Device Depth (m)	Current Velocity (m/s)	Current Heading (°)
			Latitude	Longitude
Outbound	1	09.45	1° 0′ 1.4″	123° 0′ 11.9″	10.31	0.18	336.0
		09.50	1° 0′ 8.6″	123° 0′ 7.6″	10.44	0.16	334.8
	2	10.45	1° 5′ 57.2″	123° 0′ 10.4″	10.73	0.03	280.8
		10.54	1° 6′ 0.8″	123° 0′ 7.9″	10.85	0.10	305.1
	3	12.01	1° 12′ 4.8″	123° 0′ 5.4″	10.55	0.10	344.5
		12.09	1° 12′ 3.7″	123° 0′ 3.6″	10.85	0.09	314.5
Inbound	2	13.14	1° 5′ 53.5″	123° 0′ 4.7″	10.09	0.08	385.7
		13.20	1° 5′ 52.8″	123° 0′ 4.3″	10.48	0.08	373.2
	1	14.22	0° 59′ 56.6″	123° 0′ 8.6″	9.70	0.11	306.6
		14.28	0° 59′ 56.8″	123° 0′ 8.6″	9.92	0.08	303.7

.

The amplitude of the current velocity is the biggest at or near the surface and gradually weakens as it gets deeper down the water column. Therefore, the on-site data acquisition is carried out at a depth that has the highest potential of sea current velocity, which is around 15 m (mixed layer zone). Maximum current velocity potential exists in the mixed layer because the water temperature in this layer is greater than that of the thermocline layer and the wind increases current velocity [21].

At a depth of approximately 10 m, the current measurement results indicate the influence of latitudes towards current velocity. Current velocity increases as the points are situated further to the north. This is caused by the Trans-Indonesian Current that flows north of Gorontalo Province, heading into the Celebes Sea. Therefore, the further north the points are, the stronger the influence it acquires from the Trans-Indonesian Current. In addition, the higher current velocity also diminishes the influence of friction with the seabed[NO_PRINTED_FORM] [8]. The data obtained from the on-site survey are processed further, and the results of the data processing are listed in

Table 5. Processing for data acquired from the on-site survey.

Time	Point	Validation
		Coordinates		Device Depth (m)	Current Velocity (m/s)	Current Heading (°)
		Latitude	Longitude
09.00	1	1° 00′ 05.2″	123° 00′ 10.9″	-	0.02	335.4
10.00	2	1° 06′ 04.3″	123° 00′ 10.9″	-	0.06	293.0
12.00	3	1° 12′ 11.2″	123° 00′ 03.1″	-	0.09	329.5
13.00	2	1° 06′ 04.3″	123° 00′ 10.9″	-	0.08	379.5
14.00	1	1° 0 0′ 05.2″	123° 00′ 10.9″	-	0.10	305.2

.

The results that are acquired on-site are then analyzed to determine the quantitative difference between the results obtained via modeling and the ones that are obtained during the on-site survey. The results of the analysis stage are described in

Table 6. Current velocity and direction validation.

Model					Velocity Difference (m/s)	Heading Difference (°)
Coordinates		Device Depth (m)	Current Velocity (m/s)	Current Heading (°)
Latitude	Longitude
1° 00′ 05.2″	123° 00′ 10.9″	-	0.05	329.02	0.03	−6.4
1° 06′ 04.3″	123° 00′ 10.9″	-	0.07	330.61	0.01	37.7
1° 12′ 11.2″	123° 00′ 03.1″	-	0.17	351.30	0.08	21.8
1° 06′ 04.3″	123° 00′ 10.9″	-	0.06	357.89	−0.03	−21.6
1° 00′ 05.2″	123° 00′ 10.9″	-	0.04	353.36	−0.05	48.2
−					−0.05	−27.9
+					0.09	107.7
Σ Difference					0.14	135.6
Average Difference					0.03	27.1
RMS of Difference					0.16	5.2

.

From the results of the validation, it can be inferred that the average difference between the velocity figure obtained from the modeling and the on-site survey is 0.03 m/s. Meanwhile, the average difference between the current heading obtained from the modeling and the on-site survey is 27.1 degrees. This shows that the modeling yields optimal results for velocity modeling, but they are not adequate enough for current heading modeling, which is caused by the instability of the vessel used during the survey, inevitable due to conditions out on the sea. The significant discrepancy might also be caused by the process in the modeling that averages heading and velocity in a uniform manner for every depth data point, where it does not consider the effect of real-world external factors such as wind and wind-generated waves that apply to the on-site survey. Another factor contributing to the discrepancy between the modeling and the on-site survey results is the fact that the areas that are surveyed on-site do not represent the whole spatial distribution of FADs that are modeled off-site. A single pixel from the modeling results represents 0.1° × 0.1°, roughly equaling 11.1 km × 11.1 km of the sea surface in the real world. In general, the current modeling results can be considered good enough and do not require a process of re-calibration.

3.3. Compliance of Existing FADs with RZWP3K

Figure 9 illustrates the compliance of existing FADs that are deployed on the northern coast of Gorontalo with the regulations set by the Gorontalo Regional Law No. 4/2018 concerning the Zoning Masterplan for Coastal Areas and Small Islands or RZWP3K.

The Zoning Masterplan for Coastal Areas and Small Islands Rencana Zonasi Wilayah Pesisir dan Pulau-Pulau Kecil (RZWP3K) is one of the legal instruments set by the Indonesian government to regulate and govern the usage of spaces above surfaces of water. The provincial governments of Indonesia have the right to exercise supervision and governance on the utilization and development of coastal and maritime areas. Out of several types of space utilization zoning, there are some types in which no other utilization or activities apart from the intended purpose are allowed. Examples of these zoning types are maritime conservation zones, P3K1 conservation zones, and navigation channels for the purpose of seafaring vessels. In the map above, it can be seen that no FADs on the northern coast of Gorontalo interfere with shipping navigation channels or conservation zones. Hence, it can be inferred that the deployment of FADs in Northern Gorontalo is compliant with regulations set by the provincial government.

3.4. Suitability of Sites of Existing FADs on the Northern Coast of Gorontalo with The Condition of Sea Currents

The suitability between existing FADs on the northern coast of Gorontalo and the conditions of sea currents in the Celebes Sea is determined by creating an overlay between the spatial distribution of the FADs and the monthly average velocity of sea currents. From the overlay, further analysis is conducted to determine the percentage of FADs that are affected by the knife currents. The results of the analysis are stated in

Table 7. Percentage of FADs affected by the knife current.

Month	Number of Affected FADs	Percentage of Affected FADs
January	3	7%
February	0	0%
March	4	10%
April	5	13%
May	5	13%
June	4	10%
July	2	5%
August	1	2%
September	2	5%
October	2	5%
November	3	7%
December	2	5%

.

The results of the spatial distribution of knife currents use the current velocity data generated by the current modeling in the form of a threshold value, which defines the minimum velocity at which a current can be considered a knife current. The threshold obtained is 0.48 m/s, which equals the value of the highest ten percent average current velocity on the northern coast of Gorontalo Province. From Table 7 it can be seen that the months of April and May have the highest percentage of affected FADs relative to other months throughout the year. The following image shows the spatial distribution of knife currents and FADs for the months of April and May. The knife current velocity value is used to predict or gather information on when and where the knife current appears. Therefore, a map of the historical occurrences of knife currents in the waters of the Celebes Sea was made, as shown in Figure 10.

The results of the current modeling align with the statements gathered from fishermen via interviews that were conducted directly on-site, from which it can be concluded that the month of May has the strongest influence on knife currents relative to other parts of the year.

3.5. Suitability of Sites of Existing FADs on the Northern Coast of Gorontalo with Indonesian Fishery Potential Zones (ZPPI)

The suitability of the spatial distribution of the existing FADs between the ZPPI (Indonesian Fishery Potential Zones) data in the Celebes Sea is determined by creating an overlay between the distribution of the FADs and the monthly ZPPI data issued by LAPAN (Indonesian Aeronautical and Space Agency). After the overlay, it is further analyzed to determine the percentage of FADs that are within a 1 km radius of a fishing hotspot issued in the ZPPI. The results of this analysis are stated in

Table 8. Percentage of FADs adjacent to ZPPI-based fishing hotspots.

Month	Number of FADs Adjacent to ZPPI Hotspots	Percentage of FADs Adjacent to ZPPI Hotspots
January	0	0%
February	1	2%
March	1	2%
April	0	0%
May	3	7%
June	2	5%
July	2	5%
August	0	0%
September	1	2%
October	1	2%
November	0	0%
December	0	0%

.

Figure 11 is a map that illustrates the spatial distribution of existing FADs on the northern coast of Gorontalo relative to the fishing hotspots generated by the ZPPI during the month of May.

3.6. Recommendations for the Most Suitable Sites for FAD Deployment

The multiple overlay-based analysis that has been carried out in this study is aimed at determining the sites most suitable for FAD deployment on the northern coast of Gorontalo. Those sites are visualized in the form of a map in Figure 12.

From the results of the overlay process, it can be inferred that there are five FADs that are situated in areas categorized as hazardous. The five FADs should be a concern of the local authorities to prevent further losses caused by knife currents. In addition, the overlay process also determines areas that are considered very suitable for FAD deployment. These areas satisfy three criteria: free of known knife currents, provide year-round fishery hotspots, and are well within the territorial waters of Gorontalo Province. The map above also illustrates some FADs that are category-wise still in areas designated as safe but might need extra caution due to the effects caused by the knife currents.

4. Discussion

4.1. Analysis of the Effects of FAD Deployment from an Economic Standpoint

In coastal areas, the welfare of the fishermen’s community greatly depends on the fishery yields within a period of time. Based on data issued by the Fisheries Department of Gorontalo Province, in the year 2021, the North Gorontalo District produced a fishery yield that is dominated by the pelagic (pelagic) type of fish. The total 2021 yields amounted to 21,535,604 kg or 83.37% of the gross tonnage of the caught fisheries sector as a whole. The fish species with the highest yields are namely the tuna skipjack (Katsuwonus pelamis) at 19.03%, Indian Scad (Selaroides leptolepis) at 11.92%, tongkol komo (Euthynnus affinis) at 8.94%, yellowfin tuna (Thunnus albacares) at 7.18%, and mackerels (Mackerel) at 3.81%, with percentages denoting the relative proportion of per-species yields to the whole caught fishery sector yields.

The data visualized in the study are the data regarding overall fishing yields. The FADs deployed by fishers around Gentuma Port are aimed at catching pelagic (pelagic) fish that normally dwell in waters far away from the coastline. On the contrary, demersal (demergere) fish, which usually roam in waters adjacent to the coastline, do not normally require the assistance of FADs to catch. Based on the fishery yields data issued by the government of Gorontalo Province, pelagic (pelagic) fish constitute 83.37% of the province’s total fishing yields, making the role of FADs especially vital to the livelihood of fishers in the province.

From the analysis, it can be inferred that the majority of preferred targets of fishery activities in North Gorontalo District are very dependent on the existence of FADs, where most of the ones deployed there are aimed at gathering pelagic (pelagic) fish. Hence, the FADs are an inseparable part of the local economy in the North Gorontalo District. The steady and ample supply of pelagic (pelagic) fish also cements the role of the FADs as a long-term supporting aspect of the coastal communities scattered throughout the northern coast of Gorontalo.

4.2. Spatial Analysis of the Effects of FAD Use in the Sulawesi Sea

The deployment of FADs should be based on the fishery yield potential and availability of the targeted fish species. This can be done by considering both the natural and anthropogenic factors associated with the targeted species of fish. The FADs that are already deployed need to be supervised in order to generate the optimal amount of fishing yields without catching other species of fish that are not suitable for fishing activity [23].

The utilization of geospatial information in this study was used to analyze the most suitable sites for FAD deployment. The suitability is determined by various factors, such as the presence of areas that are restricted by existing policies, locations with low hazard levels, and locations with a high potential of fishery yields in the North Gorontalo District. Based on existing regulations, such as RZWP3K, there are only two FADs that are in the pelagic (pelagic) and demersal (demergere) zones. Meanwhile, the other FADs are situated well within the Indonesian Exclusive Economic Zone (EEZ). Apart from that, all FADs are compliant in terms of regulations, and none of them are situated within conservation zones and navigation channels for seafaring vessels.

The placement of FADs should also consider the possibility of unwanted bycatch. As a measure to alleviate this situation, the government, via the Ministry of Fisheries and Maritime Affairs of Indonesia, prohibits the usage of both on-surface and submersible types of FADs that use nets or wires of any sort [2].

The hazardous factor at the sites where the FADs are deployed is obtained from the results of sea current modeling, which in turn is used to determine the probability of occurrences of the knife current. The knife current is one of the leading causes of the loss of FADs deployed in the waters of the North Gorontalo District. Overall, there are five FADs that are the most affected by the knife current on the northern coast of Gorontalo Province, with April and May being the months with the highest number of occurrences of the currents. This is in agreement with the statements gathered via interviews conducted with the local fishermen in the community, where it is stated that May is the month with the most occurrences of drifted or lost FADs. The areas with high potential fishery yields are detected via the ZPPI program. Out of all FADs in the district, only five are adjacent (within a 1 km radius) to fishing hotspots that are determined by the ZPPI monthly issue.

From the several factors above, it is then determined that the most suitable sites for FAD deployment are divided into three classes. Category 1 is defined as areas that are not suitable for FAD deployment due to it being a restricted area or a part of a navigation channel for vessels. Category 2 is defined as moderately suitable as it is not affected by restrictions and dangerous currents but does not provide a high yield of fishery. Category 3 is the most suitable for FAD deployment as it checks all criteria in terms of limitations, fishery yields, and safety from dangerous currents. This study determines that there are 5 FADs situated in Category 1 areas, 27 in Category 2 areas, and 6 in Category 3 areas. The five most vulnerable FADs are namely BRAVO, AWAL, B3YSNT, 73GTAT, and SASY.

The analysis shows that the existing FADs are situated very far away from the coastline. Considering the types of FADs that are deployed in the area, it can be inferred that the main preferred fishing targets of the fishermen of North Gorontalo District are tuna (Thunnus sp.), skipjack (Katsuwonus pelamis), and other species of fish that normally dwell in depths up to 60 m below the sea surface. This aligns with the data released by the Fisheries Department of Gorontalo Province regarding the yields of fishery products from the North Gorontalo District. Pelagic (pelagic) fish dominate 83.37% of the total catch fishery yields in the district, or equivalent to 21,535,604 kg. It can be inferred that the FADs are a vital and long-term supporting component of the economy of coastal communities in the Northern Gorontalo District.

In recent developments of FAD usage, the arrival of portable types of FADs that are dynamic in nature alleviates a significant portion of issues commonly encountered in older FADs. However, there are some possible negative outcomes of deploying portable FADs, such as (1) the decrease in fishing yields per intended target species of fish; (2) increased unwanted bycatch that, in the long run, may create instability of the marine ecosystem; and (3) destructive changes to the migration patterns of fish species that are associated with the usage of FADs [24].

In the waters of the North Gorontalo District, portable FADs are not commonplace. The fishermen of the community mainly still rely on conventional fishing aids, with an average of 10 FADs deployed and owned by a single fisherman. This situation does not comply with government regulations that state fishermen are only allowed to manage and own 3 FADs [2]. However, this limitation is still perceived as difficult to enforce due to strong feedback from the fishermen community. Even by deploying FADs above the legal limit set by the government, the fishermen still struggle financially.

5. Conclusions

The most suitable sites for FAD deployment are determined by considering factors that affect the sustainability and durability of the FADs after being deployed at a certain site. Based on the study carried out in the waters of Gorontalo Province, it can be inferred that:

There are three categories of site suitability for FAD deployment. Category 1 means that the area is not recommended for FAD deployment due to it being a restricted area, interfering with shipping routes, or being affected by knife currents. Category 2 is defined as an area with moderate suitability due to it not being a restricted area and not being affected by knife currents but lack the fishery potential to make it financially feasible. Category 3 is defined as an area with high suitability because it satisfies all of the three main criteria that are put into consideration.

In this study, it is found that there are 5 FADs situated in Category 1 areas, 27 situated in Category 2 areas, and 6 situated in Category 3 areas. This indicates that there are several FADs that need to be re-evaluated and treated with caution in terms of safety and sustainability.

The distribution of FADs in the waters of Northern Gorontalo are vital to the local fishermen community, as they assist in catching pelagic (pelagic) fish, which make up 83.37% of the total caught fishery yields in the district.

From the study regarding the possible safety hazards caused by knife current in the Celebes Sea, it is hoped that future FAD deployments are within areas that are designated to be safe, as this can minimize the risk of FADs being lost or drifting away into the sea, as well as enhancing the fishing yields as the recommended zones also take account of parts of the sea where fish are the most abundant.