CORAL FISH POPULATION CHANGES IN THE SURROUNDING ARTIFICIAL REEFS OF THE LEBAH COASTAL WATERS , KARANGASEM , BALI

The Directorate General of Fisheries deployed some artificial reefs in 1991 for habitat enhancement in Lebah coastal waters that have already degraded. This study was the 4th time monitoring of fifteen years of restoration period that aimed to identify changes of coral fish population in the vicinity of the artificial reefs. The study used dual approaches of old and new data gathering. The changes occurred in terms of restoration processes of the artificial reefs addressed to some progresses of coral fish taxonomy and ecological indices. These included increases in the number of species from 41 to 192, density from 0.20 to 23.32, diversity indices from 1.69 to 2.99, diversity number (N1) from 5.44 to 19.85, and diversity number (N2) from 4.23 to 12.31. Fish communities were in the relatively steady condition throughout the period of restoration with in which there was no extremely population dominance. A small number of target fishes were unseen from the study sites on the other hand a large number of major fishes were existed.


INTRODUCTION
During 1991 and 1994, the collaboration between Central Research Institute for Fisheries and Bali Provincial Fisheries Services was established to develop a technical assistance project for the Lebah coastal waters rehabilitation (Ilyas et al., 1993).The project aimed to deploy some artificial reefs and enhance aquatic habitats that might recover aquatic resource degradation and resolve fisheries management conflicts.The project was implemented for three fiscal years of 1990/1991, 1991/ 1992, and 1992/1993 respectively and this activity was terminated in 1994 (Anonymous, 1993).
Monitoring and evaluation to identify impacts of the project to the habitat rehabilitation have been carried out during the period of restoration in the vicinity of the artificial reefs.The project monitoring carried out for the first time of artificial module deployment in 1991.A small number of coral fishes aggregated around the artificial reefs were identified successfully (Wasilun et al., 1991).The affinity of the coral fish to the artificial reefs was gradually increased in number, especially in 1994, the second year monitoring activity was conducted (Edrus et al., 1996).Ten years after the project terminated, a thorough evaluation was carried out by an external evaluator to study on the impacts of the project to the social and institutional aspect of community participation as well as to the ecological aspect of the artificial reefs restorations.In 1994, a shift in the dominance of benthic lifeforms was detected, from the pioneer organisms to hard corals (acropore and non acropore) and other organisms with the average of percent cover of 19.25.In 2001, additional nine new life forms were noted among the artificial reefs.The data show that the percent cover of the benthic lifeforms was sharply increased from 21.5% in 1991 to 88.3% in 2001.This implied that the diversity of organisms around the artificial reefs could approximately reach the diversity in the natural reefs within a decade (Appendix 1).The study also shows that a tremendous growth of fish population was identified in the vicinity of the artificial reefs in 2001 (Edrus, 2002).
The last project impact evaluation was carried out in 2006 with the aim to identify the artificial reef capabilities in the preservation of coral fish biodiversity closer to the natural process in a coral reefs community.Their biological impacts depend relatively on the migration of the surrounding resources to the artificial reefs or FADs to which many fish and non fish species have a strong affinity (Chou, 1991;FAO, 1991).For this reason, a study of fish population growth is essential to assess periodically the nature process of artificial reefs rehabilitation.The results provide basic knowledge to increase the positive impacts and decrease the negative impacts of introducing technology, especially for the future artificial reefs development.
The objective of this study is to identify changes of coral fish population in the vicinity of the artificial reefs by using the indices of ecology, such as diversity, richness and evenness.

MATERIALS AND METHODS
Field data collections were conducted in July 2006.Study sites located at the coastal waters of Lebah Hamlet, Purwakerti Village, Karangasem, Bali (Figure 1).Two ways of data collection was implemented, field measurement and collection of secondary data.The later was employed to compile and analyse the initial data related to similar indicators at the similar sites.Table 1 shows geographic positions of the transect sites in which data of coral fish were obtained by using visual census on a line transects with the 250 square meter census area (English et al., 1994).Four units of artificial reefs per sites have been observed.An observer swam using SCUBA diving equipment along 5 m length of the line transects.All fish species found and their individual numbers were counted and recorded on a waterproof sheet.Fish photograph guide books were used to identify the fish species (Kuiter, 1992;Lieske & Myers, 1997).
Geographic positions of the transect locations and materials of artificial reefs Coral fishes were grouped into three categories showing their status, the major fish species group, the target species and the indicator species (English et al., 1994).The major species group consisted of the coral fishes that are usually found in coral reefs areas as a resident species and are known as ornamental fish.The target species are the coral fishes that are caught for consumption purposes.Most of the target species are having a great affinity to the coral reefs as intruders.
Indicator species are coral fishes include in the family of Chaetodontidae that are usually used to determine the health of the coral reef community.

Data Analysis
The computed values of the coral fish on the artificial reefs derived from the calculation of Margalef Richness where: S = total number of species H = Shannon index The significant changes of selected indicators in terms of different years of measurement of 1991, 1994, 2001, and 2006 were analyzed by using the graphical trend.Some indicators to be compared in different times consisted of number of family, genera, species, diversity indices, dominance indices, evenness indices, diversity numbers, and fish density.

RESULTS AND DISCUSSION
It was likely that changes of quality and quantity of coastal resources, particularly coral fish as a result of the restoration process of the artificial reef ecosystem between 1991 and 2006 were occurred.Comparison of coral fish population changes in the vicinity of the artificial reefs based on the analysis of primary data and compiled secondary data have also been detected (Table 2).
For fifteen years of the period of restoration there were at least 267 species of coral fishes attracted to the three locations of artificial reefs in the study areas (Appendix 2).A study in 2001 at the same region for five locations found a number of 314 species of coral fishes (Edrus & Suprapto, 2005).It was likely that the natural reef functions have well replaced by the artificial reefs since in the certain areas there has no longer natural reefs for long time.
The restoration project has succeded to enhance the coral fishes poor areas into the high population density areas for around one decade.From taxonomic data in term of the total number of species, genus, and family it is appear that the identified fishes from 1991 to 2001 has increased and a moderately decrease occurred from 2001 to 2006 (Table 2 and 3, Figure 2).These phenomena were also happened to the richness indices and density (Table 2 and Figure 3).
Possible reasons of the last five declining numbers of those were due to the lost some species in visual census activities; vulnerable artificial reefs fish during fishing seasons and temporal migration of fish.Some species of the target fishes group like grouper (Serranidae), soldierfishes (Holocentridae), sweetlips (Haemulidae), emperors (Lethrinidae), fusiliers (Caesionidae), parrotfishes (Scaridae), surgeonfishes (Acanthuridae), and trevallies (Carangidae), were likely temporary missing from the census area and no longer found in the 2006.Figure 4 shows that the percentage composition of the target fishes group tend to decrease after an increasing number during 1991 to 2001.Table 3.
Total number of species, genus, and family found in the artificial reefs in the Lebah coastal waters, Bali Unlike major fishes group that always settle inside and on surface of the modules, most of the target fishes group were moving across the areas with high mobility and migrated far away from the modules.It was known that behaviour of a small number of target fishes group, and certain major fishes provide a cryptic and or nocturnal (Kakimoto, 1979;1984), such as rock cods, grouper (Serranidae), and soldierfishes (Holocentridae).Groupers or rock cods migrate from reef areas to deeper waters.Some of those have strong affinity to natural reefs and artificial reefs protecting them, especially for juveniles, and other several fishes.Most groupers or rock cods were possibly avoided census taker and others hiding from crowded or went out to seek favorable areas at deeper waters.
In addition, one unit volume of the artificial reefs is probably too narrow for accomodating a great number of coral fish species.Unlike natural reefs that have more ecological niches and capable to support the higher biodiversity of fishes, artificial reefs quite depend on building volumes and surfaces area occupied by hard corals and other fauna to provide spaces, niches, and services.Table 2 also indicates that the dominance indices were low for the whole sites and time.However the relatively higher indices occurred for the first time the modules deployed compare with the following period.When the modules were deployed first time in 1991, spaces of the modules were dominated by colonies of Anthias spp., Apogon spp., and Dascyllus trimaculatus.Afterward, there were new arrivals of a large number of fish populations to make evenness of community occupied in the artificial reefs.
The diversity number (N1 and N2) indicated more present of colony species within the artificial reef community, especially for major species that linearly improved their composition during 1991 and 2006 (Table 2, Figure 3 and 4 For this reason the evenness indices of fish community seem to be steady in the whole sites during the last period restoration.As a result, diversity index was significantly improved during 1991 and 2006 (Figure 3).The increase of the diversity index also show that there was no extreme population dominance in the last five period of restoration compare with those in the first time.The lower dominance was likely due to the limited spaces and niches for growing major fish populations inside the modules.Some major fishes group had the same opportunity to grow and to occupy some favorable conditions of the inside part of artificial reefs.It likely that there was no extreme conditions of waters in the vicinity of the artificial reefs leading to a blooming individual of a certain major fishes that will force the others.
The major fishes have a ceratin territory and have more access to ecological niches inside the modules than that of the target fishes.Mostl of the major fishes spatially aggregated in the inside part of the modules and guarded the areas enclosing one or more resources of food, shelter, or potential mates or nesting sites.Major fishes that roam in the inside areas of artificial reefs are known as home ranging species.They may carry with them a portable territory with in which competitors are not allowed.Many home ranging species may be relatively sociable during times of abundant resources, but become aggressive during lean times.Hence, both major fishes and target fishes might share territories while some species may need quite large territories.However, a growing population of major fishes might have limited living spaces in the inside part of the modules.Certain target fishes might be attracted outwardly and distribute in a wider scale of territories (Lieske & Myers, 1997).
Target fishes and indicator fishes group (butterfly fish), may have territories for certain purposes or search some favourable proper areas.Certain target fish may become a daily intruder or a daily moving around in artificial reef areas as well as indicator fishes.Like indicator fishes, certain target fishes may also be unseen during the census activities especially in times of unfavourable conditions.
Low present of indicator fishes (Figure 4) in the artificial reef areas might indicate low diversity and kind numbers of hard corals (Nash, 1989).There were five butterfly fish species still existed in the areas while the other twelve species were missed.This possibly due to recruitment progresses of benthic lifeform of the artificial reefs have not yet fully attracted for butterfly fishes or major fish that have already blooming, and these become a non favored site for them.Mostl of the indicator fishes are the primary omnivores.Many indicator fishes feed on a variety of small invertebrates and coral polyps, taking a little of each over a large home range (Lieske & Myers, 1997).

CONCLUSION
The typical changes of coral fish population engaged during fifteen year period of restoration in the artificial reefs are as follows: 1.The number of species, genus and family of coral fish increased during 1991 until 2001 and moderately decreased during 2001 until 2006.The similar fluctuations were also occurred in richness indices and density.
2. Diversity index of coral fish was improved significantly during 1991 until 2006.
3. Several species of major fishes had continued to occupy the artificial reefs for fifteen years with gradually increases in percent compositions.However, a small number of target fishes and indicator fishes were no longer found in the artificial reefs in 2006.
4. From evenness indices of coral fish community indicated a steady condition for all sites during the last period of restoration and it was due to a low dominance of fish populations.
Some implications are as follows: 1.The artificial reef volume capacity is likely too narrow for growing coral fish communities.
2. The typical niches or habitats are continuously structuring the artificial reefs to support bio diversity.
3. The artificial reefs are vulnerable toward disturbances such as over-exploitation or any physical destruction.
4. Small scale areas of artificial reef development may offer low expected impacts for coral fish population growth and fishing productivity.The number of artificial reef units was insufficient to support high production of target fishes.Furthermore, the artificial reefs have not fully performed yet to reach the optimum level of restoration.These were indicated by the absences of large numbers of major fish and indicator fish species, such as Chaetodontidae, Pomacanthidae, Pomacentridae, Labridae that provide the main families generally inhabit the natural reefs.However, artificial reefs have a specific capability to preserve marine biodiversity on their substrate coverage as well as progressive recruitment of benthic lifeforms.
5. It is important to develop artificial reefs in a larger scale of areas and purposes, such as fishing ground, conservation, and marine tourism.A higher fish productivity in artificial reef areas quite depends on a large number of artificial reef units tol be deployed on degraded areas.
Appendix 2. Fish species identified in the vicinity of the artificial reefs in Lebah Costal waters, Karangasem, Bali.Families and species

Measurement years
Families and species

Measurement years
Families and species

Rainbow selebensis (Telmatherina celebensis
Boulenger) is one of the endemics fish type found in Lake Towuti.This lake, Lake Matano, and Lake Mahalona provide the three stream connected lake consecutively located in the downstream, upstream and midlle stream of the Holahola river of the Malili-Soroako region.The three lakes have been specified as recreational parks due to their panoramic beauty and virgin forest around them (pursuant to the Minister of Agriculture Decree No.274/Kpts/Um/1979).
Sixteen fish species of Telmatherina have been recorded and all are including endemic fish (Kottelat et al., 1993).T. celebensis has beautiful body color, especially male.Its body is rust colored chocolate on the top and there are three or more vertical ribbon which appear more contrast compared to the female.Therefore the fish has potentially economic value as ornamental freshwater fish.
T. celebensis is also taken for consumption purposes by people around the lake (Nasution, 2006).This fish has been included in a vulnerable species group or potentially endangered species (IUCN, 2003).Therefore, protection against the destruction and threat to the fish population decreases due to the effect of environmental changes will be needed.Degradation of environmental quality is indirectly related to the deforestation and anthropogenic activities.
To prevent population decrease better informations is needed to support sustainability of fish resources management.Two aspects of the biological informations needed are growth and condition factor.Condition factor is one of the important growth element to show fish physical capacities and the relationship with survival and reproduction.The aim of this research is to study growth and condition factor of the fish, that can be used as a basic information for fish conservation and fishery resources management of T. celebensis in Lake Towuti, South Celebes.

Study Site
The study was carried out in Lake Towuti, South Celebes (Figure 1) from March 2002 to April 2003.Sampling of fish was carried out every two months.Research stations are grouped into four sites as follow: 1. Station I: located in the area where there are aquatic plant with depth of water of 1 to 3 m (Cape of Bakara nearby sawmill).2. Station II: located in the area where there are no aquatic plant within the depth of 1 to 5 m (inlet of Lake Towuti that come from Tominanga River).3. Station III: located in the area where there are no aquatic plant within the depth of more than 10 m (close to Loeha Island).4. Station IV: located in the aquatic plant area within the depth of 1 to 3 m (close to the outlet of Lake Towuti into the Hola-Hola River).
Research station in Lake Towuti.

Fish Sampling
Sample of fishes (Figure 2) was obtained using experimental gill net with four mesh size: ¾ inch (19 mm), 1 inch (25 mm), 1¼ inch (32 mm), and 1½ inch (38 mm).The total length of the net was 200 m (each of 4x50 mesh size).Net equipped with the float at the top and sinker at the bottom.Net was positioned perpendicular to the coastline in each station and start operated in the late afternoon at 16.00 until 07.00 in the next morning.Operation time of net was more or less 15 hours.
The catches obtained from each station were sorted and separated sexually.Length of fish sampled were measured by using calliper with the accuracy of 1 mm and were weighted by digital balance with accuracy of 0.01 g.All the fish were grouped into some length class to obtain normal population size structure.

Length Weight Relationship
The length weight relationship of fish is calculated following the Ricker (1975) 2According to Carlander (1969) in Effendie (1979), most of b value laid between 2.4 to 3.5.When b value is equal to three (b=3), fish growth in length is equal to the growth in weight and is referred as isometric growth.If b is bigger or smaller than three then the growth is allometric.If b value less than three (b<3), indicating a thin or slender fish, where length growth is faster than weight growth.If b value is bigger than three (b>3) indicating that the fish is fat or plump, where weight growth is faster then length growth.

Growth
Growth pattern of T.celebensis can be predicted by Von Bertalanffy equation as follows: 3where: L t = length of fish in t year (mm) L ∞ = average maximum length t o = theoretical fish age at length=0 mm K = Von Bertalanffy growth coefficient Parameter of K and L ∞ were estimated with ELEFAN I from software FiSAT II (Sparre & Venema, 1998) and t o was estimated from Paully emphirical formula, as follows: Log (-t o )=-0.3922-0.2752log L ∞ -1.038 log K …... (4

Condition Factor
Condition factor that reflecting the plumpness of fish were calculated following Effendie (1979): 5where: K t = condition factor W = the average weight of fish (g) L = the average length of fish (mm) 10 5 = the specified value as K t approaching one

Length frequency distribution
The size of T. celebensis caught using experimental gill net with four mesh size (¾, 1, 1¼, and 1½ inch), ranging from 55.9 to 103.2 mm for male and 51.1 to 98.4 mm for female (Figure 3).The highest frequency occurred within the length class of 74.8 to 79.5 mm in male and female, respectively.
All fish samples were caught by the net with the mesh size of ¾ and 1 inch.Whereas at mesh size of 1¼ and 1½ inch net no fish were caught.This data indicate that the fish caught by the net provide the biggest fish size group within the fish population and has reached the matured gonad.Length frequency of male and female T. celebensis in Lake Towuti.

Length Weight Relationship
Analysis of length weight relationship will lead to the identification of the growth pattern of the fish.From growth pattern, it is possible to determine the fish body form wether the fish is fat or thin.According to Brown (1957) Length weight relationship of male and female T. celebensis follow the equations W=0.00078 L 3.0729 ; r 2 =0.81, and W=0.00005 L 3.04756 ; r 2 =0,90 (Figure 4).According to Carlander (1969) in Effendie (1979), most of b values ranging from 2.4 to 3.5.Base on t test indicate that the male with b=3.1 and female b=3.1 or approximately equal to 3.0.For b value equal to 3.0 indicates that the growth follow isometric pattern, while for b¹3 follow allometric pattern.Length weight relationship of male (♂) and female (♀) T. celebensis.
From the isometric growth pattern of both male and female it can be concluded that growth in length is equal to growth in weight.Morphologically the fish body is fusiform and slender.Due to its physical it is anticipated that this fish provide a fast swimmer and migratory type.The body form influence movement of fish in water.T. celebensis body form is also like tuna (thunniform) according to Jobling (1995) that provide a fast swimmer type.T. celebensis fish found spread over the lake of Towuti ranging from littoral zone to the middle of lake close to the island of Loeha (Wirjoatmodjo et al., 2003;Nasution et al., 2004).

Growth
Growth is one of the important biological parameters to understand the population dynamics of T. celebensis more thoroughly.Based on length data measurements, the average maximum length of fish (L ¥ ) and the growth coefficient (K), the Von Bertalanffy growth equation of male and female of T. celebensis are L t =119.7 [1-e -3.40  (t+0.03)]; L ¥ =119.7 mm and K=3.40 and L t =119.7 [1-e -1.60 (t+0.07)]; L ¥ =119.7 mm and K=1.60, respectively (Figure 5).The Von Bertalanffy growth curves of male and female T. celebensis.
From growth coefficient of male of 3.40 year -1 and female 1.60 year -1 , it is likely that growth rate of male is higher than female or the male growth is faster than female.Base on the growth curves, the male T. celebensis reach maximum total length at 0.8 years and the female at 1.6 years old.So far, no data that have been published to comparing growth coefficient of the same species or family in Lake Towuti.

Condition Factor
Condition factor provide an indirect indicator of the environmental influence to the fish physical condition, formulated in body weight function that compare to fish body length.Condition factor of T. celebensis show in Figure 6.
Monthly condition factors of T. celebensis varied between male and female.Male between 0.93 to 1.21 and female between 1.09 to 1.26.There are some fluctuation in the curve of condition factor, but the highest value occured in November which were equal to 1.21 for male and 1.26 for female.Peak curve is interpreted as the accuracy of environmental conditional change toward better condition in November.The changing of environmental condition was mainly due to the frequent rainfall occurring after September.Some allochtonous materials from land enter the lake and enrich food-stuff in the lake.Fish abundance was also increased in November in line with the increasing height of water level (Nasution et al., 2004).
Figure 6 also shows the change of environmental quality that happened in November, where condition factor of male and female reach its highest.Beside as environmental quality indicator, increasing condition factor reflecting an increasing activity of T. celebensis reproduction resulted from the improvement of gonad development.In November improvement of gonad maturity of T. celebensis were occurred as reflected by the large number of matured fish composition found both male and female (Nasution, 2005).
Figure 1.Map showing the study sites in Lebah Hamlet, Purwakerti, Bali.
Figure 2. Percentage composition of species, genus, and family of the target fish group during 1991 and 2006.

Figure 4 .
Figure 4. Average composition of coral fish groups in the vicinity of the artificial reefs during 1991-2006.
, the factors influencing growth pattern consisted of internal and external factors.Example of internal factors are diameter of egg size and genetic factor whereas external factors include temperature, light, chemical factors, water current, and niche.
Figure 5.The Von Bertalanffy growth curves of male and female T. celebensis.

Table 2 .
Ecological indices showing improvement of coral fish population in the vicinity of the artificial reefs in Lebah coastal waters