Sponge Farming Trials : Survival , Attachment , and Growth of Two Indo-Pacific Sponges , Neopetrosia sp . and Stylissamassa

Sponges, an important part of the reef ecosystem, are of commercial value for public aquaria, pharmacology and chemistry. With the growing demand for sponges, natural resources are at risk of being overexploited. Growing of sponges in artificial or semi natural farms is an alternative. In this study different farming methods were tested on two Indo-Pacific sponge species, Neopetrosia sp. and Stylissa massa. Survival, growth and attachment ability were observed with different substrates (suspended ropes, coral boulders and artificial substrate), two types of aquaria with different water volume and two different field sites in Indonesia. The two species responded differently to their individual locations and environmental stresses. Survival, growth and attachment rates of Neopetrosia sp. at the field site are depending on the cultivation method, we found highest volume increment (27–35%) for a horizontal line in the field. Whereas the volume increase for S. massa did not show any differences for the different transplantation methods, Neopetrosia sp. generally showed higher rates than S. massa. Further aquaria experiments, for example, on nutrient supply, should be tested to receive more detailed data about sponges, particularly because almost all fragments of both species showed a decline or steady state in mean length.


Introduction
As filtering organisms, sponges play an important role in reef ecosystems.Like other benthic suspension feeders, they are responsible for a large share of the energy flow from the pelagic to the benthic system [1,2].Sponges actively move water through their body, an advantage that enables sponges to inhabit different types of habitats also at different depths [1].Sponges were used successfully for reef rehabilitation because of their ability to clean water by filtering small particles like detritus, algae, and bacteria, and they also hold rubble and corals together [3,4].Sponges are also commercially important for farming (bath sponge) and ornamental trade and as a new resource of chemicals for pharmacology [5][6][7][8][9].
Sponges are very effective active filter feeders.The volume of water passing through a sponge can be enormous; a sponge with 10 cm in length and four cm in diameter can filter 80 L water in 24 h [10].Due to that sponge farms are tested for their potential as biofilters near fish farms or landbased sewage water [9].
Their ability to reproduce from small pieces makes sponges very attractive for commercial farming [11,12].Sponges with symbiotic relationships are of high interest for public aquaria because they are more colourful, less expensive to keep, and easier to sell than asymbiotic species [11].Some sponge species have symbionts supplying the host with nutrients and need less additional feeding [13].In the last 20 years numerous bioactive metabolites synthesized by symbiotic microorganisms in sponges have been extracted and identified [9].Bioactive substances in some sponges have potential uses in medicines, bactericides, pesticides, cosmetics, fungicides and antifoulings [11,14].Aldisine alkaloids, discovered in S. massa, were shown to be a potent inhibitor of mitogen-activated protein kinase kinase-1 that can inhibit the growth of certain tumor cells [15].Peptides from a sponge in Papua New Guinea (Neopetrosia sp.) can inhibit amoeboid invasion by human tumor cells [16].

Material and Methods
2.1.Sponge Species.Neopetrosia sp.(de Laubenfels 1949) is a hard blue sponge with a rough porous surface [21] (Figure 1).It has a branching growth form with more than one osculum per branch.It can grow from 10 to 20 cm in length and two to four cm in diameter [21].
Stylissa massa (Carter 1881) is a soft yellow sponge, ranging in size from 7 to 20 cm in length and 5 to 11 cm in diameter [21] (Figure 1).It has an irregular but vertically extending growth form, which narrows to the attachment surface and several oscula are distributed irregularly over the sponge body.

Sampling and Study Sites
Pulau Pari-Sampling and Study Site 1. Pulau Pari is an island located in the Thousand Island reef complex north of Jakarta, the capital of Indonesia.The sampling and study site (5 • 51 55 S, 106 • 36 18 E) was located in the sandy lagoon of Western Pulau Pari, at a depth of 2.5 m and about 200 m from the main island (Figure 2).24 donor sponges of S. massa with a length of 10 to 20 cm were collected in the lagoon and on the reef flat at a depth between 0.5 and two meters during low tide.These specimens were growing in the sandy bottom of the transition zone between sea grass and the outer reef as well as on coral boulders in the sandy lagoon.At this site survival, growth and attachment only of S. massa fragments on horizontal ropes and live rock were tested (      site (8 • 08 12 S, 114 • 36 E) was located in the Bay Teluk Pegametan in the north of Bali, on one of the sandy reef flats at a depth of 2.5 m at low tide and about 750 m from the mainland (Figure 3).50 donor sponges of Neopetrosia sp. and 41 of S. massa with a length of 10 to 20 cm were collected at the same depth as in Pulau Pari.S. massa was found growing on coral boulders while Neopetrosia sp. was more often found on live corals.
Two types of aquaria were used, a glass tank and a cement tank.The glass tanks, located at CV. Dinar Goris, Northern Bali, were 80 cm in length, 40 cm high, and 25 cm deep.The cement tank was 2400 cm in length, 70 cm high, and 140 cm deep with a water depth of 40 cm.Both types of aquaria were supplied daily with seawater from the Bay Teluk Pegametan, five km from the facility.The seawater was filtered through a   biological filter consisting of coral boulders.No additional nutrients were added to the aquarium seawater.A roof protected the aquaria from rainfall and solar radiation.

Substrata and Attachment Methods
Neopetrosia sp.The fragmentation of Neopetrosia sp. was always done in the same way by cutting horizontally to provide fragments of approximately equal size.Before the fragments were attached, the length and diameter of Neopetrosia sp.fragments were measured with a plastic calliper.Stylissa massa.The fragmentation of S. massa was done in two different ways; half of the collected sponges were cut horizontally, the other half vertically.The fragments were approximately of equal size.Before the fragments were attached, the length, diameter, and cutting width of S. massa were measured with a plastic calliper.
Nine specimens of each species were left uncut as controls; three of each were placed in the field (Teluk Pegametan), glas and cement tank.
Three horizontal nylon ropes (Ø = 0.7 cm) with the same length (120 cm) were tied on each end at two bamboo sticks (length = 130 cm, Ø = 4 to 5 cm) that were fixed on the seabed.The three horizontal nylon ropes were tied to the sticks at different depths, the first rope started 30 cm above the sea bottom and the next two ropes followed every 30 cm.Twelve sponge fragments were tied to the rope with cable ties, four in each line (Figure 4, Table 1).To prevent shifting of the fragments every cable tie was tied to the nylon rope with a plastic ribbon.
Two horizontal ropes were placed parallel in one aquarium.Four sponge fragments of each of the two species were placed on each of the horizontal ropes in the same way as in the field (Table 1).
Four sponge fragments of each of the two species were attached to coral boulders by using toothpicks and elastics (Figure 4, Table 1).The toothpick was pushed through the fragment, placed on the boulder, and tied to it with elastics.In the aquaria eight sponge fragments of each of the two species on coral boulders were placed in two parallel lines.
In the field, four sponge fragments were tied on artificial substrates (round plates, made of pebbles and cement) in the same way as on coral boulders and placed on a metal grid (200 × 100 × 50 cm) that was set on the sea bottom (Figure 5, Table 1).
In the cement tank the four sponge fragments on artificial substrates were placed on the bottom (Table 1).
Three replicates were used for each method, both in the aquaria and at the field sites.

Abiotic Factors.
Temperature, salinity, pH, and dissolved oxygen were measured with a WTW MultiLine P4 set (conductivity cell, pH combined electrode, D.O. probe) and turbidity with a turbidity meter (Lovibond).Ammonium, carbonate, and nitrite content were measured with a standard test from Merck (Merck KGaA, Darmstadt).Surface currents were measured in the field by recording the time required for an object to float a distance of 6 m.

Analyses of Data.
Growth was expressed in an increase in volume; therefore, length, diameter and cutting width (cutting width only for Stylissa massa) of the fragments were measured with a plastic calliper every two weeks when visiting the sites at Teluk Pegametan and Goris, Northern Bali, and every two months at the site Pulau Pari.
For both species the volume was calculated with the following factors and formulas: for Neopetrosia sp.: (l = length, r = radius of the fragment), for S. massa:  In addition, every time visiting the study site, the following state of the fragments was considered: alive, partly dead, dead, damaged, fretted, covered with sediment, overgrown by, for example, fouling organisms, and attached to the substrates.The survival (S) and attachment (A) rate of S. massa and Neopetrosia sp. was calculated as the percentage of the number of fragments being alive and accordingly attached at the beginning of the study (N 0 ) divided by the fragments left in the end of the study (N E ): To calculate the mean percentage growth (V ) of fragments, the following formula was used, where N 0 is the initial volume and N E is the volume of fragments in the end of the study period: Before starting statistical analysis, the homogeneity of variances of the different variables (species, places, methods, and cutting method) was tested with Levene's test.It was decisive for the statistical analysis method.
To identify significant differences in monthly length and volume growth between the different species, places, methods, and in case of S. massa different ways of cutting (horizontal cuttings and vertical cuttings) a t-test for independent samples with separate variance estimates was done with Statistica-Student Version (Version'99).In Table 1 the days of culture and number of fragments associated with the tested methods at the different study sites are listed.

Growth Rates
In Table 2 mean initial and end values of length and volume are compared for the two species Neopetrosia sp. and S. massa, from aquaria and field sites and for S. massa between horizontal and vertical cuttings.

Mean Length and Volume Increase at the Field Sites Teluk
Pegametan and Pulau Pari Neopetrosia sp.Neopetrosia sp.fragments grew better on horizontal ropes.Mean length increase for all three transplantation methods was significantly different (P = 0.02), as was mean volume increase on horizontal rope and live rock or live rock and artificial substrate (P = 0.002).
Stylissa massa.At both field sites (Teluk Pegametan and Pulau Pari) S. massa fragments did not show any significant differences for mean length and volume increase for the different transplantation methods (horizontal rope, coral boulders, and artificial substrate) or for the different cutting methods (horizontal and vertical way of cutting) (Figures 6(b) and 6(c)), but fragments on horizontal ropes and coral boulders at Pulau Pari showed significant differences in both mean length and volume increment (P = 0.001).
On horizontal ropes those fragments grew better.When comparing the two field sites, fragments at Teluk Pegametan showed higher mean length (P = 0.02) and volume (P = 0.01) increase than those at Pulau Pari.S. massa fragments on horizontal ropes at Pulau Pari showed an initial decline in mean length and volume at the second measuring interval (after one week) before they finally started to grow.
For both mean volume and length increase S. massa and Neopetrosia sp.fragments on the individual depths of the horizontal ropes did not show any significant differences.
Comparing mean length and volume increase at the field sites, the two species S. massa and Neopetrosia sp.showed significant differences in mean length (P = 0.001) and volume (P = 0.01) increase, Neopetrosia sp.fragments grew faster.

Mean Length and Volume
Increase in the Aquaria.In the aquaria the fragments of both species showed a decline or steady state in mean length, except fragments of S. massa on artificial substrate in the big tank.Neopetrosia sp.fragments were significantly different (P = 0.01) in mean length for all three transplantation methods, horizontal rope and coral boulders, horizontal rope and artificial substrate and coral boulders and artificial substrate and in mean volume increase only for fragments on horizontal ropes and coral boulders.The fragments of Neopetrosia sp. on horizontal ropes developed small tentacle-like branches on top of the fragments in horizontal directions.
The mean length and volume increase differed greatly between the study sites.Significant differences (P = 0.001) in mean length and volume increase were determined for Neopetrosia sp.fragments between field Teluk Pegametan and aquaria and S. massa fragments between Teluk Pegametan and Pulau Pari, whereas a significant difference was found in mean length increase between Teluk Pegametan and aquaria and in mean volume increase between Pulau Pari and aquaria (Figure 7).
In general, Neopetrosia sp. and S. massa fragments showed significant differences in mean length, but not in volume increase.

Mean Final Growth, Survival, and Attachment Rates.
Similar results for mean length and volume increase were found in mean final growth rates (end volume minus initial volume) (Figure 7).As shown in Figure 8, no significant differences for final survival and attachment were calculated between fragments of Neopetrosia sp. and S. massa.In the field Teluk Pegametan Neopetrosia sp.fragments on horizontal ropes and artificial substrate showed the highest final survival and attachment rates, but only one of 12 fragments on coral boulders survived (Figure 8(a)).In the case of S. massa the horizontal cuttings showed higher survival and attachment rates than the vertical cuttings for coral boulders and artificial substrate, while vertical cuttings survived and attached more easily on the horizontal ropes (Figures 8(b) and 8(c)).
For both species, the survival and attachment rates of fragments on the artificial substrate at the field site Teluk Pegametan and in the aquaria were highest at 100%, with the exception of 50% attachment of Neopetrosia sp.fragments in the big tank.
The mean final survival rates for horizontal ropes and coral boulders in the aquaria for both species were very low, zero percent for Neopetrosia sp.fragments and vertical cuttings of S. massa on coral boulders and horizontal cuttings of S. massa on horizontal ropes, 21% of Neopetrosia sp. and 12.5% of horizontally cut S. massa fragments on coral boulders, 8% vertically cut on horizontal ropes survived.Almost all fragments of S. massa (horizontal ropes: 50% for horizontal cuttings, 92% for vertical cuttings; coral boulders: 58% for horizontal cuttings, 83% for vertical cuttings) in the aquaria which did not survive until the final checking were already dead two weeks after the installation; in the case of Neopetrosia sp.fragments, 58% were dead one month after installation.
Although Neopetrosia sp. and S. massa fragments did not show significant differences in survival and attachment, it is obvious that almost all fragments of Neopetrosia sp. which survived attached well.
Already two weeks after Neopetrosia sp.fragments were placed on horizontal ropes, coral boulders, and artificial substrates in the field, all of them did attach.In contrast, S. massa fragments did not attach as easily.
One month after starting the experiments, 30 of 33 remaining Neopetrosia sp.fragments started to encrust the cable ties, attaching them to the horizontal ropes.After encrusting the cable ties, those fragments started to close the gap between them and the encrusted cable tie, while they were also starting to grow vertically towards the horizontal rope.The first two of 36 fragments reached the rope already two weeks after installation.At the end of the study, after 108 days, all 33 remaining fragments had closed the space between fragment and cable tie and 30 of 33 had attached to the rope, eight of them already grew along the rope and 20 above it.
One month after starting the experiments, eight of 12 Neopetrosia sp.fragments on artificial substrates at Teluk Pegametan started to encrust their substrate.After two months all of the 12 fragments had already encrusted parts of the artificial substrate.3 salinity, oxygen, pH, turbidity, and temperature for the four different study sites did not vary significantly at the different measuring times.Even daily profiles did not show a great difference between the values.The aquaria showed a higher mean salinity value

Method
Neopetrosia sp. S. massa Hippo spongia lachne [32] Dysidea avara [6] Latrunculia welling-tonensis [17] Polymastia croceus [17] Hippo spongia and Spongia spp.[33] Mycale hentscheli [18] HL-field 106 TP 34 TP sp.For S. massa all three methods (horizontal lines, coral boulders, and artificial substrate) have almost equal growth rates.The horizontal rope method was already successfully tested by de Voogd [8] on Callyspongia biru, an encrusting sponge, and for bath sponges by Corriero et al. [5] and C ¸elik et al. [22].The success of farming sponges requires the knowledge of the optimal environmental conditions of the cultured sponge species [9].The position of Neopetrosia sp.fragments on horizontal ropes and artificial substrates (50 cm above the sea bottom on a metal grid) is similar to its position in natural habitats, where they were observed to grow on other sponges or corals in order to have an exposed position to filter water.Neopetrosia sp.fragments on coral boulders had a low survival rate (8.33%).It is assumed that the fragments of this species probably did not like to be placed directly on the sea bottom.The higher sedimentation near the bottom leads to the block-up of ostia, causing the sponge to stop feeding and forcing all its energy toward cleaning [23].For other sponge species like Spongia officinalis no significant differences in growth could be observed for different positions of transplants in the water column [22].
S. massa prefers to live on coral boulders or on the sandy bottom, in the transition zone between coral reef and sea grasses or mangroves.Its habitat preferences are therefore more widespread than those of Neopetrosia sp.In the Bay Teluk Pegametan, all S. massa fragments showed higher growth, attachment, and survival rates than in the Pulau Pari lagoon.Fragments on coral boulders were totally covered with sand due to wave action and crab movements.Newly built crab burrows can cover sponges and corals in nonelevated positions.Currents ripped off 38% of the fragments on horizontal ropes.Too much wave action and unsuitable farming methods for those areas were the reasons for the low survival and attachment rates.As described by Duckworth [9], in areas with strong water flow or partly stormy conditions the growth behavior of sponge species could be different, there is less feeding, and more energy is used to form a stronger skeletal structure.
Generally, Neopetrosia sp.showed higher growth rates on the different substrates than S. massa, but they had almost the same attachment and survival rates.It is unclear if there are any differences in growth rates of wild stocks between these two species.There were no differences in growth between uncut sponges (donor sponge) of both species at the field site Teluk Pegametan, but Neopetrosia sp.displayed signs of being fed on or taken away by current, while S. massa does not show any signs of damage.Like in the study by De Caralt et al. [6], it is difficult to compare the uncut sponges of the studied species due to their much more complex growth forms in comparison to the regularly cut fragments.
Fragments of Neopetrosia sp.grow faster after cutting than uncut sponges.Duckworth [24] tested regeneration and growth on wounded Latrunculia wellingtonensis sponges and demonstrated that damaged specimens showed better growth than nondamaged ones, which possibly invest their energy in reproduction.Wound size is greater for S. massa (massive growth form) than for Neopetrosia sp.(branching growth form) due to different morphological appearance.Wound recovery needs energy and reduces growth, and wound size influences survival [24].Duckworth et al. [25] demonstrated wounds on some fragments of a massive sponge Psammocinia hawere, which did not fully recover after 80 days.This could be a reason why S. massa grew slower.However, it is difficult to compare both tested species for growth and attachment due to different skeletal structure, habitat preference, growth forms of the donor sponge, and wound size of the fragments.
Other studies have shown that growth of fragments on specific substrates can depend on sponge skeletal structure [7,8,18], soft sponges grow better in mesh bags and hard sponges on rope lines [17,20].In comparison with other studies (Table 4) growth rates of S. massa showed similar results using the suspended rope method, but Neopetrosia sp.showed even higher growth rates, 106% growth rate per month.Fragments in mesh bags showed higher growth (over 400% per month [18]) and survival rates, but the ones on the ropes showed higher attachment rates [19].Fragments can be placed gently in mesh bags, which partly protect fragments against damage, especially for soft and fleshy sponges, but the bags also limit fragments growth and feeding possibility, especially if biofouling species like algae, bryozoans and ascidians were growing on the mesh [9,17,19].In contrast the exclusion of sediment can lead to smaller growth rates; those sponges can grow 50% less than others [23].In a different study, Duckworth and Battershill [17] showed that two different New Zealand sponge species (L.wellingtonensis and P. croceus) differed in survival and growth rates, greatest survival rates for meshes (meshes: 61 and 96%, ropes: 22 and 59%) and greatest growth for ropes (weight increment meshes for P. croceus: about 50%, ropes: over 100%) after nine months.4.2.Growth, Attachment, and Survival Rates at the Aquaria Sites.Sponge fragments in the field showed higher growth, attachment, and survival rates than the ones in the aquaria.Inhibition of sponge growth can be caused not only by high sediment loads but also by low loads [2], which can be the problem in the aquaria.The aquaria fragments of both species even show a decrease in length and volume.Shrinking is evidence for insufficient food supply.Even if food like bacteria is present, also too gentle aeration, infrequent change of seawater, and the slowness of inflow can affect the availability of food [9,26].The higher survival and attachment rate, observed in the cement tank in comparison to the glass tank, may have been due to a higher water volume exchange.
Neopetrosia sp.fragments on horizontal ropes in the glass tank were shrinking, but they also have developed little tentacle-like branches.These branches might be a sign of starvation.However, adding food can help sponges to survive for several years in aquaria [26].
Another reason for the low survival rates in the glass tank is dying sponge fragments, which can be toxic [27] and affect the survival rate of healthy fragments.
Nickel et al. [28] suggest removing specimen with little parts of the attached substrate to have a better recovery in aquaria, because attachment will promote survival [19].Further proper collection and transportation and more experiments in aquaria, for example, on light, current, and nourishment could help to discover how to maintain, feed, or reproduce species like Neopetrosia sp. and S. massa in aquaria.
Artificial substrate is easy to install and in the case of Neopetrosia sp. the encrusting growth form helps to cover the substrate in order to gain a marketable appearance in the aquaria.

4.3.
Recommendations.The precision of calculated volumes is limited due to the nonuniformly morphological appearance.For a new series of studies it is recommended to use additional measuring methods to determine growth, for example weight [8], or other calculation methods depending on the growth form of the sponge species as described by Sipkema et al. [29].Page et al. [18] measured the growth of sponge fragments of Mycale hentscheli via the surface area in correlation with wet weight.
As Neopetrosia sp.showed high survival, growth, and attachment rates, it would be interesting to test smaller cuttings.Duckworth [9] has shown that some sponge species (Latrunculia wellingtonensis and Polymastia croceus) can also recover totally from 90% removal of biomass.As other studies [5,22] have also shown, growth rates of Spongia officinalis between different sized explants are similar.Neopetrosia sp.fragments show nearly no algae cover.In case few algae attach, they are easily removable and do not affect their state of health.
As sponges are important for the reef ecosystem, this study could help to reproduce sponges to help a reef to recover, to make it more attractive again to other marine species after destruction, further offering new and hopefully sustainable exploitation for local fishermen.
To minimize harvesting impact, it would be advisable not to harvest the whole sponge stock, but to leave a small part behind to give the sponge the chance to recover and regenerate into a new sponge.Sampling sponges for export trade does not only place extreme pressure on the wild stock, also the attachment surface will be damaged, which could cause injury to other living creatures, mostly corals.The knowledge of how to reproduce sponges on different substrates will contribute to the reduction of wild harvest and its damage to the environment.

Figure 1 :
Figure 1: Neopetrosia sp.(a) and Stylissa massa (b) fragments attached to the artificial substrate and placed on a metal grid at Teluk Pegametan.

Figure 2 :
Figure 2: Map of Pulau Pari, Thousand Islands, North of Jakarta [30] the X marks the study site.

Figure 3 :
Figure 3: Map of the sampling and study site (X) at Teluk Pegametan, Northern Bali [31].

Figure 4 :Figure 5 :
Figure 4: A scheme of Neopetrosia sp.(top) and S. massa (down) fragments on horizontal lines and live rock at the field sites.
. The experiment started in the field at Pulau Pari on 2 November 2002, at Teluk Pegametan, in the glass tanks at Goris in the facility of CV.Dinar on 3 December 2002, in the field at Teluk Pegametan with the artificial substrate on 28 December 2002 and on 03 February 2003 with the artificial substrate in the cement tank.The last measurement of sponge fragments at Pulau Pari took place on 12 March 2003 and at Teluk Pegametan on 20 March 2003.

Figure 6 :
Figure 6: Fragments mean length at the field sites for different attachment methods and different places.

Table 1 :
List of application methods of fragments at the different study sites associated with the number of transplants and the days of culture.

Table 2 :
Mean initial and end values of length and volume of fragments at the different sites for all methods.

Table 3 :
Mean values for abiotic factors.