Algal community composition within territories of the damselfish nigricans (, Labroidei) in Fiji and the Cook Islands

Lindsey W Zemke-White, Emma L Beatson Division of Applied Sciences, Auckland University of Technology, Private Bag 92006, Auckland,New Zealand [email protected]

ABSTRACT The species composition of algal assemblages inside territories of the damselfish Stegastes nigricans was examined from Nananu-i-ra Island, Fiji and Rarotonga, Cook Islands. A total of 32 algal species were identified, the majority of them Rhodophytes (24 species). Seven species were unique to Rarotonga, six were unique to Fiji, with the remaining 19 species recorded at both locations. The biodiversity of territories was similar at both locations. The red alga, Herposiphonia secunda, was the most abundant species found, comprising 40.33% ± 3.51SE and 33.94% ± 3.58SE from Rarotonga and Fiji respectively. Gelidiopsis intricata, Polysiphonia sp., Aglaothamnion cordatum, and various Cyanophytes were also abundant at both locations. As a group, filamentous accounted for 70.35% of the algae found within the territories at all sites. We suggest that if S. nigricans are actively weeding their territories, they are selecting filamentous algae, and particularly Herposiphonia sp. and Polysiphonia sp.

1 INTRODUCTION If territorial pomacentrids are obtaining the bulk of Herbivory is the main route by which photosynthetic their nutritional requirements from detritus or other fine energy is transferred to marine organisms (Elton 1927). fraction components, then these fish are selectively The primary productivity of coral reefs is quickly weeding for a purpose other than removal of algal species consumed by herbivores (Lubchenco and Gaines 1981); with low nutritional value. Wilson and Bellwood (1997) primarily herbivorous fish (Choat and Bellwood 1985). suggest algal species might be selectively retained in the This rapid consumption results in the observed low algal territories for their ability to trap detritus, perhaps due to biomass that is commonly associated with undisturbed their morphology or the ability to produce extra-cellular ecosystems (Huston 1985). Many species of mucilages. Whether the fish are retaining the algae for nominally herbivorous damselfishes (Family their nutritional components, or for their ability to trap Pomacentridae) defend territories against other fish detritus, it is necessary to identify the composition of the (Letourneur et al. 1997; Wilson and Bellwood 1997; territories to species level so we can further determine the Letourneur 2000; Hata and Kato 2003) and maintain algal qualities of these preferred species. In this study we “farms” in these territories. These algal farms support high determined composition of benthic algal assemblages in primary production (Klumpp et al. 1987) and in some the territories of S. nigricans, from coral reef flat habitats cases high numbers of invertebrates (Lobel 1980; Klumpp of Nananu-i-ra Island, Fiji, and Rarotonga, Cook Islands. et al. 1988: Hata and Nishihira 2002 ). Stegastes nigricans is a small territorial pomacentrid, 2 MATERIALS AND METHODS no more than 14cm in length, and is widespread in the Algal turf samples were collected from within the Indo-Pacific region (Allen and Emery 1985). This species territories of S. nigricans from three sites along the inhabits reef flats and lagoon reefs and is frequently found fringing reef of Nananu-i-ra Island, Fiji (17°10’S, in colonies associated with staghorn corals, spp. 178°29’E), and five sites from the fringing reef (Allen and Emery 1985). Stegastes nigricans has been surrounding Rarotonga, Cook Islands (21°14’S, found to actively weed the algae in its territories, biting 159°46’W). At each site five S. nigricans territories were unwanted algae off and discarding it outside of their selected and five replicate algal turf samples of territories (Lieske and Myers 1994; Wilson and Bellwood approximately 5 cm² were collected from each territory by 1997; Hata and Kato 2002). scraping the turf and underlying dead coral with a knife. Stegastes nigricans is normally regarded as The algal turf samples were stored in formalin for later herbivorous and so the basis for the selection of certain examination. Appropriate controls would entail collection species of algae in the territories has been explained by the of turf from areas adjacent to S. nigricans territories and nutritional quality of the farmed algae (Montgomery 1980; determination of the algal species composition in the Lobel 1980; Russ 1987; Hata et al. 2002; Hata and Kato controls. However, this was note possible as the areas 2002). Recent work has attempted to test the ability of the outside the defended territories were denuded of all turfing fish to assimilate weeded vs. non-weeded algal species and algae; turfing algae is found inside the territories because found that non-weeded species were more nutritious to the S. nigricans keeps other herbivorous fish out fish (Hata and Kato 2002). In contrast to these results, The method used to quantify the samples was modified Wilson and Bellwood (1997) found that S. nigricans are from a diet analysis methodology (sensu Moran and primarily detritivores, and appear to be targeting not the Clements 2002). A series of 10 transects, each 10 cm long, algae, but the detrital material that becomes trapped in the was printed on acetate and glued underneath a clear complex three dimensional structure of the algal turf perspex tray. Each sample was rinsed in seawater and the within the territories. mat of algae was separated into individual species.

43 The South Pacific Journal of Natural Science, Volume 23, 2005

50 45 )

% 40 (

n Rarotonga

o 35 i t i Fiji s 30 o p 25 m o c

20 y r

o 15 t i r r 10 e T 5

0 . 3 a a 2 a s 1 s a a e 1 2 a a 1 l p

r m

t

l s s n . d a . d . . . . s i i e u m a e o o a n

p i p p t p p p t g s h c n i g a s s r s u i s t a s s i s n r m

r

a y

s e c d p t e i u O a a a r n a a l l v s i s k i e i i s n i i i

i o o r

r p a n m c n s n i

g s

i

c e a e o a o m o n s D

a t t l p s v a i e i i r h

h i u h n c c f o f i s m n i s p

a p n p o a a i i i p i a p p d a a p o m i b b s i s s r o n l h y o o h a i t o o y d y y r e r e l i y l l i l p H d m o n n r r i i e o o l o h u G a a a y s B m C e p a P P P h y y p r o t C G e p m C C S o r p a A e l S g H A

Taxa Figure 1. Mean proportions (% ± SE) of algal species in the territories of Stegastes nigricans. Open bars = Rarotonga, Shaded bars = Fiji. “Other” comprised less than 1% of territories.

The algae were identified to species level (where identified (Table 1), the majority of them Rhodophytes (22 possible) and the sample was suspended in sea water and species). Of these 32 species, 7 were found only in evenly distributed on the tray. Four transects were Rarotonga, 6 were found only in Fiji, and the remaining 19 randomly selected and the length and species of each alga species were recorded in both locations. The species touching the transect was recorded. From this we richness of territories was comparable at both locations, calculated the mean percentage of each alga within each with 10.83 ± 0.55 SE species recorded from Rarotongan sample. sites and 7.80 ± 0.56SE species from Fijian sites. There Algal species were classified into five functional was no difference in the Shannon-Weaver index between groups adapted from Steneck and Watling (1982): (1) sites (Rarotonga sites F=1.26, p>0.32; Fiji site F=1.18, Cyanophytes; (2) Filamentous algae - delicately branched p>0.35) or locations (F=3.65, p>0.063). filamentous algae with no or light cortices; (3) Foliose Herposiphonia secunda, a filamentous rhodophyte, was algae - thin tubular and foliose algae; (4) Coarse by far the most abundant species found within S. nigricans macrophytes - coarsely branched algae with or without territories, comprising 40.33% ± 3.51SE and 33.94% ± cortices; and (5) Calcareous algae. These categories are 3.58SE of the total algae collected from Rarotonga and Fiji functional in the sense that their morphological differences respectively (Figure 1). Gelidiopsis sp., a corticated may affect a) the ability of the fish to digest them, or b) the rhodophyte, Polysiphonia sp., and Aglaothamnion amount of detritus trapped in the territory. Thus algae such cordatum, both filamentous rhodophytes and various as Caulerpa spp., while not corticated, were placed into cyanophytes were also abundant at both locations. Spyridia “coarse macrophytes”. Each sample was characterized by filamentosa, a filamentous rhodophyte and Bryopsis its species diversity using the Shannon-Weaver index plumosa, a foliose chlorophyte were abundant only at (Shannon and Weaver 1949). The data were transformed Fijian sites, whereas Hypnea spinella, a corticated using the arcsine transformation, which is appropriate for rhodophyte was only abundant at Rarotongan sites. proportional data (Fowler et al. 1998). ANOVA (Statistica In terms of functional groups, the territories of S. 5.7, Statsoft, Tulsa, U.S.) was used to test for differences nigricans were similarly structured in Rarotonga and Fiji, between sites and locations in terms of mean numbers of the majority being filamentous algae at all sites in both species and Shannon-Weaver indices. locations (Figure 2). Filamentous algae comprised 67.78% (± 2.59SE) and 67.24% (± 4.15SE) of the total algae found 3 RESULTS within territories in Rarotonga and Fiji respectively. In both locations, Fiji and Rarotonga, the territories of Corticated macrophytes were also relatively abundant, the damselfish S. nigricans were colonised by diverse accounting for approximately 20% of the total algae found assemblages of turf algae. A total of 32 algal species were within territories at both locations. 44 Algal community composition within territories: Zemke-White and Beatson

Table 1. Benthic algal taxa found within Stegastes nigricans territories in Rarotonga and Fiji, and the mean abundance (% ± SE) of each taxon. C = Cyanophyta; G = Chlorophyta; B = Phaeophyceae; R = Rhodophyta. (see text for explanation of Functional Groups)

Functional Taxa Rarotonga Fiji Group

Cyanophyte C Cyanophyte sp. 1 6.60 ± 1.29 1.79 ± 0.59 C Cyanophyte sp. 2 2.00 ± 0.82 1.86 ± 0.65

Filamentous R Herposiphonia secunda (C. Agardh) Ambronn 40.33 ± 3.51 33.94 ± 3.58 R Polysiphonia sp. 1 Greville 6.95 ± 3.03 3.09 ± 3.09 R Spyridia tasmanica (Kützing) J. Agardh 0.96 ± 0.27 11.18 ± 6.44 R Ceramium vagans P.C. Silva 4.35 ± 1.42 1.51 ± 0.55 R Aglaothamnion cordatum (Børgesen) Feldmann-Mazoyer 4.29 ± 2.34 4.57 ± 2.15 R Polysiphonia sp. 3 Greville 3.50 ± 0.98 7.99 ± 1.88 R Spermothamnion sp. 1 Areschoug 2.61 ± 0.96 0.07 ± 0.07 R Polysiphonia sp. 2 Greville 2.20 ± 0.45 4.89 ± 1.23 R Dasya sp. C. Agardh 1.57 ± 1.56 R Heterosiphonia crispella (C. Agardh) M.J. Wynne 0.32 ± 0.22 R Ceramium sp. Roth 0.24 ± 0.15 0.08 ± 0.06 G Entocladia viridis Reinke 0.11 ± 0.10 R Ceramium cimbricum f. flaccidum (H.E.Petersen) Furnari & Serio 0.03 ± 0.03 R Balliella subcorticata (Itono) Itono & Tanaka 0.74 ± 0.77 R Griffithsia subcylindrica Okamura 0.04 ± 0.04 G Boodlea vanbosseae Reinbold 0.33 ± 0.31

Foliose R Bryopsis plumosa (Hudson) C. Agardh 0.64 ± 0.25 9.98 ± 3.49 R Hypoglossum simulans M.J. Wynne, I.R. Price & D.L. Ballantine 0.07 ± 0.07 0.42 ± 0.31 G Rhipidosiphon javensis Montagne 0.02 ± 0.02 B Dictyota dichotoma (Hudson) J.V. Lamouroux 0.64 ± 0.06

Corticated R Hypnea spinella (C. Agardh) Kützing 8.66 ± 2.03 0.44 ± 0.23 R Gelidiopsis intricata (C. Agardh) Vickers 7.46 ± 1.79 12.72 ± 3.80 R Gelidiopsis sp. Schmitz 1.07 ± 0.56 1.14 ± 0.81 R Champia vieillardii Kützing 0.19 ± 0.09 R Laurencia intermedia Yamada 0.04 ± 0.04 G Caulerpella ambigua (Okamura) Prud'homme van Reine & Lokhorst 2.94 ± 0.89 4.01 ± 2.46 R Gracilaria edulis (S.G. Gmelin) P.C. Silva 1.64 ± 1.73

Calcareous R Amphiroa fragilissima (Linnaeus) J.V. Lamouroux 1.69 ± 0.51 0.95 ± 0.95 R Tricleocarpa fragilis (Linnaeus) Huisman & Townsend 0.97 ± 0.46 0.00 G Halimeda sp. J.V. Lamouroux 0.10 ± 0.10 0.55 ± 0.55

4 DISCUSSION similar species richness index and composition of algal These findings contrast with those of Hata and Kato functional groups was recorded in all territories, at both (2003) who found that S. nigricans from the northern locations. These findings are important as they provide the Pacific selectively weed their territories to maintain basis for future investigations of the nutritional ecology of monocultures of easily “digestable” algae. If S. nigricans S. nigricans. in Rarotonga and Fiji are actively weeding their territories, Previous studies have reported productive, species rich they are not selecting for the growth of a single alga, but algal communities within territories actively defended by are instead allowing up to 10 other species to grow resident damselfish (Montgomery 1980; Klumpp et al. alongside the dominant species Herposiphonia secunda. 1987; Russ 1987; Klumpp and Polunin 1989). This has Further studies at these locations are needed to determine frequently been related to high proportions of filamentous active weeding in these fish. algae found in gut contents of the fish with the conclusion That there were differences in species composition of that S. nigricans selectively feeds on the more easily S. nigricans territories between Rarotonga and Fiji was digested algal species (De Loma and Ballesteros 2002; expected given their geographical separation. But it is Hata et al. 2002). However, simply comparing algal significant that the same dominant species, as well as a biomass inside and outside of territories, and the relative abundance of algal functional groups found within the gut 45 The South Pacific Journal of Natural Science, Volume 23, 2005 fails to recognize the importance of other resources, such algae, combined with thorough gut content analysis of as detritus, within the territories. resident fish. The dominance of Herposiphonia sp. The volume of algae observed inside territories may Gelidiopsis sp., and Polysiphonia sp. in the territories of S. give the illusion that it is the dominant nutritional nigricans points to the need to determine the ability of component for the fish. Wilson and Bellwood (1997) these algal genera to trap detritus, particularly in demonstrated that epilithic algae contribute less than 25% comparison to other algae. Such information may provide of the organic matter available within S. nigricans a critical link in the understanding of the nutritional territories, while detrital material accounts for more than ecology of these fishes. half. They also analysed the gut contents of S. nigricans, taking into account the entire contents not just the algal REFERENCES fraction. They found the relative proportions of major Allen, G.R. and Emery, A. R. 1985. A review of the dietary components in the gut contents to be comparable to Pomacentrid fishes of the genus Stegastes from the Indo- their occurrence in territory samples, i.e. a high proportion Pacific, with descriptions of two new species. Indo- of detrital material. There is therefore sufficient evidence Pacific Fish 3, 31 pp. to argue that S. nigricans may in fact be predominantly Choat, J. H. and Bellwood, D. R. 1985. Interactions detritivorous, with the species gaining a higher proportion amongst herbivorous fishes on a coral reef: influence of of its organic material from fine particles trapped within spatial variation. Marine Biology 89, 221-234. the algal turf than from the algae itself. Clayton, M. N. and King, R. J. 1990. Biology of marine 80 plants. Longman Chesire, Melbourne.

) 70 % De Loma, T. L. and Ballesteros, E. 2002 Microspatial (

n 60 variability inside epilithic algal communities within o i t i s 50 territories of the Damselfish Stegastes nigricans at La o p Réunion (Indian Ocean). Botanica Marina 45, 316-323 m 40 o c

Elton, C. S. 1927. Ecology. Sidgwick and Jackson,

y 30 r o

t London. i

r 20 r

e Fowler, J., Cohen, L. and Jarvis, P. 1998. Practical T 10 Statistics for Field Biologists. Chichester, John Wiley 0 and Sons Ltd. Cyanophytes Filamentous Corticated Calcareous Foliose algae algae macrophytes algae Hata, H. and M. Kato. 2002. Weeding by the herbivorous Functional group damselfish Stegastes nigricans in nearly monocultural algae farms. Marine Ecology Progress Series 237, 227- Figure 2. Mean proportions (% ± SE) of algal functional 231. groups in the territories of Stegastes nigricans. Open bars Hata, H. and Kato, M. 2003. Demise of monocultural algal = Rarotonga, Shaded bars = Fiji. farms by exclusion of territorial damselfish. Marine Ecology Progress Series 263, 159-167. The abundance of delicately branching filamentous Hata, H. and Nishihira, M. 2002. Territorial damselfish algae recorded within S. nigricans territories at Rarotonga enhances multi-species co-existence of foraminifera and Fiji may suggest that this group of algae, and mediated by biotic habitat structuring. Journal of Herposiphonia secunda in particular, are selectively Experimental Marine Biology and Ecology 270, 215-240. retained within territories for their ability to trap detrital Hata, H., Nishihira, M. and Kamura, S. 2002. Effects of material. Observations of S. nigricans from other parts of habitat-conditioning by the damselfish Stegastes the Indo-Pacific have also reported territories dominated nigricans (Lacapède) on the community structure of by similar species of filamentous algae such as benthic algae. Journal of Experimental Marine Biology Herposiphonia sp. and Polysiphonia sp. (De Loma and and Ecology 280, 95-116. Ballesteros 2002; Hata et al. 2002; Hata and Kato 2003). It Huston, M. A. 1985. Patterns of Species Diversity on is possible that there are particular characteristics of Coral Reefs. Annual Review of Ecology and Systematics filamentous algae that assist this functional group to 16, 149-177. accumulate larger volumes of detritus than the other algal Jan, R.Q., Ho, C.T. and Shiah, F.K. 2003. Determinants of groups. As Wilson and Bellwood (1997) point out, the territory size of the dusky gregory. Journal of Fish delicately branching stems and intricate rhizoid systems of Biology 63, 1589-1597. filamentous algae provide an increased surface area for Klumpp, D.W., McKinnon, A.D. and Daniel, P. 1987. settlement of particles from the water column. Damselfish territories: zones of high productivity on Furthermore, many species have the added ability to trap coral reefs. Marine Ecology Progress Series 40, 41-51. detritus via production of extra-cellular mucus (Clayton Klumpp, D.W., McKinnon, A.D. and Mundy, C.N. 1988. and King 1990). Motile cryptofauna of a coral reef: abundance, Detritivory has long been recognized as a significant distribution and trophic potential. Marine Ecology feeding strategy of invertebrates in temperate and tropical Progress Series 45, 95-118. marine environments. However, the role of marine fish as Klumpp, D.W. and Polunin, N. V. C. 1989. Partitioning detritivores, and coral reef fish in particular continues to be among grazers of food resources within damselfish overlooked. The trophic position of territorial territories on a coral reef. Journal of Experimental pomacentrids could be confirmed with future study of the Marine Biology and Ecology 125, 145-169. amount of detritus accumulated among fine filamentous

46 Algal community composition within territories: Zemke-White and Beatson

Letourneur, Y., Galzin, R. and Harmelin-Vivien, M. 1997. from the Gulf of California, Mexico. Journal of Temporal variations in the diet of the damselfish Experimental Marine Biology and Ecology 47, 9-24. Stegastes nigricans (Lacepè) on a Reunion fringing reef. Moran, D. and Clements, K. D. 2002. Diet and endogenous Journal of Experimental Marine Biology and Ecology carbohydrases in the temperate marine herbivorous fish 217, 1-18. Kyphosus sydneyanus. Journal of Applied Phycology 60, Letourneur, Y. 2000. Spatial and temporal variability in 1190-1203. territoriality of a tropical benthic damselfish on a coral Russ, G.R. 1987. Is the rate of removal of algae by grazers reef (Reunion Island). Environmental Biology of Fish 57, reduced inside territories of tropical damselfishes? 377-391. Journal of Experimental Marine Biology and Ecology Lieske, E. and Myers, R. 1994. Collins Pocket Guide. 110, 1-17. Coral reef fishes Indo-Pacific and Caribbean including Shannon, C. E. and Weaver, W. 1949. The mathematical the . Haper Collins Publishers, 400 pp. theory of communication. University of Illinois Press, Lobel, P.S. 1980. Herbivory by damselfishes and their role Chicago. in coral reef community ecology. Bulletin of Marine Steneck, R. S. and Watling, L. Feeding capabilities and Science 30, 273-289. limitation of herbivorous molluscs: A functional group Lubchenco J, and Gaines, S.D. 1981. A unified approach approach, Marine Biology 68, 299 – 319. to marine plant-herbivore interactions: I. Populations and Wilson, S. and Bellwood, D. R. 1997. Cryptic dietary communities. Annual Review of Ecology and Systematics components of territorial damselfishes (Pomacentridae, 12, 405–437. Labroidei). Marine Ecology Progress Series 153, 299- Montgomery, W. L. 1980. Comparative feeding ecology of 310. two herbivorous damselfishes (Pomacentridae: Teleostei)

47