Impacts of Foraging Behavior of Initial and Terminal Phase Sparisoma Viride on a Caribbean Coral Reef

PAGE 1

IMPACTS OF FORAGING BEHAVIOR OF INITIAL AND TERMINAL PHASE SPARISOMA VIRIDE ON A CARIBBEAN CORAL REEF BY SAMANTHA ORTIZ A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requirements for th e degree Bachelor of Arts Under the sponsorship of Dr. Gordon Bauer Sarasota Florida April 2012

PAGE 2

ii Dedication This is dedicated to the two strongest women I know: My mother Janet and my grandmother Antoinette. I love you both more than anything

PAGE 3

iii Acknowledgements Thank you to my academic advisor and thesis sponsor Dr. Gordon Bauer. You have been a major source of support and motivation for me and I have enjoyed every minute working with you throughout my academic career at New College. I c ould not ask for a better advisor and sponsor! Thank you to Dr. Al Beulig. You are more than just a professor to me. You are my comrade. You helped me immensely over the past years and I am indebted to you for your unending support and guidance. My s ummers at ITEC were incredible and I could not have done any of this without you. Thank you to Dr. Heidi Harley. Your words of motivation made me realize that this process is one to truly be proud of and embrace. Thank you to Duff Cooper for your ass istance with running my analyses and for always having a bright, shining attitude. Thank you to the Institute for Tropical Ecology and Conservation for accommodating me and providing support throughout the past two summers. My experiences there were un forgettable and helped mold me into the person I am today. Thank you to my parents family friends and loved ones particularly my mother and my Uncle Dave whose unconditional love helped me through my ups and downs. You are both a huge inspiration to me. Thank you for appreciating every step of my growth both academically and as a person. I love you.

PAGE 4

iv Table of Contents Dedication ................................ ................................ ................................ .......................... ii Acknowledgements ................................ ................................ ................................ ........... iii Table of Contents ................................ ................................ ................................ .............. iv List of Figures and Tables ................................ ................................ ................................ ... v Abstract ................................ ................................ ................................ ............................. v i Introduction ................................ ................................ ................................ ......................... 1 Methods ................................ ................................ ................................ ............................. 20 R esults ................................ ................................ ................................ ............................... 23 Discussion ................................ ................................ ................................ .......................... 27 References ................................ ................................ ................................ .......................... 33 Appendix A ................................ ................................ ................................ ....................... 37 Appendix B ................................ ................................ ................................ ...................... 46

PAGE 5

v List of Figures and Tables Figure 1 ................................ ................................ ................................ .............................. 20 Figure 2 ................................ ................................ ................................ .............................. 21 Table 1 ................................ ................................ ................................ ............................... 23 Figure 3 ................................ ................................ ................................ ............................. 25 Table A1 ................................ ................................ ................................ ............................. 40 Figure A1 ................................ ................................ ................................ .......................... 41 Table B1 ................................ ................................ ................................ ............................ 46

PAGE 6

v i IMPACTS OF FORAGING BEHAVIOR OF INITIAL AND TERMINAL PHASE SPARIS OMA VIRIDE ON A CARIBBEAN CORAL REEF Samantha Ortiz New College of Florida 2012 ABSTRACT The stoplight parrotfish, Sparisoma viride is considered a major reef herbivore because the regular consumption of algae by S. viride helps prevent coral algal phase shifts on reefs. However, r ecently S. viride was accused of consuming and destroying live corals most notably the boulder star coral Montastraea annularis and in some cases the finger coral Porites porites S. viride live coral consumption be haviors are displayed agonistic in nature. Live coral consumption by S. viride sometimes results in reef destruction through partial or full mortality of the corals. Because of this destruction caused by the fish, the conservation of stoplight parrotfish is questioned. The foraging and agonistic behaviors of S. viride were studied while SCUBA diving using the focal animal all occurrences method. It was hypothesized that initial phase S. viride had less of an impact on the reef than terminal phase S. viride and that behavioral differences occurred depending on the vertical distribution of the fish. Results did not support the hypotheses. Initial phase members consum ed more of both algae and live coral. Vertical distributions in the water column had no main effects on consumption of any food types. No significant agonistic behaviors among the stoplight parrotfish were observed. These results suggest that territoria l defense may not be necessary for the population studied due to the abundance of resources. Additionally

PAGE 7

v i studying bite counts may not be a complete measure because yield per bite of different food types among individuals may vary. Conservation of this p opulation should not be neglected. This is because their impact is more positive than negative; however this impact may differ among different ecosystem populations. Dr. Gordon Bauer Division of Psychology

PAGE 8

Running head: IMPACTS OF FORAGING BEHA VIOR 1 Impacts of foraging behavi or by Initial and Terminal phase Sparisoma viride on a Caribbean coral reef Increasing attention is currently devoted to the effects herbivorous and corallivorous fish have on coral reefs throughout the world. Specifically members of the Scarid family are a main concern. Scarids are herbivorous fishes that regularly consume algae as the main source of their diet (Bruggemann van Oppen & Breeman 1994). This herbivory prevents coral reefs from shifting to an algae dominated alternate stable state ( Bru ggemann Kessel van Rooij & Breeman 1996 ). Recently Sparisoma viride or the stoplight p arrotfish was accused of destroying and consuming live corals (Bruckner Bruckner & Sollins 2000). Because of these behaviors as well as their regular consumpt ion of dead coral rock and algae they are considered a major source of bioerosion on the reefs (Bruggemann et al. 1996). S. viride can be considered both a vital member and a destructive member of reef communities. This poses the issue of whether it is worth promoting conservation of the species if they are destructive to the reefs. A pilot study was conducted in 2010 on Conch Reef in Bocas del Toro, Panama to determine the impacts of foraging behavior by Initial Phase (IP) and Terminal Phase (TP) s toplight parrotfish (Appendix A). The focal animal all occurrences method was used to determine frequencies of bites on algae and live coral, as well as to assess agonistic and territorial behaviors displayed by S. viride Agonistic and territorial behav iors were studied because past literature notes live coral consumption as being

PAGE 9

IMPACTS OF FORAGING BEHAVIOR 2 territorial in nature (Bruggeman, et al. 1994; van Rooij, Bruggeman, Videler, & Breeman, 1995). The pilot study was inconclusive due to lack of sufficient data; however, some noteworthy trends were observed. S. viride consumed 85% algae and 15% live coral over the course of the study. Additionally, the most consumed live coral was Montastraea annularis These findings were consistent with those of other studies (Bruckner, et al., 2000; Sanchez, Gil, Chasqui, & Alvarado 2004). However, there were almost no agonistic or territorial behaviors observed, inconsistent with findings of past research (Mumby & Wabnitz, 2002; van Rooij et al., 1996). Mumby & Wabnitz (2002) define pa rrotfish territorial and agonistic behaviors as occurring when one fish chases another fish or conspecific, or when there was a fin display between two or more individuals. The results of the pilot study suggested that territoriality was not utilized. A s tudy on another reef ecosystem in Bocas del Toro, Panama, noted algae dominance on the reef (Cooper, unpublished). Conch Reef may also have abundant algae available for consumption. It may not have been worth defending territories because exerting additi onal energy to defend would not yield additional resources. The current study aims to assess the role of the population of S. viride on Conch Reef in Bocas del Toro Panama. It was based on the aforementioned pilot study conducted in 2010. Life phase and depth were assessed as potential predictors of consumption rates of algae and live coral to determine whether the population plays a positive or negative role on the reef. This information can lend insight as to the role that S. viride may play worldwide as well as provide evidence for or against

PAGE 10

IMPACTS OF FORAGING BEHAVIOR 3 conservation of the species. To put this study of the impact of S. viride in perspective, the typical diet, role on the reef, and social system of the species is reviewed. I. Diet The typical diet of reef herb ivores including S. viride consists of endolithic and epilithic algae (Bruggemann et al 1994). This algae grows on various surfaces and substrates such as on sandy bottoms, dead corals, and live corals. Scarids either excavate or scrape off algae fr om corals and substrate s with their beak like mouths. By scraping and excavating these surfaces they also consume sand and pieces of dead and live corals. Bruggemann et al. (1996) studied the bioerosion and sediment ingestion of S. viride and Scarus vet ula the queen parrotfish. The authors found that parrotfish consume large amounts of sediment when foraging. This sediment is typically composed of calcium carbonate. The consumption of sediment results in the parrotfish excreting the calcium carbonate back on to the reef ; therefore, much of the sand on reefs is parrotfish excrement. Contrary to previous assumptions not all Scarids employ the same feeding strategy. Due to morphological and behavioral differences parrotfishes can be separated into dist inct categories. Bellwood & Choat (1991 ) analyzed these differences among 24 Scarid species inhabiting the Great Barrier Reef. They focused on whether scarids were morphologically uniform or whether differences existed in jaw structure. Additionally t hey questioned the biological and ecological implications given Scarid structural differences. To address this they studied jaw morphology shape of feeding marks

PAGE 11

IMPACTS OF FORAGING BEHAVIOR 4 degree of excavation of substrata bites per minute number of bites per feeding foray bi te speed of single bites and patterns in microhabitat utilization while foraging. After assessing the results the researchers found that Scarids can be divided into two separate categories: excavators and scrapers. They defined excavators as species t hat remove pieces of substratum while feeding. Excavating species utilize their strong jaws and are capable of damaging or fracturing coralline substrata. Sparisoma viride falls under this category. Scrapers were defined as species that employ non excav ating bites only removing material from surfaces. Because of their weak bones and muscles they cannot easily excavate coralline substrata. The researchers noted that because Scarids do not form one uniform group it should be noted that the different g roups have different effects on the reef particularly in terms of bioerosion. Excavating parrotfish such as S. viride frequently forage and remove pieces of material from convex surfaces such as dead and live corals. Occasionally parrotfish will consu me large quantities of live corals. This behavior is typically considered to be a territorial behavior usually by adult TP males (Bruggemann et al. 1994; van Rooij Bruggemann Videler and Breeman 1995). Rotjan and Lewis (2005) found parrotfish to h ave selective corallivory or consumption of corals. They studied the interactions between parrotfishes and colonies of Porites astreoides on a backreef habitat. Colonies occurring within 5 x 60 m 2 transects were surveyed and the presence or absence of p arrotfish grazing scars were recorded. Coral mortality was also assessed. The authors reported that 13% of the colonies surveyed experienced either partial or complete mortality due to the grazing. These results suggest that parrotfish may play a

PAGE 12

IMPACTS OF FORAGING BEHAVIOR 5 detrime ntal role in the surv ival of some coral species. One major issue with Rotjan and causative factor of the coral mortality. They were not able to control for all possi ble confounding variables, which suggests a flawed conclusion. Parrotfish grazing may have contributed to coral damage and mortality, however, claiming that the mortality was solely due to grazing stretches the conclusion too far. A more appropriate conc lusion would be that the population of parrotfish they studied selectively preyed on specific species of corals, and that parrotfish predation may detrimentally affect those coral species. P arrotfish do indeed selectively prey on live corals; however thes e preferences are particular to the location of the study. Parrotfish selectivity may not be the same in all reef habitats. Rotjan & Lewis (2006) addressed selective corallivory in several types of reef habitats in Belize. They studied six major habitats : backreef lagoon upper spur and groove lower spur and groove inner reef slope and outer ridge crest. Parrotfish and coral abundance were recorded in each habitat within five 30 m x 2 m belt transects. Fourteen species of parrotfish and nine corals were recorded. Spot and focused grazing scars which were indicated by the presence of paired scars were counted and reported. The results of the study found that Montastraea annularis or the boulder star c oral was the coral of choice targeted by par rotfish across all reef habitats particularly in shallow areas with over 55% of all colonies showing evidence of grazing. Montastraea cavernosa Agaricia agaricites Diploria strigosa P orites astreoides and P orites porites were not preferred. Siderast rea siderea was preferred in spur and groove habitats. These results are particularly interesting especially in relation to the aforementioned study

PAGE 13

IMPACTS OF FORAGING BEHAVIOR 6 pertaining to P. astreoides (Rotjan & Lewis, 2005). Parrotfishes in Rotjan & Lewis study (2006) did not prefer to consume live P. astreoides unlike the parrotfishes in the previous study (Rotjan & Lewis, 2005) This suggests that selective corallivory can be highly dependent on habitat type and field location. S. viride and sometimes Sparisoma aurofrenatum are the major coral predators on a reef in the Florida Keys particularly P porites and Porites divaracata (Miller & Hay 1998). It was concluded that the coral consumption by these fishes counterbalances the benefits of their herbivory. Clumps of seaw eed were attached to cinder blocks that were then left bare (no cage treatment) completely enclosed in large plastic mesh (full cage treatment) or partially enclosed with two walls and a top (half cage treatment). Corals ( P. porites and P. divaracata) w ere then attached to the blocks through slits in the mesh. Coral consumption of P. porites and P. divaracata was measured. Fifty six percent of the corals in the no cage and half cage treatments were missing appearing to be bitten off from the base. Th e scars that remained on the base of the corals looked like parrotfish feeding scars. S. viride inferred that the fish were trying to access the corals that were protected. S.viride bit off la rge portions of the coral branches while S. aurofrenatum only scraped the surface of the corals. Miller and Hay concluded that direct feeding by the parrotfishes can substantially reduce the abundance and distribution of corals. However they made sure t o point out that the results cannot be generalized to all corals and that it will depend on species specific traits of the corals and fishes. Due to their coral consumption parrotfish have been argued as either herbivores or corallivores. Yoshioka (2008) surveyed algae and invertebrates (octocorals tunicates

PAGE 14

IMPACTS OF FORAGING BEHAVIOR 7 anemones oysters and sponges) found on buoy lines that were placed on a reef as artificial substrates. The lower ends of the buoy lines were grazed upon much more than the upper portion of them. Additionally Sparisoma viride the redband parrotfish Sparisoma aurofrenatum and the striped parrotfish Scarus croicensis were found at all sites of the buoy lines however they remained close to the bottom of the lines as grazers. This indicated tha t some fish might have been responsible for the various vertical patterns of epibiota distributions. To test this they lowered the upper portions of the buoy lines to about a meter above the bottom of the reef. After a week most epibiota particularly algae were severely reduced on the portions of the buoy lines that were lowered. It was also found that some parrotfish such as S. aurofrenatum would spit out fragments of sponge after biting off a piece Thi s raised questions about whether or not to consider some Caribbean parrotfishes as herbivores which is what they are typically regarded as or whether they should be considered omnivores. The researchers suggested that they should be considered omnivores rather than herbivores because of their grazing activities despite the fact that they typically do not consume animals. This issue is an important one because it addresses the problem of the conservation of the species. If parrotfish are considered an herbivore rather than an omnivore or even corallivore they may be more actively conserved because omnivorous and corallivorous activities are considered destructive. The role of a parrotfish as an important herbivore on reefs needs to be maintained if they will be conserved whereas if they become considered an omnivore or corallivore this may not be the case. Corallivory can be considered harmful to reefs due to the destruction of the coral skeletons. It can be

PAGE 15

IMPACTS OF FORAGING BEHAVIOR 8 argued that although they partake in a if they do not actually consume the animals (corals) they are herbivores not true omnivores or corallivores. However there have been several reports made that parrotfish specifically S. viride do indeed prey on and consume corals ( Bruckner et al 2000; Bruggeman et al. 1995; Bruggeman et. al. 1996; Miller & Hay 1998; Rotjan & Lewis 2005 2006; van Rooij et al. 1995). II. Role on Reef The consumption of algae on reefs by parrotfish among other herbivorous fishes prevents a reef phase shift from a coral reef to an algal reef (Bruggemann et al. 1996; McClanahan Bergman Huitric McField Elfwig Nystrm & Nordemar 2000; McManus & Polsenberg 2004; Mumby 2009). The ecological and environmental aspects of coral algal pha se shifts were assessed by McManus & Polsenberg ( 2004 ) The authors indicated that coral algal phase shifts occur when reefs suffer from low coral cover and high cover of fleshy macroalgae. Typically a phase shift begins after some sort of trauma such or recovery back to a coral reef is dependent on multiple factors such as increased nutrients and herbivory. Loss of herbivory was considered a major causative factor in phase shifts from coral to algal reefs. Both herbi vorous fish and invertebrates were considered. Diadema antillarum the long spined sea urchin is a keystone species on many reefs as a n herbivore consuming algae and preventing reef phase shifts similar to the role of herbivorous fishes such as parrotf a massive dieoff of D antillarum occurred. Approximately 93% of the species was reduced in the Caribbean. Reefs seriously suffered after the dieoff many of which shifted from coral reef to algal reef. If

PAGE 16

IMPACTS OF FORAGING BEHAVIOR 9 herbivorous fish such as pa rrotfish were not present on coral reefs similar outcomes could potentially occur. If late succession algae dominate coral reefs the abundance of different species of herbivores can be reduced. Reduction of fleshy erect algae on patch reefs in Belize r esulted in a rapid increase of the abundance of the blue headed wrasse the biomass of the blue tang and abundance and biomass of the stoplight parrotfish. Bite rates and inter and intra specific aggressive encounters also increased in these species aft er algal removal (McClanahan et al. 2000). Algae were removed from sections of the reefs by trimming with hedge trimmers and wire brushes. Observations of aggression and bite counts were conducted in which all bites of substrate were recorded as well as all interactions and attacks on other fishes. Each fish was haphazardly selected and observed for one minute. Fish counts were made and biomass was recorded using size frequency data. There was a significant difference in biomass between the control and experimental patches in the major herbivores. For example the stoplight parrotfish had a biomass six times higher in the trimmed patches than on the control patches. Aggression rates of the major herbivores increased in the manipulated patches as well. Therefore because the domination of late succession algae on coral reefs negatively affects the abundance biomass and behavior of important herbivores once late succession algae takes over the coral reef will likely become an algal reef due to the lack of herbivore grazing. Although parrotfish are beneficial in their consumption of algae they are also destructive. They are a major source of bioerosion on reefs. Bruggemann et al. (1996)

PAGE 17

IMPACTS OF FORAGING BEHAVIOR 10 studied the bioerosion and sediment ingestion of S. viri de and Scarus vetula on a reef in Bonaire Netherland Antilles. They measured the foraging behavior of the two species by calculating the proportion of bites that left grazing scars on substrates erosion rates mass and sources of the sediments ingested and the spatial patterns of their bioerosion. The results indicated that the main bioeroder at the site was S. viride Grazing scar volume from S. viride was negatively correlated with density of the different substrates; therefore low density substrat es such as M. annularis are eroded faster by stoplight parrotfish than those of higher density such as Acropora palmata and Acropora cervicornis Additionally S. viride removed more carbonate on substrates with endolithic algae than those covered in cr ustose corallines. Interestingly t his population of S. viride rarely consumed living corals. The researchers noted the lack of density of other excavating grazers which included sea urchins such as the previously discussed D. antillarum This paired wi th the lack of live coral consumption is notable. Many other study sites suggested that S. viride consumes a significant amount of live corals. The amount of corallivory by these fish at the sites created debate among the coral reef research community. The reef dynamics at the sites vary so perhaps there are indicators to the corallivorous impact of the parrotfish such as abundance of endolithic and epilithic algae as well as density of other grazing species such as D. antillarum S. viride members have an interesting set of behaviors called spot and focused biting. Bruckner et al. (2000) address these types of biting distinguishing spot biting as occurring when random individual bites are taken off the surface of the coral while focused biting occurs when there are repeated spot bites that overlap on the coral. These behaviors particularly focused biting cause considerable damage to the coral skeleton

PAGE 18

IMPACTS OF FORAGING BEHAVIOR 11 and tissue. However Bruckner et al. assures that it is rarely fatal and that any tissue damaged by the fish typically regenerates. Predation by parrotfish may exacerbate issues for corals caused by environmental stressors such as coral bleaching (Rotjan et al. 2006). Coral colonies of Montastrae a faveolata and Montastrae a franksi were s ampled on a reef in Belize. The reef was disturbed by several environmental factors including a recent bleaching event unseasonably high temperatures destruction from Hurricane Ivan and low zooxanthellae densities. Grazing pressure was measured by co unting bite marks on the coral heads. Temperature of the water was recorded and zooxanthellae were measured on the Montastraea Symbiodinium were analyzed Th ere was a greater diversity of Symb iodinium and a decreased zooxanthellae density on grazed corals compared to intact corals. Therefore coral predation by parrotfish when coupled with bleaching can affect the coral zooxanthellae symbiosis and intensify coral stress. Parrotfish grazing i s associated with reduction in the nutritional quality and suppressed nematocyst densities of corals (Rotjan & Dimond 2010). Montastraea spp. were sampled to determine coral nutritional quality nematocyst densities and grazing pressure in Belize. Chain link and PVC cages were placed over selected Montastraea spp. colonies in order to keep adult parrotfishes from grazing and feeding on the corals. The experimental reduction of grazing was carried out for 8 months with the conditions being either grazed or intact and the treatments being either caged or exposed (control).

PAGE 19

IMPACTS OF FORAGING BEHAVIOR 12 Parrotfish grazing reduced the coral nutritional quality and decreased nematocyst density. The significance of these findings goes back to the ongoing debate of the conservation of th e species. The findings in this study suggest that parrotfish do have a negative impact on the reefs ; however the impact that they have may not be significant enough to be profound. Both the destructive and beneficial qualities of parrotfish on coral re efs is the topic of much debate within the coral reef academic community. Mumby (2009) reviewed the literature to date concerning whether or not parrotfish were beneficial to coral reefs and came to the conclusion that although they clearly are impacting the reefs through corallivory any attempts to avoid the conservation of the species are unwarranted at this point because their role as major herbivores is highly beneficial. He addressed several different factors including coral recruitment partial a nd whole coral colony mortality and coral growth physiology and reproduction. Coral recruitment was positively impacted by parrotfish due to the increase of grazing and the decrease of macroalgae. This conclusion was drawn despite past evidence that parrotfish may reduce recruitment due to damage to juvenile corals (Bak & Engel 1979 ; Birkeland 1977). Partial colony mortality was shown to vary significantly among different species of corals with the star corals M annularis and M faveolata as the most highly preyed upon species. These species have a high capacity for tissue regeneration therefore although they are heavily grazed and preyed upon the effects do not totally impair recovery. Whole colony mortality was not considered to be a major issue because few if any studies have shown parrotfish to be the sole cause of colony mortality. The only

PAGE 20

IMPACTS OF FORAGING BEHAVIOR 13 cases in which parrotfish were accused of full mortality were faulty in that the cause of the mortality could not be supported by the observations made (Rotjan & Lewis 2005). Coral growth may be affected by parrotfish predation (Rotjan & Lewis 2006). Bite lesions may contribute to stress for corals particularly in Montastraea spp. Reproduction on the other hand may be benefitted by parrotfis h grazing because it reduces macroalgae therefore alleviating the impacts due to macroalgae on coral fecundity. This review highlights several issues that must be considered when addressing the role that parrotfish particularly S. viride play on the r eef. Although they are the main corallivores in the Scaridae family remaining as a coral reef but also for the benefits to the coral colonies themselves with the removal of macroalgae that can impai r recruitment and reproduction. III. Social System S. viride is a sexually dichromatically dimorphic species meaning that males and females differ in coloration. They are also protogynously sequent ially hermaphroditic meaning Initial Phase (IP) female s of the species are capable of becoming T erminal P hase (TP) males. Many members of the Scarid family are both dichromatically dimorphic and protogynously sequentially hermaphroditic (Choat & Robertson 1975). A list of Scarid species can be found in App endix A. Hermaphroditic fishes employ different sex change strategies. S. viride undergo an early sex change (Munday Buston & Warner 2006). Munday et al. explain that early IP to TP sex changes occur when there is a higher chance of taking over a ha rem if

PAGE 21

IMPACTS OF FORAGING BEHAVIOR 14 another TP male dies. Changing sex early allows the fish to be in a better position to take over a territory An early sex change strategy can yield a higher reproductive value for the species by taking over territories The social system of S. vir ide is quite complex. It is se parated into one male and multi male groups. van Rooij et al. (1996) took an in depth look at the S. viride social and mating system on a reef off Bonaire Netherlands Antilles. In order to assess these systems they took records of behaviors and movements of the fish for three years. Some fish wer e observed for the entire three year period. The researchers took censuses in order to assess the vertical distribution of the fish. IP members were caught and tagged for recog nition purposes. TP members were identified by individual markings. Territories size of foraging ranges exclusivity of range stability of range and status social interactions reproductive activity and abundance of the members were recorded. Results found that all recognized fish both IP and TP demonstrated a strong site attachment. Due to juveniles frequently losing their tags they were not studied. There was a marked vertical distribution of the TP. TP members can be broken down into two cate gories: multi male groups occur in water shallower than 3 m and one male groups in water 3 m and deeper. Multi male groups were defined as groups including multiple TP males and several IP members whereas one male groups were defined as groups including one to two TP males and multiple IP members (van Rooij et al. 1996). In one male groups an individual TP male controlled a harem of several IP members. Occasionally there would be more than one territorial TP male in the group. There was very little overlap of ranges. One male members were found to deter conspecifics not belonging to their group but not members of other species from their

PAGE 22

IMPACTS OF FORAGING BEHAVIOR 15 territories. They would actively patrol and control the borders of their territories. At grazing sites especi ally in locations where grazing scars on corals were found interesting interactions were observed. Aggressive and territorial interactions were followed by the territorial TP taking bites from the scars and then other group members in order of decreasing size biting from the same scar. This suggested a hierarchical structure in the territories. In multi male groups many TP members lived together in a harem with many IP members. Ranges of most individual TP members from the group overlapped by 40% 100 %. Conspecifics from other groups were patrolled against and kept out. A dominance hierarchy was also observed in the multi male members. This was concluded from observations of smaller group TP males withdrawing from the area where larger group TP male s were found. These results suggest that one male TP members are more territorial against other individual TP members than those in multi male TP members. Because of this social interactions including social foraging may differ by type of group. Due to their social foraging multi male groups create a higher grazing pressure on their group areas than grazing pressure by one male groups in their territories. Social foraging behavior is also mediated by other factors such as different structural attri butes across coral reef habitats. Auster and Lindholm (2008) found that there were higher rates of social foraging in low complexity habitats such as in rubble sand habitats where there were few protective areas from predators and the distribution of prey items were patchy. The reef observed was located in the Florida Keys called Conch Reef. It was divided into four different habitats: continuous reef coral rubble sand reef

PAGE 23

IMPACTS OF FORAGING BEHAVIOR 16 edge and spur and groove. Social foraging bouts during surveys were observed in each of these habitats in 5 m x 5 m patches. Bouts were recorded as either a single or mixed species foraging group. Each survey was 20 minutes and 90 species of fishes were recorded. Thirty six point seven percent of species participated in singl e and/or mixed species foraging bouts. There were variable rates of social foraging across the different habitats despite there being similar species and community composition among them. Sparisoma viride or the stoplight parrotfish only participated i n two bouts or .57% of social foraging bouts. Both bouts were in mixed species foraging. One issue with the study was that they only recorded in depths between 15 20 feet. This is problematic because in some species of fish the social structure may di ffer at different depths as in S. viride Perhaps the researchers would have found more social foraging by S. viride (or other species as well) if they had recorded in shallower depths where there are multi male groups that would potentially forage toget her more often. IV. Present Study Behaviors of the Stoplight Parrotfish may vary depending on phase depth and social structure (Auster & Lindholm 2008). Terminal Phase members are more territorial than Initial Phase members because they are the domi nant members of their groups. Additionally one male groups are more territorial than multi male groups (van Rooij et al. 1996). This suggests that feeding behaviors may be different between one male and multi male groups and IP and TP members if they are more territorial.

PAGE 24

IMPACTS OF FORAGING BEHAVIOR 17 Spot biting and focused biting are considered territ orial behaviors (Bruckner 2000). If one male groups are more territorial than multi male groups then it is safe to assume that the destructive behaviors of spot biting and focu sed biting would more likely occur in one male groups. Additionally this can be said about IP and TP members with TP members more likely to spot and focused bite. Spot and focused biting are destructive behaviors as discussed earlier. Therefore the t erritorial behaviors of the species may have a negative impact on the reef. The c urrent study aimed to address the difference in behavior of S. viride between TP and IP members. Specifically the feeding and territorial behaviors of the species will be a ssessed. This will give an idea of how vertical distribution and phase of the fish impact the reef. The hypotheses are : 1) TP males have more of an impact than IP members. 2) Members in deeper areas of the reef (one male) have more of an impact than those in s hallower depths (multi male). Impact is measured as either positive or negative. Positive impact is consumption of algae and negative impact is consumption of live coral. One implication of this study may be the conservation of S. viride Is it worth promoting the conservation of the species if it is indeed destructive to the reef? Additionally this study may provide some insight as to whether or not S. viride is a keystone species. According to Mills Soule and Doak (1993) keystone species are b roadly understood as species that maintain a vital role in ecological communities. Loss or removal of these species can potentially cause detrimental effects in their respective ecosystems. There are several types of keystone species: predator prey pla nt link and

PAGE 25

IMPACTS OF FORAGING BEHAVIOR 18 modifier. S. viride could potentially be placed in the modifier category. Keystone modifiers are species that modify their habitats in ways that maintain stable states. Their impacts can prevent other species from entering the ecosystem as well as keeping the conditions of the ecosystem conducive to the survival of other species. One well known keystone species is D. antillarum or the long spined sea urchin. The widespread loss of populations of the urchin caused immense macroalgal growt h on Caribbean reefs (Aronson and Precht 2000). D. antillarum is considered a reef modifier (Mills et al., 1993). Considering S. viride are also reef herbivores and may have similar modifying impacts their role can potentially be compared to that of D. antillarum Above all the current study will provide a greater understanding and insight to the role S. viride plays on the reef. Understanding the interactions of S. viride with different ecosystems can play a large part in the overall conservation o f the species. Roth (2005) also studied a population of herbivores including S.viride on Coral Key in Bocas del Toro Panama. The author found that the population of herbivores on the reefs off the island consumed 30 100% of the algal turf production. It was also determined that fish size was directly correlated to higher algae consumption therefore larger reef herbivores such as S. viride are vitally important in maintaining the stable state of a coral reef. The author implied that if the population sizes of these large reef herbivores were reduced a phase shift would likely occur therefore management and conservation of these herbivores is necessary to preserve the coral reefs. Another parrotfish species Bolbometopon mericatum the bumphead parr otfish is currently on the NOAA species of concern list (NOAA Species of Concern: Bumphead Parrotfish 2010). Populations of this typically corallivorous and sometimes herbivorous species have rapidly declined due

PAGE 26

IMPACTS OF FORAGING BEHAVIOR 19 to overexploitation destructive fishing techniques and loss of coral reef habitats. If S. viride encounters a similar fate to that of B. mericatum coral reefs may be detrimentally affected due to the vital role of the fish in grazing potentially harmful algae.

PAGE 27

IMPACTS OF FORAGING BEHAVIOR 20 Methods Subjects and Study Sit e Eighty stoplight parrotfish (40 IP 40 TP) were observed in the study No juvenile members were included All fish were observed off Conch Reef near the Institute for Tropical Ecology and Conservation (ITEC) in Bocas del Toro Isla Colon Panama (Fig 1) Isla Colon is an island in the greater archipelago of Bocas del Toro located in the North Western area of Panama Conch Reef is located in the Caribbean Sea Figure 1 Map of Panama Outlined in the red box is the Bocas del Toro archipelago re gion where the study was conducted Materials and Procedure The study required both SCUBA diving and snorkeling in early mornings to late afternoons While diving one 50 m transect was run horizontally across the reef A random point was predetermined through a random numbers table From this random point a second 50 m transect ran vertically down the reef A predetermined random point on the vertical transect was used as a reference point (Fig 2). From the reference point the nearest adult S vi ride was observed When observation was completed for a fish another predetermined random point on the vertical transect was used as a reference point and the nearest S viride from there was observed Once 10 fish were observed

PAGE 28

IMPACTS OF FORAGING BEHAVIOR 21 along the vertical tra nsect it was moved further across the reef by using another predetermined random point on the horizontal transect The procedure was repeated from there until 80 fish (40 IP and 40 TP) were observed Figure 2 Diagram of sampling method Horizontal y ellow line represents the first transect Red dot in the middle of the horizontal yellow line represents the first randomly determined point Vertical yellow line represents the second transect Red dot in the middle of the vertical yellow line represen ts the second randomly determined point from which the nearest S viride subject was found and observed Focal animal all occurrences sampling was used in observations meaning only one randomly determined subject was observed at a time and every occurr ence of the predetermined behaviors displayed by the subject were recorded in frequency tally counts This method was adopted because it is particularly effective in field studies to record every occurrence of particular behaviors for specific individual subjects (Altmann 1974 Lehner 1996) The aforementioned pilot study used an almost identical method of observation (Ortiz unpublished) Additionally Roth (2005) adopted a similar method in order to determine consumption rates of reef herbivores Five minutes of observation time were allotted to each fish A standard wait period of one minute was allotted to find a fish if it went out of sight If the fish went out

PAGE 29

IMPACTS OF FORAGING BEHAVIOR 22 of sight for longer than one minute observation for that fish was terminated Al l behavioral observations were recorded on underwater PVC dive tablets Feeding behaviors observed included consumption of substrate algae coral algae live coral and sponge These behaviors were recorded as individual bite counts Other behaviors re corded included color changes chasing behaviors dorsal fin raises cleaning and defecation Each individual bite or occurrence of other behaviors were tallied and counted for each fish Data were inputted into an Excel file after observations were compl ete Bite frequencies were converted into bites per minute rates. Statistical analysis was run using SAS Due to skewed distributions and unequal sample sizes, ANOVAs could not be run on the raw rate data Log transformations were attempted; however, di stributions did not change much Therefore, bite rate scores were converted to ranks and 2x2 ANOVAs were performed on these ranks, an essentially non parametric ANOVA (Conover and Iman, 1981).

PAGE 30

IMPACTS OF FORAGING BEHAVIOR 23 Results Depth and Phase Depth was initially recorded in fe et. Data were converted from feet to meters When analyzed depth was divided to make a dichotomous v ariable of either shallow (<4.6 m ) or deep ( 4.6 m ) (c.f. van Rooij, et al., 1996) Of the 80 total subjects in the study 51 fish were found in shallow waters and 29 fish found in deep waters When analyzed by phase 22 Initial Phase members were found in shallow areas and 18 were found in deep areas Twenty nine Terminal Phase males were found in shallow waters and 11 found in deep waters (Table 1) Mean depth for IP members was 4.7 m ( SD = 1.7 ) Mean depth for TP members was 4.1 m ( SD = 1.7 ) Phase and depth had no significant correlation ( p = 09) Table 1 Numbers and Percentages of Sparisoma viride Members in Shallow and Deep waters Initial Pha se Terminal Phase Total Depth N % N % N % Shallow 22 55% 29 72% 51 63.75% Deep 18 45% 11 27% 29 36.25% Total 40 100% 40 100% 80 100%

PAGE 31

IMPACTS OF FORAGING BEHAVIOR 24 Consumption Median consumption rates and interquartile ranges were determined for each food type Medians are reported because the distribution for consumption was skewed. A comparison of median bites per minute for each food type by IP and TP members is in Figure 3. Substrate algae and coral algae were combined into a single variable nam ed When appropriate statistics were run on each type of algae All other behaviors including color changes chasing behaviors dorsal fin raises cleaning and defecation behaviors were discarded due to lack of sufficient data For overall alg ae consumption, a main effect was found for phase F(1 76)= 9.10 p =0.0035. IP members consumed more algae than TP members There was no statistically significant main effect found for depth on bites of algae F(1 76)= 0 00 p = 0 9890 showing that diffe rent depths did not significantly affect the consumption of algae No statistically significant interaction was found between depth and phase F(1 76)=1 68 p =0 1984 The effect becomes clearer when types of algae are separated. In order to determine w hat type of algae resulted in the main effect two separately on substrate algae and coral algae No significant main effects were found for substrate algae bites when analyzed across depth (F(1 76)= 1.03 p =0 3135 ) phase (F(1 76)= 11 p =0 7373 ) and the interaction between depth and phase (F(1 76)= 16 p =0 6873 ) results showed that none of the variables could predict differences in consumption However, a significant main effect was found for coral algae bites across phase F(1 76)= 10.26 p =0 002 0 IP members consumed sign ificantly more coral algae

PAGE 32

IMPACTS OF FORAGING BEHAVIOR 25 than TP members No significant main effect was found for depth (F(1 76)= 30 p =0 5838 ) and no significant interaction was found between phase and depth (F(1 76)=1 01 p =0 3184 ) F igure 3: Medians of Food Type Bite Rate per Minute per fish by Initial and Terminal Phase Sparisoma viride There was a significant main effect found for live coral bites across phase F(1 76)= 12 19 p =0 0008. IP members consumed more live coral th an TP members There was no statistically significant main effect found for depth on consumption of live coral F(1 76)= 0 p = 0 9711 No statistically significant interaction was found between depth and phase F(1 76)=1 54 p =0 2188 In order to determi ne whether there was another depth more appropriate to divide shallow and deep residents a median test was run on the depth data. Four meters was determined as the median depth for the sample studied so the results were rerun where the shallow residents were those found in <4 m. and deep residents were found in 4m. Bite Rate per Minute F(1,76) = 7.87, p = .02; ** F(1,76) = 9.71, p = .01; *** F(1, 76) = 12.56, p = .001.

PAGE 33

IMPACTS OF FORAGING BEHAVIOR 26 However the distributions were still similarly skewed and the ANOVA on ranks results did not differ much from the original results. Despite changing the depth threshold for social structu re differences depth st ill did not predict consumption rates.

PAGE 34

IMPACTS OF FORAGING BEHAVIOR 27 Discussion It was initially proposed that Terminal Phase members would have more of an impact on the reef than Initial Phase members. Positive impact was measured as the consumption of algae because herbivory is beneficial for coral reefs to maintain a coral dominated stable state (Bruggeman, et al., 1996). Negative impact was measured as the consumption of live coral, because the destruction of the corals by corallivory is considered by so me researchers to be detrimental to the reefs (Bruckner, et al. 2000). The hypotheses were not supported by the data collected in the study. Although phase did predict consumption IP members consumed more of al l food types than TP members. Therefore I P members had both a greater positive and negative impact than TP members on the reef. This is assuming that positive impact is me asured by the bite counts or consumption rates of algae and negative impact is measured by the bite counts or consumption rates of live coral. Additionally depth did not predict consumption across any food type also not supporting the hypotheses. The trends of the current study were consistent with those found in the pilot study on Conch Reef (Appendix A). Algae was the m ost consumed food type, and very little live coral was consumed. Additionally, M. annularis was the most consumed live coral. Very few territorial and agonistic behaviors were observed, as well, which is why they were not reported. Although the pilot st udy was inconclusive, the findings of the current study lend greater insight to the dynamics and factors contributing to the impact S. viride has on Conch Reef.

PAGE 35

IMPACTS OF FORAGING BEHAVIOR 28 Depth T here was a skewed vertical distribution of Sparisoma viride on the reef. More IP and TP individuals were found in shallower than 4.6 m than in 4.6 m and deeper Additionally t he analyses run on the data indicated that consumption of live coral and algae particularly coral algae could be predicted by phase but not by depth. T here were no interactions of phase and depth across any of the variables. This posed a question as to whether or not the social structure of S. viride varied based on vertical distribution with this particular population or whether or not the social structure cha nged at 4.6 m This presents questions as to why they were not distributed more evenly One explanation may be that there were more resources available in the shallow depths; therefore it was more advantageous for the fish to reside there. However reso urce availability has not been documented on Conch Reef so this cannot be confirmed. Another possible explanation may be the distribution of different groups. As stated earlier one male groups are typically found in deeper areas whereas multi male grou ps are found in shallow areas. This may in turn result in less overall individuals in deeper depths because the territories of one male groups usually do not overlap (van Rooij et al 1996). However there is a question of the validity of these assum ptions as far as applying this social structure to all populations of the species Although a median test was run to determine if there was a more appropriate depth to divide deep and shallow residents the results were skewed similarly to the original di stribution. This suggests that the assumption that coral consumption behaviors were territorial in nature with deeper residents in one male groups being more territorial may not be a valid assumption. Alternatively it can also suggest that territorial ity is not

PAGE 36

IMPACTS OF FORAGING BEHAVIOR 29 necessary in this population. Overall resources may be abundant enough to support the population making it less necessary for territorial defense. Krebs and Davies (1993) discussed the economics of resource defense. If resources are high in an ecosystem it is not advantageous for animals to defend them because the energy exerted for defense is not worth sacrificing foraging time. The animals do not gain any additional resources by defending them because the resources are so abundant. Again future studies would need to assess resource distribution in the area to determine this; however, Cooper (unpublished, 2010) noted that another reef ecosystem on Bocas del Toro was algae dominated. This may also apply to Conch Reef in Bocas del Toro. I f this were the case, the abundance of early succession macroalgae would likely make it less necessary for resource defense by S. viride which supports the findings of the current study. Phase Although depth did not predict consumption phase did. Howe ver contrary to the hypotheses IP members consumed more of all food sources than TP males particularly coral algae and live coral. One explanation that was not taken into account during the current study was nutritional yield per bite of food. Brugg e m an et al. (1994) conducted a study that assessed food preferences of S. viride in socially determined habitats. They looked at both the size of the grazing scars and the nutritional value of the different food types consumed in order to determine nutriti onal yield per bite. They were able to determine that differences in substrate types in deeper reef regions may result in a higher nutritional yield per bite in S. viride However they did not indicate whether one phase had a higher nutritional yield pe r bite than the other. The results of the current study

PAGE 37

IMPACTS OF FORAGING BEHAVIOR 30 present a question that can be answered by future studies as far as whether or not TP males have a higher yield per bite than IP members. Perhaps TP males were simply taking larger bites than IP me mbers. Future Directions As demonstrated in this study bite counts may not be a complete measure of food consumption for S. viride A possible alternative to this method is to measure and record random bites taken by S. viride in order to determine bit e size and food type. This would account for yield per bite because larger bite sizes may in turn result in higher nutritional yields. Additionally taking samples of the food sources and determining their nutritional quality would lend valuable insight S. viride may be taking different sized bites depending on the nutritional quality of the food. Future studies should also assess whether or not TP males have a higher yield per bite than IP members. Other factors to take into account are ecosystem di fferences between differ ent reefs. For example Conch R eef may have higher or lower sedimentation than other reefs worldwide due to runoff from land based activities resulting in different rates of microalgal and macroalgal growth. Additionally resourc e abundance also varies across reefs which may result in behavioral differences among different populations. Individual differences among ecosystems may play a large part in the behavior of different species on reefs. Another question that should be ans wered is whether or not S. viride is a keystone species. The results of the current study indicate that the population studied consumed much more algae than live coral indicating that the destruction caused by the fish was

PAGE 38

IMPACTS OF FORAGING BEHAVIOR 31 minimal compared to their posit ive impact However the algae consumption of S. viride should be compared to that of other keystone species such as D. antillarum D. antillarum was a major herbivore on Caribbean reefs. They maintained the stable state of coral reefs and prevented ph ase shifts to algal reefs. However was detrimental to reefs particularly in Jamaica. Macroalgal cover increased significantly and a phase shift occurred (Aronson and Pretcht 2000). However Rogers and Lorenzen (2008) discus sed that an increase in the population of D. antillarum on these reefs would potentially help restore the reefs to a coral reef state. It is reasonable to believe that the consumption of algae of the population studied in the current study is great enough to pla y a large part in preventing a phase shift from a c oral reef to an algal reef. Additionally similarities and differences between the current population and other he population in the current study may be beneficial enough to be a keystone species they may be much more destructive in other locations. Implications S. viride is clearly a major reef herbivore on Conch Reef. There was little corallivory displayed i ndicating that they were not as destructive as they were beneficial. This provides evidence o f their importance on this reef. This species is worth conserving in order to maintain a stable state on the reef Ways to protect this species include prevent ing possible overfishing as well as avoiding increases in sedimenta tion from land based activities If sedimentation continues or increases the reef may reach its threshold resulting in a phase shift where

PAGE 39

IMPACTS OF FORAGING BEHAVIOR 32 late succession macroalgae will grow (Hughes et al. 2010). S. viride consumes mostly early succession macroalge therefore they are less likely to graze if late succession macroalge dominates the reef ( McClanahan et al. 2000) If the population of S. viride declines it poses further implication s for the reef including other reef herbivores as well as the state of the reef itself. Potentially a major destructive phase shift may occur as in Jamaica with the dieoff of D. antillarum in the ea S. viride is imperative to t he conservation and resilience of coral reefs.

PAGE 40

IMPACTS OF FORAGING BEHAVIOR 33 References Altmann, J. (1974). Observational Study of Behavior: Sampling Methods. Behavior 49: 227 265. Aroson, R., Precht, W. (2000). Herbivory and Algal Dynamics on the Coral Reef at Discovery Bay, Jamaica. Limnology and Oceanography 45 : 251 255. Auster, P., Lindholm, J. (2008). Variation in social foraging by fishes across a coral reef landscape. Proceedings of the 11 th International Coral Reef Symposium : 286 290. Bak, R., Engel, M. (1979). Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Marine Biology 54: 341 352. Bellwood, D., Choat, H. (1991). A functional analysis of grazing in parrotfishes (f amily Scaridae): the ecological implications. Environmental Biology of fishes 28 : 189 214. Birkeland, C. (1977). The importance of rate of biomass accumulation in early successional stages of benthic communities to the survival of coral recruits. Proceed ings of the 3 rd International Coral Reef Symposium: 16 21. Bruggeman, J., van Oppen, M., Breeman, A. (1994) Foraging by the stoplight parrotfish, Sparisoma viride Food selection in different, socially determined habitats. Marine Ecology Progress Series 106 : 41 55. Bruckner, A. Bruckner, R., Sollins, P. (2000). Coral Reefs 19 : 50.

PAGE 41

IMPACTS OF FORAGING BEHAVIOR 34 Bruggeman, J. van Kessel, A, van Rooij, J., Breeman, A. (1996) Bioerosion and sediment injestion by t he Caribbean parrotfish Scarus vetula and Sparisoma viride : implications of fish size, feeding mode and habitat use. Marine Ecology Progress Series 134 : 59 71. Bykbanashian. (2011, August 7). Fish Identification: Find Species Family: Scaridae Parrotfish es. [Electronic List]. Retrieved from http://fishbase.org/identification/specieslist.php?famcode=364 and areacode= Choat, J ., Robertson, D. (1975). shes of the family Scari dae. 263 283 in R. Reinboth, ed. Intersexuality in th e animal kingdom Springer, Hei delberg. Conover W., Iman, R. (1981). Rank transformations as a bridge between parametric and nonparamentric statistics. The American Statistician 35 : 124 129. Hugehes, T., Graham, N., Jackson, J., Mumby, P., Steneck, R. (2 010). Rising to the challenge of sustaining coral reef resilience. Trends in Ecology and Evolution 25 : 633 642. Krebs, J., Davies, N. (ed.) (1993). Behavioural Ecology, an Evolutionary Approach. Oxford: Blackwell Scientific. Lehner, P. (1996). Handbook of Ethological Methods. Cambridge, UK: Cambridge University Press. McClanahan, T., Bergman,K., Huitric, M., McField, M., Elfwing, T., Nystr m, M., Nordemar, I. (2000). Response of fishes to algae reduction on Glovers Reef, Belize. Marine Ecology Progress S eries 206 : 273 282.

PAGE 42

IMPACTS OF FORAGING BEHAVIOR 35 McManus, J., Polsenberg, J. (2004). Coral algal phase shifts on coral reefs: ecological and environmental aspects. Progress in Oceanography 60 : 263 279. Miller, M., Hay, M. (1998). Effects of fish predation and seaweed competition on th e survival and growth of corals. Oecologia 113 : 231 238 Mills, L., Soule, M., Doak, D. (1993). The Keystone Species Concept in Ecology and Conservation BioScience 43 : 219 224. Mumby, P. (2009). Herbivory versus corallivory: are parrotfish good or bad f or Caribbean coral reefs? Coral Reefs 28 : 683 690. Mumby Peter J. and Wabnitz Colette C.C. (2002) Spatial patterns of aggression, territory size, and harem size in five sympatric Caribbean parrotfish species Environm ental Biology of Fishes 63 : 265 279 Munday, P., Buston, P., Warner, R. (2006). Diversity and flexibility of sex change strategies in aminals. TRENDS in Ecology and Evolution 21: 89 95. National Oceanic and Atmospheric Administration, National Marine Fisheries Service (2010). Species of Con cern: Bumphead Parrotfish (Bolbometopon muricatum). Rogers, A., Lorenzen, K. (2008). Recovery of Diadema antillarum and the potential for active rebuilding measures: modeling population dynamics. Proceedings of the 11 th International Coral Reef Symposium, Session 20. Roth, M. (2005). Does Fish Size Matter? Evaluating the Effect of Herbivorous Fish Size on Coral Reefs (Unpublished Masters Thesis). Center for Marine Biology and Conservation, Scripps Institution of Oceanography, San Diego, California.

PAGE 43

IMPACTS OF FORAGING BEHAVIOR 36 Rotjan R. and Dimond, J. (2010). Discriminating causes from consequences of persistent parrotfish corallivory. Journal of Experimental Marine Biology and Ecology 390: 188 195. Rotjan R., Lewis S. (2005) astreoides. Marine Ecology Progress Series 305: 193 201 Rotjan, R., Lewis, S. (2006). Parrotfish abundance and selective corallivory on a Belezian coral reef. Journal of Experimental Marine Bi ology and Ecology 335 292 301. Sanchez J. Gil, M. Chasqui, L. Alvarado, E. (2004) Grazing dynamics on a Caribbean reef building coral. Coral Reefs 23 : 575 583. van Rooij, J., Bruggeman J., Videler, J., Breeman, A. (1995) Ontogenetic, social, spatial and seasonal variations in condition of the reef herbivore Sparisoma viride Marine Biology 123 : 269 275. Yoshioka, P. (2008). Fish mezograzers as gatekeepers of the species composition of coral reefs. Proceedings of the 11 th International Coral Reef Sympos ium, Session 10.

PAGE 44

IMPACTS OF FORAGING BEHAVIOR 37 Appendix A IMPACTS OF TERMINAL AND INITIAL PHASE SPARISOMA VIRIDE AGONISTIC AND FEEDIN G BEHAVIOR ON A REEF ECOSYSTEM IN BOCA DE L DRAGO PANAMA Abstract Recently Sparisoma viride the Stoplight Parrotfish was accused of destroying an d consuming live corals most notably Montastraea annularis and in some cases Porites porites (Hixon 1997). Because of this as well as their regular consumption of dead coral rock and algae they are considered a major source of bioerosion on the reef s (Hixon 1997). The foraging and agonistic behaviors of S. viride were studied on the Far Mangroves reef off the coast of Bocas del Toro Panama while SCUBA diving using the focal animal method. It was hypothesized that initial phase S. viride had less of an impact on the reef than terminal phase S. viride Other than algae being the most consumed resource and M. annulari s being the most consumed live coral t he results of the study were inconclusive. However there were several notable behaviors and p henomena observed including no terminal phase members observed in depths greater than 15 feet no notable agonistic behaviors and the terminal phase members being much more timid than the initial phase members. Although there was no statistical signific ance there were many unanswered questions that can lead to future research including whether or not S. viride is a keystone species on the reef. Keywords: Sparisoma viride Stoplight Parrotfish corallivore algae grazing territorial behavior.

PAGE 45

IMPACTS OF FORAGING BEHAVIOR 38 Introduc tion Increasing attention has been devoted to the role of herbivorous and corallivorous reef fish on the character of coral reefs. There has been a recent ongoing debate concerning live coral consumption by parrotfishes specifically S. viride and its im pact on the reef. In 1988 as cited in Charles Kiene researched bioerosion on the Great Barrier Reef. He found that scarids are a main source of bioerosion. Choat (1991 as cited in Sale) found similar results in that he regarded scarids as major bioer oders of the reef which he documented in his paper on the biology of herbivorous fishes. Mc Afee and Morgan (1996) investigated resource use in parrotfishes off the San Blas Islands of Panama. They found that S. viride consumes primarily dead coral and that aggressive non sexual encounters among Scarids are atypical; however they also mention that in time that may change as resources decrease and competition increases. This may have occurred given that the article was written 14 years ago. In 1996 an article by van Rooij et al. discussed the social mating systems of S. viride in one male vs. multi male groups. They noted that in one male groups the terminal phase males control up to 77% of the reef and have more access to higher yield food patc hes. Bruckner (2000) examined S. viride on M. annularis live corals. He defined and differentiated between the two types of biting and coral lesions caused by S. viride He also noted that tissue affected by S. viride typic ally regenerates. Sanchez et al. (2004) published a paper on grazing dynamics on Caribbean reef building coral in which they discussed the grazing of S. viride on M. annularis. They found that M. annularis was able to tolerate the grazing pressure cause d by S. viride Nemeth (2008) noted the population of S. viride in La Paraguera Puerto Rico. He concluded that S. viride was

PAGE 46

IMPACTS OF FORAGING BEHAVIOR 39 most commonly found in the shallow fore reef zone which was 3 meters or less and that this was likely because there were more preferable food resources at a shallower depth due to higher algae production. Currently the issue of whether S. viride is causing more destruction on the reef than benefiting it is a hot topic within the field. Parrotfish have always been seen as a vit al mesograzer (Yoshioka 2008). They take on the role of helping to control algae on the reef along with other herbivorous species. So what would happen if the parrotfish S. viride Is this a keystone species? Although the argument that S. viride is destroying some live corals may be valid it can also be counter argued that S. viride is vitally important on the reef and therefore is causing more good than harm. A behavioral reperto ire of parrotfish was established by producing an ethogram in order to obtain a better more qualitative understanding of generalized grazing on the reef. It was hypothesized that the terminal phase S. viride would be more dominant than the initial phase S. viride and therefore has more access to and defends resources specifically corals from initial phase conspecifics. Methods The species studied was the Stoplight Parrotfish Sparisoma viride The location of the study was at the Far Mangroves (Con ch Reef) site in Boca del Drago Isla Colon Bocas del Toro Province Panama. Data were collected using the focal animal method (Lehner 1996). A down slope transect line was set in place on the reef as a guiding point to ensure that a repeated observat ion and record of the same group of fish was not conducted. Each individual fish

PAGE 47

IMPACTS OF FORAGING BEHAVIOR 40 was given two minutes of observation time. Observation was completed in the late mornings and early afternoons. The behaviors recorded included feeding behaviors such as a lgae grazing live coral biting and defecation as well as agonistic behaviors such as chasing being chased raising of dorsal fin retreat and hiding. These behaviors were recorded using frequency tally counts. In the study the independent variable was fish phase and the dependent variable was agonistic behavior. The level of measurement was the ordinal level so the Mann Results Twenty four initial phase members were record ed while only six terminal phase members were recorded (see Table B1). No terminal phase members were found beyond 15 ft. It was found that the majority of foraging within both phases was algae grazing and biting (see Figure B1). However when S. virid e consumed live corals M. annularis was the live coral that was most consumed. Table A 1: Sums and averages of data collected from both phases. The Mann Whitney U test indicated that there was neither statistical significance e coral (IP mean: 4.1; TP N Mean D epth Algae B ites Live Coral B ites A gonistic behaviors Initial Phase 24 15.6 283 40 2 Terminal P hase 6 12.5 59 2 0

PAGE 48

IMPACTS OF FORAGING BEHAVIOR 41 mean: 2). Because of this differences of impact on the reef between the two phases were not determined. Figure A 1: Percentages of food selection of Sparisoma viride Both initial and terminal phases are represented. Fish members found in shallow areas of the reef (<15 ft.) and deep areas of the reef (>15 ft.) however the test could not be calculated due to a lack of sufficient data. Discussion The study was inconclusive given the limited amount of data; therefore the hypothesis could not be supported or disproved. However although there was no statistical significance found there were several noteworthy behaviors and phenomena observed durin g the study. As the data suggested live corals were rarely consumed and substrate algae were the most consumed resource. If live coral was consumed M. annularis findings (Bruckner Bruckner, and Sollins, 2000; Sanchez, Gil, Chasqui, Alvarado 2004)

PAGE 49

IMPACTS OF FORAGING BEHAVIOR 42 however the frequency of occurrence was much less than anticipated. This is likely because of the abundance of algae on the reef. Cooper (unpublished 2010) found algae dominated area s of reef within Bocas del Toro. Although the Far Mangroves site was not studied for algae dominance it is probable that the conditions are similar. It is reasonable to assume that S. viride did not consume significant amounts of coral at this site due to the unnecessary expense of energy required to do so. It was also observed that terminal phase males were rarely found beyond 15 feet and there were no terminal phase males recorded beyond that point. Literature of studies conducted in Bonaire noted the population and social dynamics of S. viride (van Rooij et al., 1996). They found that there are two types of social groups: one male groups and multi male groups. In one male groups one terminal phase male and several females live amongst each other in a harem. In multi male groups several terminal phase males live together in a group with multiple females. One male groups were found in deeper reef regions whereas multi male groups were typically found in shallower reef regions. This would suppor t the observation of the reduction of terminal phase males observed in deeper areas. However these same studies note that terminal phase males defend their home range against intruders. In the present study there were no significant agonistic behaviors observed. There were only two occurrences of agonistic behaviors recorded and both were between two initial phase members. This observation contradicts that of others that noted clear territoriality within S. viride (van Rooij et al 1996; Mumby and Wabn itz 2002). Again this may be due to the abundance of resources available on the reef. It can be assumed that S. viride would exploit the most readily available resources

PAGE 50

IMPACTS OF FORAGING BEHAVIOR 43 on the reef. Because algae were so abundant it was not necessary to actively def end areas of the reef. There were enough food resources for all members of the population. Lastly terminal phase males were much more timid than initial phase members. This contributed to the difficulty of finding and observing terminal phase males bec ause they were more likely to quickly flee and hide from the observer. On the other hand the initial phase members continued what they were doing without any clear concern about the observer close by. This was particularly interesting because past litera ture suggests that S. viride as a species is typically not shy. Although this study was inconclusive it can be viewed as a pilot study for future research. Terminal or initial phase members may differ in the consumption of more algae or more corals on t he reef. Given that S. viride may be an important part of maintaining the reef if it consumes more algae than coral the reef may maintain itself as a coral reef. However if it were to consume more coral than algae the reef may then shift to an altern ate state and become an algal reef. This would be interesting to study in future research both at Bocas del Toro and at other reef areas where this species is found. One burning question is whether or not S. viride is a keystone species. Is the reef larg ely dependent on this species? What would happen if it were to be removed from the ecosystem? These points and questions are important to the advancement of knowledge in this field because this species more than likely has a large impact on reefs. Lessi os (1988a) discussed the massive die off of Diadema antillarum in 1983 due to a waterborne disease. After this phenomenon there was a large influx of algae on reefs. The maintenance of the reef was largely dependent on D. antillarum If S. viride under went a similar die off there may be a major effect of an algae takeover on the reefs.

PAGE 51

IMPACTS OF FORAGING BEHAVIOR 44 Despite the inconclusive results of the present study there were clearly many observations that lead to unanswered questions about the species pertaining to resource u se depending on abundance phase distribution agonistic behavior and territoriality and the possibility of this species being a keystone species. This species is very likely a vital member on the reef and both phases may have a large impact on the reef However each phase may impact the reef differently due to the clearly different distributions and behaviors between the phases. Acknowledgements I would like to thank Dr. Al Beulig for his guidance with this study. I would also like to thank Loren Hi ll for assisting me with data collection and the Institute for Tropical Ecology and Conservation in Bocas del Toro Panama for providing the facilities and equipment to carry out my project. References Bruckner, A. Bruckner, R., Sollins, P. (2000). Parro Coral Reefs 19 : 50. Choat J.H. and D.R. Bellwood. (1991). The biology of herbivorous fishes on coral reefs In P.F. Sale (ed) The ecology of fishes on coral reefs San Diego CA: Academ ic Press de Girolamo M. et al. (1999) Social organization and sexual pattern in the Mediterranean Parrotfish Sparisoma cretense (Teleostei: Scaridae). Marine Biology 135 : 353 360. DeLoach Ned (1999) Reef Fish Behavior: Florida Caribbean Bahamas Jack sonville FL: New World Publications Inc.

PAGE 52

IMPACTS OF FORAGING BEHAVIOR 45 Kiene W.E. 1988. A model of bioerosion on the Great Barrier Reef in Birkeland Charles (ed) Life and Death of Coral Reefs New York NY: Chapman and Hall. Lehner P.N. (1996) Handbook of ethological methods N ew York NY: Cambridge University Press. Lessios H.A. 1995 Diadema antillarum 10 years after mass mortality still rare, despite help from a competitor Proc. R. Soc. Lond. B 259 : 331 337. Mc Afee S.T. Morgan S.G. (1996) Resource use by five sympatri c parrotfishes in the San Blas Archipelago, Panama. Marine Biology 2 5 : 427 437 Mumby Peter J. and Wabnitz Colette C.C. (2002) S patial patterns of aggression, territory size, and harem size in five sympatric Caribbean parrotfish species. E nvironmental Biology of Fishes 63 : 265 279. Neemeth M. (2008) Differential depth effects upon biomass in patterns in an herbivorous coral reef assemblage. Proceedings of the 11 th International Coral Reef Symposium, Ft. Lauderdale, FL van Rooij J.M. et al. (1996c ) The social and mating system of the herbivorous reef fish Sparisoma viride : One male versus multi male groups. Env ironmental Biology of Fishes 47: 353 378. Sanchez J. Gil, M. Chasqui, L. Alvarado, E. (2004) Grazing dynamics on a Caribbean reef building coral. Coral Reefs 23 : 575 583. Yoshioka P.M. (2008) Fish mesograzers as gatekeepers of the species composition of coral reefs Proceedings of the 11 th International Coral Reef Sympos1ium, Ft. Lauderdale, FL.

PAGE 53

IMPACTS OF FORAGING BEHAVIOR 46 Appendix B Table B1 Scaridae Genus Species and Common Name List Genus Species Common Name 1 Bolbometopon muricatum Green humphead parrotfish 2 Calotomus carolinus Carolines parrotfish 3 japonicus Japanese parrotfish 4 spinidens Spinytooth parrotfish 5 viridescens Viridescent parrotfish 6 zonarchus Yellowbar parrotfish 7 Cetosaurus bicolor Bicolor parrotfish 8 ocellatus Spotted parrotfish 9 Chlorurus atrilunula Bluemoon parrotfish 10 bleekeri Bleekers parrotfish 11 cyanescens Blue humphead parrotfish 12 enneacanthus Captain par rotfish 13 frontalis Pacific slopehead parrotfish 14 genazonatus Sinai parrotfish 15 gibbus Heavybeak parrotfish 16 japanensis Palecheek parrotfish 17 microrhinos Steephead parrotfish 18 oedema Knothead parrotfish 19 perspicillatus Spectacled parrotfish

PAGE 54

IMPACTS OF FORAGING BEHAVIOR 47 20 Chlorurus rhakoura (N/A) 21 strongylocephalus Steephead parrotfish 22 troschelli 23 Cryptotomus roseus Bluelip parrotfish 24 Hipposcarus harid Candelamoa parrotfish 25 longiceps Pacific longnose parrotfish 26 L eptoscarus vaigiensis Marbled parrotfish 27 Nicholsina denticulata Loosetooth parrotfish 28 usta colletti (N/A) 29 Scarus altipinnis Fillament finned parrotfish 30 arabicus Arabian parrotfish 31 caudofasciatus Redbarred parrotfish 32 chameleon Ch ameleon parrotfish 33 chinensis (N/A) 34 coelestinus (N/A) 35 coeruleus Blue parrotfish 36 collana Red Sea parrotfish 37 compressus Azure parrotfish 38 dimidiatus Yellowbarred parrotfish 39 dubius Regal parrotfish 40 falcipinnis Sicklefin p arrotfish 41 ferruguneus Rusty parrotfish 42 festivus Festive parrotfish

PAGE 55

IMPACTS OF FORAGING BEHAVIOR 48 43 Scarus flavipectoralis Yellowfin parrotfish 44 forsteni 45 frenatus Bridled parrotfish 46 fuscicaudalis Darktail parrotfish 47 fuscipurpureus Purp le brown parrotfish 48 ghobban Blue barred parrotfish 49 globiceps Globehead parrotfish 50 gracilis (N/A) 51 guacamia Rainbow parrotfish 52 hoefleri Guinean parrotfish 53 hypselopterus Yellow tail parrotfish 54 iseri Striped parrotfish 55 k oputea Marquesan parrotfish 56 longipinnis Highfin parrotfish 57 maculipinna Spot fin parrotfish 58 niger Dusky parrotfish 59 obishime (N/A) 60 oviceps Dark capped parrotfish 61 ovifrons Knobsnout parrotfish 62 perrico Bumphead parrotfish 63 persicus Gulf parrotfish 64 prasignathos Singapore parrotfish 65 psittacus Common parrotfish

PAGE 56

IMPACTS OF FORAGING BEHAVIOR 49 66 Scarus quoyi 67 rivulatus Rivulated parrotfish 68 rubroviolaceus Ember parrotfish 69 russelii Eclipse parrotfish 70 scaber Five saddle parrotfish 71 schlegeli Yellowband parrotfish 72 spinus Greensnout parrotfish 73 taeniopterus Princess parrotfish 74 tricolor Tricolor parrotfish 75 trispinosus (N/A) 76 vetula Queen parrotfish 77 viridifucatus Roundhead parrotfish 78 xanthopelura Red parrotfish 79 zelindae (N/A) 80 zufar Dhofar parrotfish 81 Sparisoma amplum (N/A) 82 atomarium Greenblotch parrotfish 83 aurofrenatum Redband parrotfish 84 axillare (N/A) 85 choati West African parrotfish 86 chrysopterum R edtail parrotfish 87 cretense (N/A) 88 frondosum (N/A)

PAGE 57

IMPACTS OF FORAGING BEHAVIOR 50 89 Sparisoma griseorubrum (N/A) 90 radians Bucktooth parrotfish 91 rocha Trindade parrotfish 92 rubripinne Redfin parrotfish 93 strigatum Strigate parrotfish 94 tuiupiranga (N/A) 95 vi ride Stoplight parrotfish Note : List compiled from fishbase.us Fish Identification: Find Species Family: Scaridae Parrotfishes.