Impact of Dugong Grazing and Turtle Cropping on Tropical Seagrass Communities
Total Page:16
File Type:pdf, Size:1020Kb
,, - Impact of Dugong grazing and turtle cropping on tropical seagrass communities 1 1 LEMNUEL ARAGONES ·~ and HELENE MARSH •3 The impact of grazing by two megaherbivores, the Dugong Dugong dugon and the Green Turtle Chelonia mydas on the community structure of intertidal seagrasses was investigated experimentally over two time frames (shorter term: 1-4 months; longer-term: 10 and 13 months), at three levels of grazing intensity (leaf cropping, light grazing, intensive grazing), at two seagrass meadows in tropical Queensland, Australia: (1) a mixed species bed of Zostera Ca.pricorni, Halophila ova/is, Halodule uninervis, Cymodoce~ rotundata and Cymodocea seriulata, and (2} a monospecific bed of Halodule uninervis; From the parspective of the- megah-erbivores,-grazing- improved the structure and dynamics of the tropical seagrass communities by altering their biomass, volume of detritus, ..net abovegrour.d biomass productivity and the species composition of. the mixed-species bed. Recovery from grazing disturbance occurred after several months tQ_ .f!y~ru . Key words: Seagrass, Disturb~nce, Grazing, Dugong, Green Turtle. INTRODUCTION subtropical ~vforeton Bay the growth of Halophila ovalis, a species preferred by Dugongs, is IN the tropics and subtropics, seagrasses promoted while the expansion of a less favoured provide food for threatened megafauna: the species, Zostera capricorni, is controlled (Preen Green Turtle Chelonia mydas (Garnett et al. 1985; 1995). Bjorndal et al. 1991; Read 1991; Brand 1995), and· three species of sirenians (sea cows): the The studies conducted to investigate the effects West Indian Manatee Trichechus manatus, the of turtle cropping on seagrass communities have West African Manatee Trichechus senegalensis, and been carried out mostly in the Caribbean region the Ougong Dugong dugon (Marsh et al. 1982, (e.g., Bjorndal1980, 1985; Ogden et al.l983; 1999; Reynolds and Odell 1991 ). Despite the •· Zieman et al. 1984). There are few similar sizes . of these megaherbivores (adult Dugongs studies for other tropical systems, particularly 350--400 kg; Manatees 350-1 600 kg; Green in Australia where Dugongs and Green Turties Turtles 200 kg), Thayer et al. (1984) assumed are found in large numbers (Ogden 1980; that their current densities are too low for Thayer et al. 1984; Marsh et al. 1995). The only them to make ~ignificant impact on seagrass studies of Dugong grazing in the tropics have communities. This assumption is unjustified in been in Indonesia (de Iongh et al. 1995) and areas where densities of Dugongs and Green Thailand (Supanwanid 1996) where Dugong Turtles are high. For example, Preen ( 1992) densities are low. estimated that about 600 Dugongs consumed We used an experimental approach to investi 28% of the total seagrass production in favoured gate the impact of grazing by Dugongs and areas· in Moreton B~y, Queensland, Australia. Green Turtles on the community structure of Dugongs and Manatees graze destructively by seagrasses at two mainland sites in the Great uproo~ing seagrasses when the rhizomes are Barrier Reef region in tropical Australia. The accessible (Marsh et al. 1982, 1999; Preen 1995; study was conducted over two time frames: Anderson 1998). When the rhizomes are not shorter-term (one to four months at Cardwell) accessible Dugongs and Manatees feed on and longer-term ( 10 months at Ellie Point and seagrasses by cropping their leaves (Anderson 13 months at Cardwell), and at several levels of 1981; Marsh et al. 1999), the mode of se~grass grazing intensity: intensive grazing, light feeding used routinely by . Green Turtles grazing, and cropping. Our experiments showed (Bjorndal 1980; Lanyon et al, 1989). These that herbivory influenced the structure and modes of foraging can have important effects dynamics of tropical seagrass communities by on the structure of seagrass communities. Green altering their biomass, net aboveground biomass Turtles in the Caripbean maintain "grazing productivity, and the species con1position of a plots" wherein individuals regularly recrop multi-species meadow. Recovery from grazing patches of seagrass for the younger leaf growth disturbance occurred in several months to a (Bjorndal 1980). The grazing disturi?ance created year. We conclude that the three most important by Dugongs may make individual seagrass properties used in assessing the feeding niches beds more desirable as Dugong habitats. In of herbivores, the quality, quantity, and botanical 'School of Tropi.:al Environment Studies and Geography, James Cook Univer~ity, Townsville, Au~tralia 4811, 1l'resent addre5s: Institute of Biological Sciences, University of Philippines, Los B~nos College, Laguna, 4031, Philippines. 'For correspondence. l'ACIFIC CONSERVATION BIOLOGY Vol. 5: 277-RS. Surrey Beatty & Sons, Sydn~:y. :woo. 278 PACIFIC CONSERVATION BIOLOGY compoSition of vegetation {Bell 1971; Jarman an average removal of about 70%, 80% and 1974), are positively correlated with the intensity 85% of the aboveground biomass of H. ovalis, of grazing by Dugongs and Green Turtles in Zostera/Cymodocea and H. uniner·uis, respectively.. tropical sea.grass ecosystems. This treatment simulated the level of removal by small groups of grazing Dugongs at sites MATERIALS AND METHODS in tropical Queens-land, where Dugongs have been recorded as removing .a!l. average of Experime!1ts were perfonned in the intertidal 63-:86% of seagrass biomass from feedipg trails regions of the two seagrass meadows, both of through various species incluc[ing, H. bvalis, which have 2 tidal range of approximately 3 m. Z. caprico·rn:i, and H. uninervis (Wake 1975). (3) 0 C ihec~mead_QW _a_t_ CEllie f>_ojn_t c(1fi 5_3~S_, _ 14_5° 4_6'£) LeaLc:rnp~ping: Jea~es_ ·cut l=-2~cm ahoYe __gxound was dominated-by Zostera capricorni (wide variety) throughout each experimental unit (except interspersed with some Halophila ovalis, Halodule for H . ovalis for which entire aboveground uninervis, Cymodocea rotund.ata and C. serrulata. biomass removed). This treatment simulated the The intertidal meadow at CardweH (l8°l5'S, feeding behaviour of Green Thrtles. (4) Controi: 146°01 'E) consisted of Halodule uninervis. Some undisturbed. Halophila ovalis occurred sub-tidally in this meadow. As discussed below evidence of Dugong Shorter-term experiments grazing was observed during the experiments at The shorter-term experiment at Cardwell was Cardwell, but not at Ellie Point monitored for four months. This experiment was basically similar tq the longer-term experiment Longer-term experiments but used four 6 x 6 Latin squares with the Grazing simulations were carried out at each f9Jlowiqg treatments: .( 1) Ltgl:J.t gt!.l?ip.g h~e~ted of .four haphazardly chosen locations within after four months; (2) Control for Treatment I; each meadow in May-June 1993 and monitored (3) Cropping 1 harvested after one month; for 11 months (Ellie Point) and 13 months (4) .Control for Treatment 3; (5) Cropping 2 (Cardwell). The four grazing treatments were harvested after two months; ·{6) Control for arranged in a 4 X 4 Latin square at each Treatment 5. Each experimental unit was reduced locae:ion. The treatments of the ·first row and to 0.4356 m 2 to accommodate an increased -.column of each Latin square were assigned number of'Sampling units.(JN = 144). ~ntensive randomly. PVC pipes hammered into the grazing was not included in the trearments substrate marked the corners of the 1 m 2 because the rnonit9ring period was insufficient experimental unit for each replicate. Each for recovery from this treatment. pipe had no more than 3 em protruding to Monitoring the grazing pl11ts allow relocation while avoiding the "halo effect" that tends to occur around structures on these Changes in seagrass species composition and seagrass beds (pers. obs.). A movable quadrat abundance (leaf biomass only) of the subunits fitted tightly on to these markers ·ensuring within each experimental unit were monitored exact relocation of each experimental unit systematically by Aragones using a vertically during the monitoring that followed the held video camera (Sony video 8 Handycam; experiments. This quadrat was divided into CDD-TR305E) while standing on a movable 0.11 m 2 subunits using colour-coded strings platform which fitted over the quadrat as enabling us to exactly relocate and re-photo detailed in Aragones (1996). Monitoring was graph subunits within each experimental unit. performed monthly for both longer-term experi Buffer zones at least 1 m wide were maintained ments; twice monthly for two months, then along all sides of each experimental unit monthly for the shorter-term experiment. to · reduce the likelihood of edge effects. We Evidence of natural disturbance e.g., Dugong. actively avoided disturbing the seagrass in these grazing was recorded and quantified using the buffer zones during our fieldwork. video camera. Grazing from fish and inverte brates was considered negligible and ignored. The four grazing treatments were as follows: (1) Intensive grazing: all aboveground material The aboveground biomass (g dry weight) removed from experimental unit; some below for each species in each frame of the video ground left. This is close to the level of removal recording (which matched the relevant experi recorded in favoured Dugong feeding sites mental subunit as delineated by the colour in sub-tropical Moreton Bay, where herds of coded strings within the movable quadrat) was Dugongs removed 96% of the aboveground estimated using a scoring scale and regression biomass and 95% of