WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 Diversity, abundance and distribution of intertidal invertebrate species in the Ningaloo Marine Park Final Report 30 May 2011 WAMSI Project 3.2.2b 1 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 1. Executive Summary 1.1 Date 30 May 2011 1.2 Project Title & Number 3.2.2b Diversity, abundance and distribution of intertidal invertebrate species in the NMP 1.3 Project Leader Dr. Robert Black, School of Animal Biology, University of Western Australia 1.4 Project Team Prof. Michael S. Johnson, School of Animal Biology, University of Western Australia; Dr. Jane Prince, School of Animal Biology and Oceans Institute, UWA; Dr. Anne Brearley, School of Plant Biology and Oceans Institute, UWA; 1.5 Dates Covered July 2007 to May 2011 1. 6 Summary (for even shorter account see section 2.1) Aims and approach A quantitative pilot study of the composition of the benthic community of macro- invertebrates on intertidal rocky platforms was undertaken to (A) provide detailed information on variation in biodiversity along the length of the Ningaloo Marine Park and (B) determine the appropriate design of a monitoring protocol powerful enough to determine predefined levels of change. These general overall aims were in the context of the Ningaloo Marine Park Draft Management Plan of 2004, which set out a vision of maintaining the ecological values in the Park, and protecting it from adverse human impacts. The design of research and monitoring schemes must include several crucial features: (1) adequate, replicated sampling for each combination of time, location and any other controlled variable; (2) adequate, replicated sampling in areas with and without human impacts; and (3) pre-defined, quantitative criteria for what constitutes an important, continuing temporal trend or concerning difference between the sanctuary zones and impacted areas, or between 2 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 3 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 2 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 sanctuary zones some time after their establishment and their initial conditions. Even well- designed studies have to overcome the challenges of i) natural variability and patchiness at different temporal and spatial scales, and ii) natural events that overwhelm, obscure, or counteract the effects of human impacts. The Ningaloo Marine Park Draft Management Plan of 2004 seemed to fail to appreciate, comprehend, or even acknowledge the existence and magnitude of natural spatial and temporal variability against which to judge what is major rather than minor, foreseen rather than accidental loss, or prolonged rather than transient. Of course, defining the size of differences or changes that are critical (the maximum acceptable impact or effect) in any precise quantitative way will be difficult because any general definition cannot apply to all components of an ecological community, at all times, and all places within an area as large as Ningaloo Marine Park. Therefore, cases probably need individual attention in setting appropriate effect sizes indicative of concern. One view is that critical effect sizes in advance with reference to the local environment, yet the Draft Management Plan was silent about this issue. Specification of effect sizes is most often in the context of power analysis and the importance of both kinds of errors: Type I (rejecting a null hypothesis when it is true), and Type II (accepting a null hypothesis when it is false). Both require specification in studies about potential impacts that use an approach involving formal hypothesis testing, and both are relevant to managers trying to make decisions to minimize harm. However, an alternative approach, parameter estimation with confidence intervals emphasizes that the confidence intervals serve the same function as hypothesis testing and show power automatically. There is no way of avoiding careful evaluation of effect size. Sites This project examined the assemblages of macroinvertebrates at 36 sites on rocky intertidal platforms from Mildura Wreck in the north to 3 Mile in the south of Ningaloo Marine Park during 2007 to 2010, visiting 18 sites twice. There were two or more sites in seven of the 3 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 4 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 3 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 Sanctuary Zones, and one site each in two other Sanctuary Zones, totaling 22 sites. Eleven sites were close to but outside Jurabi, Bateman Bay, Pelican, Gnarraloo Bay, and 3 Mile Sanctuary Zones, and the remaining three, outside sanctuary zones, improved the geographic distribution of the sites (Table 1 - next page). Within the array of sites, and times they were sampled, there are sets of sites that are suitable for making comparisons among geographical regions in the Park, Inside and Outside of Sanctuary Zones and Inside and Outside Sanctuary Zones at different times, and for contrasting spatial with temporal variability. These comparisons are addressed for specific cases in other parts of this summary, and in detail in the research chapters. Sampling and macroinvertebrates At each site, our standard sampling scheme involved careful searches of 20 1-m2 quadrats in order to count the number of individuals of each species. Overall, the 31059 individuals in the 1744 1-m2 quadrats were allocated to 289 species of invertebrates of which most were gastropods (127 species), but included cnidarians (50), echinoderms (33), crustaceans (28), bivalves (19), chitons (12), and unusual taxa (20). Ninety-two or 32% of these species occurred as 1 individual, so additional sampling will continue to discover new species, and it is unlikely that any sampling program will ever be able to extensive enough to reveal all the species living on rocky platforms in the Park. Sites north of Yardie Creek, and sites south of Bateman Bay shared many species, but there were species found in the north only or in the southonly, indicating that future studies wishing to include a complete view of the macroinvertebrates must include sites along the length of the Park. The wider distribution of the 102 species for which we have precise identifications (mostly but not restricted to gastropods) suggests that few species are restricted the Park, and many have distributions that extend to other states. 4 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 5 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 4 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 Table 1. Sites in relation to sanctuary zones, and nature of protection. Regions (shown in Figure 1) and sites are listed in order from north to south. * indicates sanctuary zones which do not extend to shore, so the sites are not strictly in the sanctuary zone; ** indicates that the shoreline is a “Special Purpose (Shore-based Activities) Zone; *** indicates a site used to study giant clams only. (Table 1 in Research Chapter 1). Location and Region Sanctuary Zone Sites Inside Zone Sites Outside Zone Mildura Wreck Mildura Wreck West Surfers Lighthouse Bay** North Surfers South B North of northern boundary: Jurabi In 1 Jurabi Out 2 Jurabi In 2 Jurabi** Jurabi Out 3 Jurabi In 4 Jurabi Out 4 Jurabi In 3 Jurabi Out 1 Mangrove Bay C Mangrove Bay Mangrove Point Mandu South Cobble South of southern boundary: Mandu Mandu South Flat Pilgramunna D South of southern boundary: Osprey Bay Yardie Creek North Yardie Creek South North of northern boundary: Bateman Bateman Bay In Bateman Bay Out 2 Bateman Bay Out 1 Coral Bay North E Maud** Coral Bay North no map*** Coral Bay South South of southern boundary: Pelican** Elle’s In Elle’s Out Gnarraloo Bay In 2 Gnarraloo Bay Out 2 Gnarraloo Bay* Gnarraloo Bay In 1 Gnarraloo Bay Out 1 F 3 Mile North South of southern boundary: 3 Mile* 3 Mile In 2 3 Mile Out 1 3 Mile In 1 3 Mile Out 2 5 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 6 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 5 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 Focal species - cowries Several of our analyses focused on particular species because of their abundance or status as a targeted or iconic species. Two potentially targeted species, the cowries Cyrpaea caputserpentis (serpent-head cowry) and C. moneta (money cowry) have wide geographic distribution, and within the Park one or both occurred at 26 sites, but in low abundance, rarely exceeding 1 m-2. Based on patterns of densities over time at these sites, recruitment and mortality of cowries varied considerably among sites, especially for C. moneta, even over the short period of this study. Comparisons of samples of cowries from four sites inside and four sites outside Jurabi Sanctuary Zone in February 2010 had post hoc powers to detect a two-fold difference in abundance of cowries inside and outside the zone of 0.113 and 0.399 for the two species (i.e., probability of making a Type II error (concluding there is no difference when there is) 0.887 and 0.601). Expressed in a different way, to achieve a power of 0.80, one of the convention values, these comparisons would need many more replicate sites (43 or 9) than seems possible, either logistically, or from lack of suitable sites. Cowries during daytime low tides are often hidden, and we tried to determine which microhabitat they preferred. At all the scales of our sampling, from sites within the Ningaloo Marine Park, to individual belt transects and 1-m2 quadrats to microhabitats within those, the occurrence of an individual cowry probably depends on more factors than we considered, and it is difficult to specify characteristics of prime habitat that apply to all sites. Focal species - giant clams The small giant clam, Tridacna maxima, is an iconic, tropical species with its brightly- colored mantle a conspicuous feature of many platforms at Ningaloo Marine Park, where it is unusually abundant. We investigated giant clams as a focal species because they provided a tractable system to investigate some general ecological issues about dynamics of populations. Understanding variability of recruitment and mortality is essential for assessing changes due to perceived disturbances or attempts to conserve populations. In the absence of long-term 6 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 7 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 6 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 studies, we examined population density and size-frequency distributions of T. maxima at 20 sites where the clams are abundant on discontinuous, intertidal rocky platforms attached to the shoreline. Abundance ranged over two orders of magnitude (0.05 - 8 m-2), and size ranged from 1.5 to 31.0 cm. The shapes of the size-frequency distributions varied substantially, indicating variability in recruitment and mortality, including failures of cohorts to recruit and catastrophic events of mortality. Consistency of recruitment, as indexed by the coefficient of variation of the size-frequency distribution, was greater toward the north of the park, on intertidal platforms with greater complexity across their widths, and with smoother surfaces in the part of the platform occupied by the clams. The average turnover time was estimated at 5.5 years, giving a median age of 13 years. Variation among sites was large, however, highlighting the importance of variability of the dynamics of local populations, and the need for long-term studies to understand any particular population. Focal species - abundant macroinvertebrates demonstrate spatial and temporal variability We selected 15 species for detailed analyses of spatial and temporal variability because they were most abundant overall in our sampling by 1-m2 quadrats. Ten were gastropods, three were bivalves, one a coral and one a sea urchin. The coral and small giant clam (T. maxima), get some of their energy from endosymbiotic zooxanthellae, as well as from small particles in the water. The bivalves are suspension feeders, depending on particles in the water. The vermetid uses mucus threads to capture particles from the water. The ceriths and the stromb probably feed on small organic particles in sediments. The turban shell, trochid, and sea urchin are herbivores. The thaids and cone are predators. Thus, this selection of species includes examples of most kinds of feeding by marine organisms. The coral, the bivalves, and the vermetid are permanently attached to the surface of the platform, and the urchin is usually associated with a depression in the surface, while the remaining species can move. The most important feature of the abundances of these 15 species of macroinvertebrates on rocky intertidal platforms at Ningaloo Marine Park is the pervasive spatial variability. Spatial 7 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 8 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 7 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 variability at a regional scale was obvious among four sanctuary zones (Jurabi, Bateman Bay, Gnarraloo Bay, and 3 Mile) for three of the seven species analyzed because the variance component exceeded 50% and the Zone term was statistically significant in an analysis of variance. However, the variance component ranged from 15 to 30 % in the other four, thus adding substantial variation over that present within the sites among the quadrats. Large- scale spatial variability was also obvious among sites in general because the magnitude of effect (ω2, one measure of effect size) was greatest for Site for 14 of the 15 species when compared with the magnitude of effect for Year and Year x Site, with highest values ranging from 36 - 70 % in six species and lowest values ranging from 15 to 1% in six others. Spatial variability at this scale, therefore, could be large, but was not universal, and depended on the species considered. Spatial variation was very much less between adjacent sites as judged by the small effect size of Status (In or Out of the sanctuary zone, kms apart). However, this would be an expected result if the platforms at adjacent sites were matched in physical attributes, as we attempted to do. Temporal variability between 2007 and 2009 was very much less than spatial variability. The magnitude of effect associated with Year ranged from 0 to only 5.4%. The two species with the highest values, Tectus pyramis and Turbo haynesi, were both encountered, mainly in 2007, as small, newly-recruited individuals. Perhaps this temporal variability reflects variation in the abundance of cohorts of recruits that do not survive well on the platforms. The species with no temporal variability were Echinometra mathaei, Conus sponsalis, and Morula uva. The sea urchin is known to have sporadic recruitment and long-lived adults, so this might explain our observations of exclusively adult urchins with little variation between years. The two gastropods belong to groups that can live for several to many years. Judging temporal variability by the effect sizes (measured as r, which ranges from 0.0 to 1.0) associated with Year in analyses of variance tables provides the same view of temporal variability as the magnitude of effects. The effect sizes in the analyses of the seven species at the eight sites inside and outside sanctuary zones were tiny, the largest only 0.081, and thus explaining only 0.66% of the variation. The effect sizes for these seven species were larger in the analyses of all the sites, ranging from 0.011 to 0.187, but these are still small effects by 8 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 9 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 8 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 conventional standards. The other eight species had effect sizes of 0.003 to 0.284, the largest approaching a “medium” effect size. For the 15 species the magnitude of effect is highly positively correlated with (effect size)2, so these measures provide almost identical views of temporal variability. These evaluations of spatial and temporal variability suffer from the same difficulty because in almost all the analyses there were statistically important interactions between places and time, meaning that there are extra additive effects of the individual combinations of levels of the factors associated with space and time over and above the main effects of the levels of space and time. However, given the dominating size of the spatial variability in most cases, the influence of the extra variability due to the interaction would be relatively small. There are some logical reasons why spatial variability far exceeded temporal variability in these data. The first is that the number of sites is so much greater than the two times; there was much more opportunity to find spatial variation. The second, related reason is that the two years are close together, and the processes that produce temporal variation, variations in recruitment and mortality with time, did not have long to act. Related to this is the dependence on life history characteristics of individual species in determining their population dynamics, the frequency and extent of numerical changes. Some species, such as Echinometra mathaei and Tridacna maxima, are known to be long-lived, so their populations show inertia, changing little from year to year, unless they experience catastrophic conditions. Many gastropods have life-spans of several years, and few species on the platforms would have annual life cycles. One suggestion about judging whether populations are stable over time has been to observe populations long enough that there is a complete turn-over of individuals in it. By this criterion, studies such as ours, or continuing monitoring schemes, probably need to involve 5 to 10 or more years. 9 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 10 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 9 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 Spatial and temporal variability in multivariate analyses of assemblages of invertebrates To analyze the composition of the assemblages of macroinvertebrates using multivariate methods for comparisons among sites and among times, we used the sum of each species sampled in the quadrats adjusted to be equivalent to the sum in 20 in the 1-m2 quadrats and transformed by log (x+1) to reduce the influence of extremely abundant species. We 10 calculated Bray-Curtis similarity matrices for all pairs of sites in several subsets of the data to estimate and compare spatial and temporal variability. We used non-metric multidimensional scaling ordinations (nMDS), and canonical analyses of principal co-ordinates ordinations (CAP) to visualize the relationships among sites, and permutational multivariate analysis of variance for formal statistical tests. The samples involved three years, 2007 (18 sites in north and south), 2008 (10 sites in the north), and 2009 (32 sites in the north and south of the Park). The composition of the assemblages of varied at all geographic scales that we considered. There were latitudinal differences, but considerable overlap between north and south parts of the Park. Sites in four categories of management (outside sanctuary zones, in sanctuary zones, in Special Purpose Area, and inshore of a sanctuary zone that does not extend to the shoreline) were intermixed in the ordinations, but sites in the south of the Park do not have fully protected intertidal platforms. The analyses of variance revealed that there were statistically significant interactions between Date (2007 and 2009) and Region (north and south) and Date and Sanctuary (nine zones in the Park). This means that the effect of Date was inconsistent between the two regions, and among the nine zones. For example, the four sites in Lighthouse Bay Sanctuary Zone differed little among each other and between years, while two sites in 3 Mile sanctuary were similar in composition of assemblage of invertebrates within years, but differed drastically between years. Spatial variability was 1.5 or 2.0 times larger that temporal variability. The detection of these statistical interactions between space and time suggest that future sampling of these sites could reveal influences of different management regimes. 10 of 30 WAMSI 3.2.2b Intertidal Invertebrates 1. Executive Summary 2. Key Findings and Recommendations 4. Communication and Outputs 30 May 2011 11 of 30