2 Spatial and Temporal Variation S in Species Composition in the Monterey Bay but Clarify

Total= 2+16+34+30+

Lit Cited= 5

[2] Spatial and Temporal Variations in Species Composition in the Monterey Bay but clarify that this is about kelp forests

Andrew Kim

Abstract. [16] An effective approach for identifying the abiotic or biotic mechanisms that determine species composition in a kelp forest involves describing the spatial variation in an area.[???] The Monterey Bay in central California is a nutrient-laden ecosystem that fuels kelp forests within which a great diversity of organisms take up residence. To identify variations in composition, we surveyed a suite of conspicuous species of fish, algae, and invertebrates at Hopkins Marine Station and Point Lobos State Reserve while accounting for the temporal variation inherent in our survey. Through our analysis, we were able to tease out the geographical features that may be the factor accounting for spatial heterogeneity in the assemblage of species at these otherwise similar sites. [and those were???]

Introduction [4, 3, 3, 4, 3, 2, 4, 2, 3, 3, 3] 34

Describing spatial or temporal variation inspecies composition of an environment[??]is an effective method of discovering the abiotic or biotic mechanisms that may determine patterns in assemblage and has been an oft used approach by marine ecologists studying kelp forest organisms(Broitman et. al. 2008, Duggins et. al. 1989, Estes et. al. 1989). Canopy-forming species of kelp provide a highly dynamic environment for an abundance of fish, invertebrate, and algal species to inhabitin the Monterey Bay. There are approximately 650 macroalgal, 500 fish, and thousands of invertebrate species that have been identified to interact within the kelp forest community in this area(Graham et. al. 2008). Species compositions in kelp forest ecosystems can vary spatially or temporally for a variety of reasons and understanding how they differ can provide insight into the intricacies of this coastal ecosystem.

Kelp forests can vary spatially within the Monterey Bay, from the Nereocystis beds that thrive in more turbulent waters to the Macrocystis beds that characterize others(Foster & Schiel 1985). Depending on the location in the bay, the communities that rely on these biogenic habitats can vary greatly as well(Anderson 1994). To explore possible reasons for spatial variability in the Monterey Bay, we quantitatively surveyed two Marine Protected Areas (MPAs), Hopkins Marine Station and Point Lobos State Reserve, for various species of fish, invertebrates and algae using the swath transect method. As per the design of the study, tThe two sites were both sampled twice over two days. We were interested in determining if the species composition differed between these two sites, while acknowledging sources of variance in the data collected.

These study sites were chosen for their similarity and proximity to one another, but also because of a generally acknowledged variance in abundance of the species that aggregate within them. Identifying these differences can aide us in teasing out the mechanisms that determine these dynamic community structures. [what about differences in exposure??]

Thusly, there were several questions to address for this observational study: 1) Do species abundances differ between Hopkins and Point Lobos? 2) Do species abundances differ between the days they were sampled 3) Do species abundances differ as a function of both the site and day? 4) Do these differences vary by taxonomic group?

Materials and Methods [2, 0, 0, 3, 3, 2, 2, 0, 2, 1, 2, 0, 3, 2, 3, 3, 2] 30

General approach???

Species Description and survey methods

Figure 1 List of all species sampled at both sites by taxa.

To determine differences in species abundances at the sites, a suite of species was chosen for fish, invertebrates and algae based on their because they arecommon and conspicuous nature and known presence at the sitesin central California kelp forests (fig. 1). Using 30m meter tapes, swath transects were conducted surveying for each of the taxonomic groups on one swath, three of which made up a single transect. There were offshore and onshore transects at Hopkins and transects in the same direction 5m apart at Lobos. Swaths were chosen as opposed to quadrats in order to encompass the range of species in the survey.8 BuddyEight pairsof divers were sent out to specific meter marks along a main line to cover a wide area. Teams were sent out to both sites on both days to account for spatial and temporal effects on variance in data. Each transect was divided into three, 10m long segments, each segment functioning as a replicate for that meter mark for that day.

Survey Sites and spatial variation [state these subheading as actual hypotheses!!]

Surveys took place in the Monterey Bay at two MPAs, Hopkins Marine Station and Point Lobos State Reserve. [lat, lon???] Both sites are known for their rocky subtidal substrate, a M. pyrifera dominated canopy and a saturation diversity of biota within them. The two sites are approximately 15km apart and very similar with a slight distinction in tidal swell exposure. Hopkins is the more protected reef and closer to a non-MPA zone.

To determine differences in species composition between the two sites, mean abundances were collected and graphed for all species at both sites. Bray Curtis dissimilarity was produced as a simple visual representation of spatial variation. PERMANOVA analysis was conducted using statistical software.

Spatial variation by taxa

Spatial effects by taxa were examined by graphing mean abundances for all species observed at each site separated by taxa. PERMANOVA was used as a means for identifying significance in any variation observed. Bray Curtis dissimilarity analyses were also produced for each of the three taxa as a visual aide for existing spatial variations.

Temporal variation

The surveys were conducted on 10/11/2011 and 10/13/2011. Both sites experienced an increase in swell activity on the second day of sampling. Bray Curtis dissimilarity was graphed for a visual representation of temporal variation. To account for temporal effects, PERMANOVA analysis was conducted using statistical software.

Temporal variation by taxa

Temporal variance was examined by taxa using mean abundances of all species observed on each day. PERMANOVA analysis revealed any temporal variation in the abundance of species by taxa. The Bray Curtis dissimilarity graph supplied visual aides in identifying temporal differences by taxa.

Interaction effects

Bray Curtis dissimilarity was produced for a visual representation of the interaction effect of site and day on the sampled species. Interaction effects between survey site and day were analyzed using PERMANOVA.

Interaction effects by taxa

To determine interaction effects by taxa, Bray Curtis dissimilarity and PERMANOVA analysis were utilized for fish, invertebrates, and algae.

Results

a)b)

Figure 2. Bray Curtis plot across all species shows a clear segregation mean abundances by site with the green triangles(Hopkins) clustered toward the right and blue triangles(Lobos) on the left side of the plot. PERMANOVA table of results of all species shows a significant effect of variation by site with a p-value of ~0.001.

Spatial Variation

The Bray Curtis analysis shows that site variation was a major source across all species. The PERMANOVA results add to this visual aide with a low p-value (~0.001) for the effect of site on variation. This is significant indication that the site was a major source of variation for all species (fig. 2).

Spatial Variation by Taxa

a)b)

c)d)

Figure 3. a)Bray Curtis analysis of invertebrates

from both sites with an indicator for the day.

b) Bray Curtis for fish across both sites for both

days. c) Bray Curtis for Algae across both sites for

both days. d) PERMANOVA table of results for the 3 taxa indicating site effects, day effects, and interaction effects. Algae show a significant site effect (p-value ~0.001), fish show significant day effect (p-value 0.043) and interaction effect (p-value 0.015), and inverts exhibit significant site effect (p-value ~0.001).

Analysis of variation for invertebrates by site using the visualization of Bray Curtis dissimilarity shows a weak site effect with some vague clustering by color (fig. 3a). PERMANOVA analysis for invertebrates implies a significant site effect with a low p-value of ~0.001(fig. 3d). Looking at mean abundances of invertebrates shows a relatively high abundance of B. elegans at both sites, with a lesser occurrence at Lobos. T. aurantia is also a relatively abundant species at both sites (fig. 4b).

Analysis of variation by site for fish as seen in the Bray Curtis analysis shows a scattered distribution with relation to site, with points from either site occurring in a random fashion (fig. 3b). PERMANOVA analysis shows no significant account of variance in assemblage by site with a relatively high p-value of 0.319 (fig. 3d). Mean abundances of fish species by site shows a high occurrence of S. atrovirens at Hopkins but no obvious site relationship among species.

The Bray Curtis analysis for algal species shows a clear distinction in the assemblages of species between the two sites (fig. 3c). PERMANOVA shows a low p-value, ~0.001, for site which implies a significant site effect on algal composition (fig. 3d). Mean abundances of invertebrates reveals a great abundance of B. elegans at Hopkins and a variance between the two sites for this species.

Analysis of variation by site for the different taxa show generally a strong effect on invertebrate and algal compositions while there is no significant effect on the assemblage of fish species.

a) Algae

b) Invertebrates

c) Fish

Temporal Variation

No pattern is obvious in the temporal variation of species from the Bray Curtis analysis across all species(fig. 2a). The PERMANOVA test gives a p-value of 0.382 for sampling day effects across all species(fig. 2b). There is no significant temporal variation in the analysis across all species.

Temporal Variation by Taxa

Variation by day is not clear in the Bray Curtis dissimilarity plot for invertebrates (fig. 3a). PERMANOVA analysis gives a p-value of 0.505 for the day effect on invert composition(fig. 3d). The graphs for mean abundances of invertebrates showed no changes in major changes in mean abundances(fig. 4b). Sampling day was not a source for variance of invert species composition.

No readily discernable patterns are available for temporal shifts in algal assemblages based on the Bray Curtis dissimilarity(fig 3c). PERMANOVA gives a high p-value of 0.926 for day effects on algal compositions(fig 4a). Mean abundance charts show a decrease in the observed number C. corymbifera at Hopkins over the two days(fig. 4a). There was no significant temporal variation in the abundance of algal species.

Sampling day effects on fish assemblages were not obvious from the Bray Curtis analysis(fig. 3b), but running PERMANOVA reveals a significant day effect with a p-value of 0.043(fig. 3d). The mean abundances graph shows a general decrease in the mean abundance across all species between the two days(fig. 4c).

Interaction Effects

Bray Curtis dissimilarity across all species indicates a clear distinction between sites, implying no general interaction effect for analysis that involving all species(fig. 2a). PERMANOVA analysis gives a p-value of 0.375 for an interaction effect between sample site an day(fig. 2b). An interaction effect is not apparent when analyzing variances across all species.

Interaction Effects by Taxa

Figure 5. Percent contribution to variance in data by sample day and site for algae, fish and invertebrates.

PERMANOVA for invertebrates gives a high p-value of 0.569 for an interaction effect, which tells us the variance explained by both sample site and day are not significant(fig. 3d). This was a similar result for algae which had an interaction p-value of 0.926(fig. 3d). The percent contribution of site and day for these two taxa shows that most of it is explainable by spatial variance.

However, there was a significant interaction effect observed in the composition of fish species as given by the PERMANOVA p-value of 0.015(fig 3d). Approximately 80 percent of the variation in fish assemblage depended on the sampling day(fig. 5).

Discussion

The results of this survey indicate several important things about the spatial and temporal variation in the species composition of marine biota in the Monterey Bay. The results make it apparent that the species assemblages are variable spatially, and temporally by taxa.

Algae, Inverts, and Spatial Heterogeneity

Most of the variability in the species composition of macroalgae and invertebrates seem to be explained spatially. This makes sense biologically due to the life histories of the organisms surveyed. Thallus forming algae generally stick to their recruitment site for the entirety their lives, which is why they would tend not to vary as much temporally, especially for the time period we were working under.

We sampled mostly sessile or sedentary species of invertebrates, which explains why there was not any real temporal variation in their abundance at a specific site. These are traits that make for good subjects in terms of sampling and describing spatial variations.

As for the adequacy of our survey for understanding the spatial variation in species composition, power analyses show us that some species were more suited to sampling than others, though the ones that were most suited may not have been the most ideal for describing heterogeneity.

Among the algal species surveyed, M. pyrifera had a very high power index, meaning there was little variation among the surveyors in the data collected. However, C. corymbifera, which was observed in the highest mean abundance for algae at Hopkins has a low power index for this study, which lends me to believe there may have been some sampling inadequacies(fig.6).P. californica came out with a significant amount of agreement between the samplers on its presence at Lobos. Some sources for the variance in data may have come from a misidentification of the species amongst the divers or missing the organism on the transects. Still, the data is compelling for spatial variation of algal species and the idea that these abundances differ is important in the fact that some organisms may be obligatorily associated with these abundant species of algae Therefore, a higher presence of a particular algae might be indicative of a higher presence of a specific associate.

Figure 6. Power index analysis of all algal species sampled. Shows high power index for M. pyrifera

Figure 7. Power index shows high power in sampling of P. miniata, T. aurantia, and U. piscivora.

For the invertebrates, P. miniata has a very high power index, which made it very good for sampling, and like M. pyrifera, their general abundance is unmistakable and their highly conspicuous nature makes them for good sampling, but their standard frequency at both sites does not make them the best subjects for identifying the mechanisms causing other differences in species composition. B. elegans was observed with the highest mean abundance and had an adequate power index(fig. 7), but comparatively low to other invertebrates. This may have been due to the life history of this proliferating cnidarian. They are found in aggregations, and patches of B. elegans falling on some transects and not others may have skewed this data. We can say though that their presence at Hopkins was observed far more than at Lobos. The presence of B. elegans as a sample species may have also skewed the collection of other data due to their sometimes overwhelming abundance.

Temperate fish

The results of fish abundances, on the other hand, were very different. Fish are highly mobile and tend to move around according to changes in their environment depending on availability of resources(Mathews 1985). It seems likely that the increasing swell resulted in a change in the abundance of fish observed on the transects on the second day of survey and resulted in our observation of high temporal variance. The generally low abundance of fish made it hard for this aspect of the survey to be very significant, and the comparatively low power indices reflect this(fig.8).

Figure 8. Power index for all species of fish observed. Generally low power index for all species.

The divers seemed to agree on the sparse yet general presence of S. atrovirens and E. lateralis particularly. These data may have turned out differently had there been no variance in ocean conditions, but this requires further testing.

Figure 9. Power index for all three species. X axis shows number of transects to achieve asymptote above power index above 2, the number needed to be considered significant.