Foraminiferal Densities and Pore Water Chemistry in the SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES * NUMBER 36 Indian River, Florida SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION Emphasis upon publication as a means of "diffusing knowledge" was expressed by the first Secretary of the Smithsonian. In his formal plan for the institution, Joseph Henry outlined a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This theme of basic research has been adhered to through the years by thousands of titles issued in series publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Contiibutions to Anthropotogy Smithsonian Contributions to Botany Smithsonian Contributions to tiie Earth Sciences Smithsonian Contributions to the Marine Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Folklife Studies Smithsonian Studies in Air and Space Smithsonian Studies in History and Technology In these series, the Institution publishes small papers and full-scale monographs that report the research and collections of its various museums and bureaux or of professional colleagues in the world of science and scholarship. The publications are distributed by mailing lists to libraries, universities, and similar institutions throughout the world. Papers or monographs submitted for series publication are received by the Smithsonian Institution Press, subject to its own review for format and style, only through departments of the various Smithsonian museums or bureaux, where the manuscripts are given substantive review. Press requirements for manuscript and art preparation are outlined on the inside back cover. Robert McC. Adams Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES • NUMBER 36 Foraminiferal Densities and Pore Water Chemistry in the Indian River, Florida Martin A. Buzas and Kenneth P. Severin SMITHSONIAN INSTITUTION PRESS Washington, D.C. 1993 ABSTRACT Buzas, Martin A., and Kenneth P. Severin. Foraminiferal Densities and Pore Water Chemistry in the Indian River, Florida. Smithsonian Contributions to the Marine Sciences, number 36, 38 pages, 32 figures, 29 tables, 6 appendices, 1993.—Two stations were established about 10 m apart at a depth of about 1 m at Link Port, Florida. One consisted of quartz sand and the other of quartz sand with a dense stand of seagrass. At the surface of each station and at a depth of 10 cm at the grass site, four replicate samples consisting of 5 ml each were taken every fortnight from 27 March to 6 November 1978 (17 sampling times, 204 samples). The taxa Quinqueloculina, Elphidium, Ammonia, Bolivina, and Ammobaculites comprising 98% of the fauna were enumerated. In addition, pore water chemistry was measured for temperature, salinity, oxygen, pH, Eh, NH3, P04, Si, N02, and N02 + N03. General linear models were used to analyze the bare surface-grass surface, and grass surface-grass 10 cm data sets. Foraminiferal densities were evaluated for differences between sites, periodicity, sites x periodicity (interaction), and environmental variables. Differences in overall density between the bare surface-grass surface sites were not significant for the three most abundant taxa (Quinqueloculina, Elphidium, and Ammonia). At the grass site the density for all taxa were significantly lower at 10 cm than at the surface (very few individuals were observed at 10 cm). Hypotheses for periodicity and interaction were significant for all taxa in all comparisons except for Bolivina in the bare surface-grass surface analysis. At the bare surface, maximum densities occurred in spring while at the grass surface in summer. Although densities were low at 10 cm, no synchronization between the grass surface and 10 cm was evident. The environmental variables were significant for all taxa in both comparisons. The environmental variables are, however, highly correlated. To alleviate this difficulty, a principal component analysis was performed on these variables. The first three components included all of the 10 variables. Subsequent multiple regression of foraminiferal densities and the principal components indicated that usually at least two components, accounting for most of the variables, were statistically significant. Thus, no simple relationship between pore water chemistry and density is apparent. The very large difference in density between the grass surface and 10 cm depth is much more strongly related to the pore water chemistry than the smaller differences with time at the surface sites. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: Seascape along the Atlantic Coast of eastern North America. Library of Congress Cataloging-in-Publication Data Buzas, Martin A. Foraminiferal densities and pore water chemistry in the Indian River, Florida / Martin A. Buzas and Kenneth P. Severin. p. cm.—(Smithsonian contributions to the marine sciences ; no. 36) Includes bibliographical references (p. ). 1. Foraminifera—Florida—Indian River. 2. Protozoan populations—Florida—Indian River. 3. Population den- sity—Florida—Indian River. 4. Pore water—Florida—Indian River. I. Severin, Kenneth P. II. Title. III. Series. QL368.F6B875 1993 593.1'2O4526323'097592-dc20 92-46076 ® The paper used in this publication meets the minimum requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48—1984. Contents Page Introduction 1 Acknowledgments 1 Methods 1 Field 1 Laboratory 2 Statistical 2 Bare Surface and Grass Surface 3 Environmental Variables 3 Species Densities, Station Differences, Periodicity, and Environmental Variables 10 Quinqueloculina 10 Elphidium 11 Ammonia 11 Bolivina 11 Ammobaculites 12 Grass Surface and Grass 10 cm 13 Environmental Variables 13 Species Densities, Station Differences, Periodicity, and Environmental Variables 15 Quinqueloculina 15 Elphidium 16 Ammonia 16 Bolivina 21 Ammobaculites 22 Comparison of Analyses 23 Comparison with Other Studies 24 Appendix 1: Bare Surface 29 Appendix 2: Grass Surface 31 Appendix 3: Grass 10 cm 33 Appendix 4: Bare Surface 35 Appendix 5: Grass Surface 36 Appendix 6: Grass 10 cm 37 Literature Cited 38 in Foraminiferal Densities and Pore Water Chemistry in the Indian River, Florida Martin A. Buzas and Kenneth P. Severin Introduction A basic variable for ecological studies is density, the number of individuals per volume or area. Densities of foraminiferal species, like those of all organisms, vary in space and time. Geographic changes in foraminiferal species densities have been documented between all marine environments from marshes to the abyss. Large differences in space like those between a marsh and the abyss or the Arctic and the tropics are easily recognized. A general qualitative correlation between observed species densities and the environment is easily accepted as an explanation for these changes. Similarly, differences over vast amounts of geologic time are easily recognized, and explained by the interplay of evolution and the environment. As we decrease the scale of our observations in space and time, however, and, at the same time, increase our effort to achieve quantitative results, differences in species densities and their explanation become much more difficult. Nevertheless, densities do vary over a matter of meters and within a time scale measured in weeks, months, or years. The present study is an analysis of quantitative measurements of species densities and environmental variables observed at two stations about 10 m apart which were sampled every fortnight for 9 months. During 1978 the chemistry group of the Harbor Branch Oceanographic Institution, Ft. Pierce, Florida monitored 10 pore water chemistry variables on a continual basis at two sites (stations) about 10 m apart at a depth of about 1 m in the Indian River (a shallow lagoon of nearly normal marine salinity on the central east coast of Florida). One station was located on bare quartz sand, the other on quartz sand with a covering of Martin A. Buzas, Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. Kenneth P. Severin, Department of Geology and Geophysics, University of Alaska-Fairbanks, Fairbanks, Alaska 99775-0760. seagrass (mostly Halodule wrightii and Thalassia testidium). We viewed their study as an ideal opportunity to conduct a study of the foraminifera with an experimental design allowing us to test statistically for differences in density between stations and with time as well as for the statistical significance of 10 environmental variables. Foraminiferal densities were also enumerated at a depth of 10 cm within the sediment at the grass station. Few living foraminifera were observed at 10 cm and most of the water chemistry variables exhibited a dramatic difference compared to the measurements made at the surface. To test the efficacy of the statistical procedures used in this study, and in others, the same statistical analyses were employed in evaluating the differences between the grass surface and at 10 cm as for the two surface stations. ACKNOWLEDGMENTS.—We thank the former Chemistry group at the Harbor Branch Oceanographic Institution, especially J. Montgomery, M. Hucks, G. Peterson, M. Price, and C. Zimmermann. In the field and laboratory K. Carle, D. Mook, and H. Sheng were most helpful. J. Jett prepared the tables, figures, and appendices. L.S. Collins, L.C. Hayek, and B.K. Sen Gupta offered helpful suggestions on the manuscript. We thank J.C. Warren for his careful copy editing. This is contribution No. 319 from the Smithsonian Marine Station at Link Port. Methods FIELD.—Two stations were established about 100 m south of the Link Port jetty at a depth of about 1 m. The stations were marked with four poles encompassing an area of about 1 m2 so that the same area could be re-occupied easily. The sediment at one station consisted of bare quartz sand with a silt-clay content of about 2%. The other was on the same substrate, but had a dense stand of Halodule wrightii with some Thalassia SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES testidium. The sediment was sampled by pushing plastic coring tubes (inner diameter 3.5 cm) into the substrate. At each station four sediment samples were taken indiscriminately (not statistically randomized) at each sampling time which con- sisted of every fortnight from 27 March until 6 November 1978 (17 sampling times). The temperature, salinity, oxygen, pH, Eh, NH3, P04, Si, N02, N02 + N03 were measured on pore waters by the chemistry group of the Harbor Branch Oceanographic Institu- tion throughout the field experiment (Montgomery et al., 1979). LABORATORY.—Immediately upon return to the laboratory (within a half-hour), 5 ml of sediment was removed from each core top, and at a depth of 10 cm. Each 5 ml sample was washed over a 63 (im sieve and preserved in 95% ETOH. Prior to examination for foraminifera the sample was stained overnight in rose bengal, dried, floated in a mixture of tetrabromine and acetone (specific gravity 2.3), and re-wet using "photo-flo" as a wetting agent. The foraminifera were enumerated while underwater, a procedure which facilitates the recognition of vividly stained protoplasm. The taxa counted were Quinqueloc- ulina (mostly Q. impressa and Q. seminulum), Elphidium (mostly E. mexicanum and E. gunteri), Ammonia (A. beccarii), Bolivina (B. striatula), and Ammobaculites (A. exiguus). These taxa are sufficiently dissimilar so that they can easily be identified under a binocular microscope and account for 98% of the foraminiferal fauna. The second replicate sample taken at the grass surface on 22 May 1978 was destroyed in a laboratory accident. The mean number of individuals from the other three replicates was used to estimate the missing data (Appendix 2). The systematics of the taxa used here are treated by Buzas and Severin (1982). Enumeration was made by Severin. STATISTICAL.—We have, then, two stations: one bare sand and the other with a stand of seagrass. For this study, the bare surface, the grass surface, and a depth of 10 cm within the sediment at the grass station were sampled for foraminifera. These three sites were sampled 17 times with four replicates of 5 ml each taken for examination. There are, then, N = 68 replicates or observations at each of the three sites. The number of individuals counted for each of five taxa and the total (the five taxa accounted for over 98% of the total) for each replicate is tabulated for the bare surface in Appendix 1, for the grass surface in Appendix 2, and for grass 10 cm in Appendix 3. We obtained the measurements made by the chemistry group for 10 water chemistry variables at each of the three sites at each sampling time. These measurements for the bare surface are tabulated in Appendix 4, for the grass surface in Appendix 5, and for the grass 10 cm in Appendix 6. The data were divided into two sets for statistical analyses, bare surface-grass surface, and grass surface-grass 10 cm. For each analysis we wished to test the following hypotheses; difference between sites (bare surface vs. grass surface or grass surface vs. grass 10 cm), differences with time (periodicity), different periodicities at each site, and the significance of the environmental variables. To accomplish this, we constructed a general linear model (GLM) similar to the one used by Buzas et al. (1977). In matrix notation the GLM is written as co: x = Z' (3 + e (nxl) (nxq) (qxl) (nxl) where x, the "dependent" variable is a vector of observed species densities for the n = 136 observations (nj = n2 = 68), Z' is a matrix of q "independent" variables, the composition of which will be discussed below, P is a vector of q parameters "regression coefficients" explaining the observations, and e is a vector of "errors" or "residuals," assumed to have a normal distribution. The original counts x were transformed to ln(x + 1) to make the data more Normal and stabilize the variance. Restricted Q. models containing s parameters are constructed by equating the appropriate individual or groups of (3 to 0. The sum of squares of the residuals, e'e, for each model is a scalar and is estimated by a least squares solution (Buzas et al., 1977, give the equations). Restricted models can be compared to the general model by the ratio = F (q - s)(n - q) (e'e^-e'e^) / (q - s) e'en / (n-q) The mglh program of "SYSTAT" was used to calculate the sum of squares of the residuals for each model, and these residuals were then used to calculate the F-ratio given above. The results of the analysis are most easily displayed in the standard ANOVA table. The composition of the matrix Z' of the co model is given in Table 1. The vector zl is composed of l's so that (3j is a constant, z2 contrasts the difference between sites by assigning +1 to one site and -1 to the other. The vectors z3 and z4 are made up of sin (m x 7t/3) and cos (m x TC/3), respectively, where m = 1, ,17. The vectors z5 and z6 are composed of sin (m x 7T./6) and cos (m x TI/6), respectively. These vectors are components of a periodic regression (Bliss, 1958) and account for a possible overall periodicity in the observations. Figures 1 and 2 illustrate these vectors over the period of our observa- tions. The possibility exists that the two sites may exhibit periodicity, but that the periodicity differs at the two sites. The interaction vectors in = z1x z3, z8 = z2 x z4, z9 = z2 x z5, and z10 = z2 x z6 account for this. We have, then, 8 vectors to examine the possible periodicity in our data. Had we constructed instrumental variables to examine the differences between the 17 sampling times and their interaction, we would have required 32 vectors, and made the model much more complicated. Finally, vectors zn through z20 contain the water chemistry variables completing the Z' matrix for the GLM. NUMBER 36 TABLE 1.—Composition of the Z' matrix. h = a vector of units h = +1 for bare surface, -1 for grass surface h = sin (m x nJ3), m = 1 ...,17 z4 - cos (m x rt/3), m = 1 17 z5 = sin (m x rc/6), m = 1 ,...,17 Z6 = cos (m x JI/6), m = 1 ,...,17 h = z2 xz3 Z8 = z2xz4 Z9 = z2 xz5 ho = z2 x z6 hi = temperature Z!2 = salinity zn = oxygen Z!4 = PH Z15 = Eh Z16 = NH3 Z17 = PO4 £18 = Si ZI9 = N02 Z20 ~ N02 + N03 tA, V ^ -& 1978 % ^> ^ I ^ § 0- 's "fc "V, 'Vv. ur\ \vo 1978 ^ \ •^ % % % % l% w 1978 FIGURE l.—n/3 periodicity. V ^ 1978 FIGURE 2.—TI/6 periodicity. Bare Surface and Grass Surface ENVIRONMENTAL VARIABLES.—The recorded temperature values are plotted in Figure 3. The results of a two-way ANOVA with one observation per cell (no replicates were measured for the environmental variables) is shown in Table 2. The hypotheses for differences in mean temperature between stations and with time are both significant at the .05 level, even though the mean temperature at the bare surface is only l°C higher. The minimum at the bare station was 23°C and at the grass station 22°C. The maximum at each was 32°C and 3l°C, respectively. Table 2 shows the mean square for time is higher than for station differences, and, as might be expected, Figure 3 shows summer temperatures were higher than spring and fall. The recorded salinity values are shown in Figure 4 and the ANOVA results in Table 2. The slightly higher salinities recorded at the bare station were not statistically significant. Differences with time, however, were highly significant with minimum salinities of 20°/oo and 22°/oo recorded at the bare and grass stations, respectively, in August. Oxygen measurements are plotted in Figure 5 and ANOVA results are presented in Table 2. Differences between stations SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES TABLE 2.—Analysis of variance for chemical variables on bare surface and grass surface. CRASS SURFACE £ = 26.61 8 = 2.96 df Variable Temperature Salinity Oxygen PH Source stations time residual stations time residual stations time residual stations time residual Sum of squares 1 16 16 11.77 281.75 8.47 6.62 1 459.62 16 41.38 16 0.39 1 161.39 16 25.26 16 0.06 1 2.98 16 0.55 16 P(F) mean square 11.77 22.24 0.00 17.61 33.28 0.00 w 0.53 < 6.62 2.56 0.13 06 28.73 11.11 0.00 5 2.59 0.39 0.25 0.63 10.09 6.39 0.00 1.58 0.06 1.77 0.20 0.19 5.42 0.00 0.03 •w 4r 'V V Eh stations 1823.56 1 1823.56 1.94 0.18 time 248676.53 16 15542.28 16.52 0.00 residual 15053.94 16 940.87 NH3 stations 597.94 1 597.94 2.84 0.11 time 9405.39 16 587.84 2.79 0.02 residual 3368.87 16 210.55 PO4 stations 9.52 1 9.52 1.27 0.28 time 127.32 16 7.96 1.06 0.45 residual 120.04 16 7.50 Si stations 117.23 1 117.23 0.91 0.35 time 46728.51 16 2920.53 22.78 0.00 residual 2051.49 16 128.22 N02 stations 0.02 1 0.02 2.25 0.15 time 0.83 16 0.05 5.53 0.00 residual 0.15 16 0.01 N02 + N03 stations 1424.70 1 1424.70 0.62 0.44 time 36134.21 16 2258.39 0.99 0.51 residual 36611.80 16 2288.24 1978 BARE SURFACE $ = 27.78 3 = 3.06 were not statistically significant, but differences with time were. As expected, there is an inverse relationship with temperature, and generally the oxygen values are higher in the spring and fall. Minimum values occurred at the bare station in May, June, and July, and at the grass station in July. Values for pH are plotted in Figure 6 and ANOVA results are presented in Table 2. Differences between recorded values between stations were small and not statistically significant. Differences with time were significant and, like oxygen, the highest values occurred in spring and fall. Minimum values at both stations were 7.5. Eh values are plotted in Figure 7 and ANOVA results are presented in Table 2. Although the difference in the mean value between stations was relatively large (Figure 7) it was not statistically significant. On the other hand, differences with time were highly significant, and fluctuated greatly between sampling times. Minimum values at the bare station and grass station were -208 and -128 mV, respectively. NH3 values are plotted in Figure 8 and results of the ANOVA »'V\ 1978 FIGURE 3.—Temperature measurements in °C. are presented in Table 2. No significant difference was observed between stations, but, once again, differences with time were highly significant. Very low to zero values were recorded in the spring and fall with maxima at both stations in the summer. P04 values are plotted in Figure 9 and ANOVA results are presented in Table 2. No statistical difference was observed between stations or with time. The measured values were generally very low to zero with the exception of the bare station in September which appears to be an outlier. The measured Si values are plotted in Figure 10 and the ANOVA results are presented in Table 2. No statistical difference was observed between stations, but differences with time were significant. Zero values were recorded at both stations in NUMBER 36 GRASS SURFACE £ =28.79 8 = 3J5 GRASS SURFACE £ = 4.47 C =1.83 3 20 %\ * % •%. *** *■>„ %. % X XI \ X 1978 1978 2 20 < BARE SURFACE \X =29.68 O ■ 4.48 z w 6 rj 6 > x o BARE SURFACE ji =4.26 8 = 2.89 \hr 1978 ^ •V i i •• i ^i -^ i <^ 1978 FIGURE 4.—Salinity measurements in %o. FIGURE 5.—Oxygen measurements in mg-at/1. •*. cu kfe> V °A. <*. *> I ^ \\\ ^ GRASS SURFACE £ = 8.01 O =0-24 % \ W i^ 1978 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES GRASS SURFACE £ = -4.06 0 = 78.29 200 150 100 SO 0 i ■SO -100 -ISO -200 -250 \ '<& i ^ % % X \\ n \%h -fei \ i v* i 1978 a 8.o BARE SURFACE ft = 8.10 S =0.41 1978 FIGURE 6.—pH measurements. 200 150 100 50 0 - -50- -100 -150 4- t>> 1978 BARE SURFACE ji = -18.71 8 = 101.75 ^'xi\i\r\ FIGURE 7.—Eh measurements in mV. NUMBER 36 100 80 - 60 - GRASS SURFACE $ =4.28 6 = 8.0S 18 16 14 12 -I ^ 10 o GRASS SURFACE £ =0.54 8 =0.57 40 1978 BARE SURFACE £ = 12.67 0 = 27.09 1978 FIGURE 8.—NH3 measurements in ug-at/1. 1978 FIGURE 9.—P04 measurements in ug-at/1. GRASS SURFACE £ = 48.96 8 = 40 32 0.6 53 75 4s w 1978 ** * SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES 0.7 -, 0.2 - GRASS SURFACE £ = 0.14 0 = 0.16 V % ij> \ *> v rv \ V V \ 1978 100 ■ 53 75 - BARE SURFACE £ = S2.67 6 = 37.73 (1.7 0.2 BARE SURFACE £ = 0.19 O = 0.19 % *c-„ 1978 ^ FIGURE 10.—Si measurements in ug-at/1. FIGURE 11.—N02 measurements in ug-at/1. NUMBER 36 300 -I GRASS SURFACE U = 19.03 3 = 66.28 250 ■ 200 150 100 ■ 50 - 4 \ 4r \ | X \\ 1 \ 1978 1978 FIGURE 12.—N02 + N03 measurements in ug-at/l. BARE SURFACE £ = 6.09 0 = 12.38 August; however, values generally increase from March to November. N02 values are plotted in Figure 11 and the ANOVA results are presented in Table 2. Significant differences were again noted only with time. In general, values are low with maxima in summer and fall at the bare station and spring, summer, and fall at the grass station. Both stations had minima values from late August until early October. N02 + N03 values are plotted in Figure 12, and ANOVA results are presented in Table 2. No significant differences were observed between stations or with time. An unexplainably high value was recorded in April at the grass station. Table 3 shows the correlation coefficients between the environmental variables. Temperature and NH3 are positively correlated with one another and negatively with oxygen, pH, and Eh which are all positively correlated with one another. Consequently, any hypothesis concerning the significance of a particular variable on species density is not independent of the others. A way to avoid this difficulty is to transform the original variables to principal components. Principal component analy- sis is a technique which produces a succinct parsimonious summarization of many correlated (non-zero covariance) variables by transforming the original variables to independent (zero covariance) variables called principal components. An additional advantage of the technique is that the first principal component will account for most of the variability in the data, the second less, and so on (Seal, 1964). Eigenvalues were calculated from the correlation matrix (Table 3) using the SYSTAT statistical package. The first three eigenvectors account for 63.43% of the variability. The factor score coefficients (standardized vectors which when multiplied by the original standardized variables produce the principal components) indicate that all the environmental variables contribute substan- tially to the first three principal components (Table 4). The coefficients (Table 4) indicate that the first principal compo- nent (PCI) accounting for 31% of the variability contrasts TABLE 3.—Correlation matrix for chemical variables for bare surface and grass surface. 0.05 level is underlined. Temperature Salinity Oxygen PH Eh NH3 P04 Si N02 N02 + N03 Temperature 1.00 Salinity -0.14 1.00 Oxygen -0.68 0.07 1.00 pH -0.36 0.03 0.69 1.00 Eh -0.52 -0.08 0.69 QM 1.00 NH3 0.40 0.29 -0.48 -0.44 -0.52 1.00 P04 0.14 -0.24 0.03 0.03 -0.07 -0.03 1.00 Si -0.06 -0.39 -O.05 -0.33 0.04 0.04 0.13 1.00 N02 -0.15 -004 -0.08 -0.14 0.09 0.32 -0.05 0.21 1.00 N02 + N03 -0.26 0.24 0.15 0.10 -0.06 0.02 -0.07 -0.23 0.12 1.0010 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES TABLE 4.—Factor score coeffecients for chemical variables for bare surface and grass surface. Chemical Factor variable PC 1(31%) PC2(18%) PC3(14%) Temperature -0.24 0.05 -0.30 Salinity 0.01 -0.45 0.04 Oxygen 0.29 0.00 0.04 pH 0.24 -0.07 -0.19 Eh 0.25 0.12 0.13 NH3 -0.22 -0.18 0.22 P04 -O.02 0.24 -0.16 Si -0.05 0.39 0.34 N02 -0.03 0.01 0.59 N02 + N03 0.06 -0.31 ^0.17 GRASS SURFACE QUINQUELOCULINA ji = 254.68 0 = 350.65 S IOOO Z >* rv v «> i x 1978 DARE SURFACE QUINQUELOCULINA £ =328.59 = 585 .57 U 100 -: 2 iooo H 7. z K O z W 100 - * i \\ ■% 1978 FIGURE 13.—Mean number of individuals of Quinqueloculina per 5 ml of sediment (density). temperature and NH3 with oxygen, pH, and Eh. The second principal component (PC2) accounting for 18% of the variability consists mainly of salinity, Si, and N02 + N03, although P04, NH3, and Eh also contribute and a line of demarkation is not as clear cut as for PCI. The coefficients (Table 4) indicate that the third principal component (PC3) accounting for 14% of the variability consists mainly of N02, Si, and temperature, but again there is no dramatic demarka- tion. The most highly correlated variables (Table 3) are all concentrated on PC 1. SPECIES DENSITIES, STATION DIFFERENCES, PERIODICITY, AND ENVIRONMENTAL VARIABLES.—Quinqueloc- ulina: Quinqueloculina was the most abundant taxon making up about 75% of the total living foraminifera at the surface stations. A plot of the mean densities observed at the bare and grass surface at the two stations is shown in Figure 13. The ANOVA table for six hypotheses is shown in Table 5. We recall that each hypothesis is formulated by equating the desired (3 to zero. For example, the Q. model used to test for station differences deletes B,, for 7i/3 periodicity and inter- action P3 = p4 = (i7 = (3g = 0, and so on. Table 5 indicates that all hypotheses except for station differences are significant. The overall mean density at the bare station is higher than at the grass station, but is not statistically significant at the chosen 0.05 level. Figure 13 shows that the bare station had high densities in spring while the grass station had high densities in summer. The environmental variables are significant as a group. Because they are not independent, testing for the significance of each individually is risky. Nevertheless, using the standard errors of the (3's from the co model for calculating confidence limits indicates oxygen, Eh, P04, Si, and N02 + N03 are significant. Again, emphasizing that the variables are not independent, the ease of calculation using "SYSTAT" made the calculation of simple regressions for each of the variables vs. density irresistible. The results shown in Table 6 indicate the regressions for oxygen, pH, P04, Si, and N02 are significant. A way of avoiding the correlations between variables is to calculate a multiple regression using principal components instead of the original variables. Because the TABLE 5.—Statistical analysis of GLM for Quinqueloculina for bare surface and grass surface. Variability on Sum of df Mean F P(F) account of squares square Stations 1.86 1 1.86 3.64 0.06 7t/3 periodicity 22.53 4 5.63 11.04 0.00 and interaction n/6 periodicity 27.50 4 6.88 13.38 0.00 and interaction rc/3 interaction 9.88 2 4.94 9.68 0.00 JI/6 interaction 20.45 2 10.23 20.05 0.00 Environmental 37.34 10 3.73 7.32 0.00 variables Residual 59.17 116 0.51 NUMBER 36 11 TABLE 6.—Values of F-ratio's for simple regressions on species densities and environmental variables at bare surface and grass surface. (+ indicates signficant (.05 level) positive value of p; - significant negative value of (J.) i $ Environmental variables / / Temperature 0.05 5.20- 0.29 2.37 4.09" Salinity 0.20 21.10+ 5.77+ 49.23+ 6.79" Oxygen 7.66+ 13.22- 0.08 0.54 4.37+ PH 15.44+ 16.99+ 0.10 0.83 3.10 Eh 0.15 0.08 0.08 0.41 0.19 NH, 1.64 0.00 1.98 1.22 7.09" PO4 4.00" 2.03 1.19 11.21+ 1.59 Si 7.23- 7.60" 0.25 17.88" 3.36 N02 4.38" 0.05 9.60* 0.67 1.95 N02 + NO, 0.07 1.02 0.22 6.64+ 0.27 principal components are orthogonal, each hypothesis is independent and the analysis is similar to a one-way ANOVA. The results of an analysis on the log densities of Quinqueloc- ulina and the first three PC's of the environmental variables are shown in Table 7. The overall F-ratio is significant and the test for the significance of each PC, which are independent, indicates that PCI and PC3 are significant at the 0.05 level while PC2 is nearly so. We recall from Table 4 that temperature, oxygen, pH, Eh, and NH3 all contribute nearly equally to the first PC, and the third PC is weighed heavily on temperature, Si, and N02. Thus, the analysis using principal components requires seven of the 10 environmental variables for two PC's and 10 for three to explain the results. All of the above analyses indicate that the identification of one or two variables as solely significant for the observed densities is impossible. Elphidium: Elphidium constitutes about 14% of the total living population and mean densities at the bare and grass stations are plotted in Figure 14 and the results of the GLM TABLE 7.—Regression of Quinqueloculina and principal components for bare surface and grass surface. Variable Coefficient Standard error t P(2 tail) R2 = 0.11 Constant 5.00 0.09 57.41 0.00 PCI 0.20 0.09 2.28 0.02 PC2 -0.17 0.09 -1.94 0.06 PC3 -0.24 0.09 -2.69 0.01 Analysis of variance Source Sum of .r squares Mean square F-ratio P Regression Residual 16.68 3 136.13 132 5.56 1.03 5.39 0.00analysis is presented in Table 8. Although the mean density at the bare station is again higher than at the grass station, it is not statistically significant. The TC/3 periodicity and interaction hypotheses are significant, but the 7t/6's are not. Figure 14 indicates a spring high at the bare station, but no pronounced summer high at the grass station as was observed with Quinqueloculina. The group of environmental variables are significant, and the p's for Eh, P04, N02, and N02 + N03 were significant. Regressions of density vs. each individual environ- mental variable yielded a significant F-ratio for temperature, salinity, oxygen, pH, and Si (Table 6). The results of multiple-regression of Elphidium densities and the first three PC's of the environmental variables are shown in Table 9. The first two PC's are significant and these account for all of the environmental variables except N02. Here, we note that individual tests of the P's from the co model and individual P's from simple regressions do not agree testifying to the difficulty encountered when variables are highly correlated. Ammonia: Ammonia makes up about 8% of the total living population and mean densities at the bare and grass stations are plotted in Figure 15. The results of the GLM analysis are presented in Table 10. The overall mean density is slightly higher at the bare station, but not significantly so. The xc/3 periodicity and interaction hypotheses are significant and Figure 15 indicates an early spring maximum at the bare station while both stations have summer and fall maxima. Environ- mental variables are significant as a group and the (3's for Eh, N02, and N02 + N03 were significant. Individual simple regressions on density vs. environmental variables yielded significant F-ratios for salinity and N02 (Table 6). The results of the multiple regression on densities of Ammonia and the PC's of the environmental variables are shown in Table 11. The results present us with a small dilemma because the F-ratio for the overall analysis has a probability of 0.06 which is slightly above our chosen level, and, therefore, is not significant. On the other hand, PC3 is significant (Table 11). If we choose to regard the third principal component as significant, then temperature, Si, and N02 are the major contributors, especially N02 (Table 4). The F-ratio for environmental variables in the co model, while significant, is the smallest encountered for any of the five taxa analyzed. Ammonia appears, then, to be the least influenced by the 10 variables measured in this study. Bolivina: Bolivina constitutes about 1% of the total living population and mean densities at the bare and grass stations are plotted in Figure 16. The results of the GLM analysis are shown in Table 12. Although the differences in densities between stations are small, they are, nevertheless, statistically signifi- cant, and the highest density is at the grass station. None of the hypotheses for periodicity are significant. There does appear to be a decrease in density over the course of sampling, but the fluctuations in density shown in Figure 16 are small compared to those considered previously (note the difference in the scale of the ordinate). The environmental variables are significant as 12 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES V ^' S 1000 z Q z < a 10 GRASS SURFACE ELPHIDrUM £ = 30.90 CT = 17.96 1978 TABLE 8.—Statistical analysis of GLM for Elphidium for bare surface and grass surface. Variability on Sum of df Mean F P(F) account of squares square Stations 0.88 1 0.88 1.78 0.18 7C/3 periodicity 8.28 4 2.07 4.21 0.00 and interaction TC/6 periodicity 2.94 4 0.74 1.49 0.21 and interaction 7t/3 interaction 3.73 2 1.87 3.79 0.03 7t/6 interaction 2.91 2 1.45 2.95 0.06 Environmental 28.43 10 2.84 5.78 0.00 variables Residual 57.04 116 0.49 BARE SURFACE ELPHIDIUM (2= 80.96 a = 14S.S8 TABLE 9.—Regression of Elphidium and principal components for bare surface and grass surface. Variable Coefficient Standard error t P(2 tail) R2 = 0.18 Constant 3.50 0.07 50.73 0.00 PCI 0.21 0.07 3.05 0.00 PC2 -0.30 0.07 -4.32 0.00 PC3 -0.03 0.07 -0.40 0.69 1+ 1978 FIGURE 14.—Mean number of individuals of Elphidium per 5 ml of sediment (density). Analysis of variance Source Sum of .f squares Mean square F-ratio P Regression Residual 18.25 3 85.54 132 6.08 0.65 9.39 0.00 a group, and the (3's for salinity and Si were significant. F-ratios for simple regressions are significant for salinity, P04, Si, and N02 + N03 (Table 6). Multiple regression on densities of Bolivina and the first three PC's of the environmental variables are shown in Table 13. Only PC2 consisting mostly of salinity, Si, and N02 + N03 (Table 4) is significant. The relationship with salinity and to a lesser extent with Si are notable for this species. Ammobaculites: Ammobaculites also constitutes about 1% of the total living population and mean densities for the bare and grass stations are plotted in Figure 17. The results of the GLM analysis are shown in Table 14. The hypothesis for station differences is significant with the highest densities occurring at the grass surface. The rc/3 periodicity and interaction hypotheses are significant, and Figure 17 indicates the now familiar spring maximum was observed at the bare station after which time the densities remained very low (Appendix 1). At the grass surface the densities increased overall during the sampling duration (Appendix 2). The hypothesis for the environmental variables is significant. The (3's of the CO model for the variables pH, P04, Si, N02, and N02 + N03 were significant. The F-ratios for temperature, salinity, oxygen, and NH3 were significant for simple regres- sions on density and environmental variables (Table 6). The results of a multiple regression on the densities of Ammobacu- lites and the first three PC's of the environmental variables are shown in Table 15. The first two PC's are significant indicating that all the variables except for N02 (Table 4) are important contributors. Again, we note the inconsistencies obtained by testing the environmental variables individually. In summary, we recognize no station differences for the three most abundant taxa. For the rare species Bolivina and Ammobaculites which together constitute only about 2% of the total living population, densities are higher at the grass station. All of the taxa except for Bolivina exhibited periodicity. Quinqueloculina, Elphidium, and Ammonia all showed high densities in spring at the bare surface station and in summer at the grass surface. Bolivina exhibited an overall decreasing density from spring onward at both stations while Ammobacu- lites increased at the grass surface station and decreased at the NUMBER 36 13 1000 100 *A V \ \ GRASS SURFACE AMMONIA ji = 27 £4 O = 29.74 x I % 1978 % TABLE 10.—Statistical analysis of GLM for Ammonia for bare surface and grass surface. Variability on Sum of df Mean F p(F) account of squares square Stations 0.06 1 0.06 0.13 0.72 JI/3 periodicity 11.29 4 2.82 5.98 0.00 and interaction rt/6 periodicity 4.24 4 1.06 2.24 0.07 and interaction 7i/3 interaction 6.38 2 3.19 6.76 0.00 n/6 interaction 1.94 2 0.97 2.05 0.13 Environmental 16.60 10 1.66 3.52 0.00 variables Residual 54.70 116 0.47 BARE SURFACE AMMONIA 1 = 33.94 ]i- ■■ 29.78 TABLE 11.—Regression of Ammonia and principal components for bare surface and grass surface. Variable Coefficient Standard error t P(2 tail) R2 = 0.06 Constant 3.14 0.07 45.92 0.00 PCI -0.01 0.07 -0.16 0.87 PC2 -0.12 0.07 -1.72 0.09 PC3 0.15 0.07 2.18 0.03 *> M \ > ■if 1* w 1978 FIGURE 16.—Mean number of individuals of Bolivina per 5 ml of sediment (density). Variable Coefficient Standard error t P(2 tail) R2 = 0.28 Constant 1.38 0.07 19.10 0.00 PCI 0.04 0.07 0.47 0.64 PC2 -0.52 0.07 -7.14 0.00 PC3 0.01 0.07 0.19 0.85 Analysis of variance Source Sum of .r squares Mean square F-ratio P Regression Residual 36.34 3 93.66 132 12.11 0.71 17.07 0.00 NH3 values are plotted in Figure 23. The depth hypothesis is significant. The values at 10 cm are usually two orders of magnitude greater than at the surface (Appendix 5, 6). Two zero values were recorded at 10 cm, one in April and one in October. We suspect these anomalies are due to problems with instrumentation. P04 values are plotted in Figure 24. The hypothesis for depth is significant. The values for P04 were very low at the surface, and one or two orders of magnitude higher at 10 cm (Appendix 5, 6). A very high value was recorded at 10 cm in June. Si values are plotted in Figure 25 and the hypothesis for depth is significant. The values for Si at the surface are an order of magnitude smaller than at 10 cm (Appendix 5, 6). At 10 cm zero values were recorded in April, August, and October. Except for the zero value in August, the pattern is very similar to that observed for NH3 at 10 cm. N02 values are plotted in Figure 26. The hypothesis for depth is significant, and values for N02 are low. At 10 cm very low values were recorded in all months except October. N02 + N03 values are plotted in Figure 27. There is no significant difference with depth or with time. Except for one measurement at the surface in April (Figure 27, Appendix 5) all the measurements were low. The two-way ANOVA's for environmental variables at the bare surface vs. grass surface showed most significant differences are with time (Table 2). In contrast, the analysis for grass surface vs. grass 10 cm shows most of the significant differences are with depth (Table 16). Moreover, the F-ratios for the latter are usually much higher. We have, then, a situation where, except for temperature, salinity, and N02 + N03, there are much larger differences than at the surface stations. The environmental variables are highly correlated as shown in Table 17. High positive values occur between oxygen and Eh, oxygen and N02, pH and Eh, Eh and N02, NH3 and P04, NH3 and Si, and P04 and Si. High negative values occur between oxygen and NH3, oxygen and P04, oxygen and Si, pH and NH3, pH and Si, Eh and NH3, Eh and P04, and Eh and Si. The correlations differ from those at the bare and grass surface NUMBER 36 15 GRASS SURF ACE AMMOBACULITES £=5.69 S = S.80 1978 TABLE 14.—Statistical analysis of GLM for Ammobaculites for bare surface and grass surface. Variability on Sum of df Mean F P(F) account of squares square Stations 21.59 1 21.59 44.79 0.00 7t/3 periodicity 8.56 4 2.14 4.44 0.00 and interaction 7t/6 periodicity 3.13 4 0.78 1.62 0.17 and interaction JI/3 interaction 4.90 2 2.45 5.08 0.01 it/6 interaction 0.08 2 0.04 0.08 0.92 Environmental 18.26 10 1.83 3.79 0.00 variables Residual 55.87 116 0.48 BARE SURFACE AMMOBACULITES £ = 1.41 S = 2.97 1978 FIGURE 17.—Mean number of individuals of Ammobaculites per 5 ml of sediment (density). TABLE 15.—Regression of Ammobaculites and principal components for bare surface and grass surface. Variable Coefficient Standard error t P(2tail) R2 = 0.06 Constant 1.03 0.08 12.80 0.00 PCI 0.18 0.08 2.17 0.03 PC2 0.17 0.08 2.07 0.04 PC3 -0.03 0.08 -0.34 0.73 Analysis of variance Source Sum of .f squares Mean square F-ratio P Regression Residual 7.97 3 115.46 132 2.66 0.88 3.04 0.03 (Table 3) in that P04, Si, and N02 now are highly correlated so that the matrix of correlation coefficients has many more high values. To succinctly summarize the environmental variables, and remove the covariance between them, a principal compo- nent analysis was calculated on the correlation matrix. The first three eigenvalues account for about 70% of the variability. The factor score coefficients are given in Table 18. The first vector accounting for 45% of the variability has high values for oxygen, pH, Eh, NH3, P04, Si, and N02. The second vector accounting for 15% of the variability has the highest values for salinity and N02 + N03, and the third accounting for 10% of the variability has a high value for temperature. Thus, all of the water chemistry variables are accounted for by using the first three principal components, and PCI accounts for all of them except temperature, salinity, and N02 + N03 which, interest- ingly, are the variables without significant differences with depth (Table 16). SPECIES DENSITIES, DEPTH DIFFERENCES, PERIODICITY, AND ENVIRONMENTAL VARIABLES.—Quinqueloculina: Quin- queloculina mean densities are plotted in Figure 28, and analysis by the GLM is shown in Table 19. As noted earlier, the densities for Quinqueloculina at the grass surface exhibited a summer maximum in July and August (Appendix 2). At 10 cm maxima were observed in March and November. The large increase in density observed in the summer at the surface is not reflected at 10 cm. All of the hypotheses tested by the GLM were significant (Table 19). The mean square for differences with depth is very large reflecting the two orders of difference in magnitude of the mean density between the surface and 10 cm. The set of A environmental variables are significant and in the co model P's for oxygen, Eh, and NH3 were significant. Simple regressions on density vs. environmental variables indicate the F-ratios for all variables except temperature are significant (Table 20). The results of a multiple regression on the densities of Quinqueloc- ulina and the first three PC's of the environmental variables are 16 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES GRASS SURFACE £ = 26.61 8 = 2.96 TABLE 16.—Analysis of variance for chemical variables on grass surface and 10 cm. Variable Source Sum of squares df mean square F P(F) Temperature depth time residual 0.54 273.64 73.63 1 16 16 0.54 17.10 4.60 0.12 3.72 0.74 0.01 Salinity depth time residual 5.72 311.96 32.35 1 16 16 5.72 19.50 2.02 2.83 9.64 0.11 0.00 Oxygen depth time residual 87.68 96.21 45.59 1 16 16 87.68 6.01 2.85 30.77 2.11 0.00 0.07 pH depth time residual 2.95 1.03 1.05 1 16 16 2.95 0.06 0.07 45.09 0.98 0.00 0.52 Eh depth time residual 469412.50 107511.88 68712.00 1 16 16 469412.50 6719.49 4294.50 109.31 1.57 0.00 0.19 NH3 depth time residual 1148940.80 601797.43 572067.08 1 16 16 1148940.80 37612.34 35754.19 32.13 1.05 0.00 0.46 PO4 depth time residual 12189.21 22394.25 22652.76 1 16 16 12189.21 1399.64 1415.80 8.61 0.99 0.01 0.51 Si depth time residual 1477313.29 635755.12 571797.45 1 16 16 1477313.29 39734.70 35737.34 41.34 1.11 0.00 0.42 N02 depth time residual 0.11 0.19 0.23 1 16 16 0.11 0.01 0.01 7.82 0.85 0.01 0.62 N02 + N03 depth time residual 2933.23 35026.37 35275.14 1 16 16 2933.23 2189.15 2204.70 1.33 0.99 0.27 0.51 shown in Table 21. The first and third PC's are significant indicating that only N02 + N03 did not contribute substantially to the significance of the analysis. The F-ratios for the simple regressions are much larger than for the bare surface and grass surface (compare Tables 6 and 20). Similarly, the F-ratio and R2 for the multiple regression using the PC's is much larger for the grass surface and grass 10 cm than for the bare surface and grass surface (compare Tables 7 and 21). This is, of course, a reflection of the very large differences between densities and environmental variables between the surface and 10 cm. The signs of the factor score coefficients for PCI (Table 18), the correlation matrix (Table 17), and the P's of the simple regressions (Table 20) show that the analysis contrasts oxygen, pH, Eh, N02 with NH3, P04, Si. Elphidium: Elphidium mean densities at the grass surface and grass 10 cm are plotted in Figure 29. The summer maximum observed at the surface for Quinqueloculina was not observed for Elphidium (Appendix 2). Minima at the surface were observed in June and November which is similar to the V \ \ 1p I I t9- 1978 1 GRASS 10 CENTIMETERS $ = 26.86 8 = 3.60 X X *• \ \ \\\ % 1978 FIGURE 18.—Temperature measurements in °C. pattern of Quinqueloculina (Appendix 2). At 10 cm a maximum occurred in March due to high number of individuals in two of the four replicates (Appendix 3). Once again, however, the densities at 10 cm were very low, averaging less than two individuals. Table 22 indicates the largest mean square is for the hypothesis contrasting depth. The rc/3 periodicity with interac- tion is significant while the 7t/6 is not. The set of environmental variables are significant, and individual p's of the co model for Eh and NH3 were significant. F-ratios of all variables except temperature and salinity are significant for simple regressions (Table 20). The results of a multiple regression on the densities of Elphidium and the first three PC's of the environmental variables indicate that all three PC's, and, consequently, all the environmental variables contribute substantially (Table 23). Ammonia: Ammonia densities for the grass surface and grass 10 cm are plotted in Figure 30. A maximum density was NUMBER 36 17 GRASS SURFACE £ = 28.79 8 = 3.35 •^ «fc 1978 1% GRASS SURFACE \X = 4.47 X 1p X X X is 1978 35 n ^, 1A X\ %- GRASS 10 CENTIMETERS £ = 29.62 8 = 3.21 "fe ^> 'l%f\l 1978 * FIGURE 19.—Salinity measurements in %o. \ i X z C GRASS 10 CENTIMETERS £ =126 8 =235 1978 FIGURE 20.—Oxygen measurements in mg-at/1. 18 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES GRASS SURFACE ji = 8.01 8 =0.24 GRASS SURFACE j! = -4.06 8 = 78.29 V ^' X X *% x x x is 1978 x\ * ^ s X \ 1978 GRASS 10 CENTIMETERS ji = 7.42 8 = 0.27 GRASS 10 CENTIMETERS jl = -239.06 8 = 69.89 ■fc ■1>. ^ 1978 FIGURE 21.—pH measurements. V \ 1978 FIGURE 22.—Eh measurements in mV. NUMBER 36 19 GRASS SURFACE £ =4.28 1978 8 = 8.05 300 200 \ 1A % GRASS SURFACE \X =054 8 =037 % i X. 1978 \' X GRASS 10 CENTIMETERS \L = 38.41 8 = 53.06 100 - GRASS 10 CENTIMETERS £ =371.93 1978 8 = 270.74 FIGURE 23.—NH3 measurements in u.g-at/1. FIGURE 24.—P04 measurements in ug-at/1. 20 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES ^ GRASS SURFACE £ = 48.96 1978 8 = 4032 O 0.3 Z GRASS SURFACE p = 0.14 8 = 0.16 -%, K h 1978 * i X GRASS 10 CENTIMETERS p =465.85 8 =271.75 1978 * FIGURE 25.—Si measurements in ug-at/1. 0.6 -, GRASS 10 CENTIMETERS p = 0.02 8 = 0.02 0.5 ■ 0.4 ■ z 0.3 - 0.2 - 0.1 A —* ■ v * % Vx i -v MM %- 1 X l0f> 1978 FIGURE 26.—NO, measurements in ug-at/1. NUMBER 36 21 -■—i- XI *« o z + 150 c z x I 1978 GRASS SURFACE jl = 19.03 8 =66.28 ">V V>>, X IS GRASS 10 CENTIMETERS ji = 0.46 8 = 0.76 % X X s 1978 * FIGURE 27.—N02 + N03 measurements in ug-at/1. observed at the surface in July. Like Quinqueloculina and Elphidium, a minimum occurred in June (Appendix 2). Similar to the aforementioned taxa, at 10 cm the maximum was observed in March (Appendix 3). In general, densities were very low at 10 cm averaging slightly less than two individuals. Table 24 shows the results of the GLM analysis. Again the hypothesis for depth has a very large mean square and is highly significant. The Jt/3 periodicity with interaction is significant while the 7t/6 is not. The set of environmental variables are significant and the P's of the CO model for NH3 and N02 + N03 are significant. The results of a multiple regression on the densities of Ammonia and the first three PC's of the environmental variables indicate that the first and third PC's are significant (Table 25). Consequently, all the variables except for salinity and N02 + N03 are involved. The analysis for the grass surface and 10 cm stands in marked contrast to the analysis for bare surface and grass surface where the F-ratio and R2 was relatively small (compare Tables 11 and 25). Bolivina: Bolivina mean densities at the grass surface and grass 10 cm are plotted in Figure 31. At the surface the maximum occurred in May and the June minimum observed for the three previously discussed taxa is also shown by Bolivina (Appendix 2). From July to September the density steadily decreased at the grass surface (Figure 31). At 10 cm the maximum occurred, similar to the previous taxa, in March. The average number of individuals observed was slightly less than one, and at seven sampling times no individuals were observed at 10 cm (Appendix 3). Table 26 shows the statistical analysis of the GLM. The mean square for depth is once again large and highly significant. The hypotheses for rc/6 periodicity and interaction, and 71/3 interaction are significant. This combination is not observed for any of the other taxa analyzed. The set of environmental variables are significant, and the (3's of the co model for temperature and salinity are significant. The F-ratios for simple regressions on densities vs. environmental variables TABLE 17.—Correlation matrix for chemical variables on grass surface and 10 cm. 0.05 level is underlined. Temperature Salinity Oxygen pH Eh NH3 PO4 Si N02 N02 + N03 Temperature 1.00 Salinity -0.02 1.00 Oxygen -0.35 -0.17 1.00 PH -0.06 -0.02 0.37 1.00 Eh -0.26 -0.25 0.78 0.63 1.00 NH3 0.31 0.36 -0.57 -0.47 -0.79 1.00 P04 0.10 0.29 -0.42 -0.27 -0.56 0.81 1.00 Si 0.19 0.18 -0.51 -0.53 -0.77 0.85 0.57 1.00 N02 -0.25 -0.09 0.41 0.38 0.44 -0.31 -0.20 -0.30 1.00 N02 + N03 -0.23 0.23 0.31 0.31 0.18 -0.14 -0.09 -0.17 0.27 1.0022 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES TABLE 18.—Factor score coefficients for chemical variables for grass surface and 10 cm. Chemical Factor variable PC 1(45%) PC2(15%) PC3(10%) Temperature 0.08 -0.25 0.76 Salinity 0.07 0.44 0.28 Oxygen -0.17 0.11 -0.20 PH -0.14 0.16 0.51 Eh -0.21 -0.02 0.07 NH3 0.20 0.17 0.03 P04 0.16 0.24 -0.03 Si 0.19 0.08 -0.18 N02 -0.12 0.23 -0.10 N02 + N03 -0.07 0.47 0.09 GRASS SURFACE QUINQUELOCULINA |2 = 254.68 8 = 350.65 ^ 1A 1978 \ i x GRASS 10 CENTIMETERS QUINQUELOCULINA £ = 3.10 8 = 5.12 100001 iooo- are all significant (Table 20). The results of a multiple regression on the densities of Bolivina and the first three PC's of the environmental variables (Table 27) indicate that all three PC's (all the environmental variables) are significant. Compari- son of Tables 13 and 27 indicate that for the bare surface and grass surface only PC2 was significant while at the grass surface and 10 cm all three PC's were significant. Simple regressions for the bare and grass surface had four variables significant (Table 6), while at the grass surface and 10 cm all were significant (Table 20). The F-ratio for salinity at the bare surface and grass surface is significant and the p's for the co model's for salinity in both analyses is significant. We recall that at the bare surface and grass surface the densities of Bolivina decrease during the period of our observations (Figure 16) and salinity exhibits the same trend (Figure 4). Salinity, then, may be important in regulating the density of Bolivina, but at 10 cm the decrease and increase in other variables probably overshadows its importance. Ammobaculites: Ammobaculites mean densities for the grass surface and grass 10 cm are shown in Figure 32. In keeping with the low densities observed in June for the other taxa at the grass surface station, no living individuals of Ammobaculites were observed on 19 June (Appendix 2). At 10 cm, the maximum number of individuals (three) were observed, as with the other taxa, in March (Appendix 3). One living individual was observed in April and one in June (Appendix 3). Table 28 displays the results of the statistical analysis of the GLM. The hypothesis for depth has the highest mean square and is significant. All of the periodicity hypotheses are significant except for 7T./3 interaction. The set of environmental variables are significant, and the (3's of the co model for salinity, oxygen, N02, and N02 + N03 are significant. The F-ratios for simple regressions on densities vs. environmental variables are significant for all variables except temperature and N02 + N03. Multiple regression of the densities of Ammobaculites and the first three PC's of the environmental variables (Table 29) indicate that the first and third PC's are significant. Therefore, all the environmental variables except for salinity and N02 + TABLE 19.—Statistical analysis of GLM for Quinqueloculina for grass surface and 10 cm. 1978 FIGURE 28.—Mean number of individuals of Quinqueloculina per 5 ml of sediment (density). Variability on Sum of df Mean F P(F) account of squares square Depth 24.18 1 24.18 49.85 0.00 rt/3 periodicity 22.99 4 5.75 11.85 0.00 and interaction 71/6 periodicity 28.80 4 7.20 14.85 0.00 and interaction JI/3 interaction 16.26 2 8.13 16.76 0.00 TC/6 interaction 17.30 2 8.65 17.83 0.00 Environmental 15.37 10 1.54 3.17 0.00 variables Residual 56.28 116 0.48 NUMBER 36 23 TABLE 20.—Values of F-ratio's for simple regressions on species densities and environmental variables at grass surface and grass 10 cm. (+ indicates signficant (.05 level) positive value of (3; - significant negative value of (}.) Environmental variables / / / Temperature 0.47 1.57 1.29 4.39" 0.28 Salinity 4.07- 0.22 0.64 6.10+ 20.94" Oxygen 55.17- 39.62+ 29.75+ 18.93- 26.27+ PH 130.66+ 155.71+ 110.81+ 38.94+ 71.08+ Eh 182.37+ 136.34+ 108.64+ 38.74+ 73.90+ NH3 112.61" 89.91- 74.87" 23.52" 50.86" PO4 26.66- 19.75" 15.44- 4.88" 15.60" Si 132.61- 113.69- 84.40" 34.70" 48.41- N02 20.84+ 23.28+ 25.01+ 8.70+ 34.35+ N02 + N03 4.76+ 7.98+ 2.38 9.00- 0.24 N03 are contributors according to the multiple regression analysis. As with most of the other taxa the F-ratio and R2 for the multiple regression at the bare surface and grass surface are much smaller than at the grass surface and 10 cm (Tables 15 and 29). Comparison of Analyses For the three most abundant taxa, Quinqueloculina, Elphid- ium, and Ammonia, the difference in overall densities at the bare surface and grass surface stations is not significant. In contrast, at the grass surface and 10 cm the differences in overall density for all taxa are significant. This is not at all surprising. The "living zone" for foraminifera in this area was previously determined to be about 6 or 7 cm (Buzas, 1977). Clearly, a depth of 10 cm in this area is an inhospitable environment for foraminifera, and the observed living individu- als may be due to haphazard excursions or to bioturbation by TABLE 21.—Regression of Quinqueloculina and principal components for grass surface and 10 cm. Variable Coefficient Standard error t P(2 tail) R2 = 0.66 Constant 3.00 0.11 26.61 0.00 PCI -1.70 0.11 -14.91 0.00 PC2 -0.02 0.11 -0.13 0.90 PC3 0.69 0.11 6.04 0.00 Analysis of variance Source Sum of ,r squares Mean square F-ratio P Regression Residual 446.03 3 227.50 132 148.68 1.72 86.26 0.00other organisms (Severin et al., 1982; Severin, 1987; Wetmore, 1988). The statistical analyses of the GLM for the gTass surface and 10 cm confirm what is abundantly clear by glancing at the plots of densities at the grass surface and at 10 cm (Figures 28-32). In both comparisons hypotheses for periodicity are signifi- cant for all taxa except Bolivina at the bare surface and grass surface. Except for the aforementioned case, the periodicites differ (interaction hypotheses) for all taxa between the areas being compared. At the surface stations there was a spring maximum at the bare station which was not observed at the grass station where maximum densities occurred in summer. Densities were very low at 10 cm, but nevertheless, the significance of the interaction hypotheses indicates no synchro- nization between the surface and 10 cm. Two-way ANOVA's on 10 environmental variables indicate that at the bare surface and grass surface most of these differ significantly with time while at the grass surface and 10 cm with depth. The analyses of both GLMs indicate that the environmental variables are significant for all taxa. The environmental variables are highly correlated so that the significance of any one is difficult to determine. In general, oxygen, pH, and Eh are positively correlated with one another, and negatively with NH3. Principal component analyses on both data sets indicate that the first three PC's account for all of the variables. Multiple regression of the first three PC's and densities of taxa indicate higher F-ratios and R2 for the grass surface and 10 cm analysis. In this analysis, PCI and PC3 are always significant, and all three are for Bolivina. On the other hand, for the bare surface and grass surface PCI and PC2 are significant for Elphidium and Ammobaculites, and PCI and PC3 for Quinqueloculina. Although these PC's also account for most of the variables and are significant, the strength of the regression relationship as judged by R2 and F-ratios are much smaller. For Bolivina only the F-ratio for PC2 is significant and salinity contributes heavily to this PC. This species also has a very high F-ratio for salinity on simple regression and is further identified as a significant (3 in the CO model. At the bare surface and grass surface only PC3 (accounting for 14% of the variability) is significant for Ammonia. N02 is the largest contributor to this PC, and like salinity for Bolivina, is also identified by simple regression and in the co model. Thus, no simple relationship between one or two environ- mental variables and density emerges. In both analyses oxygen, pH, and Eh are contrasted with NH3, and in the grass surface and 10 cm N02 is added to the former and P04 and Si to the latter. Except for Bolivina, the most commonly considered variables, temperature and salinity, have little involvement, and any attempt to predict faunal densities from them is futile. The very large change in density between the surface and 10 cm, perhaps, equivalent to a faunal change, is much more strongly related to the environmental variables than the smaller significant change with time at the surface sites. 24 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES g Z S 10 Q Z < GRASS SURFACE ELPHIDIUM ji = 30.90 a = 17.96 TABLE 22.—Statistical analysis of GLM for Elphidium for grass surface and 10 cm. Variability on Sum of df Mean F p(F) account of squares square Depth 20.79 1 20.79 54.29 0.00 TC/3 periodicity 6.49 4 1.62 4.23 0.00 and interaction TC/6 periodicity 1.57 4 0.39 1.02 0.40 and interaction 7T./3 interaction 5.63 2 2.82 7.35 0.00 7T./6 interaction 1.07 2 0.53 1.39 0.25 Environmental 15.05 10 1.50 3.93 0.00 variables Residual 44.39 116 0.38 •w "&. ^ °o, 1978 GRASS 10 CENTIMETERS ELPHIDIUM (2 =1.68 S =3.80 1978 FIGURE 29.—Mean number of individuals of Elphidium per 5 ml of sediment (density). TABLE 23.—Regression of Elphidium and principal components for grass surface and 10 cm. Variable Coefficient Standard error t P(2tail) R2 = 0.64 Constant 1.94 0.08 24.20 0.00 PCI -1.11 0.08 -13.65 0.00 PC2 0.17 0.08 2.13 0.04 PC3 0.51 0.08 6.25 0.00 Analysis of variance Source Sum of ,r squares Mean square F-ratio P Regression Residual 200.81 3 115.32 132 66.94 0.87 76.62 0.00 Comparison with Other Studies Quinqueloculina is the most abundant taxon making up about 75% of the total living foraminifera at the surface stations. The fortnightly observations of mean densities from March until November, 1978, exhibit a pronounced periodicity. The differences between maxima and minima densities are two orders of magnitude at both stations. At the grass station the maximum was recorded in summer and at the bare station in spring. In Aransas Bay, Texas, Phleger and Lankford (1957) also observed a summer maximum for Quinqueloculina. Jones and Ross (1979) found Quinqueloculina only during the summer months in Samish Bay, Washington. The temperature regimes in Aransas Bay (19°-30°C) and in the Indian River (22°-32°C)are similar. However, the waters of Samish Bay are much colder (5°-20°C),and the maximum at Samish Bay is the equivalent to the minimum at the other two areas. Elphidium is the second most abundant taxon comprising about 14% of the total number of living individuals at the two surface stations. The genus is composed mainly of E. mexicanum and E. gunteri. This taxon did not have significant differences between stations, but did exhibit a 7t/3 periodicity and interaction. Maxima occurred at the bare surface in early spring and at the grass surface in early summer. The spring maximum at the bare surface was much larger than at the grass surface. A number of researchers have demonstrated through observation of juveniles and density changes that E. excavatum reproduces all year round, often at different rates with maxima NUMBER 36 25 u 100 P en Z w Q Z < w GRASS SURFACE AMMONIA £ = 27.54 % % V K Vi X 1978 a ioo g Z - a GRASS 10 CENTIMETERS AMMONIA j2 = 1.98 CT = 4.61 1978 FIGURE 30.—Mean number of individuals of Ammonia per 5 ml of sediment (density). TABLE 25.—Regression of Ammonia and principal components for grass surface and 10 cm. Variable Coefficient Standard error t P(2 tail) R2 = 0.52 Constant 1.85 0.08 21.65 0.00 PCI -0.96 0.09 -11.04 0.00 PC2 0.08 0.09 0.93 0.36 PC3 0.41 0.09 4.73 0.00 Analysis of variance Source Sum of .f squares Mean square F-ratio P Regression Residual 143.94 3 130.85 132 47.98 0.99 48.40 0.00 varying from year to year (Buzas, 1965, 1969; Haake, 1967; Haman, 1969; Wefer, 1976). Other species of this genus follow a similar pattern (Boltovskoy and Lena, 1969). The different timing between stations (interaction) observed here was also observed by Buzas (1969) in the Choptank River, Maryland. Thus, the results of the present study are in accord with earlier observations. The environmental variables for Elphidium were significant as a group. Simple regressions on individual variables were highest for salinity and pH, however, the first two PC's proved significant and they account for all the variables except N02. Buzas (1969) using a similar analysis concluded the environ- mental variables were significant as a group as did Wefer (1976). Ammonia, consisting of A. beccarii (most of the individuals belong to the form called tepida), constitutes about 8% of the total living population at the bare and grass surface stations. More information is available for this species than any other, much of which is summarized by Walton and Sloan (1990). In this study, A. beccarii showed a 7t/3 periodicity with interaction. The bare surface had maxima in spring, summer, and fall while the grass surface had maximum densities in summer. This species also exhibited periodicity in the Choptank River, Maryland (Buzas, 1969), Texas Bays (Phleger and Lankford, 1957), Samish Bay, Washington (Jones and Ross, 1979), and Puerto Deseado (Boltovskoy and Lena, 1969). As in this study, Phleger and Lankford (1957) and Buzas (1969) reported different densities with time between stations, while Boltovskoy and Lena (1969) between years. On the other hand, in Narragansett Bay, Rhode Island, Brooks (1967) found 26 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES 100 n GRASS SURFACE BOLIVINA ji =726 6 = 7.98 TABLE 26.—Statistical analysis of GLM for Bolivina for grass surface and 10 cm. Variability on Sum of df Mean F P(F) account of squares square Depth 6.70 1 6.70 11.05 0.00 TC/3 periodicity 4.84 4 1.21 2.00 0.10 and interaction rc/6 periodicity 6.27 4 1.57 2.59 0.04 and interaction Jt/3 interaction 4.63 2 2.31 3.82 0.02 Tt/6 interaction 0.04 2 0.02 0.03 0.97 Environmental 24.25 10 2.42 4.00 0.00 variables Residual 70.25 116 0.61 *A, -fci. ^> 1978 GRASS 10 CENTIMETERS BOLIVINA ji = 0.94 6 = 2.50 1978 FIGURE 31.—Mean number of individuals of Bolivina per 5 ml of sediment (density). TABLE 27.—Regression of Bolivina and principal components for grass surface and 10 cm. Variable Coefficient Standard error t P(2 tail) R2 = 0.33 Constant 0.99 0.08 12.75 0.00 PCI -0.51 0.08 -6.52 0.00 PC2 0.29 0.08 3.72 0.00 PC3 0.22 0.08 2.81 0.01 Analysis of variance Source Sum of .r squares Mean square F-ratio P Regression Residual 52.24 3 107.38 132 17.41 0.81 21.40 0.00 no difference in densities between monthly sampling during 1963, and, likewise, Buzas et al. (1977) found no difference in densities between monthly sampling during 1969-1970 in Jamaica. Most authors agree that reproduction takes place throughout the year. Given enough data, it seems likely that a particular station could show periodicity in some years, and not in others, while a nearby station might behave in an altogether different manner. In the present study of Ammonia, analysis by the GLM showed that the environmental variables were significant as a group. Simple regressions on density vs. individual environ- mental variables indicated that only salinity and N02 were significant. Regression analysis on the PC's indicated that only the third PC accounting for 14% of the variability in the environmental variables was significant. The largest factor score coefficients for PC3 are N02, Si, and temperature. Interestingly, the sign of the simple regression coefficient for temperature, although not significant, was negative. In the Choptank River, Maryland, Buzas (1969) found the group of measured environmental variables were significant for Ammo- nia; however, in Jamaica, they were not (Buzas et al., 1977). The laboratory work of Bradshaw (1961) and Schnitker (1974) indicate a temperature of at least 17°-22°C is required for reproduction (Walton and Sloan, 1990, give an extensive discussion). The temperatures in Jamaica and in the present study are always above these limits. In the Choptank River, however, the temperature is below 15°C for 7 months of the year (Buzas, 1969), and Ammonia exhibits significant increases in density during this period. In Narragansett Bay where temperatures vary from 0°-20°C, Brooks (1967) found no NUMBER 36 27 100 GRASS SURFACE AMMOBACULITES £ = 5.69 a = S.80 \ X X i\JTABLE 28.—Statistical analysis of GLM for Ammobaculites for grass surface and 10 cm. Variability on Sum of df Mean F p(F) account of squares square Depth 5.77 1 5.77 21.62 0.00 TC/3 periodicity 2.83 4 0.71 2.65 0.04 and interaction TC/6 periodicity 9.98 4 2.49 9.34 0.00 and interaction TC/3 interaction 0.63 2 0.32 1.19 0.31 n/6 interaction 2.51 2 1.25 4.69 0.01 Environmental 20.44 10 2.04 7.65 0.00 variables Residua] 30.91 116 0.27 1978 z u <=> 10 GRASS 10 CENTIMETERS AMMOBACULITES $ = 0.07 8 =0.32 TABLE 29.—Regression of Ammobaculites and principal components for grass surface and 10 cm. Variable Coefficient Standard error t P(2tail) R2 = 0.42 Constant 0.78 0.07 11.78 0.00 PCI -0.62 0.07 -9.22 0.00 PC2 -0.11 0.07 -1.66 0.10 PC3 0.17 0.07 2.50 0.01 , ^N 1978 FIGURE 32.—Mean number of individuals of Ammobaculites per 5 ml of sediment (density). Analysis of variance Source Sum of J* squares Mean square F-ratio P Regression Residual 55.59 3 78.05 132 18.53 0.59 31.34 0.00 significant difference in density with time over the period of a year. Both of these authors sampled with replicates to offset the variation inherent because of spatial distribution. In Puerto Deseado, Argentina, Boltovskoy (1964) and Boltovskoy and Lena (1969) report, through observations of juveniles, continu- ous reproduction of this species where the maximum tempera- ture is only 15°C. These observations suggest to us that populations of A. beccarii can adjust to lower temperatures for reproduction than were observed in the laboratory. The alternative suggested by Walton and Sloan (1990) is that seasonal trends may be an artifact due to the sampling of patchy distributions. We believe the replication and use of independent hypotheses through the analysis of variance technique (Brooks, 1967; Buzas, 1969) and observations of juveniles (Boltovskoy, 1964) make this conclusion unlikely. Bolivina, consisting mainly of B. striatula, makes up about 1% of the total living population. The mean number of individuals at the grass station was about 7 and at the bare station 3, which was statistically significantly different. At both stations there appears an overall decrease in density during the sampling period. However, none of the hypotheses for periodicity in the GLM were significant. In Jamaica (Buzas et al., 1977), B. striatula was the most abundant species and it did exhibit an overall periodicity, even though most of the species analyzed in that study did not. No other data are available. Perhaps, as more data becomes available, this species will exhibit the same unpredictable behavior cited above for Ammonia. 28 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES In the GLM for Bolivina the environmental variables are significant as a group. Simple regressions indicate salinity, P04, Si, and N02 + N03 are significant. Multiple regression of densities and PC's found only PC2 significant and the factor score coefficients indicate these same variables are the main components of PC2. Buzas et al. (1977) found no statistical relationship between environmental variables and densities for this species in Jamaica. Ammobaculites, consisting mainly of A. exiguus, makes up about 1% of the total living population. At the grass surface the mean density is about 6 and at the bare surface 1, which is statistically significantly different. The 7t/3 overall periodicity and interaction hypotheses are also significant. At the grass surface there is an increasing trend in density during the sampling period while at the bare surface a spring maximum is followed by very low densities during the sampling period. In the Choptank River, Maryland, Buzas (1969) also found a TC/3 overall periodicity and interaction hypotheses significant for this species. Maximum densities occurred in all seasons. Based on size distribution histograms which showed mixed age groups throughout the year, Phleger and Lankford (1957) also concluded that A. salsus (probably the same or a closely related species) reproduced during all seasons of the year. The null hypothesis for the environmental variables in the GLM for Ammobaculites was rejected. Simple regressions on density and individual environmental variables indicated temperature, salinity, oxygen, and NH3 were significant. Regression analyses of density and the PC's indicated PCI and PC2 were significant. These two components account for all the variables except N02. Buzas (1969) also found the environ- mental variables were significant for this species in the Choptank River, Maryland. The life cycle of some foraminifera such as Elphidium crispum and Glabratella ornatissima produces a marked yearly (often spring and summer) increase in density (Myers, 1942; Erskian and Lipps, 1987). The majority of species such as those studied here apparently reproduce continuously, although they need not do so at the same rate. The interplay of life cycle behavioral characteristics and environmental variables may be more favorable at some particular times than at others. Furthermore, predation severely depletes foraminiferal densi- ties (Buzas, 1978), and differential predation may decrease density at a particular location. Indeed, the density of potential predators often varies from season to season and year to year (Young et al., 1976). On the other hand, these causes of density regulation may offset each other and the same species may sometimes not exhibit periodicity. The observations of various researchers cited above for Ammonia which sometimes shows periodicity and sometimes not is best explained in this fashion. Appendix 1 Bare Surface Number of individuals per 5 ml of sedimenL Date (1978) / A? 27 Mar 712 140 30 2 4 890 1154 395 95 9 3 1661 598 194 71 1 6 873 2232 900 123 10 19 3304 10 Apr 1815 229 53 12 3 2112 2683 606 111 18 8 3434 2537 411 75 7 7 3038 2021 328 55 13 11 2434 24 Apr 412 118 26 5 2 566 404 118 25 7 558 514 86 32 1 634 482 94 36 1 615 8 May 499 78 20 3 604 275 38 24 4 342 280 101 22 6 2 419 221 36 18 5 0 283 22 May 81 50 11 1 0 148 94 45 31 8 2 185 91 38 9 2 0 140 29 9 26 2 0 68 5 Jun 150 24 22 9 1 220 307 35 39 9 1 398 78 48 20 3 0 151 82 50 25 2 0 162 19Jun 44 45 28 3 0 125 47 45 44 7 0 156 30 38 26 8 0 106 155 51 51 9 0 279 3 Jul 97 35 70 3 0 208 108 50 84 6 0 251 92 30 30 5 1 167 138 33 82 7 0 265 17 Jul 38 78 41 3 1 167 41 12 31 1 0 91 56 20 25 2 0 106 90 28 29 8 1 163 31 Jul 290 32 13 6 1 351 231 15 26 2 0 279 175 23 11 3 2 215 187 14 23 3 1 235 14 Aug 109 18 22 0 0 155 261 27 18 1 2 312 98 34 21 1 0 155 164 24 26 1 0 220 29 30 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES Appendix 1.—Continued. Date (1978) 28 Aug 44 11 6 1 0 64 53 12 16 0 0 86 93 14 8 1 1 118 65 14 3 0 2 85 11 Sep 102 6 11 0 0 121 137 24 11 1 0 174 51 20 26 0 0 101 73 24 29 1 4 131 25 Sep 59 37 16 0 1 118 86 19 11 0 0 119 41 22 15 2 0 82 82 10 11 0 1 104 9 Oct 27 15 8 0 2 53 55 14 3 0 0 72 27 7 1 1 0 38 46 6 14 2 1 79 23 Oct 253 77 120 0 0 451 70 9 14 0 0 93 233 44 28 1 0 306 104 24 18 0 0 146 6 Nov 77 29 22 0 0 129 69 27 41 0 1 138 82 44 55 0 0 182 213 165 129 1 0 508Appendix 2 Grass Surface Number of individuals per 5 ml of sediment. Date (1978) / ■4? £ *? 27 Mar 70 20 24 19 i 151 95 28 24 13 i 166 140 29 38 24 9 269 136 58 17 11 2 233 10 Apr 301 49 35 8 5 411 189 27 22 20 1 273 137 38 29 28 0 251 85 25 21 16 2 154 24 Apr 213 41 16 22 2 305 98 43 9 8 0 163 260 46 15 5 2 335 145 40 19 7 4 224 8 May 77 14 15 7 2 122 97 26 7 11 0 147 103 29 18 6 1 164 129 30 17 7 0 190 22 May 156 72 54 33 11 365 132 68 39 21 6 283 126 77 47 31 6 297 115 55 15 0 2 187 5 Jun 80 26 10 9 1 137 30 7 9 1 0 47 113 34 24 13 3 197 40 21 9 3 0 91 19Jun 52 7 3 2 0 69 63 8 7 4 0 87 55 11 4 0 0 73 18 12 8 3 0 42 3 Jul 91 59 40 14 3 216 202 39 55 11 2 315 79 40 33 7 10 176 111 44 60 18 4 246 17 Jul 1063 30 55 10 2 1164 1018 33 73 5 2 1151 2163 83 222 14 7 2507 1412 38 58 11 1 1534 31 Jul 427 26 20 5 3 502 467 28 24 9 2 540 524 30 17 3 4 605 624 49 79 2 1 767 14 Aug 715 23 46 3 18 828 526 25 15 1 5 592 399 21 19 2 6 448 783 19 26 3 7 849 31 32 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES Appendix 2.—Continued. Date (1978) 4 / 28 Aug 163 19 14 0 5 208 196 17 10 2 2 248 196 3 14 1 2 221 185 33 4 2 6 247 11 Sep 117 31 11 1 12 178 77 24 11 1 11 128 76 18 20 0 4 127 95 41 19 1 6 164 25 Sep 208 24 18 0 16 270 162 29 3 0 17 217 131 18 22 0 9 187 199 20 15 2 7 255 9 Oct 152 33 27 3 16 239 159 19 25 5 16 227 145 17 19 0 10 207 67 10 4 0 8 108 23 Oct 224 25 30 7 22 329 304 81 56 5 18 480 197 34 54 3 19 321 238 25 37 7 16 336 6 Nov 115 20 33 2 14 208 19 17 19 0 1 57 17 6 6 0 3 39 17 9 5 2 9 55Appendix 3 Grass 10 cm Number of individuals per 5 ml of sediment. Date (1978) / / 27 Mar 2 1 2 0 0 5 27 7 14 4 0 53 9 18 15 5 2 54 22 24 32 17 1 100 10 Apr 5 0 1 0 0 6 2 0 2 0 0 4 1 3 1 0 0 5 6 2 2 0 0 11 24 Apr 5 4 7 2 0 20 0 0 2 0 1 4 5 4 0 0 0 10 0 0 2 0 0 3 8 May 0 1 2 0 0 3 1 1 0 0 0 2 3 1 0 0 0 4 1 0 1 0 0 2 22 May 3 2 1 0 0 8 1 1 5 4 0 11 11 3 0 1 0 16 5 4 5 0 0 16 5Jun 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 4 1 0 0 0 0 1 19Jun 14 6 5 4 0 29 3 3 4 5 0 15 0 0 1 0 0 1 4 4 4 1 1 15 3 Jul 0 0 0 0 0 0 1 0 0 0 0 1 1 0 1 0 0 2 0 3 1 0 0 4 17 Jul 0 1 0 0 0 2 1 0 2 3 0 7 2 2 0 0 0 5 4 5 2 7 0 19 31 Jul 1 0 3 0 0 4 4 0 1 0 0 5 4 0 1 0 0 5 2 0 1 0 0 3 14 Aug 1 1 1 0 0 4 1 1 0 0 0 2 2 1 0 0 0 3 3 3 2 3 0 11 33 34 SMITHSONIAN CONTRIBUTIONS TO THE MARINE SCIENCES Appendix 3.—Continued. Date 5>° £ ■<$> (1978) / # *? 28 Aug 1 0 0 0 0 1 1 0 0 0 0 1 1 1 0 0 0 2 0 1 0 0 0 1 11 Sep 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 25 Sep 2 1 1 0 0 4 5 3 2 3 0 13 2 0 5 0 0 7 0 0 0 0 0 0 9 Oct 2 0 0 0 0 2 0 0 0 0 0 0 7 0 1 1 0 9 1 0 0 0 0 1 23 Oct 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 1 1 0 3 6 Nov 2 0 1 0 0 3 2 0 0 0 0 2 6 0 0 0 0 6 20 2 0 0 0 22Appendix 4 Bare Surface Pore water chemistry. Date (1978) Temperature Salinity Oxygen pH Eh NH, P04 Si NOz N02 + N03 27 Mar 24.5 32.0 7.4 8.63 31 0.1425 0.7354 13.3037 0.0721 4.3109 10 Apr 26.5 32.5 9.1 8.70 85 0.0920 0.2202 26.7167 0.0609 3.0221 24 Apr 23.5 32.5 10.2 8.75 30 0.1149 0.0320 7.1574 0.0113 4.9155 8 May 27.0 31.5 0.4 7.50 -208 0.1204 0.5373 11.5527 0.0218 27.1770 22 May 28.3 39.0 1.8 8.30 -70 21.2810 0.0412 49.3007 0.0229 3.8981 5Jun 32.0 35.0 0.6 7.50 -151 95.9300 0.8373 70.7123 0.3944 4.5691 19Jun 27.0 33.0 5.1 8.10 184 2.1058 0.0560 29.4110 0.3152 3.3008 3 Jul 30.0 32.0 0.4 7.55 -187 65.8357 1.1206 32.1760 0.3722 47.3260 17 Jul 31.0 30.5 3.3 7.50 47 1.4197 0.0524 68.8000 0.3134 0.1385 31 Jul 30.0 28.0 4.4 8.30 -90 0.1717 0.1915 55.7310 0.2811 1.3133 14 Aug 31.0 20.0 2.3 8.08 -28 21.3100 1.6034 51.5230 0.2359 0.5315 28 Aug 30.5 26.0 2.4 8.21 -34 0.1721 0.5327 0.0000 0.0224 0.0622 11 Sep 30.5 28.0 3.4 8.06 -46 1.8610 2.2904 104.0400 0.0215 0.0668 25 Sep 29.5 26.0 5.4 8.12 -30 0.1638 16.4580 66.4087 0.0225 0.1841 9 Oct 24.0 25.5 6.0 7.97 50 0.1790 0.0742 133.5466 0.0218 0.0844 23 Oct 24.0 26.0 5.3 8.00 44 1.2926 0.4990 105.5033 0.3724 0.6684 6 Nov 23.0 27.0 4.9 8.35 55 3.1530 1.8657 69.4803 0.6236 1.8702 35 Appendix 5 Grass Surface Pore water chemistry. Dale (1978) Temperature Salinity Oxygen PH Eh NH3 PO4 Si N02 NOz + N03 27 Mar 22.0 32.5 4.9 8.03 -4 0.1425 0.8665 13.7943 0.0549 20.8870 10 Apr 24.5 31.0 7.6 8.25 105 0.0920 0.1329 22.6467 0.0535 3.2316 24 Apr 22.8 33.0 7.3 8.35 7 0.9383 0.1563 12.0900 0.2704 275.4266 8 May 27.0 31.0 4.2 8.00 -56 0.1204 0.0391 14.7607 0.0218 5.3229 22 May 26.5 33.0 3.6 8.15 -67 7.8668 0.0412 32.5660 0.0229 10.0146 5 Jun 29.0 31.0 4.1 7.60 -128 33.7027 0.3404 50.9217 0.1105 2.2312 19Jun 25.2 32.0 5.6 8.00 185 0.1486 0.0560 1.3497 0.1303 1.2319 3 Jul 29.5 29.0 0.5 7.50 -115 5.5559 0.4846 69.2380 0.2146 1.5599 17 Jul 28.0 28.0 4.1 7.74 38 0.0000 0.0000 70.8490 0.0313 0.0000 31 Jul 29.0 31.0 3.6 8.08 -50 4.6810 0.4182 43.1990 0.1782 0.0000 14 Aug 30.0 22.0 2.1 8.17 -47 6.9164 0.9911 16.9515 0.1416 0.0639 28 Aug 31.0 26.0 3.0 8.21 -12 0.1721 0.3936 0.0000 0.0224 0.1458 11 Sep 29.5 25.5 2.8 8.03 -33 2.9416 1.1962 98.0553 0.0215 0.0668 25 Sep 28.5 25.5 5.4 8.16 -37 0.3979 0.4561 72.2520 0.0225 0.0676 9 Oct 23.5 24.5 5.3 7.72 52 0.1790 0.5758 138.5800 0.0218 0.0844 23 Oct 23.3 28.0 5.9 8.00 43 4.6589 0.7639 106.3733 0.5176 1.4438 6 Nov 23.0 26.5 6.0 8.19 50 4.2475 2.2476 68.6020 0.5030 1.7514 36 Appendix 6 Grass 10 cm Pore water chemistry. Date (1978) Temperature Salinity Oxygen pH Eh NH3 PO4 Si N02 N02 + N03 27 Mar 24.2 27.0 0.0 7.71 -196 96.2270 12.2930 241.4466 0.0095 0.0475 10 Apr 23.8 30.7 0.0 7.67 -235 0.0000 0.0000 0.0000 0.0000 0.0000 24 Apr 23.5 33.0 0.0 7.60 -232 285.6600 28.0770 440.2833 0.0113 0.0382 8 May 27.2 33.5 0.0 7.30 -286 312.6400 26.0440 583.6932 0.0218 0.0589 22 May 27.2 34.5 0.0 7.45 -289 463.4599 29.3460 515.6699 0.0229 0.0540 5 Jun 28.5 33.5 0.0 7.50 -307 836.8533 51.8203 688.2266 0.0254 2.2813 19Jun 25.5 32.5 0.0 7.60 -331 929.6766 32.1800 659.5033 0.0230 2.2518 3 Jul 36.2 32.0 0.0 7.52 -312 745.4833 70.4513 616.1066 0.0235 1.4169 17 Jul 27.7 30.5 0.0 7.35 -282 518.7266 46.8573 572.8933 0.0313 0.1127 31 Jul 27.8 30.5 0.0 7.44 -288 387.2833 36.5710 559.7299 0.0282 0.0800 14 Aug 21.0 24.0 0.0 7.62 -223 198.0666 19.7187 553.3200 0.0209 0.5521 28 Aug 30.0 26.5 0.0 7.46 -223 233.1100 13.1470 0.0000 0.0224 0.3995 11 Sep 31.0 27.5 0.0 7.49 -227 417.0299 25.0397 917.0634 0.0215 0.0668 25 Sep 29.0 27.5 5.4 7.49 -225 470.0866 26.2023 785.4966 0.0225 0.2034 9 Oct 24.0 26.0 5.3 7.35 -199 240.5966 12.2607 461.1333 0.1000 0.0844 23 Oct 26.5 27.3 4.7 7.04 -33 0.0000 0.0000 0.0000 0.0000 0.0000 6 Nov 23.5 27.0 6.0 6.57 -176 187.9799 22.9157 324.8800 0.0232 0.0817 37 Literature Cited Bliss, C.I. 1958. Periodic Regression in Biology and Climatology. Connecticut Experiment Station Bulletin, 615:3-55. Boltovskoy, E. 1964. Seasonal Occurrences of Some Living Foraminifera in Puerto Deseado (Patagonia, Argentina). Journal du Conseil International pour I'Exploration de la Mer, 29(2):136-145. Boltovskoy, E., and H. Lena 1969. Seasonal Occurrences, Standing Crop, and Production in Benthic Foraminifera of Puerto Deseado. Contributions from the Cushman Foundation for Foraminiferal Research, 20(3):87-95. Bradshaw, J.S. 1961. Laboratory Experiments on the Ecology of Foraminifera. Contribu- tions from the Cushman Foundation for Foraminiferal Research, 12(3):86-106. Brooks, A.L. 1967. Standing Crop, Vertical Distribution, and Morphometries of Ammo- nia beccarii (Linne). Limnology and Oceanography, 12(4):667- 684. Buzas, M.A. 1965. The Distribution and Abundance of Foraminifera in Long Island Sound. Smithsonian Miscellaneous Collections, 149(1): 89 pages. 1969. Foraminiferal Species Densities and Environmental Variables in an Estuary. Limnology and Oceanography, 14(3):411-422. 1977. Vertical Distribution of Foraminifera in the Indian River, Florida. Journal of Foraminiferal Research, 7(3):234-237. 1978. Foraminifera As Prey for Benthic Deposit Feeders: Results of Predator Exclusion Experiments. Journal of Marine Research, 36(4):617-625. Buzas, M.A., and K.P. Severin 1982. Distribution and Systematics of Foraminifera in the Indian River, Florida. Smithsonian Contributions to the Marine Sciences, 16: 73 pages. Buzas, M.A., R.K. Smith, and K.A. Beem 1977. Ecology and Systematics of Foraminifera in Two Thalassia Habitats, Jamaica, West Indies. Smithsonian Contributions to Paleobiology, 31: 139 pages. Erskian, M.G., and J.H. Lipps 1987. Population Dynamics of the Foraminiferan Glabratella ornatissima (Cushman) in Northern California. Journal of Foraminiferal Research, 17(3):240-256. Haake, F.-W. 1967. Zum Jahresgang von populationen einer Foraminiferen-Art in der westlichen Ostsee. Meyniana, 17:13-27. Haman, D. 1969. Seasonal Occurrence of Elphidium excavatum (Terquem) in Llan- danwg Lagoon (North Wales, U.K.). Contributions from the Cushman Foundation for Foraminiferal Research, 20(4): 139-142. Jones, G.D., and C.A. Ross 1979. Seasonal Distribution of Foraminifera in Samish Bay, Washington. Journal of Paleontology. 53(2):245-257. Montgomery, J.R., C. Zimmermann, G. Peterson, and M. Price 1983. Diel Variations of Dissolved Ammonia and Phosphate in Estuarine Sediment Pore Water. Florida Scientist, 46(3/4):407-414. Montgomery, J.R., C. Zimmermann, and M. Price 1979. The Collection, Analysis, and Variation of Nutrients in Estuarine Pore Water. Estuarine and Coastal Marine Science, 9:203-214. Myers, E.H. 1942. A Quantitative Study of the Productivity of the Foraminifera in the Sea. Proceedings of the American Philosophical Society, 85(4): 325-342. Nyholm, G.-G., and I. Olsson 1973. Seasonal Fluctuations of the Meiobenthos in an Estuary on the Swedish West Coast. Zoon, 1:69-76. Phleger, F.B, and R.R. Lankford 1957. Seasonal Occurrences of Living Benthonic Foraminifera in Some Texas Bays. Contributions from the Cushman Foundation for Foraminiferal Research, 8(3):93-105. Schnitker, D. 1974. Ecotypic Variation in Ammonia beccarii (Linne). Journal of Foraminiferal Research, 4(4):216-223. Seal, H.L. 1964. Multivariate Statistical Analysis for Biologists. 207 pages. New York: Wiley. Severin, K.P. 1987. Laboratory Observations of the Rate of Subsurface Movement of a Small Miliolid Foraminifer. Journal of Foraminiferal Research, 17(2):110-116. Severin, K.P, S.J. Culver, and C. Blanpied 1982. Burrows and Trails Produced by Quinqueloculina impressa Reuss, a Benthic Foraminifer, in Fine-grained Sediment Sedimentology, 29:897-901. Walton, W.R., and B.J. Sloan 1990. The Genus Ammonia Brunnich, 1772: Its Geographic Distribution and Morphologic Variability. Journal of Foraminiferal Research, 20(2):128-156. Wefer, G. 1976. Environmental Effects on Growth Rates of Benthic Foraminifera (Shallow Water, Baltic Sea). In C.T. Shaffer and B.R. Pelletier, editors, First International Symposium on Benthonic Foraminifera of Continental Margins, Part A: Ecology and Biology, pages 39-50. Halifax, Nova Scotia: Editoral Board of Maritime Sediments under the auspices of the Coordinating Committee of Benthonics "75." Wetmore, K.L. 1988. Burrowing and Sediment Movement by Benthic Foraminifera, as Shown by Time-lapse Cinematography. Revue de Paleobiologie Special Volume, 2:921-927. Young, D.K., M.A. Buzas, and M.W. Young 1976. Species Densities of Macrobenthos Associated with Seagrass: A Field Experimental Study of Predation. Journal of Marine Research, 34(4):577-592. 38 REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION Manuscripts intended for series publication receive substantive review (conducted by their originating Smithsonian museums or offices) and are submitted to the Smithsonian Institution Press with Form SI-36, which must show the approval of the appropriate authority designated by the sponsoring organizational unit. Requests for special treatment—use of color, foldouts, case-bound covers, etc.—require, on the same form, the added approval of the sponsoring authority. Review of manuscripts and art by the Press for requirements of series format and style, completeness and clarity of copy, and arrangement of all material, as outlined below, will govern, within the judgment of the Press, acceptance or rejection of manuscripts and art. Copy must be prepared on typewriter or word processor, double-spaced, on one side of standard white bond paper (not erasable), with 17V margins, submitted as ribbon copy (not carbon or xerox), in loose sheets (not stapled or bound), and accompanied by original art. Minimum acceptable length is 30 pages. Front matter (preceding the text) should include: title page with only title and author and no other information; abstract page with author, title, series, etc., following the established format; table of contents with indents reflecting the hierarchy of heads in the paper; also, foreword and/or preface, if appropriate. First page of text should carry the title and author at the top of the page; second page should have only the author's name and professional mailing address, to be used as an unnumbered footnote on the first page of printed text. Center heads of whatever level should be typed with initial caps of major words, with extra space above and below the head, but no other preparation (such as all caps or underline, except for the underline necessary for generic and specific epithets). Run-in paragraph heads should use period/dashes or colons as necessary. Tabulations within text (lists of data, often in parallel columns) can be typed on the text page where they occur, but they should not contain rules or numbered table captions. Formal tables (numbered, with captions, boxheads, stubs, rules) should be submitted as carefully typed, double-spaced copy separate from the text; they will be typeset unless otherwise requested. If camera-copy use is anticipated, do not draw rules on manuscript copy. Taxonomic keys in natural history papers should use the aligned-couplet form for zoology and may use the multi-level indent form for botany. If cross referencing is required between key and text, do not include page references within the key, but number the keyed-out taxa, using the same numbers with their corresponding heads in the text. Synonymy in zoology must use the short form (taxon, author, yearpage), with full reference at the end of the paper under "Literature Cited." For botany, the long form (taxon, author, abbreviated journal or book title, volume, page, year, with no reference in "Literature Cited") is optional. Text-reference system (author, yearpage used within the text, with full citation in "Literature Cited" at the end of the text) must be used in place of bibliographic footnotes in all Contributions Series and is strongly recommended in the Studies Series: "(Jones, 1910:122)" or "...Jones (1910:122)." If bibliographic footnotes are required, use the short form (author, brief title, page) with the full citation in the bibliography. Footnotes, when few in number, whether annotative or biblio- graphic, should be typed on separate sheets and inserted immedi- ately after the text pages on which the references occur. Extensive notes must be gathered together and placed at the end of the text in a notes section. Bibliography, depending upon use, is termed "Literature Cited," "References," or "Bibliography." Spell out titles of books, articles, journals, and monographic series. For book and article titles use sentence-style capitalization according to the rules of the language employed (exception: capitalize all major words in English). For journal and series titles, capitalize the initial word and all subsequent words except articles, conjunctions, and prepositions. Transliterate languages that use a non-Roman alphabet according to the Library of Congress system. Underline (for italics) titles of journals and series and titles of books that are not part of a series. Use the parentheses/colon system for volume (number):pagination: "10(2):5-9." For alignment and arrangement of elements, follow the format of recent publications in the series for which the manuscript is intended. Guidelines for preparing bibliography maybe secured from Series Section, SI Press. Legends for illustrations must be submitted at the end of the manuscript, with as many legends typed, double-spaced, to a page as convenient. Illustrations must be submitted as original art (not copies) accompanying, but separate from, the manuscript. Guidelines for preparing art may be secured from the Series Section, SI Press. All types of illustrations (photographs, line drawings, maps, etc.) may be intermixed throughout the printed text. They should be termed Figures and should be numbered consecutively as they will appear in the monograph. If several illustrations are treated as components of a single composite figure, they should be designated by lowercase italic letters on the illustration; also, in the legend and in text references the italic letters (underlined in copy) should be used: "Figure 9b." Illustrations that are intended to follow the printed text may be termed Plates, and any components should be similarly lettered and referenced: "Plate 9b." Keys to any symbols within an illustation should appear on the art rather than in the legend. Some points of style: Do not use periods after such abbrevia- tions as "mm, ft, USNM, NNE." Spell out numbers "one" through "nine" in expository text, but use digits in all other cases if possible. Use of the metric system of measurement is preferable; where use of the English system is unavoidable, supply metric equivalents in parentheses. Use the decimal system for precise measurements and relationships, common fractions for approximations. Use day/month/ year sequence for dates: "9 April 1976." For months in tabular listings or data sections, use three-letter abbreviations with no periods: "Jan, Mar, Jun," etc. Omit space between initials of a personal name: "J.B. Jones." Arrange and paginate sequentially every sheet of manuscript in the following order: (1) title page, (2) abstract, (3) contents, (4) foreword and/or preface, (5) text, (6) appendices, (7) notes section, (8) glossary, (9) bibliography, (10) legends, (11) tables. Index copy may be submitted at page proof stage, but plans for an index should be indicated when the manuscript is submitted.