Abundance and distribution of testate amoebae bearing siliceous plates in freshwater lakes and ponds along the east coast of North America: Importance of water depth and pH

Peter A. Siver1,2, Anne M. Lott1,3, and Paula Torres1,4

1Department of Botany, Connecticut College, 270 Mohegan Avenue, New London, Connecticut 06320 USA

Abstract: Testate amoebae are a highly diverse and polyphyletic group of heterotrophic, free-living amoeboid pro- tists that consist of a single cell enclosed within a shell, or test. These organisms inhabit a broad range of habitats, including lakes, ponds, rivers, bogs, wetlands, and peatlands, where they prey on bacteria, algae, other protists, and in some cases small micrometazoans. One group of testate amoebae produce the test out of overlapping siliceous plates that are formed individually within the cell and then secreted and glued together to make an organized shell. Upon amoebae death, the siliceous plates can accumulate in lake sediments and become part of the fossil record. The goals of the current study were to document the concentrations of siliceous plate morphotypes in waterbodies along the east coast of North America, examine distributional patterns, and determine the relationships between environmental variables and abundances of plates. Seven siliceous plate morphotypes representing remains of tes- tate amoebae belonging largely to the Order Euglyphida were enumerated in surface sediments from 125 water- bodies situated from North Carolina, USA, to Newfoundland, Canada. Circular-shaped plates were the most widely distributed morphotype, found in 95% of the waterbodies, and along with oval-shaped plates, they accounted for 75% of all specimens enumerated. Other plate morphotypes, including quadrangular, rectangular, rhomboidal, and scutiform, were also common, and all morphotypes exhibited distinctive distribution patterns. We used best subset regression to evaluate the relationships between environmental variables and the concentration of plates/dry mass of sediment and to assess differences among sites. Water depth and pH were consistently shown to explain the most variation in testate plate abundance. Concentrations of potassium and sulfate were of lesser importance. Higher con- centrations of siliceous plates were found in shallow and acidic waterbodies, indicating the possibility of recon- structing estimates of these 2 variables in ancient waterbodies. Key words: Euglyphida, North America, pH, siliceous plates, testate amoebae, water depth

Testate amoebae, often referred to as thecamoebians, form form a covering. Differences in the test, type of pseudopo- a highly diverse and polyphyletic group of heterotrophic, dia (e.g., filose or lobose), and other cellular features reflect free-living amoeboid protists that consist of a single cell the polyphyletic nature of the group (Mitchell et al. 2008). enclosed within a structure called the test (Mitchell et al. Testate amoebae within the Class Imbricatae in the Phy- 2008). Pseudopodia, which emerge from the test through lum Cercozoa form tests out of siliceous components, or an opening known as the pseudostome, are used for move- plates, that are produced within the cell, secreted to the out- ment and feeding. Some testate amoebae construct the test side of the cell membrane, and glued together to form an or- from organic or inorganic components (e.g., silica plates) ganized covering. The shapes of the siliceous plates on the secreted by the cell, whereas others scavenge material and body of the organism may be circular, square, rhomboid, particles from the environment and glue them together to or scutiform (shield-shaped), depending on the species

E-mail addresses: [email protected]; [email protected]; [email protected]

DOI: 10.1086/711691. Received 11 January 2020; Accepted 5 May 2020; Published online 27 October 2020; Associate Editor, Seth Wenger. Freshwater Science. 2020. 39(4):000–000. © 2020 by The Society for Freshwater Science. 000 000 | Abundances of freshwater testate amoebae P. A. Siver et al.

(Ogden and Hedley 1980). Some siliceous plates may bear found in the Giraffe core, Barber et al. (2013) discussed a protruding spine, and the apertural plates that surround the possibility of using these euglyphid remains to aid in re- the pseudostome are usually differentiated from body plates constructing the history of the waterbody, which is repre- where the anterior end is lined with teeth (Ogden and Hed- sented by >65 m of mudstone sediments. Since the Barber ley 1980, Ogden 1981). The most common and ecologi- et al. (2013) study, we have uncovered euglyphid specimens cally important thecamoebians forming siliceous plates are in many additional strata from the Giraffe site and believe taxa within the Order Euglyphida Copeland, 1956 emended they may serve as a valuable group of organisms for recon- Cavalier-Smith, 1997, especially species encompassing the structing lake water depth, pH, and possibly other environ- genera Euglypha Dujardin, 1841; Scutiglypha Foissner and mental variables. Schiller, 2001; Assulina Ehrenberg, 1872; and Trinema Du- Most, if not all, ecological studies of thecamoebians in- jardin, 1841 (Wylezich et al. 2002). These organisms are clude members of both Orders Euglyphida and Arcellinida commonly referred to as euglyphids. Most non-euglyphid (Whittle et al. 2018) and relate individual species to envi- testate amoebae belong to the Order Arcellinida within the ronmental variables. Because the fossil remains in the Gi- Phylum Amoebozoa Lühe, 1913 emended Cavalier-Smith, raffe Pipe locality represent isolated siliceous plates and not 1998. Tests of most arcellinids are formed from organic com- entire tests, especially after extraction from the mudstone ponents, and these organisms have lobose pseudopodia in rocks, we have abundance estimates for all plate morphol- contrast to the filose type possessed by euglyphids. ogies but not for specific species. The primary goals of this Testate amoebae inhabit a broad range of terrestrial and study were to: 1) identify and quantify siliceous plate morpho- aquatic habitats, including organic-rich soils, moss beds, types in 125 waterbodies from 5 regions along the east lakes, ponds, rivers, bogs, wetlands, peatlands, and even brack- coast of North America, 2) document distribution patterns ish environments (Ogden and Hedley 1980, Escobar et al. between regions, and 3) examine relationships between con- 2008, Mitchell et al. 2008, Barnett et al. 2017, Amesbury centrations of siliceous testate plates and environmental et al. 2018), where they prey on bacteria, algae, other protists, variables. and small micrometazoans (McKeown et al. 2019). Water- table depth is an important factor regulating testate amoe- bae species abundances and vertical distributional patterns METHODS in wetlands and peatlands, whereas in terrestrial habitats ad- In an earlier work, Siver and Lott (2012) reported on the ditional variables such as soil thickness and pore space are biogeographic distributions of scaled chrysophytes in fresh- also important (Ogden and Hedley 1980, Mitchell et al. 1999, water bodies spread along the east coast of North America. Booth 2001, Charman 2001, Booth and Jackson 2003). Con- Because the remains of scaled chrysophytes are siliceous, centrations of organic matter in sediments, along with other the preparation methods used by Siver and Lott (2012) also factors such as pH, conductivity, and trophic status, are also provided remains of siliceous testate amoebae plates. Water- important variables controlling the presence/absence, and bodies from 5 regions (n 5 125) included in that work— often concentrations, of individual species (Patterson and coastal North Carolina, the Pinelands National Preserve Kumar 2002, Booth and Zygmunt 2005, Payne et al. 2006, in New Jersey, Connecticut, Nova Scotia, and Newfound- Escobar et al. 2008, McKeown et al. 2019). Species responses land—were used in the current study. We took sediment to specific environmental conditions, coupled with high re- cores from each waterbody with a Glew gravity corer (Glew productive rates, make thecamoebians sensitive bioindica- 1988) and sectioned the cores into 1-cm units using a me- tors of environmental change (Ogden 1981, Schönborn 1992, chanical extruder (Glew 1989). We used the 0 to 1-cm sec- Charman 2001). As a result, the remains of tests in sediments tion from the surface of each core to identify and quantify and peats are useful in reconstructing historical conditions siliceous plates of testate amoebae from each site. Surface (Warner 1988, Charman 2001), such as paleoclimate (Booth sediment samples are commonly used to study organism 2001, 2002, 2008, Charman 2001), peatland hydrology (Payne remains because they effectively integrate growth of the et al. 2006, Amesbury et al. 2018), land-use changes (Patter- organisms over the course of a year or more (Smol 1995). son et al. 2002), sea-level changes, and salt enrichment (Geh- We measured 20 environmental variables for each study rels 2000, Gehrels et al. 2001, Whittle et al. 2018). site: water color, maximum depth, Secchi disk depth, alka- The oldest known fossil specimens representing mem- linity, pH, specific conductivity, chlorophyll a, total phos- bers of the Order Euglyphida are from the early Eocene Gi- phorus, total nitrogen, chloride, sulfate, potassium, sodium, raffe Pipe fossil locality near the Arctic Circle in Canada calcium, magnesium, latitude, and mean maximum and (Barber et al. 2013). Euglyphid fossil remains have also been minimum temperatures in January and July. We derived uncovered from Middle Eocene (Loeblich and Tappan 1964), maximum depth from existing bathymetric maps, govern- Middle Miocene (Williams 1985, Foissner and Schiller 2001), ment databases, or estimates made at the time of collection. and Pliocene (Boeuf and Gilbert 1997) deposits. Given the Details for all other parameters, including all chemical anal- high abundances of well-preserved siliceous testate plates yses, were according to Canavan and Siver (1994), Ahrens Volume 39 December 2020 | 000 and Siver (2000), Lott and Siver (2005), and Siver and Lott Georgia). We coated samples with a mixture of gold and (2010) and are summarized in Siver and Lott (2012). palladium for 2 min with a Polaron Model E sputter coater To prepare samples for examination with light and scan- (Quorum Technologies, East Sussex, UK). ning electron microscopy, we processed surface sediment We used PRIMER-e (version 7; Quest Research Limited, from the 125 cores as follows: 1) We thoroughly mixed wet Auckland, New Zealand; Clarke and Warwick 2001, Ander- sediment from the 0 to 1-cm section of each core and added son et al. 2008) to conduct non-metric multidimensional a known amount to a beaker along with a mixture of sulfuric scaling (NMDS) to examine differences in testate plate mor- acid–potassium dichromate according to the procedure of photypes between sites and study regions. The organism Marsicano and Siver (1993). The amount of wet mass used/ matrix consisted of concentration data for each testate plate sample varied between samples, with a mean of 1.3 g. We morphotype at each site. We square-root transformed plate gently heated the mixture to facilitate oxidation of organic abundances to downweight the influence of abundant mor- matter. Once oxidized we transferred the material to a cen- photypes and calculated a Bray–Curtis similarity matrix, trifuge tube and washed it with deionized water a minimum which was then used in the NMDS. The Bray–Curtis metric of 5Â. We transferred the resulting slurry to a glass vial and is commonly used for comparing sites with species count brought the volume to 10 mL by adding deionized water. data. 2) We placed a 2nd wet-mass sample from the 0 to 1-cm sec- We used simple linear regression, coupled with scatter- tion of the core onto an aluminum weighing boat and dried it plots, and best subset regression analysis to assess relation- in an oven at 1057C to a constant mass. We calculated the % ships between the abundances of testate plates/site and en- dry mass and used that value to estimate the amount of dry vironmental variables. We used the Shapiro–Wilk test to mass of material used to derive the slurry. 3) We diluted a evaluate normality, the Durbin–Watson statistic for inde- known volume of the slurry to 30 mL and slowly poured it pendence of residuals, and Spearman rank correlation to into a Battarbee tray containing 5 wells, each well fitted with test for constant variance. The best subset regression anal- a 22-mm diameter circular glass cover slip (Battarbee 1986). ysis evaluates all possible combinations of the independent We covered the Battarbee trays and placed them on a variables and selects subsets of variables that best predict vibration-free table, and we allowed the solution to air dry the dependent variable. We used the Mallows’ Cp as the cri- such that the microfossils became affixed to the cover glasses terion for inclusion of an independent variable into the fi- within the tray. 4) We permanently mounted the cover glasses nal model. Cp estimates the bias added to the dependent- containing the dry sediment material onto glass slides using variable estimate when independent variables are removed Hyrax or Naphrax mounting medium and labeled the slides from the model. Cp accounts for the difference between the with a diamond knife. 5) We scanned the permanent slides true population mean for the dependent variable and the at 40Â magnification with an Olympus BX51 microscope population mean based on the independent variables in- (Waltham, Massachusetts) using a phase contrast lens (nu- cluded in the model. For a given model, the optimum value merical aperture 5 0.65) and recorded the numbers of each of Cp is equal to the number of variables in the model plus testate plate morphotype within a known number of fields. the constant (Cp 5 k 1 1, where k 5 the number of inde- The surface area of a field of view under 40Â magnification pendent variables in the model). A value close to k 1 1 in- was 0.29 mm2. 6) Given the above parameters, we calcu- dicates a low bias of the model regression coefficients used lated the amount of dry mass of sediment/field at 40Â mag- to infer the dependent variable. Following best subset re- nification and reported concentrations of testate plates on gression, we then used multiple regression analyses to de- a lg-dry-mass basis. The mean across all samples was 0.29 lg velop models for each of the top 10 subsets of variables. dry mass/field. We further used the predicted residual sum of squares Prior to enumeration of testate plates, we first examined (PRESS) statistic, which we computed using leave-1-out samples with a Leica DMR light microscope (Leica Micro- cross validation, to evaluate regression models for each of systems, Buffalo Grove, Illinois) using a 100Â plan apo- the top 10 subsets of variables identified in the best subset chromat lens (numerical aperture 5 1.4) coupled with a regression analysis. PRESS provides a measure of the po- Zeiss Axiocam 503 color camera (Carl Zeiss Microscopy, tential predictive power of a regression model. Once a model White Plains, New York) and with either a Zeiss Leo-982 has been developed, each observation of the dependent var- or a Nova NanoSEM™ 450 (FEI™, Hillsboro, Oregon) iable is removed, 1 at a time, and the model is refitted using field emission scanning electron microscope. These analy- the remaining observations. The model is then used to esti- ses yielded initial qualitative estimates and images of the mate the value of the observation that was removed, and the plate morphotypes found in each sample. For scanning process is repeated across all observations. The PRESS statis- electron microscopy, we air dried an aliquot of the oxidized tic is calculated as the sum of squares of the residuals. Lower slurry onto a piece of heavy-duty aluminum foil, trimmed PRESS values indicate models with better prediction ability it, and attached it to an aluminum scanning electron micro- and that are less likely to be overfitted. We used SigmaPlot scope stub with Apiezon® wax (M&I Materials, Suwanee, (version 12.5; Systat Software Inc., San Jose, California) for 000 | Abundances of freshwater testate amoebae P. A. Siver et al. all regression analyses and to calculate the PRESS statis- Secchi disk depths than waterbodies in the other 4 regions. tic and standard error of estimation for each model as well In contrast, sites on the Atlantic Coastal Plain are more as Pearson correlation coefficients between environmental acidic and poorly buffered, have high concentrations of col- variables. ored dissolved organic matter, and have low Secchi disk depths. A mixture of clear water and humic-stained water- RESULTS bodies characterize those in the Canadian Maritime re- Physical and chemical properties of the study lakes gions. Although the pH ranged from 3.5 to 8.6, a total of < The sites from North Carolina and the Pinelands of New 64 Canadian Maritime waterbodies had a pH 6. Lakes in Jersey are situated in non-glaciated areas on the Atlantic Nova Scotia and Newfoundland have lower concentrations Coastal Plain, whereas those from the 3 remaining regions of sulfate and potassium than those in the other regions. were previously glaciated. Waterbodies on the Atlantic Based on chlorophyll a, total phosphorus, and total nitro- Coastal Plain are more similar to each other, as are those gen concentrations, the most eutrophic lakes are situated in Nova Scotia and Newfoundland, than they are to those in North Carolina and New Jersey, whereas the more north- in Connecticut (Table 1). Although different from the other ern and glaciated sites become more oligotrophic with lower 4 regions, waterbodies in Connecticut are more similar to nutrient and chlorophyll a concentrations. those in the Canadian Maritime than those along the Atlan- tic Coastal Plain. The mean maximum and mean minimum Diversity and abundances of siliceous shell plates January temperatures ranged from 12.8 to –1.77C and 1.1 to Siliceous shell plates were separated into 7 different –9.47C, respectively, across all regions (Table 1). There is morphotypes for quantitative purposes (Table 2, Figs 1A– ~10 to 127C difference in the mean maximum and mean Y, 2A–H, 3A–I). Circular shaped plates (Figs 1H–K, 3E, minimum temperatures during July in all regions. F) were the most abundant type, ranging from 32.6 (North On average, the Connecticut waterbodies have higher Carolina) to 64% (Connecticut) of the total/region and ac- pH and alkalinity and are clearer waterbodies with deeper counting for 45% of plates from all sites. Except for North

Table 1. Mean values of chemical, physical, and climate-related variables for waterbodies from 5 regions included in this study. Latitude values are also given. NC 5 North Carolina, NJ 5 New Jersey Pine Barrens, CT 5 Connecticut, NS 5 Nova Scotia, NFL 5 Newfoundland, SD 5 Secchi disk depth, ALK 5 alkalinity, SC 5 specific conductivity, – –2 Chl a 5 chlorophyll a, Color 5 platinum–cobalt water color, TP 5 total phosphorus, Cl 5 chloride, SO4 5 sul- 1 1 1 1 fate, K 5 potassium, Na 5 sodium, Ca 2 5 calcium, Mg 2 5 magnesium, TN 5 total nitrogen, Tmin 5 mean minimum temperature, Tmax 5 mean maximum temperature. Parameter NC NJ CT NS NFL

SD (m) 0.7 1.19 3.32 2.68 1.46 ALK (leq/L) –6 58 431 –13 89 pH 4.31 5.16 7.06 4.99 5.71 SC (lS) 100 75 98 35 53 Chl a (lg/L) 23.5 9 6.5 1.7 1.5 Color 537 118 30 92 74 TP (lg/L) 47.3 17.8 33.1 5.1 12 TN (mg/L) 1.25 1.02 0.44 0.18 0.37 – Cl (meq/L) 0.31 0.34 0.27 0.19 0.33 –2 SO4 (meq/L) 0.2 0.19 0.16 0.04 0.04 1 K (meq/L) 0.02 0.03 0.03 0 0 1 Na (meq/L) 0.38 0.28 0.29 0.2 0.29 1 Ca 2 (meq/L) 0.16 0.1 0.36 0.02 0.1 1 Mg 2 (meq/L) 0.11 0.13 0.2 0.04 0.05 JanTmax (7C) 12.8 5 1.7 0.4 –1.7 JanTmin (7C) 1.1 –5.6 –7.8 –7.3 –9.4 JulyTmax (7C) 31.7 29.4 28.6 21.6 20.8 JulyTmin (7C) 21.7 18.3 16.9 12 10.6 Latitude 34.9 39.4 41.4 44.5 48.1 Depth (m) 1.7 1.4 10.7 4.9 2.2 Table 2. Relative abundances of 7 euglyphid plate morphotypes in surface sediments of waterbodies from 5 regions along the east coast of North America. The percentage of waterbodies containing each plate type/region is also given, and data is summarized across all regions. Types include plates that are circular (cir), oval (ova), square (squ), rhomboid or rectangular (rho), scutiform (scu1, scu2) and toothed (too). The regions include Connecticut (CT), North Carolina (NC), New Jersey (NJ), Nova Scotia (NS), and Newfound- land (NFL). % plate types % of waterbodies

Region cir ova squ rho scu1 scu2 too cir ova squ rho scu 1 scu 2 too

CT 64.1 11.3 0.1 2.8 11.8 8.8 1.0 92 53 2 25 49 55 11 NC 32.6 39.7 12.3 3.5 1.4 8.1 2.4 96 100 74 74 48 91 61 NJ 45.9 37.0 8.4 1.4 0.2 6.2 1.0 100 100 100 89 22 100 33 NS 61.3 11.7 10.0 2.6 5.6 6.7 2.0 100 82 71 47 71 94 47 NFL 43.0 23.8 5.2 3.1 10.2 11.4 3.3 95 95 62 52 95 96 52 All regions 45.3 30.0 8.2 2.7 4.2 8.0 1.6 95 76 42 46 58 76 38

Figure 1. Light micrographs of siliceous plate morphotypes from freshwater testate amoebae. Morphotypes include quadrangular (A–C), oval (D–G), circular (H–K), denticulate (L, M, T), scutiform morphotype 1 (N–P), rectangular to rhomboid (Q–S), and scutiform morphotype 2 (U–Y). Scale bar 5 10 lm. 000 | Abundances of freshwater testate amoebae P. A. Siver et al.

Scutiform plates with a bilateral symmetry were separated into 2 groups based on the shape along the wider portion of the plate. The wider end of the plate, referred to as the aboral end, faces the posterior of the test, whereas the narrower end faces the oral end of the test. The 1st scutiform-shaped plate, morphotype 1, has strongly undulating margins along both the aboral and oral ends of the plate, resulting in 3 projections on each end (Figs 1N–P, 2A, B). The 2 outer projections on the aboral end of the plate are highly accentuated and extend slightly further than the central process or projection, yield- ing a butterfly shape. The margins of the plate connecting the aboral and oral ends are relatively straight to slightly curved and tapering, completing the bilateral design. Except for the central projection, the lateral margins along the ab- oral end of morphotype 2 are broadly rounded, and they are not as accentuated and projected out from the plate as they are on morphotype 1 (Figs 1U–Y,2D,E,G,H).Narrowscuti- form morphotype 2 plates have a shape that resembles a lemon rather than a butterfly (Figs 1U, V, Y, 3G). Scutiform

Figure 2. Scanning electron micrographs of siliceous plate morphotypes from freshwater testate amoebae. Morphotypes include scutiform morphotype 1 (A, B), denticulate (C, F), and scutiform morphotype 2 (D, E, G). Denticulate plates illustrated have 11 (C) and 9 (F) teeth. The lemon-shaped plate with 7 teeth (H) is situated on the test just behind the row of denticu- late plates that line the pseudostome. Scale bars 5 3 lm (F, G), 4 lm (E) and 5 lm(A–D, H).

Carolina localities, circular plates were the most abundant type in all other regions. Circular plates were widely distrib- uted, recorded in 95% of all study lakes. Oval plates (Figs 1D– G, 3G) were the 2nd most abundant plate type, with maximum abundances recorded from waterbodies along the Atlantic Coastal Plain in New Jersey and North Carolina, where they accounted for 39.7 and 37% of all plates, respectively (Ta- ble 2). Collectively, circular and oval morphotypes com- prised 75% of all plates. Square- or quadrangular-shaped plates (Figs 1A–C, 3H, I) accounted for 8% of the total and were relatively evenly distributed between regions, but they were noticeably rare in Connecticut lakes. In contrast, qua- Figure 3. Scanning electron micrographs of siliceous plate drangular plates were present in all of the New Jersey sites. morphotypes from freshwater testate amoebae. Morphotypes Rhomboidal and rectangular-shaped plates (Figs 1Q–S, include rectangular to rhomboid (A–D), circular (E, F), oval 3A–D) were found in all regions and in 46% of the study (G), and quadrangular (H, I). Scale bars 5 2 lm (A, C, F, H) sites; however, they accounted for only 2.7% of all plates. and 3 lm (B, D, E, G, I). Volume 39 December 2020 | 000 morphotypes 1 and 2 accounted for only 4 and 8% of all among regions are supported by the NMDS visualization. plates enumerated, respectively. However, morphotype 1 First, quadrangular plates have the most limited distribution, was significantly more abundant and accounted for between are clearly more abundant in New Jersey and North Carolina, 5.6 and 11.8% of the total in the 3 northern regions, Con- and are largely lacking in Connecticut sites (Fig. 5D). Second, necticut, Nova Scotia, and Newfoundland. They were much the distributions and abundances of scutiform morphotype 2 less abundant in lakes situated on the Atlantic Coastal Plain and rectangular plate types largely overlap (Fig. 5C, F). Third, where they accounted for <1% of all plates. Despite low butterfly-shaped scutiform morphotype 1 plates have greater abundances, morphotype 1 was present in 48 and 22% of abundances in northern regions and are noticeably less abun- the waterbodies in North Carolina and New Jersey, respec- dant in North Carolina and New Jersey (Fig. 5E). Fourth, cir- tively. Morphotype 2 was much more evenly distributed cular plates exhibit the widest distribution and can be found among regions (Table 2). in high abundances in lakes from multiple regions, but less Plates with projecting teeth that surround the pseudo- so in Connecticut (Fig. 5A). Fifth, oval plates are also widely stome opening varied in the number and position of the distributed with highest abundances in more southern re- teeth (Figs 1L, M, T, 2C, F). Denticulate plates with 5 to gions and lower concentrations in many Connecticut local- 13 teeth were observed. Although combined in abundance ities (Fig. 5B). estimates, denticulate plates can be separated into 2 types. One type typically had 5 to 7 teeth, each of which originated from the undersurface of the plate, and possessed a more Siliceous plate abundance and environmental variables accentuated aboral end (Fig. 1L, M). The 2nd type of dentic- Water depth had a substantial influence on testate plate 2 ulate plate usually had more (e.g., 9–11) and smaller teeth concentration (linear regression, p < 0.001; r 5 0.43), that projected from the margin of the plate in the same where deeper lakes had substantially lower abundances of plane as the body of the plate. This plate type has a shorter plates (Fig. 6A). The highest abundances of plates were projecting aboral margin (Figs 1T, 2C, F). found in waterbodies <3 m in depth, with the greatest num- Siliceous plate types differed among regions, as illustrated bers in sediments from shallow ponds <1 m deep. The re- by NMDS. The ordination of sites based on plate morpho- lationship between pH and testate plate concentration was 2 type composition (Fig. 4) clearly shows the high degree of sim- also important (linear regression, p < 0.001; r 5 0.45), where ilarity between New Jersey and North Carolina, the 2 south- abundance of plates increased with a decrease in pH (Fig. 6B). ernmost localities included in the study. Most Connecticut Most waterbodies with pH >7 had <1 plate/lg dry-mass sites also separate from the more southern regions and are sediment. In contrast, waterbodies with a pH <5oftenhad more similar to other northern regions (Nova Scotia and 3toover10plates/lg dry-mass sediment. Newfoundland). Differences in the distributions of plate types The best subset regression analyses confirmed the im- portance of pH and water depth in determining testate plate abundance and highlighted other potentially impor- tant environmental variables. All of the top 10 models iden- tified with best subset regression included the variables pH, log water depth, and potassium, and the model with only 2 these 3 variables yielded an R of 0.56 and a Cp value of 3.95, almost equal to the number of variables in the model 1 1. Each of the 3 variables was highly important ( p < 0.001) and the model yielded a low standard error of estimation (0.31). In addition, of the 10 best subset regression models, this model yielded the lowest PRESS statistic (18.9) and best relationship for predicting plate concentration (Fig. 6C). Another model identified in the best subset analysis included a4th variable, sulfate. However, this model yielded a slightly larger standard error of estimation (0.34) and a larger PRESS statistic (19.1). Several of the remaining models added ad- ditional variables, including specific conductivity and total nitrogen, but these models did not improve R2 values and Figure 4. Result of a 2-dimensional non-metric multidimen- yielded higher C values (indicating increased bias), higher sional scaling (NMDS) analysis indicating the ordination of p 125freshwatersitesfrom5regionsalongtheeastcoastofNorth standard error of estimation values, and higher PRESS val- America based on the concentrations of different testate amoebae ues. Although water color was strongly correlated with other plate morphotypes. Relationships with important environmental variables included in the models (Table 3), it was not included vectors (pH, water depth, water color, potassium, and sulfate) are in any of the top models identifiedinthebestsubsetregres- shown. sion analysis. Based on the best subset regression analysis, the 000 | Abundances of freshwater testate amoebae P. A. Siver et al.

Figure 5. Non-metric multidimensional scaling (NMDS) results depicting the occurrences and concentrations of 6 testate amoebae plate morphotypes in freshwater sites representing 5 regions along the east coast of North America. Morphotypes include circular (A), oval (B), rectangular to rhomboid (C), quadrangular (D), scutiform morphotype 1 (E), and scutiform morphotype 2 (F). See Fig. 4 for locations of sites in each of the 5 regions. Keys to concentrations are given in the lower right of each panel and reflect the num- ber of plates/lg dry mass of sediment. Stress for all ordinations is 0.16.

simpler model including pH, log depth, and potassium was Testate amoebae assemblage characterization deemed best supported. from plate morphotypes We are confident that except for the quadrangular- shaped plates, the remaining specimens in our work repre- DISCUSSION sent species belonging to the Order Euglyphida. Because We documented the distributions and concentrations of the plates are disarticulated from the original test when un- siliceous plate morphotypes in surface sediments of fresh- covered from the sediment samples, assigning them to spe- water lakes along the east coast of North America. We fur- cies and even to genus level is difficult. It is especially dif- ther investigated the relationships between environmental ficult to determine from which taxa the circular or oval variables and concentrations of siliceous plates, with a par- plates were derived. Plates with these shapes are common ticular focus on pH and water depth. Because similar plate morphotypes formed by some species belonging to Eu- morphotypes have been uncovered in lake mudstones dat- glypha Dujardin, 1841; Assulina Ehrenberg, 1872; Trinema ing to the warm Eocene, we discuss the future possibility of Dujardin, 1841; Sphenoderia Schlumberger, 1845; Corythion using these remains to reconstruct ancient environmental Taranek, 1918; Tracheleuglypha Deflandre, 1928; and pos- conditions. sibly Puytoracia Bonnet, 1970. We believe that most of the Volume 39 December 2020 | 000

margins likely belong to Euglypha compressa Carter, 1864, especially because denticulate plates matching this species were also found in the samples. The quadrangular-shaped plates belong to species in the genus Quadrullela Cockerell, 1909, which also form tests with a highly organized arrange- ment of siliceous plates. However, unlike taxa in the Order Euglyphida that produce filose pseudopodia, Quadrullela belongs to the Order Arcellinida Kent, 1880 that includes species with lobose pseudopodia (Ogden and Hedley 1980). The bilateral and scutiform-shaped plate morphotypes belong to the genus Scutiglypha Foisner and Schiller, 2001. Scutiglypha was erected to include species of Euglypha that bear bilateral, shield-shaped plates (Foissner and Schiller 2001), and currently the genus includes at least 6 species (De Smet and Gibson 2009). The tests of Scutiglypha species also include plates that are intermediate in shape between those bearing teeth that surround the pseudostome, the typ- ical shield-shaped body plates, and smaller and more circu- lar plates that cover the posterior end of the test (Foissner and Schiller 2001, De Smet and Gibson 2009, Schiller and Wuttke 2015). Based on our findings, and especially because both morphotypes were not always uncovered in the same collections, the 2 shield-shaped morphotypes probably rep- resent 2 different Scutiglypha species. Although some authors question the validity of the ge- nus Scutiglypha, the arguments for separating it from Eu- glypha were reviewed by De Smet and Gibson (2009). In ad- dition to the obvious differences in plate morphology, our findings indicate that there is also a major difference in the structure of the denticulate plates that surround the pseudostome. The denticulate plates of Euglypha usually have 5 to 7 teeth and a large and prominent median tooth, and the teeth clearly originate from the undersurface of the plate. In our study, this type of denticulate plate was always found in samples with circular or oval plates, which indi- cates that many of the latter plate morphotypes may indeed belong to Euglypha because other genera in the Euglyphi- dae besides Scutiglypha lack these distinctive denticulate plates. On the other hand, denticulate plates found in sam- ples with Scutiglypha body scales usually have 7 to 11 smaller teeth and a less prominent median tooth, and the teeth

Table 3. Pearson correlation coefficients for the envi- Figure 6. Abundance of testate amoebae plates vs lake depth ronmental variables pH, log water depth (depth), log (A) and pH (B) in 125 freshwater sites along the east coast of North color (color), and concentrations of potassium (K) America. Predicted vs actual abundances of testate amoebae plates and sulfate (SO4). *p < 0.001 using a multiple-regression model based on 3 variables: pH, water depth, and concentration of potassium (C). Coefficientofdetermi- Variable pH Depth Color K SO4 5 2 5 5 nation r , p-value p, standard error of estimation SEE. pH 1 Depth 0.52* 1 rectangular-shaped or rhomboidal scales belong to mem- Color –0.34* –0.62* 1 bers of the genus Euglypha (e.g., E. strigosa Ehrenberg, 1848), K 0.09 –0.25* 0.16 1 and possibly the genus Assulina (e.g., A. scandinavica Pen- SO4 0.1 0.10 –0.07 0.18 1 ard, 1890). The more rectangular-shaped plates with rounded 000 | Abundances of freshwater testate amoebae P. A. Siver et al. originate at and project from the margin of the plate. As ad- Another important variable controlling both abundance ditional species are described, it will be interesting to see if and species composition of testate amoebae is pH (Escobar the differences in the structures of denticulate plates form et al. 2008, Patterson et al. 2013). On a broad scale, many an additional character that can be valuable for distinguish- testate amoebae species are limited by pH, with some taxa ing between the 2 genera. being found to be more abundant in acidic habitats and Although rare, 2 additional plate types are worth men- others in more alkaline sites (Ogden and Hedley 1980). A tioning and likely indicate the presence of additional taxa similar result was reported by Patterson and Kumar (2002), in the collections. First, a few of the numerous oval-plate who found some species of euglyphids thriving at low pH specimens had distinctively thickened rims. Plates that match and other species, mostly of Centropyxids, more abundant this morphotype have been illustrated for Assulina mus- at higher pH sites. In a study of subtropical lakes in Florida, corum Greef, 1888 (Ogden and Hedley 1980). Another rare USA, Escobar et al. (2008) reported the highest diversity of type of plate uncovered in a few samples was a small oval testates in lakes with a high pH near 8. Although our find- plate bearing a single small tooth. This plate morphotype ings of increasing silica plate abundance with decreasing pH is typical of plates that align the pseudostome of species in largely agree with those of Patterson and Kumar (2002), they the genus Trinema. contrast with those of Escobar et al. (2008). Many of our study lakes with a pH <~5.5 had abundances of silica plates 5to10Â higher than in lakes with a pH >7, but a difference Siliceous plate abundance and environmental variables in species composition between the Escobar et al. (2008) Numerous studies have reported that the abundance, study and ours (silica plate-forming euglyphids vs species diversity, and distribution of testate amoebae in freshwa- that do not construct tests out of idiosomes) is likely the ters are related to a combination of environmental vari- reason for the difference between the 2 studies. Another ables including but not limited to water depth, pH, trophic difference is that our study was based on the abundances status, conductivity, organic content, temperature, mois- of plates and not species diversity. ture content, and substrate type (e.g., Ogden and Hedley Lake trophic status can be another important variable de- 1980, Collins et al. 1990, Ju et al. 2014, Roe and Patterson termining the assemblage of testate amoebae species found 2014, Amesbury et al. 2018, Tsyganov et al. 2019). Of these in a given waterbody (Schönborn 1992, Patterson et al. 2012, factors, water depth is often reported as the most impor- Tsyganov et al. 2019). As a result of such relationships, Rein- tant variable controlling diversity, abundance, and species hardt et al. (2005) and Drljepan et al. (2014) used these or- composition in freshwater lakes and peatlands (Mitchell ganisms to track shifts in trophic status. Highest species di- et al. 1999, Booth 2002, Patterson et al. 2012, McKeown versities of testate amoebae have been found in mesotrophic et al. 2019, Tsyganov et al. 2019). Our findings confirm to eutrophic lakes (Escobar et al. 2008, Ju et al. 2014), and that water depth is strongly associated with testate plate greater abundances of specimens are usually associated with abundance. Given the importance of water depth, Sonnen- organic-rich sediments (Patterson and Kumar 2002, Roe and burg et al. (2013) attempted to use species assemblages to Patterson 2006). Although we did not observe a relationship infer this variable over time but concluded that additional between plate abundance and the trophic-related variables data relating specific species to specific water depths were total phosphorus and chlorophyll a concentration, there were needed. In a more recent study, Tsyganov et al. (2019) de- greater concentrations of testate plates in sites from North scribed distinctive assemblages of testate amoebae species Carolina and New Jersey with higher total nitrogen levels relative to lake depth, further demonstrating that these or- and elevated water color relative to most localities. These ganisms have great promise for inferring historical lake are also sites that most likely have higher concentrations of water levels. organic matter. Using testate amoebae to reconstruct paleohydrological Our results indicate that higher abundances of siliceous conditions in peatlands is more advanced than using these testate plates, regardless of species diversity, are found in organisms to infer lake depth (Charman 1997, Mitchell et al. shallow acidic ponds and lakes. This finding supported our 1999, Booth 2002). By combining regional-scale datasets, original hypothesis that the number of testate amoebae Amesbury et al. (2018) developed transfer functions appli- would increase as the surface area for attachment/unit vol- cable for inferring peatland palaeohydrology for the Holarc- ume increased. Because the surface-area-to-volume ratio tic. In a work based on New Zealand peatlands, McKeown increases with decreasing water depth, it makes sense that et al. (2019) showed that testate amoebae-based inference the highest abundances of siliceous plates would be found models could be improved by dividing species into subsets in sediments from shallow sites. Water depth across our based on size. In their study, smaller species were related study region was, in turn, correlated with pH and water color. to different environmental variables than larger taxa, dem- Because the vast majority of sites were in woodland set- onstrating an even greater utility to use these organisms in tings, the shallower sites were typically more darkly stained historical reconstruction efforts. with dissolved humic matter and had lower pH. This raises Volume 39 December 2020 | 000 questions such as: 1) are the concentrations of testate plates ern lake study. In addition, the concentrations of plates related more to water depth, pH, color, or a combination of range widely over the length of the core, including periods these correlated variables? and 2) do the numbers of testates of extensive numbers alternating with periods with few to with siliceous plates increase in shallow sites simply because no testate remains (Barber et al. 2013). Preliminary results there is more surface area, because of lower pH, or both? indicated that concentrations of testate plates in the Giraffe Water color did not add to any of the models if either pH core were positively correlated with remains of acidic and or water depth were included; however, all models included periphytic diatoms, sponge sclereids, heliozoans, and spe- both pH and water depth. These findings support the idea cific types of chrysophyte cysts, whereas low concentrations that both water depth and pH play an important role in the were found concurrent in strata where planktonic diatoms growth and occurrence of these organisms. dominate (Barber et al. 2013, Siver 2019). Based on these re- A major difference between our approach and previous sults, our current hypothesis is that abundant concentra- works using testate amoebae to infer environmental condi- tions of testate plates correspond to time periods represented tions is that our results are based on: 1) only species that by a shallow waterbody and vice versa. Inference models for produce siliceous plates, and 2) abundances of siliceous water depth and pH based on concentrations of testate plates plates rather than numbers of individual species. None of would yield much needed independent verification of recon- the previous studies relating testate amoebae to specific en- structions based on other fossil proxies. vironmental variables, or studies focused on inferring his- In summary, remains of siliceous plates from testate torical conditions, relied on abundances of siliceous plates. amoebae are a common type of microfossil found in many Nor are there studies that include only the subset of species lakes and ponds, including those distributed along a wide that produce siliceous plates, although some studies do rely expanse of eastern North America. Most of the plate mor- solely on Arcellacea taxa (Patterson et al. 2012, Roe and photypes represent taxa in the Order Euglyphida. Greater Patterson 2014). Our interest in focusing on abundance of concentrations of plates are substantially associated with siliceous plates is because abundance is the metric we can shallower and more acidic waterbodies, indicating that this best estimate in modern lake sediments as well as in fossil metric could be used to infer historical conditions. The plate mudstones, especially if acidic oxidation procedures are morphotypes found in modern waterbodies have all been needed to prepare and extract the microfossils. Despite not uncovered in fossil localities. This finding indicates that the using a metric based specifically on species, our model still lineages of organisms producing these plate morphotypes had accounted for a majority of the variation in plate concentra- already evolved by at least the Eocene period (Barber et al. tions, which is comparable to previous works based on spe- 2013), further supporting their use in reconstructing condi- cies abundances. Undoubtedly, if other variables, such as tions found in these ancient waterbodies. biotope, food supply, and predator concentration, were in- cluded, a greater percentage of variation could probably be explained. With additional study, it may also be possible to ACKNOWLEDGEMENTS improve the models by linking specific plate morphotypes Author contributions: PAS developed the concept for the proj- to specific conditions in a similar fashion used by McKeown ect, collected and analyzed data, imaged specimens, and wrote et al. (2019) to relate testate size classes to specific conditions. and edited the manuscript. AML collected and analyzed data, imaged specimens, and edited the manuscript. PT collected and An ultimate long-term goal is not only to use the abun- analyzed data. dances of siliceous testate plates to reconstruct water depth This work was funded by the United States National Science and pH in modern waterbodies using recently deposited Foundation under grant EAR-1725265 and EAR-1940070 to PAS. sediment, but also in fossil waterbodies, including the ex- The authors thank the many students who helped with field col- tensive fossil lake from the Giraffe Pipe fossil locality (Siver lections over the years and Xuanhao Sun for assistance with the and Wolfe 2009, Barber et al. 2013, Wolfe et al. 2017). This scanning electron microscope facilities. fossil site is of particular interest because it represents an important deep-time freshwater analog of an Arctic lake that existed under a warm greenhouse climate (Siver and LITERATURE CITED Wolfe 2009, Wolfe et al. 2017). Reconstructed mean annual Ahrens, T. D., and P. A. Siver. 2000. The physical properties, tro- temperature and mean annual precipitation values for the phic conditions, and water chemistry of 60 lakes on Cape Cod, Giraffe locality are 177C higher and 4Â greater, respectively, Massachusetts, USA. Lake and Reservoir Management 16: 268–280. than present, and the region supported a warm mixed forest Amesbury, M. J., R. K. Booth, T. P. Roland, J. Bunbury, M. J. Clif- (Wolfe et al. 2017). Tracing the history of the Giraffe water- ford, D. J. Charman, S. Elliot, S. Finkelstein, M. Garneau, body can potentially help us understand how freshwater P. D. M. Hughes, A. Lamarre, J. Loisel, H. Mackay, G. Magnan, Arctic habitats will respond to future warming scenarios. E. R. Markel, E. A. D. Mitchell, R. J. Payne, N. Pelletier, H. Roe, The Giraffe core contains numerous siliceous testate plates, M. E. Sullivan, G. T. Swindles, J. Talbot, S. van Bellen, and including most of the morphotypes uncovered in our mod- B. G. Warner. 2018. Towards a Holarctic synthesis of peatland 000 | Abundances of freshwater testate amoebae P. A. Siver et al.

testate amoeba ecology: Development of a new continental- cancabur Caldera Lake, Central Andes. Acta Protozoology scale palaeohydrological transfer function for North America 48:119–126. and comparison to European data. Quaternary Science Re- Drljepan, M., F. M. McCarthy, and J. B Hubeny. 2014. Natural views 201:483–500. and cultural eutrophication of Sluice Pond, Massachusetts, Anderson, M. J., R. N. Gorley, and K. R. Clarke. 2008. U.S.A., recorded by algal and protozoan microfossils. Holo- PERMANOVA1 for PRIMER: Guide to Software and Statis- cene 24:1731–1742. tical Methods. PRIMER-E, Plymouth, UK. Escobar, J., M. Brenner, T. J. Whitmore, W. F. Kenney, and J. H. Barber, A., P. A. Siver, and W. Karis. 2013. Euglyphid testate Curtis. 2008. Ecology of testate amoebae (thecamoebians) in amoebae (Rhizaria: Euglyphida) from an arctic Eocene water- subtropical Florida lakes. Journal of Paleolimnology 40:715– body: Evidence of evolutionary stasis in plate morphology for 731. over 40 million years. Protist 164:541–555. Foissner, W., and W. Schiller. 2001. Stable for 15 million years: Barnett, R. L., T. L. Newton, D. J. Charman, and W. R. Gehrels. Scanning electron microscope investigation of Miocene eugly- 2017. Salt-marsh testate amoebae as precise and widespread in- phid thecamoebians from Germany, with description of the new dicators of sea-level change. Earth Science Reviews 164:193–207. genus Scutiglypha. European Journal of Protistology 37:167–180. Battarbee, R. W. 1986. Diatom analysis. Pages 527–570 in B. E. Gehrels, W. R. 2000. Using foraminiferal transfer functions to Berglund (editor). Handbook of Holocene Paleoecology and produce high-resolution sea-level records from saltmarsh de- Paleohydrology. John Wiley & Sons, Hoboken, New Jersey. posits, Maine, USA. The Holocene 10:367–376. Boeuf, O., and D. Gilbert. 1997. Présence de thécamoebiens du Gehrels, W. R., H. M. Roe, and D. J. Charman. 2001. Foraminif- genre Trinema, au pliocène supérieur, découverte à Chilhac era, testate amoebae and diatoms as sea-level indicators in UK (Haute-Loire, France). Academy of Sciences Reports: Series IIA: saltmarshes: A quantitative multiproxy approach: Journal of Earth and Planetary Sciences 325:623–627. Quaternary Science 16:201–220. Booth, R. K. 2001. Ecology of testate amoebae in two Lake Supe- Glew, J. R. 1988. A new trigger mechanism for sediment samplers. rior coastal wetlands: Implications for paleoecology and envi- Journal of Paleolimnology 2:241–243. ronmental monitoring. Wetlands 21:564–576. Glew, J. R. 1989. A portable extruding device for close interval sec- Booth, R. K. 2002. Testate amoebae as paleoindicators of surface tioning of unconsolidated core samples. Journal of Paleolim- moisture changes on Michigan peatlands: Modern ecology and nology 1:225–239. hydrological calibration. Journal of Paleolimnology 28:329–348. Ju, L., J. Yang, L. Liu, and D. M. Wilkinson. 2014. Diversity and Booth, R. K. 2008. Testate amoebae as proxies for mean annual distribution of freshwater testate amoebae (Protozoa) along water-table depth in Sphagnum-dominated peatlands of North latitudinal and trophic gradients in China. Microbial Ecology America. Journal Quaternary Science 23:43–57. 68:657–670. Booth, R. K., and S. T. Jackson. 2003. A high-resolution record of Loeblich, A., and H. Tappan. 1964. Sarcodina, chiefly “Theca- late Holocene moisture variability from a Michigan raised bog. moebians” and Foraminiferida. Pages 16–54 in R. C. Moore The Holocene 13:865–878. (editor). Treatise on invertebrate paleobiology. Volume 1. Booth, R. K., and J. R. Zygmunt. 2005. Biogeography and com- Geological Society of America and University of Kansas Press, parative ecology of testate amoebae inhabiting Sphagnum- Lawrence, Kansas. dominated peatlands in the Great Lakes and Rocky Mountain Lott, A.-M., and P. A. Siver. 2005. An inventory of scaled chryso- regions of North America. Diversity and Distributions 11:577– phytes from North Carolina, USA, and their relationships to 590. environmental variables. Nova Hedwigia, Beiheft 128:211–229. Canavan, R. W., and P. A. Siver. 1994. Chemical and physical Marsicano, L. J., and P. A., Siver. 1993. A paleolimnological assess- properties of Connecticut lakes. Lake and Reservoir Manage- ment of lake acidification in five Connecticut lakes. Journal of ment 10:173–186. Paleolimnology 9:209–221. Charman, D. J. 1997. Modelling hydrological relationships of tes- McKeown, M. M., J. M. Wilmshurst, C. Duckert, J. R. Wood, and tate amoebae (Protozoa: Rhizopoda) on New Zealand peatlands. E. A. D. Mitchell. 2019. Assessing the ecological value of small Journal of the Royal Society of New Zealand 27:465–483. testate amoebae (<45 lm) in New Zealand peatlands. European Charman, D. J. 2001. Biostratigraphic and palaeoenvironmental Journal of Protistology 68:1–16. applications of testate amoebae. Quaternary Science Reviews Mitchell, E. A. D., A. Buttler, B. G. Warner, and J. M. Gobat. 1999. 20:1753–1764. Ecology of testate amoebae (Protozoa: Rhizopoda) in Sphag- Clarke, K. R., and R. M. Warwick. 2001. Change in marine com- num peatlands in the Jura mountains, Switzerland and France. munities: An approach to statistical analysis and interpreta- Ecoscience 6:565–576. tion. 2nd edition. PRIMER-E, Plymouth, UK. Mitchell, E. A. D., D. J. Charman, and B. G. Warner. 2008. Testate Collins, E. S, F. M. G. McCarthy, F. S. Medioli, D. B. Scott, and amoebae analysis in ecological and paleoecological studies of C. A. Honig. 1990. Biogeographic distribution of modern theca- wetlands: Past, present and future. Biodiversity and Conserva- moebians in a transect along the Eastern North American coast. tion 17:2115–2137. Pages 783–792 in C.Hemleben,M.A.Kaminski,W.Kuhnt,and Ogden, C. G. 1981. Observations of clonal cultures of Euglyphi- D. B. Scott (editors). Paleoecology, biostratigraphy, paleoceanog- dae (Rhizopoda, Protozoa). Bulletin of the British Museum raphy and taxonomy of agglutinated foraminifera. Springer, of Natural History (Zoology) 41:137–151. Dordrecht, The Netherlands. Ogden, C. G., and R. H. Hedley. 1980. An atlas of freshwater tes- De Smet, W. H., and J. A. E. Gibson. 2009. On a new species of tate amoebae. British Museum of Natural History and Oxford euglyphid testate amoeba, Scutiglypha cabrolae, from the Li- University Press, London and Oxford, UK. Volume 39 December 2020 | 000

Patterson, R. T., A. Dalb, A. Kumar, L. A. Henderson, and R. E. A. Siver, P. A., and A. M. Lott. 2012. Biogeographic patterns in scaled Boudreau. 2002. Arcellaceans (thecamoebians) as indicators chrysophytes from the east coast of North America. Freshwa- of land-use change: Settlement history of the Swan Lake area, ter Biology 57:451–467. Ontario as a case study. Journal of Paleolimnology 28:297–316. Siver, P. A., and A. P. Wolfe. 2009. Tropical ochrophyte algae from Patterson, R. T., and A. Kumar. 2002. A review of current testate the Eocene of northern Canada: A biogeographic response to rhizopod (thecamoebian) research in Canada. Palaeogeography, past global warming. PALAIOS 24:192–198. Palaeoclimatology, Palaeoecology 180:225–251. Smol, J. P. 1995. Application of chrysophytes to problems in paleo- Patterson, R. T., E. D. R. Lamoureux, L. A. Neville, and A. L. Ma- ecology. Pages 303–329 in C. Sandgren, J. P. Smol, and J. Kris- cumber. 2013. Arcellacea (testate lobose amoebae) as pH indi- tiansen (editors). Chrysophyte algae: Ecology, phylogeny and cators in a pyrite mine-acidified lake, Northeastern Ontario, development. Cambridge University Press, Cambridge, UK. Canada. Microbial Ecology 65:541–554. Sonnenburg, E. P., J. I. Boyce, and E. G. Reinhardt. 2013. Multi- Patterson, R. T., H. M. Roe, and G. T. Swindles. 2012. Develop- proxy lake sediment record of prehistoric (Paleoindian– ment of an Arcellacea (testate lobose amoebae) based transfer Archaic) archaeological paleoenvironments at Rice Lake, On- function for sedimentary phosphorus in lakes. Palaeogeog- tario, Canada. Quaternary Science Reviews 73:77–92. raphy, Palaeoclimatology, Palaeoecology 348–349:32–44. Tsyganov, A. N., E. A. Malysheva, A. A. Zharov, T. V. Sapelko, and Payne, R. J., K. Kishaba, J. J. Blackford, and E. A. D. Mitchell. 2006. Y. A. Mazei. 2019. Distribution of benthic testate amoeba as- Ecology of testate amoebae (Protista) in south-central Alaska semblages along a water depth gradient in freshwater lakes peatlands: Building transfer-function models for paleoenvi- of the Meshchera Lowlands, Russia, and utility of the micro- ronmental studies. The Holocene 16:403–414. fossils for inferring past lake water level. Journal of Paleolim- Reinhardt, E. G., M. Little, S. Donato, E. Findlay, A. Krueger, C. nology 62:137–150. Clark, and J. Boyce. 2005. Arcellacean (thecamoebian) evi- Warner, B. G. 1988. Testate amoebae (Protozoa). Methods in dence of land-use change and eutrophication in Frenchman’s Quaternary ecology: Geoscience Canada 5:65–74. Bay, Pickering, Ontario. Environmental Geology 47:729–739. Whittle, A., M. J. Amesbury, D. J. Charman, D. A. Hodgson, Roe, H. M., and R. T. Patterson. 2006. Distribution of thecamoe- B. B. Perren, S. J. Roberts, and A. V. Gallego-Sala. 2018. Salt- bians (testate amoebae) in small lakes and ponds, Barbados, enrichment impact on biomass production in a natural popu- West Indies. Journal of Foraminiferal Research 36:116–134. lation of peatland dwelling Arcellinida and Euglyphida (testate Roe, H. M., and R. T. Patterson. 2014. Arcellacea (testate amoe- amoebae). Microbial Ecology 78:534–538. bae) as bio-indicators of road salt contamination in lakes. Mi- Williams, J. L. 1985. Thecamoebian scales from a miocene lacus- crobial Ecology 68:299–313. trine deposit in Northern Idaho. Pages 67–71 in C. J. Smiley Schiller, W., and M. Wuttke. 2015. First record of siliceous shell (editor). Late Cenozoic history of the Pacific northwest: Inter- plates of Scutiglypha sp. (Amoebozoa: Thecamoeba: Euglyphi- disciplinary studies on the Clarkia fossil beds of northern Idaho. dae) from Lake Enspel (upper Oligocene, Westerwald Moun- Pacific Division of the American Association for the Advance- tains, Germany). Palaeobiodiversity and Palaeoenvironments ment of Science, San Francisco, California. 95:513–519. Wolfe, A. P., A. V. Reyes, D. L. Royer, D. R. Greenwood, G. Doria, Schönborn, W. 1992. The role of protozoan communities in fresh- M. H. Gagen, P. A. Siver, and J. A. Westgate. 2017. Middle Eo- water and soil ecosystems. Acta Protozoologica 31:11–18. cene CO2 and climate reconstructed from the sediment fill of a Siver, P. A. 2019. Potential use of chrysophyte cyst morphometrics subarctic kimberlite maar. Geology 45:619–622. as a tool for reconstructing ancient lake environments. Nova Wylezich, C., R. Meisterfeld, S. Meisterfeld, and M. Schlegel. 2002. Hedwigia, Beiheft 148:101–112. Phylogenetic analysis of small subunit ribosomal RNA coding Siver, P. A., and A. M. Lott. 2010. The scaled chrysophyte flora regions reveal a monophyletic lineage of euglyphid testate from the Pinelands National Preserve of southern New Jersey, amoebae (Order Euglyphida). Journal of Eukaryote Microbiol- U.S.A. Nova Hedwigia, Beiheft 136:167–181. ogy 49:108–118.