Proxies that use changes in the composition of ecological communities to reconstruct temporal changes in an environmental covariate are commonly used in paleoclimatology and paleolimnology. Existing methods, such as weighted averaging and modern analog technique, relate compositional data to the covariate in very simple ways, and different methods are seldom compared systematically. We present a new Bayesian model that better represents the underlying data and the complexity in the relationships between species’ abundances and a paleoenvironmental covariate. Using testate amoeba–based reconstructions of water-table depth as a test case, we systematically compare new and existing models in a cross-validation experiment on a large training dataset from North America. We then apply the different models to a new 7500-year record of testate amoeba assemblages from Caribou Bog in Maine and compare the resulting water-table depth reconstructions. We find that Bayesian models represent an improvement over existing methods in three key ways: more complete use of the underlying compositional data, full and meaningful treatment of uncertainty, and clear paths toward methodological improvements. Furthermore, we highlight how developing and systematically comparing methods lead to an improved understanding of the proxy system. This paper focuses on testate amoebae and water-table depth, but the framework and ideas are widely applicable to other proxies based on compositional data.
AmesburyMJSwindlesGTBobrovA, et al. (2016) Development of a new pan-European testate amoeba transfer function for reconstructing peatland palaeohydrology. Quaternary Science Reviews152: 132–151.
2.
BattarbeeRW (1984) Diatom analysis and the acidification of lakes. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences305(1124): 451–477.
3.
BattarbeeRWJonesVJFlowerRJ, et al. (2002) Diatoms. In: SmolJ (ed.) Tracking Environmental Change Using Lake Sediments. New York: Springer, pp. 155–202.
4.
BirksHJB (1998) Numerical tools in palaeolimnology – Progress, potentialities, and problems. Journal of Paleolimnology20(4): 307–332.
5.
BirksHJBSimpsonGL (2013) ‘Diatoms and pH reconstruction’(1990) revisited. Journal of Paleolimnology49(3): 363–371.
6.
BirksHJBter BraakCLineJ, et al. (1990) Diatoms and pH reconstruction. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences327(1240): 263–278.
7.
BoothRK (2002) Testate amoebae as paleoindicators of surface-moisture changes on Michigan peatlands: Modern ecology and hydrological calibration. Journal of Paleolimnology28(3): 329–348.
8.
BoothRK (2008) Testate amoebae as proxies for mean annual water-table depth in Sphagnum-dominated peatlands of North America. Journal of Quaternary Science23(1): 43–57.
9.
BoothRK (2010) Testing the climate sensitivity of peat-based paleoclimate reconstructions in mid-continental North America. Quaternary Science Reviews29(5–6): 720–731.
10.
BoothRKJacksonST (2003) A high-resolution record of late-Holocene moisture variability from a Michigan raised bog, USA. The Holocene13(6): 863–876.
11.
BoothRKZygmuntJR (2005) Biogeography and comparative ecology of testate amoebae inhabiting sphagnum-dominated peatlands in the Great Lakes and Rocky Mountain regions of North America. Diversity and Distributions11(6): 577–590.
12.
BoothRKHotchkissSCWilcoxDA (2005) Discoloration of polyvinyl chloride (PVC) tape as a proxy for water-table depth in peatlands: Validation and assessment of seasonal variability. Functional Ecology19(6): 1040–1047.
13.
BreimanL (2001) Random forests. Machine Learning45(1): 5–32.
14.
CarpenterBGelmanAHoffmanM, et al. (2016) Stan: A probabilistic programming language. Journal of Statistical Software20(2): 1–37.
15.
CharlesDF (1985) Relationships between surface sediment diatom assemblages and lakewater characteristics in Adirondack lakes. Ecology66(3): 994–1011.
16.
CharmanDJBlundellA and ACCROTELM members (2007) A new European testate amoebae transfer function for palaeohydrological reconstruction on ombrotrophic peatlands. Journal of Quaternary Science22(3): 209–221.
17.
CharmanDJHendonDPackmanS (1999) Multiproxy surface wetness records from replicate cores on an ombrotrophic mire: Implications for Holocene palaeoclimate records. Journal of Quaternary Science14(5): 451–463.
18.
CliffordMJBoothRK (2013) Increased probability of fire during late Holocene droughts in northern New England. Climatic Change119(3–4): 693–704.
19.
DavisRBAndersonDSBergeF (1985) Palaeolimnological evidence that lake acidification is accompanied by loss of organic matter. Nature316(6027): 436.
20.
EdwardsKJFyfeRMJacksonST (2017) The first 100 years of pollen analysis. Nature Plants3: 17001.
21.
GajewskiK (1987) Environmental history of Caribou Bog, Penobscot Co., Maine. Naturaliste Canadien114: 133–140.
GelmanALokenE (2014) The Statistical Crisis in Science. American Scientist. 102: 460–465.
24.
GelmanASimpsonDBetancourtM (2017) The prior can often only be understood in the context of the likelihood. Entropy19(10): 555.
25.
GneitingT (2011) Making and evaluating point forecasts. Journal of the American Statistical Association106(494): 746–762.
26.
GoringSWilliamsJBloisJ, et al. (2012) Deposition times in the northeastern United States during the Holocene: Establishing valid priors for Bayesian age models. Quaternary Science Reviews48: 54–60.
27.
HakimGJEmile-GeayJSteigEJ, et al. (2016) The last millennium climate reanalysis project: Framework and first results. Journal of Geophysical Research: Atmospheres121(12): 6745–6764.
28.
HefleyTJBromsKMBrostBM, et al. (2017) The basis function approach for modeling autocorrelation in ecological data. Ecology98(3): 632–646.
29.
HobbsNTHootenMB (2015) Bayesian Models: A Statistical Primer for Ecologists. Princeton, NJ: Princeton University Press.
30.
HoetingJAMadiganDRafteryAE, et al. (1999) Bayesian model averaging: A tutorial. Statistical Science14: 382–401.
31.
HootenMBHobbsN (2015) A guide to Bayesian model selection for ecologists. Ecological Monographs85(1): 3–28.
32.
HuFSDavisRB (1995) Postglacial development of a Maine bog and paleoenvironmental implications. Canadian Journal of Botany73(4): 638–649.
33.
HustedtF (1939 [1937]) Systematische und ökologische unter-suchungen über die diatomeen-flora von Java, Bali und Sumatra nach dem material der deutschen limnologischen Sunda-Expedition. I. Archives of Hydrobiology37: 15–16.
34.
Hydro2kP (2017) Comparing proxy and model estimates of hydroclimate variability and change over the Common Era. Climate of the past13(12): 1851–1900.
35.
ImbrieJKippN (1971) A new micropaleontologial method for quantitative paleoclimatology: Application to a late Pleistocene Caribbean core. In: FlintRF (ed.) The Late Cenozoic Glacial Ages. New Haven, CT: Yale University Press, pp. 71–181.
36.
ImbrieJWebbT (1981) Transfer functions: Calibrating micropaleontological data in climatic terms. In: BergerA (ed.) Climatic Variations and Variability: Facts and Theories. New York: Springer, pp. 125–134.
37.
ImbrieJVan DonkJKippNG (1973) Paleoclimatic investigation of a late pleistocene Caribbean Deep-Sea Core: Comparison of isotopic and faunal methods 1. Quaternary Research3(1): 10–38.
38.
IversenJ (1944) Viscum, Hedera and Ilex as climate indicators: A contribution to the study of the post-glacial temperature climate. Geologiska Föreningen I Stockholm Förhandlingar66(3): 463–483.
39.
JacksonST (2012) Representation of flora and vegetation in Quaternary fossil assemblages: Known and unknown knowns and unknowns. Quaternary Science Reviews49: 1–15.
40.
JacksonSTWilliamsJW (2004) Modern analogs in Quaternary paleoecology: Here today, gone yesterday, gone tomorrow?Annual Review of Earth and Planetary Sciences32: 495–537.
41.
JugginsS (2013) Quantitative reconstructions in palaeolimnology: New paradigm or sick science?Quaternary Science Reviews64: 20–32.
JugginsSBirksHJB (2012) Quantitative environmental reconstructions from biological data. In: SmolJ (ed.) Tracking Environmental Change Using Lake Sediments. New York: Springer, pp. 431–494.
44.
LiawAWienerM (2002) Classification and regression by randomForest. R News2(3): 18–22.
45.
MarkelERBoothRKQinY (2010) Testate amoebae and δ13c of Sphagnum as surface-moisture proxies in Alaskan peatlands. The Holocene20(3): 463–475.
46.
MarlonJRPedersonNNolanC, et al. (2017) Climatic history of the northeastern United States during the past 3000 years. Climate of the Past13(10): 1355.
47.
MarsicekJShumanBNBartleinPJ, et al. (2018) Reconciling divergent trends and millennial variations in Holocene temperatures. Nature554(7690): 92.
48.
MitchellEAPayneRJLamentowiczM (2008) Potential implications of differential preservation of testate amoeba shells for paleoenvironmental reconstruction in peatlands. Journal of Paleolimnology40(2): 603–618.
49.
NygaardG (1956) Ancient and recent flora of diatoms and chrysophyceae in Lake Gribso. Studies on the humic acid Lake Gribso. Folia Limnologica Scandinavica8: 32–94.
50.
OehlertGW (1988) Interval estimates for diatom-inferred lake pH histories. Canadian Journal of Statistics16(1): 51–60.
51.
OksanenJLääräEHuttunenP, et al. (1990) Maximum likelihood prediction of lake acidity based on sedimented diatoms. Journal of Vegetation Science1(1): 49–56.
52.
OverpeckJWebbTPrenticeI (1985) Quantitative interpretation of fossil pollen spectra: Dissimilarity coefficients and the method of modern analogs. Quaternary Research23(1): 87–108.
53.
ParnellA (2014) Bchron: Radiocarbon dating, age-depth modelling, relative sea level rate estimation, and non-parametric phase modelling. R Package Version4(1). Available at: https://rdrr.io/cran/Bchron/.
54.
PayneRJMitchellEA (2009) How many is enough? Determining optimal count totals for ecological and palaeoecological studies of testate amoebae. Journal of Paleolimnology42(4): 483–495.
55.
PayneRJTelfordRJBlackfordJJ, et al. (2012) Testing peatland testate amoeba transfer functions: Appropriate methods for clustered training-sets. The Holocene22(7): 819–825.
PrenticeIC (1980) Multidimensional scaling as a research tool in quaternary palynology: A review of theory and methods. Review of Palaeobotany and Palynology31: 71–104.
58.
R Core Team (2017) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Available at: https://www.R-project.org/.
59.
RenbergIHellbergTNilssonM (1985) Effects of acidification on diatom communities as revealed by analyses of lake sediments. Ecological Bulletins37: 219–223.
60.
SalonenJSIlvonenLSeppäH, et al. (2012) Comparing different calibration methods (WA/WA-PLS regression and Bayesian modelling) and different-sized calibration sets in pollen-based quantitative climate reconstruction. The Holocene22(4): 413–424.
61.
SchapireRE (1990) The strength of weak learnability. Machine Learning5(2): 197–227.
62.
ShumanBNRoutsonCMcKayN, et al. (2018) Placing the Common Era in a Holocene context: Millennial-to-centennial patterns and trends in the hydroclimate of North America over the past 2000 years. Climate of the Past14: 665–686.
63.
SimpsonGLShillandEMWinterbottomJM, et al (2005) Defining reference conditions for acidified waters using a modern analogue approach. Environmental Pollution137(1): 119–133.
64.
SmolJP (2009) Pollution of Lakes and Rivers: A Paleoenvironmental Perspective. Hoboken, NJ: John Wiley & Sons.
65.
SullivanMEBoothRK (2011) The potential influence of short-term environmental variability on the composition of testate amoeba communities in Sphagnum peatlands. Microbial Ecology62(1): 80–93.
66.
SwindlesGCharmanDRoeH, et al. (2009) Environmental controls on peatland testate amoebae (Protozoa: Rhizopoda) in the North of Ireland: Implications for Holocene palaeoclimate studies. Journal of Paleolimnology42(1): 123–140.
67.
SwindlesGTHoldenJRabyCL, et al. (2015) Testing peatland water-table depth transfer functions using high-resolution hydrological monitoring data. Quaternary Science Reviews120: 107–117.
68.
TalbotJRichardPRouletN, et al. (2010) Assessing long-term hydrological and ecological responses to drainage in a raised bog using paleoecology and a hydrosequence. Journal of Vegetation Science21(1): 143–156.
69.
TelfordRBirksH (2009) Evaluation of transfer functions in spatially structured environments. Quaternary Science Reviews28(13–14): 1309–1316.
70.
TelfordRAnderssonCBirksH, et al. (2004) Biases in the estimation of transfer function prediction errors. Paleoceanography19(4): 01072.
71.
ter BraakCJJugginsS (1993) Weighted averaging partial least squares regression (WA-PLS): An improved method for reconstructing environmental variables from species assemblages. In: Proceedings of the twelfth international diatom symposium, Renesse, 30 August–5 September, pp. 485–502. New York: Springer.
72.
ter BraakCJVan DamH (1989) Inferring pH from diatoms: A comparison of old and new calibration methods. Hydrobiologia178(3): 209–223.
73.
TiptonJRHootenMBNolanC, et al. (2019) Predicting paleoclimate from compositional data using multivariate Gaussian process inverse prediction. arXiv:1903.05036
74.
ToivonenHTMannilaHKorholaA, et al. (2001) Applying Bayesian statistics to organism-based environmental reconstruction. Ecological Applications11(2): 618–630.
75.
TrachselMTelfordRJ (2016) Estimating unbiased transfer-function performances in spatially structured environments. Climate of the past12(5): 1215–1223.
76.
TsyganovANMityaevaOAMazeiYA, et al. (2016) Testate amoeba transfer function performance along localised hydrological gradients. European Journal of Protistology55: 141–151.
77.
Van BellenSMauquoyDPayneRJ, et al. (2014) Testate amoebae as a proxy for reconstructing Holocene water table dynamics in southern Patagonian peat bogs. Journal of Quaternary Science29(5): 463–474.
78.
Van BellenSMauquoyDPayneRJ, et al. (2017) An alternative approach to transfer functions? Testing the performance of a functional trait-based model for testate amoebae. Palaeogeography, Palaeoclimatology, Palaeoecology468: 173–183.
79.
Von PostL (1918) Forest tree pollen in south Swedish peat bog deposits, translated by Margaret B. Davis and Knut Faegri, 1967. Pollen Et Spores9: 375–401.
80.
WarnerBGCharmanDJ (1994) Holocene changes on a peatland in northwestern Ontario interpreted from testate amoebae (Protozoa) analysis. Boreas23(3): 270–279.
81.
WebbTBrysonRA (1972) Late-and postglacial climatic change in the northern Midwest, USA: Quantitative estimates derived from fossil pollen spectra by multivariate statistical analysis. Quaternary Research2(1): 70–115.
82.
WoodlandWACharmanDJSimsPC (1998) Quantitative estimates of water tables and soil moisture in Holocene peatlands from testate amoebae. The Holocene8(3): 261–273.
83.
WrightHEMannDGlaserP (1984) Piston corers for peat and lake sediments. Ecology65(2): 657–659.
84.
WrightHEWinterTCPattenHL (1963) Two pollen diagrams from southeastern Minnesota: Problems in the regional late-glacial and postglacial vegetational history. Geological Society of America Bulletin74(11): 1371–1396.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
