Foraminiferal growth and test development

Accepted Manuscript Foraminiferal growth and test development Johann Hohenegger PII: DOI: Reference: S0012-8252(17)30633-5 doi:10.1016/j.earscirev.2018.06.001 EARTH 2640 To appear in: Earth-Science Reviews Received date: Revised date: Accepted date: 13 December 2017 1 June 2018 1 June 2018 Please cite this article as: Johann Hohenegger , Foraminiferal growth and test development. Earth (2017), doi:10.1016/j.earscirev.2018.06.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Foraminiferal growth and test development Johann Hohenegger University of Vienna, Department of Palaeontology A 1090 Wien, Althanstrasse 14 AC C EP T ED MA NU SC RI PT E-mail: [email protected] ACCEPTED MANUSCRIPT ABSTRACT Growth of multichambered foraminifera can be studied by investigation of chamber volumes.. This approach is applicable to the vast majority of living and fossil species. Cell growth is represented by the test volume, whereby chamber volumes exhibit the increase in cell volume by growth stages. Two models fit foraminiferal growth, the unlimited linear and the limited sigmoidal function. The growth stage, in which reproduction begins, is represented PT in sigmoidal growth by the point of inflection; this stage cannot be determined in the unlimited linear growth model. The timing of chamber building is apparently correlated with RI cell growth, thus the “chamber building rate” remains constant in linearly growing cells whereas it decreases and ultimately approximates zero in limited growth. Longevity can be SC estimated for living individuals with sigmoidal growth by the inverse of the chamber building function at the inflection point of the cell growth function. Calculating the mean chamber NU building function for a species inhabiting a distinct geographical region, its inverse can be used in this region to estimate the individual lifetime based on the final growth state (= MA chamber number). Deviations of observed chamber volumes from theoretical values determined by the chamber building function can be irregular or oscillating. Periods of oscillating functions in larger foraminifera may point to tidal, lunar and seasonal cycles. ED Based on the determination of seasonal oscillations in the test by stable isotopes, lifetime ≥ 1 year can also be estimated for fossil species. The development in foraminifera is expressed in EP T a huge variability of test forms with different wall textures depending on functional and constructional aspects. The historical-phylogenetic approach enables a differentiation into three groups, the Tubothalamea, Globothalamea and the Nodosariids. Within the Tubothalamea and Globothalamea, several subgroups feature symbiont-bearing foraminifera AC C with large tests designed to optimize the surface/volume ratio allowing the symbiotic microalgae to occupy an optimal position near the test surface. Keywords: foraminiferal cell growth; ““chamber building rate””; lifetime; test development; growth functions; phylogenetic pathways ACCEPTED MANUSCRIPT 1. Introduction Growth is a fundamental parameter characterizing life along with ‘organization’, ‘development’, ‘response to stimuli’, ‘reproduction’, ‘homeostasis’ and ‘metabolism’ (e.g. Reece et al., 2012). The lifespan of organisms is always restricted due to thermodynamic losses in a closed system of catalysts (enzymes) and chemicals (metabolites), further restricted due to evolutionary constrains. Accordingly, reproduction takes place during the adult stage PT or at the end of an individual’s life, either sexually, asexually or both (Koshland, 2002). In asexual reproduction the newborn can be uni- or multicellular (e.g. budding), while the complete information for developing complex organisms. RI sexually produced newborns generally consist of a single cell (zygote) also containing the SC Differentiation between growth and development is necessary. Growth, on the one hand, is understood as the increase in size (biomass, volume) through time (e.g. ) caused NU by a higher rate in anabolism that converts chemicals and energy into cellular components, in contrast to catabolism, which decomposes organic matter. Development, on the other hand, is MA the spatial organization of body mass through time expressed in a shape change (structure) , which is always, but not necessarily linearly, correlated with growth (e.g. metamorphosis). Development strictly follows morphogenetic programs leading to ED differentiation of cells into organs in multicellular organisms (e.g. Gilbert, 2013; Wolpert et al., 2015), but also determines shape in unicellular organisms, especially when these EP T organisms develop shells. Growth is thus the mass increase with time caused by metabolism, when anabolism excels catabolism. Young (smaller) individuals exhibit higher metabolic rates in comparison to their AC C mass than larger, grown-up individuals. Thus, the weight-specific metabolic rate decreases during growth (e.g. von Bertalanffy, 1957), explaining the rapid growth in newborns and juveniles. The balance between anabolism and catabolism indicates either resting periods during growth (discontinuous growth), or, after initial growth in juvenile (larval) stages, the balance leads in adults to constancy in mass, ranging from days (e.g. holometabolic insects) to several years (e.g. mammals). Unlike populations, where unlimited growth can be modelled by exponential functions (Malthus-growth in r-strategists; Malthus, 1798), organismic growth is always limited due to the restricted lifetime, although it can encompass hundreds of years (e.g. trees). Although growth appears unlimited in young individuals and can be fitted by exponential functions, it becomes limited in grown-up individuals. Beside the logistic function that models limited ACCEPTED MANUSCRIPT growth in population dynamics (K-strategists), several growth functions modelling limited growth can be applied for organisms (e.g. Gompertz, 1832; Bertalanffy, 1938; Janoschek, 1957; Richards, 1959; Morgan-Mercer-Flodin, 1975). They differ in the position of the inflection point in relation to the upper limit. The inflection point is important because, after this point, the growth rate decreases. Foraminifera are single-celled, heterotrophic (sometimes mixotrophic) aquatic organisms PT belonging to the group of Retaria (Cavalier-Smith) characterized by reticulopods (Adl et al., 2013). They can be divided in two groups according to the formation of tests: the athalamous RI and thalamous foraminifera, whereby the latter can be divided into monothalamous (singlechambered) and polythalamous (multi-chambered, multilocular) foraminifera. This division is SC also supported by molecular genetic research (Pawlowski et al., 2013) and influences the growth pattern in foraminifera. While growth in athalamous species can be continuous, NU interrupted or slowed down by external (environmental) factors, growth in thalamous species must be stepwise, internally controlled by growth factors and externally by the environment MA (e.g. Bradshaw, 1957; Lee et al., 1963; Lee and Bock, 1976; Röttger, 1976; Hallock, 1981; Lombard et. al., 2009; Uthike and Altenrath, 2010; Triantaphyllou et al., 2012; Langlet et al., 2013; Ross and Hallock, 2016). In monothalamous representatives with more or less spherical ED tests, the individuals leave the test and grow to larger size, then construct a new test (e.g. Bowser et al., 1995). Species with tubes do not have to leave the test but rather add additional EP T segments to this tube. Polythalamous (multichambered) species grow by building a new chamber attached to the older test, thus enlarging the test volume, rarely in a linear manner, generally with an (initially) exponential increase in cell volume. This leads to discontinuous growth with a differentiation between periods, where energy is used for chamber construction, AC C and growth. Thus, discontinuous growth can easily be represented by the mass (volume) of successive chambers in multilocular foraminifera, which represent the vast majority of marine foraminifera with tests. The form of reproduction leading to size (mass) differences in the newborns, together with the model describing cell growth and the morphogenetic programs canalizing the cell mass into test shape, are the three factors determining growth sensu lato in multichambered foraminifera. Larger benthic symbiont-bearing Foraminifera (LBF) are used as examples for foraminiferal growth because they show three generations with asexually produced gamonts and schizonts and sexually produced agamonts. All generations are characterized by the long ACCEPTED MANUSCRIPT lifespan ranging from 0.5 to 1.5 years in gamonts/schizonts and up to 3 years in agamonts (e.g. Eder et al., 2016). Therefore, environmental influences, especially seasonal changes, can easily be checked during growth. 2. Reproduction Growth starts after reproduction, if environmental conditions are favorable. This makes PT the type, place and time of reproduction important for subsequent growth. Reproduction in foraminifera is basically characterized by an alternation of generations with sexual RI reproduction (gamogony) of haploid gamonts through isogametes leading to diploid agamonts SC and subsequent growth, followed by asexual reproduction (e.g. Grell, 1973). Apogamy with meiosis in agamonts leads either to haploid gamonts or schizonts, whereby the former reproduce sexually, and the latter asexually (e.g. Dettmering et al., 1998; Goldstein, 1999). In NU multi-chambered foraminifera, the diploid agamonts can remarkably differ by larger test size from the haploid generations, especially in larger symbiont-bearing foraminifera (Fig. 1; e.g. MA Hohenegger, 2011), and in the extremely small proloculi (except gamontogamous and autogamous species). The chamber number, counted from the offset of the zygote to the size at which reproduction takes place is much higher in agamonts than in gamonts and schizonts. ED This has an important effect on the “chamber building rate”s in polythalamous foraminifera. Both asexually produced generations differ in test size (e.g. Harney et al., 1998) as well as in EP T proloculus size, both in larger benthic foraminifera (Dettmering et al., 1998) and in smaller benthics (Lehmann et al., 2006). Deviations from these life cycles occur in many species (Goldstein, 1999). On the one hand, gamogony can be completely suppressed (e.g. AC C Spiroluculina hyalina in Arnold, 1964) or, on the other, be the sole reproduction mode in planktonic foraminifera (e.g. Hemleben et al., 1988, Schiebel and Hemleben, 2017). In gamogony, growth starts after the fusion of gametes leading to diploid zygotes. The fusion of flagellate gametes outside the maternal test involves high risks. Because selffertilization is mostly avoided (except autogamy), the successful conjunction of gametes originating from different individuals depends on their contemporaneous reproduction, the hydrodynamic situation and the distance between reproductive individuals. Therefore, the solely sexually reproducing planktonic foraminifers have reacted to these requirements by high population densities, contemporaneous release of gametes due to short lifespans ending in reproduction, which depends on lunar cycles (e.g. Jonkers et al., 2015), and an enormous production of gametes (several 100,000) spread into the environment (Bé et al., 1977; ACCEPTED MANUSCRIPT Spindler et al., 1978). Growth starts in planktonic foraminifera after zygote formation. Producing a new chamber takes a few hours and the calcification of the chamber wall lasts 1.5 h (Bé et al., 1977). Because of low cell densities in propagules, they rise to the surface and continue growing in a light-flooded environment. During the short lifespan, they continuously sink to the particular depth where reproduction starts again (Erez et al., 1991). Flagellate gametes of benthic foraminifera must find partners for conjugation, a difficult PT task due to distance and the hydrodynamics in shallow-water environments, especially when reproduction periods are not fixed. This is compensated by fixed reproduction times and the RI production of an enormous number of flagellate gametes, especially in symbiont-bearing larger benthic foraminifera. Nevertheless, the small zygotes are susceptible to hydrodynamics SC that can transport them into unfavorable environments. The growth of the extremely small juveniles is also endangered in the shallow benthic habitat where reproduction takes place. NU Thus, the relation between grown agamonts and gamonts/schizonts is unbalanced in favor of the latter, with maximum proportions in LBF of one to several thousands. Survival of the MA small, sexually produced propagules adapted to fine sediments in shallow water, but transported to inconvenient habitats (e.g. deeper water), is possible by protection within cysts produced using agglutinated particles (e.g. Alve and Goldstein, 2003; Goldstein and Alve, ED 2011). This enables survival in quiescence under unfavorable conditions for up to several years (Ross and Hallock, 2016), whereby growth starts after arrival at an appropriate habitat. EP T Drawbacks in the conjugation of flagellate gametes are eliminated by gamontogony and autogamy (see Grell, 1973). In the former, non-flagellate gametes forming zygotes are protected by the parental tests (e.g. Rosalina floridana; Lee et al., 1963). In the latter, they are AC C protected within the maternal test because of self-fertilization (e.g. Rotaliella div. species; Goldstein, 1999). Due to this protection, the number of non-flagellate gametes is strongly reduced in gamontogamous and autogamous species, rarely exceeding 20 gametes (Grell, 1973). Like young gamonts produced after agamogony, the zygotes receive parts of the maternal protoplasm containing all cell organelles. Therefore, grown-up agamonts and gamonts are represented in similar numbers within a population. After zygote formation, agamont growth starts directly after leaving the parental tests: the rather large juvenile agamonts have good chances of survival because they are adapted to the parental environment. Adaptation to the environment of the maternal organism is the main advantage of asexual reproduction at the end of agamogony and schizogony. The number of young gamonts/schizonts is high, ranging from a few hundred in smaller benthics to several ACCEPTED MANUSCRIPT thousands in LBF. They are formed either within the maternal test (e.g. Rubratella intermedia; Grell, 1958), within special brood chambers in larger benthic foraminifera (e.g. Cyclorbiculina compressa; Lutze and Wefer, 1980; Amphisorus hemprichii and Neorotalia calcar; Hohenegger, 2011) or at the surface of the maternal test (e.g. Heterostegina depressa; Röttger and Spindler, 1976). They are then released to the environment after constructing several chambers and can immediately start growing, if environmental conditions remain constant. When the asexually produced gamont/schizont propagules of smaller benthic PT foraminifera (inhabiting shallow water fine sediments) are transported to hostile deeper environments, they can survive several years as cryptobionts, like the sexually produced RI agamonts (Goldstein and Alve, 2011; Ross and Hallock, 2016). Growth starts after reposition SC into a favorable environment or switching to favorable conditions. Transport of LBF’s young gamonts/schizonts by surface currents into appropriate oligotrophic shallow-water NU environments is the main factor for their distribution, enabling traversing thousands of kilometers of oceanic zone (pelagic zone beyond the continental shelf). During this travel, growth is suppressed by dormancy, enabling transport in suspension due to low body densities MA (Hohenegger, 2011). ED 3. Growth During undisturbed life conditions, foraminifera with tests wrap the complete or part of EP T the test with a thin layer of cytoplasm extruded from apertures (or canal openings in symbiont-bearing hyaline larger foraminifera) from which reticulopods are extended. The main bundle of reticulopods is concentrated around apertures. When environmental conditions AC C deteriorate, the cytoplasm withdraws into the inner test parts, emptying the final part of the test that consists of a single or a few final chambers in multi-chambered forms. The mechanism for concentrating the cytoplasm to the minimum volume needs further investigation. Because foraminifera are single-celled organisms, growth can be simplified as the increase in cytoplasm and organelles of a cell through time. In athalamous (non-chambered) foraminifera, growth can be continuous through time, with changing growth rates being influenced by the food supply. In species constructing tests, growth is divided into steps by the test building process. At distinct time intervals that can be modeled by mathematical functions, test formation is possibly initiated by growth factor proteins, which must be studied in the future. Chamber formation is described in detail by many authors (e.g. Angell, 1967, ACCEPTED MANUSCRIPT 1980; Hemleben et al., 1977, 1986; Röttger, 1981; Bender, 1992). Due to this time dependence, chamber volumes in polythalamous foraminifera reflect intensities in cell growth within the time interval between the chamber formation processes. Therefore, periodically and instantaneously changing environmental conditions are reflected in the chamber volumes (Hohenegger and Briguglio, 2014). In multi-chambered species, growth is characterized by a series of chambers built PT successively with constant or increasing time intervals between chamber constructions. This RI enables describing growth SC in two steps. (1) NU The first step explains the dependence of the cell volume V on the chamber number n by (2) MA with the chamber volume as the first derivative (3) that determines the growth rate of the cell. ED Because this function does not directly depend on time, it can also be used for fossil species, helping to explain differences in volume growth between species and genera based on factors. EP T genetics and interpreting deviations from theoretical growth caused by internal and external AC C The second function describes the relation of chamber number with time by (4) with the “chamber building rate” as the first derivative (5) The “chamber building rate” can be calculated only for living individuals by studying chamber formation during lifetime. Deviations from the theoretical function modelling the increase in chamber numbers over time can be caused by periodically changing environmental factors (e.g. tidal, lunar, seasonal cycles), or can be instantaneous due to abiotic (e.g. fracturing by hydrodynamics) or biotic events (e.g. fracturing by predators, reduced food supply; Hohenegger and Briguglio, 2014). Nevertheless, transferring “chamber building rate”s of living species to closely-related extinct species also allows interpretation of time- ACCEPTED MANUSCRIPT related influences for fossil specimens using time indicators such as stable isotopes in the chamber walls (e.g. Wefer and Berger, 1991; Evans et al., 2013). For fitting theoretical functions to the observed cell volumes and testing their qualities, we used a specimen of Palaeonummulites venosus, characterized by tests consisting of planispirally arranged chambers (Fig. 2). The relations between successive chambers demonstrate a linear or nonlinear size increase. Accordingly, the statistical estimation of PT growth functions is strongly influenced by large deviations (residuals) in the youngest (final) chambers, leading to underestimation of residuals in initial chambers. Therefore, cell volumes (6) SC RI should be linearized by calculating their cubic roots. After fitting a theoretical growth function to linearized cell volumes, calculated at each NU growth step indicated by chamber number i, the theoretical chamber volumes V’ must be MA recalculated by (7) ED Obtaining untransformed theoretical cell volumes by ∑ (8) leads to the fit of cell volumes by the theoretical growth function, which can be tested for EP T goodness of fit (Hohenegger and Briguglio, 2014). Some common theoretical growth functions (Tab. 1) are used for fitting the observed cell growth based on chamber numbers ∞ , where i = 0 represents the nepiont (often protoconch and deuteroconch) and is the upper asymptote. AC C ∞ Applying the five selected growth models for the exemplified specimen of P. venosus (Fig. 3) allows testing their quality of fit. This can be checked by the sum of absolute standardized residuals ∑| using either the cell volumes | or chamber volumes (9) . Using the sum of standardized residuals, it can be demonstrated (Tab. 2) that the best fits are obtained by the Gompertz and Richards functions (violet and red lines in Fig. 3). This ACCEPTED MANUSCRIPT preference is based not only on the exemplified specimen (Fig. 2) but also on several specimens of P. venosus and Heterostegina depressa. Preference for the Gompertz and Richards functions has already been mentioned by Hohenegger and Briguglio (2014). The location of inflection points in cell growth (Tab. 1) determines the end of increasing growth, leading to size decrease of the following chambers. This phenomenon was first described in planktonic foraminifera as ‘kummerformen’ (Berger, 1969) and explained by PT Olsson (1973) as mature individuals adding in further growth smaller chambers. Therefore, inflection points mark the maturation stage, where reproduction starts, but further growth with RI smaller chambers is possible. This can better be demonstrated by smaller benthic foraminifera because they have much fewer chambers (Fig. 4). SC In our example of P. venosus, the inflection points are low for the Logistic (n = 50.8) and the Gompertz functions (n = 54.9) and higher for the Richards function (n = 76.6). Inflection NU points of the remaining functions are located at chamber 151 (Morgan-Mercer-Flodin) and chamber 258 (Janoschek). These values confirm the quality of the fit by the Richards function, MA because the highest chamber number of P. venosus is expected to be around 87 chambers when growing under natural conditions (Kinoshita et al., 2017). Compared to smaller benthic foraminifera, differences in growth between gamonts and ED agamonts are evident in LBF due to strong size differences (Fig. 1). Here, a gamont of H. depressa is compared with an agamont (Fig. 5). Cell and chamber growth can optimally be EP T fitted for both generations by Richards functions, although they differ strongly in their function parameters. Cell growth is intense in the gamont, starting with the nepiont volume of 1.43E-03 mm3, while cell volume at the point of inflection, about where reproduction takes AC C place, is low (1.113 mm3) at chamber 60, corresponding to a maximum theoretical chamber volume of 0.046 mm3 (Fig. 5). In the agamont, the increase in cell growth is much weaker compared to the gamont, starting with the nepiont volume of 3.82E-06 mm3. Theoretical cell volume at the inflection point at chamber number 120 is much higher (4.497 mm3), corresponding to a maximum theoretical chamber volume of 0.114 mm3 (Fig. 5). In chamber volumes of both gamont and agamont, oscillations of the observed values around the theoretical growth functions are evident (Fig. 5). These oscillations can provide a basis for determining cycles, but then they must be related to time using equation 4 (Hohenegger and Briguglio, 2014). The growth model for gamonts (schizonts) can be used for all polythalamous smaller benthic foraminifera, while it is not always applicable for the agamont generation. The use of ACCEPTED MANUSCRIPT Richards (or Gompertz) function for the agamont of H. depressa and all other nummulitid agamonts (e.g. P. venosus, Cycloclypeus carpenteri; Briguglio and Hohenegger, 2014) is possible because young gamonts are generated outside the parental test after outflow of the cytoplasm (Röttger, 1974; Krüger, 1994; Krüger et al., 1996) and thus has not influenced chamber growth in the terminal test. When young gamonts are generated within the parental cells, there are two strategies to provide space for the numerous small gamonts (Hottinger, 2000). The first involves partial resorption of test material, especially of the septa and septula; PT this is sometimes accompanied by adding a large, final brood chamber that opens by breaking the test walls to hatch the young gamont/schizonts (Fig. 6A). The second strategy is a sudden RI increase in chamber growth, significantly deviating from the former growth. This results in SC larger chambers or chamberlets with thinner walls compared to the former test (e.g. Cyclorbiculina compressa; Lutze and Wever, 1980). Again, these chambers or chamberlets NU break to release numerous gamonts/schizonts (Fig. 6B). Another deviation from normal growth occurs in the final chambers by the change of life mode from benthic-semisessile to planktonic by developing float chambers (Fig. 6C). MA The strong dependence of cell volume growth on stable environmental conditions – constant or periodic – explains the decrease in growth when freshly captured individuals are ED cultured in the laboratory investigating further growth. Despite attempts to keep environmental conditions as natural as possible (e.g. Wöger et al., 2016), the 156 investigated EP T specimens of Palaeonummulites venosus and 221 specimens of Heterostegina depressa significantly reduce growth and simultaneously show irregularities (Fig. 7). Determining the growth of cell and chamber volumes over time requires determining AC C instants of time in chamber construction (equations 4 and 5). The time intervals between chamber formations depend on the dynamics of environmental conditions. Undisturbed environments can involve either constant or periodically changing conditions. Time intervals between chamber formations are constant or increase (Fig. 8). Time intervals in chamber construction could depend on metabolic rates. In this case, mass-specific metabolic rates effect growth velocities (Bertalanffy, 1957). High mass-specific metabolic rates in newborns and young individuals cause short time intervals between chamber constructions compared to large intervals in grown-up specimens. Accordingly, increasing time intervals between chamber formations characterizes the normal condition in foraminifera. Constant intervals, if they exist at all, would be restricted to tests in which ACCEPTED MANUSCRIPT chamber volumes do not increase for chamber construction (due to linearly increasing metabolic rates) (Fig. 8). Metabolic rates can be simply estimated by relating the surface of the cell to its volume (Bertalanffy, 1957), or they can be measured by the oxygen respiration rate, showing the need for oxygen in catabolism, using the equation R = 3.98 10 3 BioVol 0.88 (Fig. 9A). In this equation, respiration R is expressed in ml O2 h1 and BioVol in m3 (e.g. Geslin et al., 2011). PT The high mass-specific metabolic rate in offspring (Fig. 9B) induces a high chamber-building rate in juveniles, decreasing with time (Fig. 9C). Consequently, mass-specific metabolic rates RI and the chamber-building rate are strongly correlated, which can be fitted by logarithmic functions (Fig. 9D). SC When environmental conditions remain constant or change periodically, intervals between chamber buildings increase constantly in most polythalamous foraminifera (Fig. 8B). Due to NU the strong dependence of growth on metabolic rates, intervals between chamber building dates show small deviations from expected values based on the mathematical function (Fig. 8B). MA Strong deviations from these strict timings can reflect instantaneous effects like fragmentation due to physical (mechanical breakage) or biological (predation) effects, which strongly disturb growth both in time and chamber volume. Such damage interrupts growth and the ED surviving individuals try to attain normal growth by various mechanisms, namely by accelerating chamber-building rates (compensatory growth) or by making strong deviations EP T from the normal chamber volume growth (unbalanced outliers in Hohenegger and Briguglio, 2014), or a combination of both (Fig. 10). A few investigations have estimated the chamber-building rates of LBF in the laboratory AC C (e.g. Röttger, 1972) to follow chamber formation day by day. Changes in size (diameter or area) were taken as an indicator for growth (e.g. Röttger, 1976; Lutze and Wefer, 1980; Ter Kuile and Erez, 1984; Schmidt et al., 2011). However, the nonlinear – almost exponential – relation of size to time can bias this effort and strongly complicates time interval estimation. Additionally, test volumes cannot represent cell and chamber volumes directly as expected by Ter Kuile and Erez (1984). Especially in nummulitids such as Heterostegina, where the central test part is involute, leaving the main volume for test walls and small proportions for chamber volumes, the test becomes evolute in younger parts with the main portion taken now by the chamber volumes (Fig. 11). “chamber building rate”s based on hours or days, in contrast, represent linear relations to time and can thus better serve for estimating the timing of chamber construction. ACCEPTED MANUSCRIPT The question regarding the best fit of the nonlinear chamber-building rate can be solved based on laboratory investigations on Heterostegina depressa (Röttger, 1972). There, the search for the best fit used four functions: the Power function, the Michaelis-Menten function in its original version as an estimation of chemical kinetics, the Generalized MichaelisMenten function used as a growth function (López et al., 2000) and the Bertalanffy function (Table 3). The Generalized Michaelis-Menten Function is simplified for foraminifera because the chamber-building rate starts with the first chamber after the proloculus or nepiont . PT becoming the lower limit RI Chamber building in the laboratory was investigated on four test groups of Heterostegina depressa from Hawaii (Röttger, 1972). Growth was less constant than measured in nature, SC shown by the addition of different chamber sizes (normal and very small chambers in fig. 4 of Röttger, 1972). Only few individuals with a continuous and more or less constant chamber NU formation could be used to test the best fit with functions of Tab. 3. Fitting the chamber formations in the selected individuals of Heterostegina depressa (Fig. MA 12) with the four functions resulted in the best fit by the Bertalanffy function with one exception (Ind. 21 in Fig. 12) (checked by the Chi-Square statistic; Tab. 4). Although the Bertalanffy function gives the best fit, it does not run through the origin (chamber number 0) ED like the other functions, which is an essential postulate. Comparing fitting the sum of chamber formations by the above growth functions (Fig. 12) EP T resulted in the best fit by the Generalized Michaelis-Menten function, which was tested by the sum of standardized residuals (Equation 9). Again, the Bertalanffy function must not be used because it does not run through the origin. Although the normal Michaelis-Menten function AC C gives the weakest (yet still significant) fit, the chamber-building rate is ideal for the initial seven chambers that are formed in the laboratory day by day, which seems to be typical for the nummulitids Heterostegina and Palaeonummulites (Fig. 13). An averaged chamber-building rate under natural conditions can be obtained for shallow benthic foraminifera by the ‘natural laboratory’ approach based on population dynamics (Hohenegger et al., 2014; Kinoshita et al., 2017).This rate cannot be gained by population dynamic investigations based on test size (e.g. Zohary et al., 1980). First, densities of populations sampled at approximately equal time intervals (from weeks to months) at the same station using replicates were calculated (Fig. 14). From these frequency distributions, the calculated maximum chamber numbers (mean plus 3 standard deviations; Fig 14) were ACCEPTED MANUSCRIPT computed by the above-mentioned functions (Tab. 3), preferably using the Michaelis-Menten function (Fig. 15). Then, inverse functions for “chamber building rate”s (e.g. Michaelis-Menten of Tab. 3 for P. venosus in Figs. 2, 3) can estimate the averaged time of chamber construction, valid for all individuals of a species at a specified region (e.g. Tropical Northwest Pacific for the examples in this article). This enables representing cell growth of an individual dependent on time PT (crosses in Fig. 16) together with the applied function fitting individual growth (e.g. Richards function in Fig. 16), also adapted to time. from the expected volumes RI Deviations of chamber volumes determined by the adjusted time-dependent growth function can be calculated using SC standardized residuals ) NU ( (10) where i represents the chamber number (Fig. 17A; see Hohenegger and Briguglio, 2014). MA Standardized residuals adjusted to time oscillate around zero. These oscillations can be decomposed into m sinusoidal functions, where the sum of sinusoids fits the empirical ED oscillations by ∑ ( ⁄ ) (11) EP T Several methods can be used to decompose an oscillating function into a number of sinusoids. Because of unequal time intervals between observations caused by the changing “chamber building rate”, the normal Fourier transformation is not applicable. Here, three AC C methods are represented, which have mostly been used in paleoclimate analyses: 1) Spectral analysis using the Lomb Periodogram algorithm (Lomb, 1976; Press et al., 1992) for unevenly sampled data (Fig. 17B, D), 2) REDFIT spectral analysis (Schulz and Mudelsee, 2002) summarizing overlapping intervals (Fig. 17C, D) and 3) Sinusoidal regression using Nyquist frequencies (Grenander, 1959; Hammer, 2017) (Fig, 17D). The specimen of P. venosus pictured in Fig. 2 was used to check the quality of fitting by the above-mentioned methods (Figs. 17, 18). Lomb spectral analysis and sinusoidal regression using Nyquist frequencies gave similar results with 3 main periods around 13.4, 34 and 7.6 days, while the fourth sinusoids differ in period lengths (5.0 versus 9.4 days). The REDFIT ACCEPTED MANUSCRIPT analysis certifies positions of the main periods around 13.5 and 34 days and combines the differing shortest periods to a single sinusoid with a period length of 7.9 days (Fig. 17C). Fitting the observed standardized residuals by the three analyses is shown in Fig. 18A, while the single sinusoids and their sum obtained by the REDFIT analysis are pictured in Fig. 18B. These periods determined in a single individual of P. venosus approximate the weekly, 2-week and the 4-week cycles. Two- and 4-week cycles based on slightly differing PT estimations of “chamber building rate”s have been reported in other nummulitid species such as in 17 specimens of P. venosus from Belau and Okinawa, 7 specimens of Cycloclypeus RI carpenteri from Ishigaki-Jima, Japan (Briguglio and Hohenegger, 2014) and 10 specimens of Heterostegina depressa from Hawaii and Okinawa (Eder et al., 2016). These dominant cycles NU interpretation of the 7-day cycle is problematic. SC were interpreted as tidal and lunar cycles (Hohenegger and Briguglio, 2014), whereas the 3.1 Deviation from general growth MA In chambered foraminifera, cell growth is combined with an increase in chamber volumes that can be modelled by growth functions and their first derivative (Tab. 1, Fig. 3). Regarding diameter during growth. ED the nominal chamber shape as a spheroid, chamber volume is characterized by an increasing EP T In some cases diameters remain constant during growth, transforming the nominal chamber shape into an equipotential ellipsoid with constant diameters. Increasing growth is then represented by changes in ellipsoid length. This growth type can be modelled in AC C planispirally coiled tests by an Archimedean spiral. The tube diameter remains constant throughout test construction in megalospheres (A-generation, gamonts/schizonts) of smaller foraminifera (Fig. 19A). In microspheres (B-generation, agamonts), tests start with extremely small tube diameters that retain constancy within the first whorls. The target for attaining tube diameters of the same size as in the final stage in megalospheres is obtained by a stepwise increase of the tube diameter in sections containing a few whorls (Fig. 19B). Megalospheres (A-generation) of larger benthic symbiont bearing foraminifera (LBF) maintain chamber size during growth as constant as possible, retaining small and equally sized compartments functioning as a reactor receptacle relieving metabolic processes (e.g. Hottinger, 2000; Hohenegger, 2011). In planispirally coiled LBF (without elongations along the coiling axis like in fusiform tests) compartmentalization is accomplished by two strategies. On the one hand by dividing the increasing chamber volumes with septula into equally sized ACCEPTED MANUSCRIPT chamberlets; test margins of this type can be modelled by a logarithmic spiral (e.g. Heterostegina, Fig. 1). On the other hand, represented in the Paleogene genus Nummulites, maintaining constant chamber height during growth to provide equally sized compartments; here, the test margins can be modelled by Archimedean spirals (Fig. 20). The need for large-sized tests (> 2mm) providing optimum surface/volume ratios for the symbiotic microalgae (e.g. Hohenegger, 2009) makes a single, constant chamber size PT insufficient during chamber growth even in small-sized megalospheres (A-generation). Because cells in LBF follow sigmoidal growth that can be modelled, e.g. by Gompertz or RI Richards functions (Fig. 21A,C), this growth type is approximated in LBF using constant chamber heights in several increasing growth steps (Figs 20, 21B, D; Briguglio et al., 2013). SC The advantage of stepwise growth perpetuating constant chamber volumes within sections is demonstrated comparing the sum of standardized residuals (Equation 9) between cell NU volumes (Richards: 5.0, Stepwise: 3.6) and chamber volumes (Richards: 22.5, Stepwise: 15.1). When megalospheres of the Paleogene Nummulites attain large size by stepwise MA enlargement of chamber height in spiral sections, they attain a species-specific optimum height (Fig. 21D) in the last whorls. A stepwise decrease of chamber height in the last whorl ED demonstrates that the inflection point in sigmoidal growth has been exceeded. This type of growth is unsuitable for microspheres (B-generations) to achieve extremely large-sized tests (largest Nummulites with diameter of 19 cm; Pavlovec, 1987), in which the EP T optimum chamber height is species-specifically fixed related to its counterpart in the Ageneration. This problem was solved by adding additional spirals after attaining the optimum AC C chamber height in the initial spiral leading to multispiral tests (Fig. 22). A detailed description of multispiral growth and its biological and ecological importance is given by Ferràndez-Cañadell (2012), who discuss the onset of multispirals in detail. The advantage of multispiral growth is demonstrated by a much higher growth rate in cell volume compared to a growth rate restricted to a singular spiral (Fig. 23A). The growth in chamber volumes confirms this model (Fig. 23B). In our example (Fig. 22) the initial spiral grows in steps up to an optimum chamber height. After attaining this height, the second spiral starts with initially smaller chamber height, reaching the optimum after approximately 10 chambers (Fig. 23B). One and a half whorls later the 3rd spiral starts (Fig. 22), again attaining the optimum chamber height after several initially smaller chambers. Remarkably, oscillations around the means in our example coincide in the three spirals, ACCEPTED MANUSCRIPT confirming the common growth and response to environmentally induced oscillating factors (Fig. 23B). When it becomes possible to calculate the “chamber building rate” (e.g. Laser ablation techniques for estimating stable isotopes in the septa), then the source of chamber volume oscillations can be checked (e.g. tides, seasonality) and, consequently, the lifespan of Nummulites A- and B-generations can be correctly estimated (see discussions in Purton and PT Brasier, 1999; Evans et al., 2013; Ferràndez-Cañadell, 2012). RI 4. Generalized growth and development SC Cell growth in foraminifera can be reduced to two models, the linear and sigmoidal growth (Fig. 24). For convenience, the sigmoidal growth model is often replaced by an NU exponential model because, until the point of inflection, the sigmoidal growth can be approximated by exponential functions, especially in the initial part (e.g. Hohenegger and MA Briguglio, 2014, Eder et al., 2016). The inflection point is rarely surpassed in foraminifera. Linear cell growth (Fig. 24A) is expressed in equally sized chamber volumes at each growth step (Fig. 24B), in contrast to exponential cell growth (Fig. 24C), which is ED characterized by increasing chamber volumes (Fig. 24D). A special case of exponential growth in chamber volumes is given when the diameter remains constant, for example in EP T tube-shaped tests with constant diameters. In this case, the nominal chamber volumes deviate from spheres, transforming to rotation ellipsoids in which the rotation axis c increases in length during growth (Fig. 24E). AC C Since growth is referred to the cell volume, development describes the changes in cell organization during growth. In thalamous foraminifera, such organizational changes define the way the growing cell is canalized into mono- or polythalamous tests. According to Seilacher (1971), morphology is based on three aspects: the historical-phylogenetic, the functional and the constructional (bauplan) aspect. Following these three aspects, foraminifera have generated a huge amount of possibilities for transforming the above-mentioned growth singularities into tests (Fig. 25). The historical-phylogenetic aspect of test construction as manifested in molecular genetics is documented by the grouping of foraminifera into forms derived from tubes or spheres. This molecular-genetically supported division is the basis of the suborders Tubothalamea and Globothalamea by Pawlowski et al. (2013). The functional aspect in test formation is ACCEPTED MANUSCRIPT discussed in detail by Hottinger (e.g. 1986, 2000). Brasier (1982 a,b, 1986) introduced the MinLOC (minimum line of communication between the proloculus and the next apertures) concept for explaining different test forms, combining constructional and functional aspects. Tyszka erected a morphospace (e.g. Raup and Michelson, 1965; McGhee, 1999) for the development of globular chambers based on the ‘moving reference model’ (Tyszka and Topa, 2005). Different arrangements of spherical chambers during volume growth lead to theoretical solutions, whereby many models have been verified during the evolution of Globothalamea PT (Tyszka, 2006). The Nodosariata (Mikhalevich, 2013), with globular chambers and terminal openings, do not fit the moving reference model because possessing a fixed terminal reference RI point (e.g. Rigaud et al., 2016). Their position within the thalamous foraminifera is thus SC unclear, as documented by the outlier position of the single investigated species Glandulina arctica in the molecular-genetic tree (Pawlowski et al., 2013). NU The main division of thalamous foraminifera in respect to the historical-phylogenetic aspect is between simple forms constructing tubes that are open at both ends (Fig. 25/1) and MA forms based on a sphere with a single aperture (Fig. 25/15). In the former, involving a single tube, growth is divided into steps in which volume growth is interrupted by the test-building process. Thus, any tube is segmented by growth steps, even though these are not perceptible. ED Rectilinear tubes with open ends on both sides (Fig. 25/1) are adapted to a sessile life sticking in soft sediment, anchored by pseudopodial roots and spreading their reticulopods into the EP T water, thus acting as suspension feeders. The transition from fixed to semisessile and vagile individuals with tubes involves the closure of one opening, thus forming the initial test after reproduction. This closure is either simple (e.g. Protobotellina) or the initial test becomes globular (Fig. 25/3) with a wide opening corresponding to the inner diameter of the following AC C tube (e.g. Hyperammina). Such tests are termed ‘bilocular’ because they seem to consist of two chambers – a spherical proloculus and a tubular deuteroloculus. Since the life style of foraminifera with bilocular tests depends on sediment composition, thus indirectly on hydrodynamics, different forms of enrolment are developed on soft sediments. These range from streptospiral (Fig. 25/5) to planispiral (Fig. 25/6) and trochospiral (Fig. 25/7), whereby uncoiled tests can be cemented on firm substrate (Fig. 25/4). Wall texture is of secondary importance in the historical-phylogenetic context because bilocular tests, as the stem group of chambered tubular tests, can have either agglutinated walls fixed with organic or inorganic cement (e.g. Ammodiscus), or test walls that are microgranular (e.g. Eotournayella), porcelaneous (e.g. Cornuspira), hyaline aragonitic (e.g. Involutina) or hyaline calcitic (e.g. Archaediscus, Spirillina). ACCEPTED MANUSCRIPT Generally, it is difficult to differentiate between the inorganic cement of agglutinated foraminifera and the pure microgranular texture because the crystallographic structure (rhomboedric) and size of the calcite crystals are identical. Therefore, the differentiation relies solely on the proportion of agglutinated particles, which can range from 0% to 90% even in the same genus or species (e.g. Orbitolina) without distinct limits within the range (compare Rigaud and Martini, 2015). PT A transition from the microgranular inorganic cement of agglutinated foraminifera to porcelaneous texture is also present, but the separation is more distinct based on RI crystallographic structure (pseudohexagonal), size (needles 1 to 2 mm long) and chemistry (high magnesium calcite content). Again, agglutinated particles are present in a few SC porcelaneous species, although this seems to be environmentally induced because it is not a differential diagnostic trait for the species (e.g. a few agglutinated particles in Peneroplis NU planatus by Rigaud et al. (2016), which are not present in most individuals of this species; personal observation). MA The wide opening of the tube in bilocular tests puts the cell at risk to be attacked from the outside. Therefore, tube openings can become slightly restricted at each growth step, forming pseudosepta (e.g. Parathikinella) leading to real septa bearing small apertures when the ED restrictions are intense. This leads to polythalamous species such as the Paleozoic Earlandinita with short chambers. Derived from these rectilinear multichambered forms, EP T streptospiral (e.g. Dainella) and planispiral enrolment (e.g. Endothyranopsis) opens the pathway to the development of large-sized tests. An optimum surface/volume ratio is advantageous for developing tests acting as microscopic glasshouses (Hohenegger, 1999, AC C 2009) harbouring symbiotic microalgae. This is obtained by developing large spindle-shaped (fusiform) tests, which are extended along the coiling axis (Fig. 26A). This pathway leading to larger benthic foraminifera is restricted to the Late Paleozoic group of fusulinids, whose tests possess short tubular chambers and are characterized by walls with a secreted microgranular texture; they represent the historical-phylogenetic base of this line. A second line, belonging to Tubothalamea, starts with bilocular tests, either stretched, sometimes attached or strepto- to planispirally enrolled (Fig. 25/6). Porcelaneous test walls are the historical-phylogenetic heritage in this line starting in the Late Paleozoic. Beside wall texture, the construction plan differs in chambered forms from fusulinids. Here, the tube is divided by long chambers. The plan starts with planispiral coiling in which the tubular part is divided into chambers differing between 1 ½ to ½ whorl lengths (Eoophthalmidium). Later, ACCEPTED MANUSCRIPT the constancy in half-whorl chambers, where the coiling plane remained at a constant 180° (Fig. 25/8), changes to ~140° steps (see Armstrong and Brasier, 2005, fig. 15.6), becoming the basic bauplan for most representatives (Fig. 25/9). This bauplan can be found in all more highly evolved species, at least in the embryonic part of microspheres (Figs. 25/11-14). Two pathways lead to large-sized tests that house symbiotic microalgae. The first involves elongation of chambers along the coiling axis, leading to fusiform tests similar to large fusulinids (Fig. 25/14). The derivation of these spindle-shaped tests from tubes with long PT chambers is manifested in the few long chambers per whorl (visible in equatorial sections; Fig. 26B). The second strategy to attain large size is by an increased encompassing of planispirally RI arranged chambers leading to annular chambers or the direct arrangement around the SC embryonic apparatus in a circular manner (annular chambers), divided into chamberlets (Fig. 25/13). NU The second line of thalamous foraminifera is represented by forms based on a sphere with a simple aperture (Fig. 25/15). Cell growth in its simplest form is confined by adding MA spherical chambers with increasing volumes to the former test in a rectilinear manner (Figs. 25/16 and 25/21). For the Nodosariata (Mikhalevich, 2013) starting in the Late Paleozoic with rectilinear, ED uniserial tests (Fig. 25/16), the wall structure becomes the main historical-phylogenetic aspect. Agglutinated particles are not present. The test forms are very restricted in variability. Starting EP T from planispiral enrolment (Fig. 25/17), secondarily rectilinear forms (Fig. 25/19) are attained by different steps of uncoiling (Fig. 25/18). Another type of test construction starts from rectilinear uniserial tests, adding chambers such that the chamber arrangement rotates along AC C the growing axis by ~140° (Fig. 25/20). This mode of test construction is present only in Nodosariana and is analogous to trochospiral coiling (represented in Globothalamea, which are also based on spherical tests). The position of forms similar to the Nodosariana but with agglutinated tests (Hormosinana according to Mikhalevich, 2013) is unclear and requires further molecular-genetic investigations. Forms with spherical chambers and agglutinating to hyaline walls are grouped as Globothalamea based on molecular genetics (Pawlowski et al., 2013). They start in the Early Paleozoic with uniserial (Fig. 25/21) and biserial (Fig. 25/22) tests. Throughout the Paleozoic and Early Mesozoic, representatives possess agglutinating tests. Hyaline wall structures start in the Late Triassic, becoming a consistent historical-phylogenetic aspect for representatives with hyaline tests. Independent of wall texture, many test forms have been developed (Figs. ACCEPTED MANUSCRIPT 26/21 to 26/35; see Armstrong and Brasier 2005, fig. 15.6). Planispiral (Fig. 25/23) and flat trochospiral (Fig. 25/24) tests are the basic forms in the test evolution of Globothalamea. The development of inflated chambers (Figs. 25/ 25 and 25/26) facilitates planktonic life. A morphological (possibly evolutionary) line from high trochospiral (Fig. 25/30) to triserial (Fig. 25/31) and biserial chamber arrangement (Fig. 25/32) leads to secondarily uniserial tests (Fig. 25/33), both in genera with hyaline or agglutinated tests. Special chamber arrangements enable fixation to hard substrates, either temporarily (Fig. 25/27) or permanently (Figs. 25/28 PT and 25/29). An important line starting from convolute planispiral forms (Fig. 25/34) leads to the development of large-sized tests by optimizing the surface/volume ratio (see Hohenegger, RI 2009) to house symbiotic algae (Fig. 25/35). Two pathways lead to large-sized tests fulfilling SC the above requirements. The first is by compartmentation of chambers in tests whose margins follow a logarithmic spiral (Fig. 25/35). This leads to cyclic chamber arrangement, when NU chambers, after a short initial spiral, surround the test margin completely (see Fig. 25/13; Hohenegger, 2011, fig. 5). The second pathway to obtain large size with compartments is by adding a series of equally sized chambers during growth, leading to Archimedean spirals as MA described above for the genus Nummulites (Figs. 20, 22; Hohenegger, 2011, fig. 5). ED 5. Conclusion The initiation of growth in foraminifera depends on environmental conditions, i.e. food EP T availability for the propagules. Abiotic (e.g. transport to unfavourable or, conversely, more favourable conditions) and biotic factors (e.g. predation, competition) influence the potential for growing. This makes the mode and timing of reproduction important for growth. AC C Especially the different size of propagules stemming from sexual and asexual reproduction leads to different transport and settlement possibilities for juveniles. Growth in foraminifera with chambered tests is partitioned into steps with continuous cell growth until becoming terminated by chamber wall construction. The timing of chamber construction seems to be genetically determined according to their growth programs, epigenetically influenced by the environmental conditions, especially when propagules start growing. Cell growth can be modelled either by linear or sigmoidal functions, where growth stages are correlated with chamber numbers (Fig. 27). The mostly used Gompertz and Richards functions lead to nearly optimal fits for sigmoidal cell growth, where the point of inflection, determines the growth state ready for reproduction. ACCEPTED MANUSCRIPT The growth rate of the cell as the first derivative of linear or sigmoidal functions defines the chamber volume at the specific growth stage. These volumes decrease behind the inflection point in sigmoidal growth. Some ‘kummerformen’ in planktonic foraminifera are not reactions to unfavorable environmental conditions (as the name implies and was often suspected), but represent the system-immanent growth rate after surpassing the inflection point. PT The timing of chamber construction follows a fixed program being induced by the form of cell growth. In linearly growing cell volumes, the “chamber building rate” corresponds to the RI growth rate (Fig. 27A). Each chamber with constant volumes requires constant time intervals under undisturbed growth. The rate is therefore also constant, becoming the first derivative of SC the linear function of chamber construction dependent on time (Fig. 27C). Correlating the rate of sigmoidal growth with the timing of chamber constructions, the NU intervals between growth steps increase until attaining a maximum value. Using unlimited growth by fitting an exponential function (Fig. 27B), the timing of chamber building can be MA modeled by functions (Fig. 27D) with the chamber-building rate as the first derivative. In limited cell growth characterized by sigmoidal functions, chamber building and the first derivative can best be modeled by the generalized Michaelis-Menten function (Tab. 3). ED The genetically fixed timing of chamber building is strongly correlated with cell growth under undisturbed environmental conditions. Weak negative environmental effects lead to EP T significant deviations of chamber volumes from normal size, but retain the “chamber building rate”. Strong negative effects such as chamber breakage cause different repair mechanisms of growth with altered chamber volumes as well as chamber-building rates. Accelerated rates AC C help attain the fixed, normal chamber size and building rate (compensatory growth). The lifespan of an individual rarely surpasses the time when maximum cell size is reached, but can be shortened by raised favorable environmental conditions. The specific size at which the cell is ready for reproduction varies insignificantly within a species, explaining the more or less homogeneous size of mature individuals. In chambered foraminifera, the known time intervals between chamber constructions help determine the lifespan of an individual under undisturbed environmental conditions. Depending on the chamber-building rate, the lifespan can range from weeks in planktonic foraminifera to years in deep-water foraminifera even though both exhibit the same cell growth correlated with growth stages having similar volumes. ACCEPTED MANUSCRIPT In foraminifera with restricted cell growth, the point of inflection represent maturity where reproduction can start. This estimation is impossible for chambered foraminifera with linear cell growth and constant intervals of chamber construction. In such cases, the timing of reproduction cannot be determined based on the test. Although this growth type is rare in living polythalamous foraminifera, it is common for the fossil larger symbiont-bearing genus Nummulites, which was extremely abundant in the Paleogene. Here, the constant chamber size based on equal time intervals changes stepwise in segments consisting of many chambers to PT approximate sigmoidal cell growth (Figs. 21, 22). No change in the “chamber building rate” due to the altered chamber size is necessary (Fig. 27C). To attain extreme sizes of several RI centimetres in Nummulites B-generations (microspheres), this growth mode involving SC increasing steps of a single spiral becomes insufficient. Therefore, after reaching the maximum chamber size in the initial spiral, additional spirals of the same chamber size are NU added, enabling extreme cell volumes (Figs. 22, 23). When growth slows in extremely large forms (e.g. N. asturicus, N. millecaputi), all spirals simultaneously change to the smaller size, MA keeping the chamber-building rate constant. Chamber volumes deviate from the fitted growth functions over time. These deviations can be instantaneous, caused by short-term external factors (e.g. reduced growth by ED starvation), or they can oscillate around the theoretical growth following periodic environmental changes such as tides, lunar and seasonal cycles. The latter enable estimating EP T lifespans in both living and fossil foraminifera because seasonal changes can be detected in chamber septa using stable isotopes. Development in thalamous foraminifera is understood as the canalization of the growing AC C cell into tests. Three historical-phylogenetically fixed types of test development have been realized in the Tubothalamea (starting from tubular tests) as well as in the Globothalamea and Nodosariana (the latter two based on spherical tests). While multilocular Globothalamea basically construct their chambers following minimization of the local communication path for apertures, the Nodosariana exhibit terminal apertures. Wall textures are of further historical-phylogenetic importance for the Tubothalamea. In that taxon, all textures found in foraminifera are represented, from agglutinated with an organic and inorganic cement to porcelaneous without pores, to hyaline-aragonitic and to hyaline-calcitic textures, both with pores. In Globothalamea, both agglutinated types with organic and inorganic cement have opened historical pathways, whereas hyaline tests with pores constitute a separate historicalphylogenetic group. Beyond the position of the terminal apertures, the Nodosariana are ACCEPTED MANUSCRIPT characterized by highly transparent hyaline tests with low magnesium calcite content and with fine pores. Moreover, they never constructed trochospiral tests. Large-sized tests adapted to house microalgae characterize the Tubothalamea and Globothalamea, but not the Nodosariana. In Tubothalamea, forms with test walls consisting of agglutinating partices fixed with inorganic cement construct either spindle-shaped (Fusulinidae), annular (e.g. Orbitopsella) or conical tests (e.g. Orbitolina). Larger PT foraminifera with porcelaneous walls are restricted to spindle-shaped (Alveolinidae), flat planispiral-embracing (Archaisinidae) or annular tests (Soritidae). Within Globothalamea, RI only forms with hyaline walls developed tests to house microalgae. These tests are either of medium size (a few millimeters) and show trochospiral enrolment (Amphisteginidae, SC Calcarinidae), or they are based on planispiral enrolment, becoming large-sized tests of a few centimeters. There are two strategies for constructing large tests with equally sized NU compartments. The first is by logarithmic chamber growth with the division into chamberlets (e.g. Heterostegina) leading to an annular chamber arrangement (e.g. Cycloclypeus). Adding MA lateral chambers is a feature found in many fossil groups (e.g. Orbitoides and Lepidorbitoides from the Cretaceous, orthophragmiids and Lepidocylcina from the Paleogene). The second strategy involves adding equally sized chambers in spirals, where the sizes increase stepwise ED in segments. Extremely large tests are obtained by adding further spirals during growth (Nummulites). Gigantic tests of more than 10 cm diameter have been developed by both Acknowledgments EP T growth strategies. AC C Primarily I thank the Austrian Science Foundation (FWF) for support in many projects, especially by the FWF grants P23459 ‘Functional Shell Morphology of Larger Benthic Foraminifera’ and P26344-B25 ‘Breakthroughs in Growth Studies on Larger Benthic Foraminifera’. Thanks are due to all coworkers in these projects, namely Antonino Briguglio, Wolfgang Eder, Shunichi Kinoshita, Julia Wöger and Erik Wolfgring. Most investigations on larger foraminifera were done in Japan supported by JSPS, the Sesoko Marine Station of the Ryukyu University Tropical Research Center and the Kagoshima Research Center for the South Pacific. Special thanks are due to my good friend Kimihiko Oki (Kagoshima University Museum). All colleagues of the WOLF-group (Working on Larger Foraminifera) inspired me to think about species, environmental dependence and growth. Thanks are due to Michael Stachowitsch (University of Vienna) for correcting the text as a professional copy editor. The ACCEPTED MANUSCRIPT reviewer Jaroslaw Tyszka and one of two anonymous reviewers were extremely helpful for improving the manuscript. References PT Adl, S., Simpson, A.G.B., Lane, C.E., Lukes, J., Bass, D., Bowser, S.S., Brown, M.W., Burki, F., Dunthorn, M, Hampl, V., Heiss, A., Hoppenrath, M., Lara. E., Le Gall, L., Lynn, RI D.H., McManus, H., Mitchell, E.A.D., Mozley-Stanridge, S.E., Parfrey, L.W., Pawlowski, J., Rueckert, S., Shadwick, L., Schoch, C.L., Smirnov., Spiegel, F.W., SC 2012. The revised Classification of Eukaryotes. Journal of Eukaryotic Microbiology NU 59, 429-493. DOI: 10.1111/j.1550-7408.2012.00644.x Alve, E., Goldstein, S. T., 2003. Propagule transport as a key method of dispersal in benthic MA foraminifera (Protista). Limnology and Oceanography 48, 2163-2170. Angell, R.W., 1967. The process of chamber formation in the foraminifer Rosalina floridana. Journal of Protozoology 14, 341-353. ED Angell, R.W., 1980. Test morphogenesis (chamber formation) in the foraminifer Spiroloculina hyalina Schulze. Journal of Foraminiferal Research 10, 89-101. MA. EP T Armstrong, H.A. and Brasier, M.D., 2005. Microfossils. Second edition. Blackwell, Malden, Arnold, Z.M., 1964. Biological observations of the foraminifer Spiroloculina hyalina Schulze. AC C University of California Publications in Zoology 72, 1-93. Banner, F.T., Pereira, C.P.G., Damini Desai, 1985. “Tretomphaloid” float chambers in the Discorbidae and Cympaloporidae. Journal of Foraminiferal Research 15, 159-174. Bé, A.W.H., Hemleben, C., Anderson, O.R., Spindler, M., Hacunda, J, Tuntivate-Chov, S., 1977. Laboratory and field observations of living planktonic foraminifera. Micropaleontology 23, 155-179. Bender, H., 1992. Chamber formation and biomineralization of Textularia candeiana d’Orbigny (Sarcodina: Textulariina). Journal of Foraminiferal Research 22, 229-241. Berger, W.H., 1969. Kummerform Foraminifera as clues to oceanic environments. Bulletin American Geological Society 53, 706. ACCEPTED MANUSCRIPT Bertalanffy, L., 1938. A quantitative theory of organic growth (inquiries on growth laws. II). Human Biology 10, 181-213. Bertalanffy, L., 1957. Quantitative laws in metabolism and growth. The Quarterly Review of Biology 32, 217-231. Bowser, S.S., Gooday, A.J., Alexander, S.P., Bernhard, J.M., 1995. Larger agglutinated foraminifera of McMurdo Sound, Antarctica: Are Astrammina rara and PT Notodendrodes antarcticos allogromiids incognito? Marine Micropaleontology 26, 7588. RI Bradshaw, J.S., 1957. Laboratory studies on the rate of growth of the foraminifer, “Streblus SC beccarii (Linné) var. tepida (Cushman)”. Journal of Paleontology 31, 1138-1147. Brasier, M.D., 1982a. Architecture and evolution of the foraminiferid test — a theoretical NU approach. In: Banner, E.T., Lord, A.R. (Eds.). Aspects of Micropalaeontology. George Allen & Unwin, London, 1-41. MA Brasier, M.D. 1982b: Foraminiferid architectural history; a review using the MinLOC and PI methods. Journal of Micropalaeontology 1, 95-105. Brasier, M.D., 1986. Form, function and evolution in benthic and planktic foraminiferid test ED architecture..In: Leadbeater, B.S.C., Riding, R. (Eds.). Biomineralisation in Lower Plants and Animals. Systematics Association Special Volume 34, Clarendon Press, EP T Oxford, 32-67. Briguglio, A., Hohenegger, J., 2014. Growth oscillation in larger foraminifera. Paleobiology AC C 40, 494-509. Briguglio, A., Hohenegger, J., Less, G., 2013. Paleobiological applications of threedimensional biometry on larger benthic foraminifera: A new route of discoveries. Journal of Foraminiferal Research 43, 72-87. Cushman, J.A., 1921. Foraminifera of the Philippine and Adjacent Seas. Smithonian Institution, United States National Museum, Bulletin 100, Wahington DC. Dettmering, Ch., Röttger, R., Hohenegger, J., Schmaljohan, R., 1998. The trimorphic life cycle in Foraminifera: Observations from cultures allow new evaluation. European Journal of Protistology 34, 363-368. Eder, W., Briguglio, A., Hohenegger, J., 2016. Growth of Heterostegina depressa under natural and laboratory conditions. Marine Micropaleontology 122, 27-43, ACCEPTED MANUSCRIPT Erez, J., Almogi-Labin, A., Avraham, S., 1991. On the life history of planktonic Foraminifera: Lunar Reproduction cycle in Globigerinoides Sacculifer (Brady). Paleoceanography 6, 295-306. DOI: 10.1029/90PA02731 Evans, D., W. Müller, W., Oron, S., Renema, W., 2013. Eocene seasonality and seawater alkaline earth reconstruction using shallow-dwelling large benthic foraminifera. Earth and Planetary Science Letters 381, 104-115. PT Ferràndez-Cañadell, C., 2012. Multispiral growth in Nummulites. Paleobiological implications. Marine Micropaleontology 96–97, 105–122. RI Geslin, E., Risgaard-Petersen, N., Lombard, F., Metzger, E., Langlet, D. and Jorissen, F., SC 2011. Oxygen respiration rates of benthic foraminifera as measured with oxygen microsensors. Journal of Experimental Marine Biology and Ecology 396, 108-114. NU Gilbert, S.F., 2013. Developmental Biology. 10th edition. Palgrave Macmillan, Basingstoke, UK. MA Goldstein, S.T., 1999. Foraminifera; A biological overview. In: Sen Gupta, B.K. (Ed.) Modern Foraminifera. Kluwer Academic Publishers, Dordrecht, 37-55. Goldstein, S. T., Alve, E., 2011. Experimental assembly of foraminiferal communities from ED coastal propagule banks. Marine Ecology Progress Series 437, 1–11. Gompertz, B., 1832. On the nature of the function expressive of the law of human mortality, EP T and on a new mode of determining the value of life contingencies. Philosophical Transaction of the Royal Society London 123, 513-585. AC C Grell, K.G., 1958. Untersuchungen über die Sexualität und Fortpflanzung der Foraminiferen. II. Rubratella intermedia. Archiv für Protistenkunde 102, 147-164. Grell, K.G., 1973. Protozoology. Springer, Berlin-Heidelberg. Grenander, U., 1959. Probability and Statistics: The Harald Cramér Volume. Wiley, New York. Hallock. 1981. Light dependence in Amphistegina. Journal of Foraminiferal Research 11, 4046. Hammer, Ø., 2017. PAST, PAleontological STatistics, Version 3.16. Reference Manual. ACCEPTED MANUSCRIPT Harney, J.N., Hallock, P., Talge, H.K., 1998. Observations on a trimorphic life cycle in Amphistegina gibbosa populations from the Florida Keys. Journal of Foraminiferal Research 28, 141–147. Hemleben, C., Bé, A.W.H., Anderson, O.R., Tuntivate, S., 1977. Test morphology, organic layers and chamber formation of the planktonic foraminifer Globorotalia menardii (d’Orbigny). Journal of Foraminiferal Research 7, 1-25. PT Hemleben, C., Anderson, O.R., Berthold, W., Spindler, M., 1986. Calcification and chamber formation in Foraminifera – a brief overview. In: Leadbeater, B.S.C., Riding, R. RI (Eds.), Biomineralization in Lower Plants and Animals. Systematics Association, SC Special Volume 30, 237-249. Hemleben, C., Spindler, M., Anderson, O.R., 1988. Modern Planktonic Foraminifera. NU Springer, New York, NY. Hohenegger, J. 1999: Larger foraminifera-microscopical greenhouses indicating shallow- MA water tropical and subtropical environments in the present and past. Kagoshima University Research Center for the Pacific Islands, Occasional Papers 32, 19-45. Hohenegger, J., 2009: Functional shell geometry of symbiont-bearing benthic Foraminifera. ED Galaxea, Journal of Coral Reef Studies 11, 81-89. Hohenegger, J., 2011. Large Foraminifera: Greenhouse Constructions and Gardeners in the EP T Oceanic Microcosm. The Kagoshima University Museum, Kagoshima, Japan. Hohenegger, J., Briguglio, A., 2014: Methods for Estimating Individual Growth of Foraminifera Based on Chamber Volumes. In: Kitazato, H., Bernhard, J.M. (Eds.), AC C Approaches to Study Living Foraminifera: Collection, Maintenance and Experimentation, Environmental Science and Engineering, Springer Japan, 29-54. Hohenegger, J., Briguglio, A., Eder, W., 2014. The Natural Laboratory of Algal SymbiontBearing Benthic Foraminifera: Studying Individual Growth and Population Dynamics in the Sublittoral. In: Kitazato, H., Bernhard, J.M. (Eds.), Approaches to Study Living Foraminifera: Collection, Maintenance and Experimentation, Environmental Science and Engineering, Springer Japan, 13-28. Hottinger, L., 1986: Construction, structure, and function of foraminiferal shells. In: . Leadbeater, B. S. C., Riding, R. (Eds): Biomineralization in lower plants and animals. The Systematics Association, Clarendon Press, Oxford, Special Volume 30, 222–235. ACCEPTED MANUSCRIPT Hottinger, L., 2000. Functional morphology of benthic foraminiferal shells, envelopes of cells beyond measure. Micropaleontology, 46, Supplement 1: Advances in the Biology of Foraminifera, 57-86. Janoschek, A., 1957. Das reaktionskinetische Grundgesetz und seine Beziehungen zum Wachstums- und Ertragsgesetz. Statistische Vierteljahreschrift 10, 25-34. Jonkers, L., Reynolds, C.E., Richey, J., Hall, I.R., 2015. Lunar periodicity in the shell flux of PT planktonic foraminifera in the Gulf of Mexico. Biogeosciences 12, 3061-3070. DOI: 10.5194/bg-12-3061-2015. RI Kinoshita, S., Eder, W., Wöger, J., Hohenegger, J., Briguglio, A., 2017. Growth, “chamber SC building rate” and reproduction time of Palaeonummulites venosus (Foraminifera) under natural conditions. Coral Reefs 36, 1097–1109. NU Koshland Jr, D.E., 2002. The seven pillars of life. Science 295, 2215-2216. Krüger, R. 1994. Untersuchungen zum Entwicklungsgang rezenter Nummulitiden: MA Heterostegina depressa, Nummulites venosus und Cycloclypeus carpenteri. Dissertation, Christian-Albrechts-Universität, Kiel. Krüger, R., Röttger, R., Lietz, R. and Hohenegger, J. 1996. Biology and reproductive ED processes of the larger foraminiferan Cycloclypeus carpenteri (Protozoa, Nummulitidae). Archiv für Protistenkunde 147, 307-321. EP T Langlet, D., Geslin, E., Baal, C., Metzger, E., Lejzerowicz, F., Riedel, B., Zuschin, M., Pawlowski, J., Stachowitsch, M., Jorissen, F. J., 2013. Foraminiferal survival after AC C long-term in situ experimentally induced anoxia. Biogeosciences 10, 7463–7480. Lee, J.J., Bock, W.D., 1976. The importance of feeding in two species of soritid foraminifera with algal symbionts. Bulletin of Marine Science 26, 530-537. Lee, J.J., Freudenthal, H.D., Muller, W.A., Kossoy, V., Pierce, S., Grosman, R., 1963. Growth and physiology of foraminifera in the laboratory: Part 3 – Initial studies of Rosalina floridana (Cushman). Micropaleontology 9, 449-466. Lee, J.J, Faber, W.W., Anderson, O.R., Pawlowski, J., 1991. Life cycles of foraminifera. In: Lee, J.J, Anderson, O.R. (Eds.). Biology of Foraminifera. Academic Press, London, UK. ACCEPTED MANUSCRIPT Lehmann, G., Röttger, R., Hohenegger, J., 2006. Life Cycle variation including trimorphism in the foraminifer Trochammina inflata from North European salt marshes. Journal of Foraminiferal Research 36, 279-290. Lomb, N. R. 1976. Least-squares frequency analysis of unequally spaced data. Astrophysics and Space Science 39, 447–462. Lombard, F., Labeyrie, L., Michel, E., Spero, H.S., Lea, D.W., 2009. Modelling the PT temperature dependent growth rates of planktic foraminifera. Marine Micropaleontology 70, 1-7. RI López, S., France, J., Gerrits, W.., Dhanoa, M., Humphries, D., Dijkstra, J., 2000. A SC generalized Michelis-Menten equation for the analysis of growth. Journal of Animal Science 78, 1816-1828. NU Lutze, G., Wefer, G., 1980. Habitat and asexual reproduction of Cyclorbiculina compressa (Orbigny) Soritidae. Journal of Foraminiferal Research 10, 251-260. MA Malthus, T.R., 1798. An Essay on the Principles of Populations. J. Johnson in St. Paul’s Church-yard, London. McGhee, Jr., G.R. 1999. Theoretical Morphology. The Concept and its Application. ED Perspectives in Paleobiology and Earth History, Columbia University Press, New York,. EP T Mikhalevich, V.I., 2013. New insight into the systematics and evolution of the foraminifera Micropaleontology 59, 493–527. AC C Olsson, R.K., 1973. What is a kummerform planktonic foraminifer? Journal of Paleontology, 327-329. Panik, M.J., 2014. Growth Curve Modelling. Theory and Applications. Wiley and Sons, Hoboken, NJ. Pavlovec, R., 1987. Svetovni rekord med foraminiferami. Proteus 49, 167–169 (in Slovenian, English Abstr. at http://www.progeo.se/news/98_1f/foramin.html). Pawlowski, J., Holzmann, M., Tyszka, J., 2013. New supraordinal classification of Foraminifera: Molecules meet morphology. Marine Micropaleontology 100, 1–10. http://dx.doi.org/10.1016/j.marmicro.2013.04.002 Pokorny, V., 1958. Grundzüge der Zoologischen Mikropaläontologie 1, VEB Deutscher Verlag der Wissenschaften, Berlin. ACCEPTED MANUSCRIPT Purton, L.M.A., Brasier, M.D., 1999. Giant protist Nummulites and its Eocene environment: life span and habitat insights from δ18O and δ13C data from Nummulites and Venericardia, Hampshire basin, UK. Geology 27, 673–768. Raup, D.M., Michelson, A. 1965: Theoretical morphology of the coiled shell. Science, 147, 1294–1295. Reece, J.B., Taylor, M.R., Simon, E.J., Dickey, J.L., 2012. Campbell Biology (seventh PT edition), Pearson, Boston, MA. Reichel, M., 1936. Étude sur les Alvéolines. Schweizerische Paläontologische Abhandlungen RI 57, 1-93. SC Richards, F.J., 1959. A flexible growth function for empirical use. Journal of Experimental Botany 10, 290-301. NU Rigaud, S., Martini, R., 2016. Agglutinated or porcelaneous tests: Where to draw the line? Journal of Foraminiferal Research v. 46, no. 3, p. 333–344, July 2016 MA Rigaud, S., Vachard, D., Martini, R., 2015. Agglutinated versus microgranular foraminifers: end of a paradigm?, Journal of Systematic Palaeontology 13:2, 75-95, DOI: 10.1080/14772019.2013.863232 ED Rigaud, S., Vachard, D., Schlagintweit, F., Martini, R., 2016. New lineage of Triassic aragonitic Foraminifera and reassessment of the class Nodosariata, Journal of EP T Systematic Palaeontology 14:11, 919-938, DOI: 10.1080/14772019.2015.1112846. Roettger, R., 1972. Analyse von Wachstumskurven von Heterostegina depressa AC C (Foraminifera: Nummulitidae). Marine Biology 17, 228-242. Roettger, R., 1974. Larger Foraminifera: Reproduction and early stages of development in Heterostegina depressa. Marine Biology 26, 5-12. Röttger, R., 1976. Ecological observations of Heterostegina depressa (Foraminifera, Nummulitidae) in the laboratory and in its natural habitat. Maritime Sediments, Special Publication 1, 75-79. Röttger, R., 1981. Die Großforaminifere Heterostegina depressa. Organisation und Wachstum der megalosphärischen Generation. C 1541 Institut für den wissenschaftlichen Film, Göttingen. Roettger, R., Spindler, M., 1976. Development of Heterostegina depressa individuals (Foraminifera, Nummulitidae) in laboratory cultures. 1st International Symposium on ACCEPTED MANUSCRIPT Benthonic Foraminifera of Continental Margins. Part A. Ecology and Biology. Maritime Sediments, Special Publication 1, 81-87. Ross, B.J., Hallock, P., 2016. Dormancy in the Foraminifera: A review. Journal of Foraminiferal Research 46, 358-368. Schiebel, R., Hemleben, C., 2017. Planktic Foraminifers in the Modern Ocean. Springer Verlag, Berlin-Heidelberg. PT Schmidt, C., Heinz, P., Kucera, M., Uthicke, S., 2011. Temperature-induced stress leads to bleaching in larger benthic foraminifera hosting endosymbiotic diatoms. Limnology RI and Oceanography 56(5), 1587–1602. doi:10.4319/lo.2011.56.5.1587. SC Schulz, M., Mudelsee, M., 2002. REDFIT: estimating red-noise spectra directly from unevenly spaced paleoclimatic time series. Computers & Geosciences 28, 421-426. NU Seilacher, A. 1971. Arbeitskonzept zur Konstruktions-Morphologie. Lethaia 3), 393–396. Spindler, M., Anderson, O.R., Hemleben, C., Bé, A.W.H., 1978. Light and electron MA microscopic observations of gametogenesis in Hastigerina pelagica (Foraminifera). Journal of Protozoology 427–433. DOI: 10.1111/j.1550-7408.1978.tb04164.x Ter Kuile, B., Erez, J., 1984. In situ growth-rate experiments on the symbiont bearing ED foraminifera Amphistegina lobifera and Amphisorus hemprichii. Journal of EP T Foraminiferal Research 14, 262-276. Triantaphyllou, M.V., Dimiza, M.D., Koukousioura, O., Hallock, P., 2012. Observations on the life cycle of the symbiont-bearing foraminifera Amphistegina lobifera Larsen, an AC C invasive species in coastal ecosystems of the Aegean Sea (Greece, E. Mediterranean). Journal of Foraminiferal Research 42, 143-150. Tyszka, J., 2006. Morphospace of foraminiferal shells: results from the moving reference model. Lethaia. 39), 1–12. Oslo. Tyszka, J., Topa, P., 2005. A new approach to modeling of foraminiferal shells, Paleobiology 31, 522–537. Uthike, S., Althenrath, C., 2010. Water column nutrients control growth and C:N ratios of symbiont bearing benthic foraminifera on the Great Barrier Reef, Australia. Limnology Oceanography 55, 1681-1696. Wefer, G., Berger, W.H., 1991. Isotope Paleontology: growth and composition of extant calcareous species. Marine Geology 100, 207-248. ACCEPTED MANUSCRIPT Wolpert, L., Tickle, C., Ariaz, A.M., 2015. Principles of Development (5th edition). Oxford University Press, Oxford, UK. Zohary, T., Reiss, Z., Hottinger, L., 1980. Population dynamics of Amphisorus hemprichii (Foraminifera) in the Gulf of Elat (Aqaba), Red Sea. Eclogae Geolicae Helveticae 73, PT 1071–109. RI Figure descriptions Fig. 1 Alternation of generations in the larger symbiont-bearing foraminifer Heterostegina SC depressa (Hohenegger 2011, modified by Wolfgang Eder). Fig. 2 MicroCT micrographs of Palaeonummulites venosus. A. Virtual equatorial section. B. NU Virtual axial section. C. Virtual chamber volumes at growth steps determined by whorl number. Micro-CT micrographs by Shuinichi Kinoshita. MA Fig. 3 Cell and chamber growth of Palaeonummulites venosus fitted by growth functions represented in Table 1. A. Cell growth. B. Cell growth up to 15 chambers. C. Chamber ED growth. D. Chamber growth of the nepiont and the 15 following chambers. Fig. 4 Volume growth of the smaller benthic Ichhtyolaria sulcata from the Early Jurassic. A. EP T Micrograph showing chamber size decrease after the inflection point. B. Growth of cell volume fitted by Gompertz function, inflection point located at chamber number 9. C. Growth of chamber volume, fitted by the first derivative of the Gompertz function. D. Graphical AC C reconstruction of theoretical growth based on the morphogenetic growth program following a logistic function for chamber volumes, an exponential function for distances between centers measured at the maximum chamber widths and a complex program for the development of ribs (after Hohenegger 1987). Fig. 5 Cell and chamber growth of an agamont and gamont of Heterostegina depressa fitted by Richards functions. Fig. 6 Deviations from normal growth in final chambers of agamonts. A. Brood chambers in Calacarina calcar with hatched gamonts/schizonts. B. Reproduction chambers in Amphisorus hemprichii with hatched gamonts/schizonts. C. Construction of a float chamber in Cymbaloporetta cifelli (after Banner et al. 1985). ACCEPTED MANUSCRIPT Fig. 7 Cell and chamber growth of a P. venosus specimen kept in culture after collection from the natural habitat. A. Chamber volumes grown under natural conditions. B. Chamber volumes added under laboratory conditions. C. Fit of cell volumes, constructed in the natural environment, by Gompertz function. D. Fit of chamber volumes, constructed in the natural environment, by Gompertz function (data based on Micro-CT graphs by Shuinichi Kinoshita). Fig. 8 Dependence of chamber formation on time under undisturbed environmental PT conditions. A. Linear dependence versus decelerated chamber formations, modeled by power function. B. Constant time intervals between chamber formations in linear dependence versus RI increasing time intervals between chamber formations in nonlinear dependence. Fig. 9 Metabolism and “chamber building rate” in a specimen of Palaeonummulites venosus. SC A. Oxygen respiration rate during growth. B. Mass-specific metabolic rate during growth based on oxygen respiration. C. Averaged “chamber building rate” for P. venosus. D. NU Correlation between mass-specific metabolic rate and “chamber building rate”. Fig. 10 Virtual equatorial section of an Operculina complanata specimen showing breakage MA of the regularly grown test (in blue), followed by irregular growth to recover regular test shape (in green) by building irregular chamber volumes (in brown) or accelerating the ED “chamber building rate” (in yellow); MicroCT graph by Julia Wöger. Fig. 11 Virtual axial section of Heterostegina depressa showing the involute central part and EP T the evolute final part; MicroCT graph by Wolfgang Eder. Fig. 12 Chamber formation in seven individuals from ‘Versuchstiergruppe 1’ of Röttger (1972) and fit of the sum of chamber formations by growth functions. AC C Fig. 13 Averaged “chamber building rate”s within 2 weeks for selected individuals of Heterostegina depressa from the ‘Versuchstiergruppe 1’ of Röttger (1972) based on different growth functions. Fig. 14 Density diagrams of chamber numbers in Peneroplis antillarum from Sesoko Island, Okinawa, Japan. The maximum chamber number defined as the arithmetic mean plus 3 times the standard deviation (from Hohenegger et al., 2014; modified). Fig. 15 Determination of the “chamber building rate” of Peneroplis antillarum using the Michaelis-Menten function based on the maximum chamber numbers (from Hohenegger et al., 2014; modified). Fig. 16 Paleonummulites venosus growth of cell volume (cubic roots) over time fitted by Richards function. ACCEPTED MANUSCRIPT Fig. 17 Decomposition of standardized residuals of chamber volumes in P. venosus into oscillating sinusoidal functions. A. Standardized residuals (3rd root). B. Power spectrum based on Lomb algorithm. C. Power spectrum based on REDFIT analysis. D. Parameters of the main sinusoids gained by the above analyses ordered by their importance. Fig. 18 Standardized residuals of chamber volumes of P. venosus based the Richards growth function using a natural chamber-building rate. A. Observed residuals and fit by spectral PT analyses and sinusoidal regression. B. Sinusoids fitting standardized residuals using the REDFIT spectral analysis with periods of 13.4, 34.7 and 7.9 days. RI Fig. 19 Ammodiscus incertus with planspiral tests following an Archimedean spiral. A. Gamont (megalosphere) with constant tube diameter. B. Agamont (microsphere) with sections SC of different tube diameters (from Cushman, 1921; modified). Fig. 20 Chamber lumina of a Nummulites fabianii megalosphere (A-generation) from the NU Late Eocene in equatorial view; spiral sections with constant chamber height in different colours; modified MicroCT-image from Briguglio et al. 2013. MA Fig. 21 Growth of Nummulites fabianii from the Late Eocene. A. Cell growth fitted by Richards function (Tab. 2). B. Cell growth fitted by stepwise changing functions. C. Chamber ED growth fitted by Richards function. D. Chamber growth fitted by stepwise changing constant functions. EP T Fig. 22 Chamber lumina of a Nummulites aturicus microsphere (B-generation) from the Middle Eocene in equatorial view. Multiple spirals in different colours, where the onset of secondary spirals are marked in relation to the growth status of the first spiral. MicroCT- AC C image from Erik Wolfgring. Fig. 23 Growth of Nummulites asturicus from the Middle Eocene. A. Cell growth fitted by three Archimedean spirals. B. Chamber growth in steps of the first spiral and constant mean chamber height in additional spirals. Fig. 24 Generalized growth models representing the nominal volumes as spheres. A. Cell volumes in linear growth steps. B. Chamber volumes in linear growth steps. C. Cell volumes in exponential growth steps. D. Chamber volumes in exponential growth steps. E. Chamber volumes in exponential growth steps with constant diameters leading to rotation ellipsoids with increasing lengths of the rotation axis. Fig. 25 Generalized pathways in the development of foraminiferal tests (independent of wall structure) and their lifestyle. Tubothalamea: 1. Bathysiphon, 2. Rhizammina, 3. ACCEPTED MANUSCRIPT Hyperammina, 4. Tolypammina, 5. Glomospira, 6. Ammodiscus (agglutinated), Cornuspira (porcelaneous), 7. Arenoturrispirillina, 8. Ophthalmidium, 9. Quinqueloculina, 10. Articulina, 11. Coscinospira, 12. Peneroplis, 13. Parasorites, 14. Borelis. Nodosariata: 16. Nodosaria, 17. Lenticulina, 18. Astacolus, 19. Frondicularia, 20. Polymorphina. Globothalamea: 15. Saccammina, 21. Reophax, 22. Textularia, 23. Haplophragmoides, 24. Ammonia, 25. Globigerina, 26. Orbulina, 27. Cibicides, 28. Planorbulina, 29. Rupertina, 30. Sitella, 31. Bulimina, 32. Bolivina, 33. Sipohogenerina, 34. Elphidium, 35. Heterostegina, (after PT Hohenegger, 2011; modified). RI Fig. 26 Spindle-shaped (fusiform) tests that yield an optimum surface/volume ratio. A. Fusulinidae with microgranular walls; B. Alveolinidae with porcelaneous walls. The SC differences in chamber number per half-whorl are marked. Modified after Pokorny (1958) and Reichel (1936). NU Fig. 27 Time of chamber building induced by cell growth; growth steps in blue. A. Cell growth based on constant time intervals. B. Cell growth based on increasing time intervals MA correlated with volume growth. C. Chamber building in linear growth. D. Chamber building AC C EP T ED in exponential growth. ACCEPTED MANUSCRIPT Tables Tab. 1 Commonly used growth functions based on chamber number (Panik, 2014) Growth function Cell growth Vlinearized Chamber growth V’linearized Inflection point n Logistic PT Gompertz MorganMercer-Flodin ( AC C EP T ED ) MA Janoschek NU [ SC RI Richards ] ( ) ( ) ACCEPTED MANUSCRIPT Tab. 2 Sums of standardized residuals of growth functions using the cell and chamber volumes of Palaeonummulites venosus. Sum of standardized residuals Volume Logistic Gompertz Richards MMF Janoschek 5.74 1.65 0.70 6.13 4.38 Chamber 5.22 3.92 3.28 10.51 8.44 AC C EP T ED MA NU SC RI PT Cell ACCEPTED MANUSCRIPT Tab. 3 Functions used for chamber-building rates in foraminifera and their inverse for estimating individual’s lifespan. Function Inverse Function Power ( ) Formation Rate ⁄ ⁄ PT Name Michaelis- Generalized ( AC C EP T ED ) ( MA Menten ( ⁄ NU Michaelis- Bertalanffy SC RI Menten ) ) ⁄ ⁄ ACCEPTED MANUSCRIPT Tab. 4 Chi-Square statistics fitting chamber formations in selected individuals from the ‘Versuchstiergruppe 1’ of Röttger (1972) by different functions. Chi-square df Individual 1 Michaelis Menten Generalized MM Bertalanffy 9 0.21 0.34 0.18 0.16 Individual 8 18 1.03 5.09 1.22 0.34 Individual 13 13 0.18 0.78 0.17 0.12 Individual 14 13 5.72 0.63 0.33 0.22 Individual 21 11 0.71 1.72 0.72 0.40 Individual 29 10 0.10 0.48 0.10 0.15 Individual 31 14 0.39 2.08 0.41 0.10 AC C EP T ED MA NU SC PT Power RI Specimens in Röttger 1972 Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27