Continuous reproduction of planktonic foraminifera in laboratory culture

Planktonic foraminifera were long considered obligate sexual outbreeders but recent observations have shown that nonspinose species can reproduce by multiple fission. The frequency of multiple fission appears low but the survival rate of the offspring is high and specimens approaching fission can be distinguished. We made use of this observation and established a culturing protocol aimed at enhancing the detection and frequency of fission. Using this protocol, we selectively cultured specimens of Neogloboquadrina pachyderma and raised the frequency of reproduction by fission in culture from 3% in randomly selected specimens to almost 60%. By feeding the resulting offspring different strains of live diatoms, we obtained a thriving offspring population and during the subsequent 6 months of culturing, we observed two more successive generations produced by fission. This provides evidence that in nonspinose species of planktonic foraminifera, reproduction by multiple fission is likely clonal and corresponds to the schizont phase known from benthic foraminifera. We subsequently tested if a similar culturing strategy could be applied to Globigerinita glutinata, representing a different clade of planktonic foraminifera, and we were indeed able to obtain offspring via multiple fission in this species. This work opens new avenues for laboratory‐based experimental work with planktonic foraminifera.

is particularly well documented.Tremendous progress in our understanding of their lifestyle and reproduction behaviors were made from the early 1970s, when a group of researchers joined forces to rear planktonic foraminifera in laboratory conditions (Bé et al., 1977;Hemleben et al., 1989).It led to the observation of gametogenesis in multiple species from different genera and allowed to characterize behaviors that are now known to precede a reproductive event, such as symbionts bleaching and spines shredding in spinose species (e.g.Adshead, 1967;Be & Anderson, 1976;Erez et al., 1991).Such studies also revealed the possibility of synchronization in the gamete release that would, in the environment, enhance chances for gamete encounters (e.g.Spindler et al., 1979).Advances in cellular biology also allowed the characterization of multiple parts of planktonic foraminifera cell composition and also led to a better understanding of their trophic (e.g.feeding) and symbiotic strategies (Anderson & Bé, 1976;Bé et al., 1981Bé et al., , 1983;;Caron et al., 1982;Hemleben et al., 1989;Spindler et al., 1979).To date, the complete life cycle of planktonic foraminifera (from the production of a zygote or offspring to their own reproduction and obtention of viable successive generations) has not been reached in laboratory conditions and most of our hypotheses for it are based on benthic species.
Although most of the evidence comes from observations of a handful of model benthic species, researchers have developed generalized concepts of the foraminifera life cycle based on these observations (Dettmering et al., 1998;Goldstein, 1999;Lipps, 1982).The schematic life cycle has been refined and amended by various authors over the years (e.g.Goetz et al., 2022) and the current concept comprises an alternation of haploid and diploid generations, with a possibility to maintain multiple generations of schizogony.The cells releasing gametes are named "gamonts" and are considered haploid.The cells resulting from karyogamy are considered diploid and referred to as "agamonts" (Goldstein, 1999 and references therein).The agamonts have been suggested to reproduce through nuclei replication (multiple mitosis) followed by meiosis and multiple fission resulting in haploid, uninucleate offspring, representing the gamont.This mode of reproduction has been often referred to as asexual, highlighting the fact that it occurs without the exchange of genetic material with other individuals.In some cases, an agamont has been observed to produce offspring that presumably remains diploid and also reproduces by multiple fission and therefore does not correspond to a gamont.This results in successive "asexual" generations, with the offspring termed "schizonts" (Rhumbler, 1911), and the process itself called "schizogony." However, observations of chromosomes to determine ploidy in foraminifera are exceedingly rare (e.g.Röttger et al., 1989) and the naming of a specimen as a gamont, agamont, or schizont relies on observation of their reproductive pathway.A specimen is called a gamont if it releases gametes, an agamont if it resulted from gamete fusion and underwent multiple fission, and a schizont if it underwent successive phases of multiple fission.In this concept, the schizont phase is the only one leading to a second generation of reproduction via multiple fission and is the only one that could lead to the production of strictly clonal generations (no recombination, no change in ploidy).
In benthic foraminifera, dimorphism in the size of the proloculus is often used as a substitute for the determination of ploidy and to distinguish between specimens produced via karyogamy or multiple fission (Lipps, 1982;Myers, 1943).Such a dimorphism in the proloculus size of planktonic foraminifera had not been clearly established until recently (Meilland et al., 2023) and planktonic species were thought to reproduce exclusively via the release of gametes (observed in multiple species, e.g.Anderson & Bé, 1976;Erez et al., 1991) followed by karyogamy (thus far unobserved), assuming obligate outbreeding because autogamy has never been documented.Multiple fission (often referred to as "asexual reproduction" in foraminifera literature) was recently observed in different groups of planktonic foraminifera (Neogloboquadrina pachyderma, Kimoto and Tsuchiya (2006), Davis et al. (2020), Meilland et al. (2023); Globigerinita uvula, Takagi et al. (2020)) and confirms that like their benthic ancestors, planktonic foraminifera life cycles probably have a biphasic basis (Darling et al., 2023).
To establish permanent planktonic foraminifera cultures, it is crucial that specimens reproduce and that their offspring reach maturity and reproduce again.
Lately, Meilland et al. (2023) described, for the first time, multiple observations of multiple fission for Neogloboquadrina pachyderma and hypothesize that despite being rare, the production of offspring via cellular division (likely following multiple fission) is crucial for the success of planktonic foraminifera in polar regions.
These observations raised hope of a possibility to produce planktonic foraminifera clones in culture and to grow them entirely under controlled conditions as has been done for benthic foraminifera (e.g. de Nooijer et al., 2014).
With this study, we aim to provide new insights into planktonic foraminifera life cycle and determine if it includes a schizont phase.Is the mode of reproduction in individuals predetermined from the moment they are spawned or established during their early developmental stages?Does it impact their morphology such that it could lead to identifiable discernible traits?To do so, we build upon our previous success with the polar species N. pachyderma (Meilland et al., 2023) which we henceforth use as a model to establish permanent culture in the laboratory.We provide a detailed sampling and culture procedure and we introduce a way to identify specimens prone to fission using traits identified in previous observations of reproduction via multiple fission (Meilland et al., 2023).Applying this strategy to specimens of N. pachyderma sampled in the Baffin Bay in September 2022 and to specimens of G. glutinata sampled in the Gulf of Eilat in March 2023, we successfully observed a total of 31 reproductive events in culture (29 for N. pachyderma and 2 for G. glutinata).Among these, second and third generations of N. pachyderma were spawned in the laboratory in Germany, validating the existence of a schizont phase in their life cycle and opening new possibilities in the field of foraminifera research, paleoceanography but also in the study of protist physiology.

M AT ER I A L S A N D M ET HOD S
For this study, 312 specimens of Neogloboquadrina pachyderma were sampled at six stations located in the Irminger Sea, the Labrador Sea, and the Baffin Bay (Figure S1; Table 1) during the expedition MSM111 onboard the R/V Maria S. Merian in Autumn 2022.In addition, 130 specimens of Globigerina glutinata were collected in the Gulf of Aqaba during a 4-day campaign in Spring 2023 at the Inter University Institute for Marine Sciences in Eilat, Israel.All specimens were picked and placed in culture following an optimized protocol described hereafter.
For the N. pachyderma material, stations 25,203 to 25,205 were characterized by temperatures ranging from 10.4 to 4.5°C in the first 100 m depth, a maximum chlorophyll-a (Chl-a) concentration between 3.5 and 2 mg/m 3 at the base of the mixed layer depths between 20 and 35 m.Stations 25,207 and 25,208 are more "transitional" with lower temperatures between 6 and 3.2°C, lower salinities between 33.4 and 34.4, and lower Chl-a (maxima between 0.6 and 2 mg/m 3 ) distributed more homogeneously in the water column.Station 25,209 and 25,210 are both located in the Baffin Bay and characterized by low temperatures between 1.5 and −1.5°C, low surface salinities of just above 30 and a not very deep but pronounced Chla maximum reaching values of up to 5.1 mg/m 3 in depths of about 24 m just below a pycnocline caused by a steep reduction in water temperature (Siccha & Kucera, 2017).
For G. glutinata, all samples were collected in the Red Sea using a WP2 net (100 μm mesh size) deployed vertically from 80 m depth to the surface at station A (Chernihovsky et al., 2023)

Optimized ship-based and marine station sampling and culture procedure for the observation of laboratory reproduction and offspring growth
The procedure we describe hereafter is the one we optimized for the observation of reproductive events and applied for both the specimens of Globigerinita glutinata and of Neogloboquadrina pachyderma (to which we also applied some "selecting picking," see next section).While based on methods previously reported in the literature (Bé et al., 1977;Hemleben et al., 1989;Huber et al., 1996;Spero, 1992), it has been standardized and adjusted such that one can collect, keep, and monitor a large number of planktonic foraminifera specimens sampled during an oceanographic research cruise or while running a culture campaign at a marine station with access to a small research boat (Figure S2).All the necessary equipment to use for the hereafter described sampling and culture procedure are detailed and provided in the Supplementary Material (Table S1).Specifically, in this procedure, all sampled specimens are placed and cultured in 12-well culture plates and not in sampling jars.
The procedure is designed such that it can be completed entirely by just two to three persons (e.g. one filtering (see 1) while one or more are selecting and isolating specimens (see 3)).It consists of the following steps (Figure 1)., optional) are deployed below the expected living depth of the targeted species in order to collect water for the culture and eventually fully characterize the environmental parameters of the inculture population initial habitat.The water should be collected according to the expected depth habitat of the targeted species (based on the literature) and/or taking advantage of the parameters recorded during the vertical downward profile of the optional CTD when the latter is readable immediately onboard.While optional, obtaining the downward profile can be useful to readjust the sampling depth to the known preferential habitat of the species.As an example, for N. pachyderma and G. glutinata, water was collected between the depth of the chlorophyll maximum and of the surface mixed layer.Once collected, the water is immediately filtered at 0.2 μm to remove most of the living communities from it.The water is then transferred to clean canisters or bottles and kept at in-situ temperature (or as close as possible).Then, 12-wells culture plates are prefilled with the filtered water and also kept at in situ temperature prior to the foraminifera collection.2. Collection of planktonic foraminifera.Planktonic foraminifera are collected at the depth(s) of their expected habitat based on the literature and using a plankton tow equipped with a mesh size sufficiently small to collect the species of interest (here, 100 μm mesh size).The towing speed should remain below 0.4 m/s for a gentle collection.When sampling in waters with high chlorophyll concentrations, the towing intervals should be adjusted to avoid net clogging (small intervals if high chlorophyll concentration) that would increase stress and also slow the recovery of the single foraminifera.3. Isolation of planktonic foraminifera.Specimens are isolated as quickly as possible to have higher chances of survival in culture (ideally within about 2 h after their collection), especially if the picking occurs at a temperature drastically different (>10°C) than the in situ water temperature.If collected with a multinet, samples should be treated one after the other starting with the one containing the apparent higher biomass to the one containing the least, unless a sufficient number of persons is available to go through them simultaneously.
To avoid predation and/or oxygen depletion (high concentration of plankton cells in a small volume) all samples to be processed are diluted into a 5-to 10-L bucket previously filled with the filtered seawater (see 1) and kept at the in situ temperature until being treated.The sample to be treated is split and transferred into glass crystallizers (250 mL), topped up with 0.2μm filtered in situ seawater such that a ratio of 1 to 4 (sample to filtered seawater) is obtained.Glass crystallizers are kept at the in situ temperature during manipulation (e.g.placed on ice when working with a polar species such as N. pachyderma).Live specimens of the species to culture are picked gently under a binocular and gently cleaned using a brush.Individuals of all sizes and with abundant and colorful cytoplasm should be preferentially selected as not only large specimens can be mature (at a stage at which reproduction could occur).Each selected specimen is transferred into a petri dish Flowchart of the six main steps of the sampling and culture procedure, from the seawater collection at the depth and site at which the plankton tow will be deployed (see 1 and 2 in the text) to the transfer of the produced offspring in 70-mL culture jars previously filled with filtered seawater.
filled with 0.2μm filtered in situ seawater for a first cleaning bath and then into a second petri dish for a second bath to remove particles attached to the shell.The water of the cleaning baths should be changed regularly.After the two consecutive baths, specimens are placed individually in a well of a previously prepared culture plate (see 1 and Figure S2).Once all wells are filled with one planktonic foraminifera, the plates are placed in their culture setting (culture room with controlled temperature or incubator) at the in situ temperature and with a light cycle and light intensity mimicking the one of the environment.All specimens are left to rest for 24 h for recovery, without being fed. 4. Feeding of the in-culture specimens.After 24 h of rest, each specimen is fed by pipetting live diatoms directly above it, in the wells.Diatom strains are selected based on their likelihood of occurrence in the environment of the target species and diatom cells should be smaller than the planktonic foraminifera in culture.Select a strain that can survive at the chosen culture temperature.The number of diatoms to add to the well should be small (about 100 cells per well).The diatoms will grow in the wells while the foraminifera will feed on them, ensuring constant live food availability without any further manipulation.Transferring too many cells to the culture wells can result in a too large a diatom population that will alter the chemical composition of the water and eventually negatively impact the foraminifera.A diatom aliquot should be kept for the entire duration of the culture, with the in situ filtered seawater (see 1) in case a specimen needs a restock.It is important to maintain the diatom culture in filtered seawater and not in a culture medium that can contain antibiotics.This could negatively impact the foraminifera in culture. 5. Monitoring.To detect a reproductive event in a timely manner, every specimen is monitored twice a day with an inverted microscope, allowing observation of the rhizopodial deployment and general cell activity.We used a Zeiss, PrimoVert, and an AxioVert inverted microscopes equipped with objectives 4X, to 40X and a camera.The camera is connected to a computer in order to allow a thorough follow-up of all specimens.Specimens should not be exposed to the microscope light longer than a couple of minutes at a time as it disturbs not only the lighting rhythm of the specimen but also often significantly increases water temperature in the wells.The high frequency of the specimen's monitoring is important to quickly notice the occurrence of a reproductive event by fission.It also facilitates the actual detection of such an event, as the offspring will not have had much time to disperse and can therefore still be found close to the empty parental shell in a condensed cloud (see "Results" section).Monitoring twice a day also enables the imaging of potential parents before their reproductive event, allowing the characterization of traits indicative of a reproductive event in preparation and which can be used subsequently to select potential parents from the plankton collection (see following section for N. pachyderma).6. Transfer of the offspring and monitoring.Once a reproductive event is noticed and completed (offspring clearly separated from one another and starting to disperse further away from the parental shell), the entire content of the well (offspring, empty parental shell, water, and diatoms) is transferred with a sterile 5-mL plastic pipette to a 70-mL culture flask, previously filled with the filtered seawater.Due to cannibalism (Meilland et al., 2023), the survival rate of the offspring is higher in a larger volume of water.We thus recommend to transfer them as soon as possible.
A large amount of food (three times what is described in 4, so ~300 cells) is added immediately to the 70-mL flask and the offspring are left to rest for 24 h.Then, the flask is monitored just once per day (to not stress the offspring with the intense light of the inverted microscope) to track the growth/chamber addition of the offspring and their food uptake.If they do not consume the diatoms available, a different strain of a smaller cell size should be added to the flask.

Neogloboquadrina pachyderma
At stations 25,204, 25,205, 25,207, and 25,208, the abundance of N. pachyderma was relatively low (no counts available) and in total, 192 specimens were selected solely based on the observation of cytoplasm in all chambers (see previous section).We will refer to these specimens as the ones "randomly picked" for N. pachyderma in the manuscript.
At stations 25,209 and 25,210, higher concentrations of N. pachyderma were found, allowing us to select a total of 120 specimens based on the presence of intense red cytoplasm in all chambers, their small size, and the high sphericity of their individual chambers and the thinness and transparency of the overall shell.These traits had been identified in the shells of the specimens that had reproduced via multiple fission in Meilland et al. (2023).These specimens will be referred to as the "selectively picked" in the manuscript.
To assess if specimens with the "aberrant" right coiling direction of N. pachyderma (Darling et al., 2006) are more likely to reproduce via multiple fission, as suggested in Meilland et al. (2023), right coiling specimens of N. pachyderma were intentionally picked and included in both the randomly picked ones (86 of 192) and in the selectively picked ones (60 of 120) using the same criteria.
All sampled and isolated specimens were cultured in the cold room of the research vessel (5-6°C) and in filtered seawater collected at three stations before the MPS deployment (Table 1).Once placed in culture, all specimens were fed once with Chaetoceros debilis, a diatom, and then monitored twice a day with an inverted microscope equipped with a camera (PrimoVert, Zeiss).Size measurements of specimens of proloculi were obtained using the ImageJ software (Abràmoff et al., 2004).

Feeding experiment for the offspring of Neogloboquadrina pachyderma produced via multiple fission onboard and in the laboratory
After each event of multiple fission, all spawned offspring of N. pachyderma were transferred to 70-mL culture flasks following the previously described procedure, and labeled with their parent individuals label (e.g.M242).To explore whether the source of food could have an impact on their growth, they were fed following the method described earlier (4) using four different strains of living diatoms known to inhabit the North Atlantic Ocean and/or Arctic region (Phaeodactylum tricornutum, Thalassiosira weissflogii, Chaetoceros debilis, or Pseudonitzschia turgidula).All diatom cultures were kept at their initial (aliquot) temperature of 15°C for Phaeodactylum tricornutum and Thalassiosira weissflogii and 5°C for Chaetoceros debilis or Pseudonitzschia turgidula.When the development of a bio-film at the bottom of the culture flask was observed during monitoring, two-thirds of the water in the flask was replaced by filtered in situ seawater.In the instances where a large number of offspring grew larger than about 10 chambers (sometimes difficult to precisely assess), they were separated into two culture flasks filled with the filtered in situ seawater.

Transport of live foraminifera, adult, and offspring
At the end of the MSM111 expedition in Autumn 2022 and of the culture trip in Eilat, Israel, in Spring 2023, all flasks containing live specimens (adults and/or offspring) were transported via airplanes in cool boxes (with ice packs for Neogloboquadrina pachyderma, without for Globigerinita glutinata, to minimize a too stark and rapid change in temperature) to the MARUM laboratories at the University of Bremen (Germany).Once in Bremen, the culture of Neogloboquadrina pachyderma was maintained at 3.5°C in a cold culture room while the culture of Globigerinita glutinata was kept at 21.8°C in an incubator.

Terminology
Due to the lack of ploidy assessment for planktonic foraminifera at the different stages of their life cycle, and the uncertainties surrounding the positions of meiosis, mitosis, and other potential reproductive pathways in planktonic species (see "Introduction" section), we choose to refrain from using the terms gamonts and agamonts.In the manuscript, we refer to foraminifera reproduction using the term "fission" (systematically multiple, instead of the commonly used "asexual reproduction"), schizogony (as defined in the introduction), and "sexual reproduction" in the observation of gamete release only, without the observation of karyogamy.

Globigerinita glutinata
After 2 days in their culture plates, three of the 130 specimens of Globigerinita glutinata isolated from the Gulf of Aqaba reproduced sexually and one specimen reproduced via multiple fission.That specimen spawned more than 80 viable offspring (Figure 2).Because of their motility and position in the well, a precise number was difficult to assess.We isolated the offspring following (6 in the method section, "Transfer of the offspring and monitoring") and successfully transported them back to our laboratory in Germany where they continued to grow (Figure 2) adding chambers at different paces.The remaining 112 living adult specimens were transported as well and on day 7 (after collection), another event of reproduction via multiple fission was observed in Bremen, producing about 70 offspring.The culture of all specimens of G. glutinata was stopped 2 weeks after the initial collection of the adult specimens in the Gulf of Aqaba, for other analytical purposes.For the offspring produced in Bremen on day 7, a few already calcified a second chamber.

Neogloboquadrina pachyderma
From the 192 specimens of randomly picked N. pachyderma, 101 died/emptied during the expedition, out of which several (>10) were directly observed to do so via reproducing sexually (e.g.video from J. Fehrenbacher of a gamete release in Neogloboquadrina dutertrei, https:// youtu.be/ iCqcK jeqR4g) and two via multiple fission (slightly less than 2%).In some instances, the shells emptied following a slow cytoplasm decay and thus, not as a result of a reproductive event.The two specimens having reproduced via multiple fission were left-coiling (Table 2).The first one to reproduce, M84, had been sampled at station 25,204, characterized by rather warm temperatures, medium Chl-a concentration, and high salinities while the second (M136) came from station 25,208, characterized by colder temperatures, low Chl-a concentration, and lower salinities (see "Material" section).
From the 120 specimens of selectively picked N. pachyderma 46 were empty before the end of the expedition among which 27 specimens had reproduced via multiple fission, accounting for 58.7% (Figure 3).Eighteen of these specimens were right-coiling and nine were leftcoiling (Table 2).Thirteen of the selectively picked specimens that reproduced were collected at station 25,209 and 14 at station 25,210, both characterized by cold water temperatures, low salinities, and rather high Chl-a concentration (see "Material" section).It is possible that among the other empty shells, some specimens had released gametes.
All events of multiple fission occurred between day 1 and 13 after the specimen collection with more than 50% (15 out of 29) of the events occurring during the first 7 days.The number of offspring produced per parent was difficult to precisely assess because of their high number and activity and because they were scattered in the well before being transferred to the culture jars.We however can estimate that about 80 to 300 specimens were produced for each parent (Figure 3).This number seemed to be proportional to the size of the parental shell and thus, the parental cell: the larger the cell, the higher the number of offspring.Altogether, these 29 reproductive events led to the production and subsequent simultaneous culturing of more than 3000 offspring.
The observation of such a high number of events for N. pachyderma allowed us to image specimens before and after their reproductive event (Figure 4).Before the event, the specimens present a bright red cytoplasm occupying most chambers of the calcite shell despite a large rhizopodial network, sometimes displaying long ectoplasmic structures (Greco et al., 2023; Figure 4).Specimens were observed to feed in large quantities (Greco et al., 2023; Figure 2) until very shortly before the reproductive event.They all bear the morphological traits identified during previous culture work and described in the method section: a very thin and translucent shell displaying almost perfect spherical chambers (Figure 4).

Growth of the offspring of Neogloboquadrina pachyderma and their response to different food treatments
After transfer and a resting phase of 24 h, we kept and fed the offspring per parental batch (all produced offspring from a parent in the same jar) to produce qualitative observation and determine the best-suited diatom strain for their growth under controlled conditions.For this, we monitored their chamber addition to determine their growth but it was not possible to track individual specimens.As a result, the findings presented hereafter are mainly qualitative and do not permit detailed numerical comparisons.
In the growth pattern of all offspring, distinct stages could be discerned.The first two chambers were always calcified in less than 24 h after the reproductive event.However, the development of the third chamber, when it occurred, appeared to be a more challenging phase in the specimens' evolution.A minimum of 2 days to up to several weeks could lie between the calcification of the second and third chamber.Once this stage was achieved, the juveniles kept adding chambers regularly until they accumulated 10 or more chambers (>80 μm, size from which we consider that specimens have reached the adult stage), or died.Despite being placed together in the same culture flask, the pace at which the offspring would add chambers differed among individuals.In each of the 29 batches of offspring produced via multiple fission, some specimens remained at a two-chamber stage for up to 2 months (after which they generally would die) while others would present more than 10 chambers during the same time interval.Clear differences in the pace of chamber addition were observed among the different food treatments.
Almost no growth was observed for specimens fed with the two strains kept at 15°C, Phaeodactylum tricornutum or Thalassiosira weissflogii.In both treatments, it took 12 days to observe the first offspring with three chambers and after 3 weeks, most of the offspring still only had two chambers.While the offspring did not seem to feed on these strains enough to grow, they did not die faster.From these batches, we altered the diet for some specimens still at the stage of three or four chambers after 3 months in an attempt to increase their growth rates to obtain more specimens large enough for further analyses (e.g.geochemistry).
Specimens fed with Chaetoceros debilis or Pseudonitzschia turgidula grew faster (four chambers observed already after 8 days) with a clear difference after 3 weeks for the specimens fed with Pseudonitzschia turgidula, presenting up to nine chambers after 14 days, in contrast to six chambers for the ones fed with Chaetoceros debilis.
Among the offspring fed with Pseudonitzschia turgidula, we recorded the longest life span with a few specimens still alive 7 months after their production, and also the largest size with specimens reaching up to 405 μm in diameter.For all offspring fed with Pseudonitzschia turgidula and belonging to different lineages, more than F I G U R E 4 Inverted light microscopy of four specimens the day before reproduction, and 12 specimens shortly after the event.The specimens before reproduction display an intense bright red cytoplasm and deployed rhizopods, round chambers, and thin shell walls.After reproduction, one can clearly see the delicate aspect of the shell wall by its transparency and the sphericity of the chambers, specially marked for specimens M269 and M287.The shells are either totally empty or bear bits of cytoplasm unused during the division (see M281, M287, M291, and M304 for which seven single shell offspring are also visible) and range from 140 (M269) to 230 μm (M242).
half could reach what would be defined as the adult stage for N. pachyderma.We measured the size of 200 of these "adult" specimens which showed a median final shell diameter of 198.5 μm (SD = 71, Table S1).From these 200 analyzed shells, 11% were right coiling and we could observe right coiling specimens in all lineages, irrespective of the initial parent shell coiling direction.
Throughout all treatments, we observed death among the specimens.The longest-lived specimen (7 months) had a diet of Pseudonitzschia turgidula.However, based on our observations we believe that the food source did not have a direct impact on the specimens' lifespan.Instead, it seemed to exert more control over their growth rather than affecting their longevity.

Neogloboquadrina pachyderma
The 29 batches of N. pachyderma's offspring were successfully transported to Germany where they arrived on October 6th.On October 17th the number of offspring in flasks M268 and M266, initially containing about 200 specimens, had doubled.The flasks contained each a mix of large specimens (seven or more chambers, first generation), dozens of one-to two-chamber specimens (second generation, with a magalospheric proloculus, Table 3), and an empty shell (parent), pointing at reproduction via multiple fission in both instances.
A second generation was also recorded for the lineages M242, M249, M252, M254 and M310 between the 9th of November and the 21st of December.There, several empty shells from the first generation were found next to dozens of one-to two-chambered second-generation specimens.In all lineages, the newly produced offspring (second generation) present proloculi similar in size to that of the previous generation and, except for the lineage M310; of the initially sampled parent (Table 3).In the lineage M310, the sampled parent had a microspheric proloculus (7.5 μm, Table 3).After multiple fission, the first and second generations of produced offspring were megalospheric (Table 3).
In total, a second generation was observed for seven of the 29 initial batches that we will henceforth refer to as lineages.
For all of these flasks, after the majority of the specimens were empty (as a result of death or reproduction), empty shells were removed and placed on micropaleontological slides for further analysis.The remaining living specimens were kept in culture and based on their size/number of chambers, assigned to different flasks labeled b for first-generation specimens or c for secondgeneration ones (Figure 5).
They were then again monitored weekly until all shells were empty and either picked (when time allowed it) or placed on the side, still in culture conditions.
On the 26th of January while monitoring the flask M249c (second generation lineage 249) which had been set aside the last week of December because all shells appeared empty and because the concentration of diatoms was too high, we noticed a large number (more than 200) of very healthy grown offspring (see video on figshare: https:// figsh are.com/s/ dca92 36f2a aff1a eed69 ) presenting a megalospheric proloculus.This constituted our first observation of a third generation.Similarly, on February 2nd we recorded the presence of dozens of one-to three-chambered specimens in the flask M254c.The flask had also been checked on January 26th and kept aside because only empty specimens were seen.The cultures were then intentionally stopped in early March 2023.
In summary, we observed 29 reproductive events via multiple fission for N. pachyderma between September T A B L E 3 Size of the proloculus of the sampled parent for the lineages in which we observed multiple generations (referred to in the first column as a = first generation, b = second generation, and c = third generation) and median size of the proloculi for the offspring produced via multiple fission.

DI SC US SION
A successful optimized sampling and culture procedure to detect and handle reproduction in nonspinose planktonic foraminifera from a ship, or inland laboratory conditions In this study, specimens from two genera (Globigerinita glutinata and Neogloboquadrina pachyderma), and from very contrasted environments, whether it is between the Gulf of Eilat and the Nordic Seas or within the Nordic Seas (four stations) reproduced via gamete release (not specifically tracked in the current study) and multiple fission in culture after having used our optimized procedure.Altogether, between August 2022 and March 2023, we observed more than 30 reproductive events of fission, which is more than all reported observations combined from the literature.The events of multiple fissions occurred within a couple of weeks after the collection of specimens in their natural environment for both species and with no synchronization, which suggests that no specific trigger (e.g.change of temperature when picked or change of diet) led to this specific type of reproductive event.
The protocol has proven helpful in growing the offspring to maturity, especially in the case of N. pachyderma for which the experiment was kept over multiple months.With this optimized procedure that we provide in detail in the manuscript, we see the potential to detect multiple fission in other nonspinose species and to grow offspring of planktonic foraminifera entirely under controlled conditions.For that, we would recommend applying the procedure to 150-200 specimens per species.Reproduction via multiple fission has still not been reported for spinose species and is still to be explored.It could, however, be that spinose species have lost the possibility to reproduce via fission throughout their evolutionary history.
To pursue the culture of N. pachyderma and G. glutinata, we had to transport the specimens and offspring from Canada or Israel to our laboratory in Germany (MARUM, Bremen).This only caused low mortality (<10%) and shows that planktonic foraminifera can sustain the stress caused by such transport, offering a promising avenue to conduct culture far from a marine research station.

Neogloboquadrina pachyderma offspring produced via multiple fission in culture by selection based on morphology
For N. pachyderma, specimens from four MPS stations and therefore four different physicochemical F I G U R E 5 Content of the flasks and processing of the foraminifera produced in culture by lineage "M XX" with XX being the number assigned to the parent when it was initially placed in culture.Each step, left to right, represents the follow-up of the multiple generations (1, 2, and 3, labeled as a, b, and c on the flasks).
environments reproduced in culture.Twenty-seven specimens reproduced within the "selectively picked" ones, all collected in the colder and more densely populated stations of the Baffin Bay (25,209 to 25,210) while only two of the "randomly picked" ones reproduced from stations 25,204 to 25,208, characterized by lower densities, warmer, and saltier but decreasing water temperatures and salinities.Within the "randomly picked" specimens, the occurrence of reproduction did not increase with decreasing temperature and salinity (i.e.stations) but rather stayed constant.This suggests that the difference in the occurrence of reproduction we observed between "randomly" and "selectively" picked does not reflect a change in the foraminifera environmental parameters (that one can however not fully exclude) but rather a real signal linked to the type of specimens selected.To a certain extent, it could also reflect or be linked to the population density.Indeed, the selectively picked specimens come from a more densely populated area in which maybe a higher proportion of the population could be prone to reproduction via multiple fission.The percentage of reproduction via multiple fission we observed in the randomly picked specimens (about 2%) is consistent with our previous findings (Meilland et al., 2023) in which 3% of N. pachyderma from the Southern Ocean and the Nordic Seas reproduced via multiple fission while 43% reproduced via gametogenesis.
Our study suggests that even if only a small portion of planktonic foraminifera reproduce via multiple fission in the natural population (Meilland et al., 2023) it is possible, using specific morphologic traits, to increase its occurrence in the laboratory to a level that one can obtain thousands of offspring from only a hundred specimens picked from plankton material.The existence of these indicative morphologic traits suggests that the determination of whether specimens of N. pachyderma will reproduce sexually or via multiple fission occurs early in its life.All specimens that reproduced via multiple fission in this study shared the specific traits described in the method section among which are very thin and translucent shell walls, in contrast to the often observed encrusted shells previously reported to be associated to gametogenesis in Neogloboquadrinids (e.g.Davis et al., 2017;Meilland et al., 2023).Indeed, in previous culture work on N. pachyderma we observed a clear thickening of specimens' shell walls before gamete release and obtained what could be referred to as encrusted shells while being in stable culture conditions and thus, irrespective of a change in depth or environmental conditions (Meilland et al., 2023;WestgÅrd et al., 2023).The specimens about to undergo multiple fission also demonstrated a tendency to accumulate a substantial amount of food before initiating reproduction.This feeding behavior is likely explained by the considerable energy requirements associated with reproduction (Komosinski et al., 2017) and stresses the importance of an adequate food source (Parfrey & Katz, 2010).
In addition, the majority of the parental specimens, 19 of 29, belonged to the "aberrant" form of N. pachyderma (Table 2), suggesting that the frequency of reproduction via multiple fission is higher in right-coiling specimens or, that it is easier to detect specimens approaching fission among the right-coiling N. pachyderma.Coiling direction could then eventually be considered another trait for selective picking.This corroborates previous observations and hypotheses formulated for benthic foraminifera, which proposed coiling direction to be associated to a specific part of the life cycle (Myers, 1936(Myers, , 1940)).This was also hypothesized for planktonic foraminifera more recently (e.g.Darling et al., 2023;Kimoto & Tsuchiya, 2006).It is, however, important to stress that this only is valid in sampling areas where Neogloboquadrina incompta is not present (such as in this study) as it would otherwise be difficult to reliably identify a specimen on the species level by morphology alone, the shell coiling direction being the only morphological trait differentiating N. pachyderma from N. incompta.
In contrast to what we previously observed (Meilland et al., 2023) but in agreement with Kimoto and Tsuchiya (2006) and Davis et al. (2020); irrespective of the parental shell coiling direction, we systematically observed right-coiling (minority) and left-coiling specimens among the offspring.This supports the suspicion that reproduction via multiple fission could introduce aberrant coiling direction in the population (Darling et al., 2023).All measured offspring produced via multiple fission presented a large proloculus (Table 3) falling in the "megalosphere" category of what had been observed in situ in the Baffin Bay in previous expeditions (Meilland et al., 2023).This supports our previous findings suggesting that, as for some species of benthic foraminifera, specimens of N. pachyderma produced by fission can be distinguished by the presence of a remarkably large proloculus.
If some specimens' traits can be relevant information to distinguish those likely to reproduce via multiple fission, the reproductive event itself happens in such a way that it cannot be timed precisely.Our observations suggest that the event could occur quickly in culture, in the 7 (50%) to 13 days following the cell collection.This could very well be a response to the culture setting and stress of the collection.

Diet as a key for successful growth
Our results on Neogloboquadrina pachyderma's offspring suggest no direct link between food treatment (different strains of live diatoms) and foraminifera lifespan.Similar proportions of long-lived specimens (more than 2 months total life span) and early death (less than a month) were observed in all treatments.While higher feeding frequency has been associated with faster growth and larger shell size (e.g.Bé et al., 1981) our study suggests that in the case of N. pachyderma, the type of food rather than its quantity is what plays a major role in the growth of the offspring.If the strain of diatoms we used for the offspring of Globigerinita glutinata (Nitzschia) was not the most adapted to them, this theory would also explain why we only observed limited growth over the few days the culture was kept.For N. pachyderma, a similar quantity of food was introduced in the offspring jars after their transfer but only very limited growth was recorded for specimens fed with Phaeodactylum tricornutum and Thalassiosira weissflogii.It is important to note though, that the two strains, of a good size for the offspring were initially kept at 15°C.It is thus also likely that the diatom cells died once placed in the colder water of the foraminifera flasks, making it a less valuable food type.Specimens grew faster when fed with Chaetoceros debilis and even more with Pseudonitzschia turgidula.C. debilis has the tendency to create spiny elongated cells, maybe making them difficult for the offspring to feed on.It has been reported that some diatom species are better sources of food for zooplankton than others (e.g.Smetacek, 1999).Their nutritional value can vary based on their size, wall composition, and nutrient content.Some diatom species have characteristics that make them highly nutritious and easily digestible and others can have a higher lipid content and provide a more energy-rich food source (e.g.Thalassiosira spp., Fahl & Kattner, 1993).This supports the previous hypothesis that food "quality" rather than "quantity" could be a controlling factor of some planktonic foraminifera species dynamic (i.e.Turborotalita quinqueloba and Globigerinita uvula), especially in subpolar to polar environments (Meilland et al., 2020).Further studies are necessary to understand and quantify the processes of planktonic foraminifera grazing on different species of diatoms and assess its impact on specimens and population growth.Single-cell metabarcoding of planktonic foraminifera will help to identify the preferred source of food of specimens in their natural environment (Greco et al., 2021) and narrow the list of potential candidate to feed live foraminifera for successful cultures.

Multiple generations via schizogony, the optional stage of the biphasic life cycle of N. pachyderma
Of the thousands of offspring successfully transported to Bremen, some died without a noticeable reproductive event and some likely did not reach a reproductive stage, probably due to inadequate food sources.However, a second generation of specimens was observed in seven lineages, and two lineages even showed a third generation.All are assumed to be the result of multiple fission because of the very large number of specimens produced, because of the large size of the offspring proloculi (Table 3), and because no evidence of gamete fusion has ever been reported.
The proportion of specimens that led to a second and then third generation via multiple fission during the experiment is lower than the 3% reported by Meilland et al. (2023) from the in situ population.This could be a culture artifact or simply reflect the selection that was made upstream, purposely selecting specimens about to reproduce via multiple fission.
Among the offspring obtained onboard and produced by multiple fission in the first instance, a second "multiple fission" generation was spawned from which specimens (schizonts, as defined in the Introduction) produced a successive "multiple fission" generation (schizogony).A look at the proloculus size of the initial parents for the lineages in which we observed multiple generations (Table 3) suggests that all except M310 were probably already in their schizont phase (large proloculus = result of multiple fission Meilland et al., 2023) when sampled.Unfortunately, it was not possible to identify the parental shell at the origin of the second and third generations and we could therefore not observe whether the specimens were right or left coiling.Except for the first generation of the lineage M310 (produced by a microspheric parent, likely resulting from gamete fusion), the specimens produced in culture during the second and third generations were most likely clonal (all have a megalospheric proloculus and result from fission) and highlight what could be referred to as schizogony, constituting in "true asexual reproduction."The frequency of schizogony, only reported for benthic foraminifera thus far, remains rare (e.g.Le Calvez, 1938, 1946) but would produce a number of schizonts sufficiently high to prevail in natural populations as well as in fossil assemblages (Dettmering et al., 1998).This corroborates our observations in culture with a very low percentage of schizonts but also our previous observations on N. pachyderma where we observed a majority of megalospheric specimens in situ (Meilland et al., 2023).
This study provides new pieces of evidence on the life cycle of N. pachyderma, and validates the presence of an optional schizont stage in its biphasic life cycle, as also suggested by Darling et al. (2023).Our latest observations of reproduction, its frequency, and the presence of a schizont phase underline the fact that establishing continuous cultures of planktonic foraminifera is feasible.

Perspectives
With this paper we show that it is possible to detect reproduction via multiple fission in various nonspinose species and genera of planktonic foraminifera.We also show that the occurrence at which one can obtain reproductive events in laboratory conditions can be improved by careful observation and description of specimen morphology, identifying morphological traits heralding an upcoming reproductive event.
of the observations and knowledge on the life cycle of planktonic foraminifera until very recently was based on tropical spinose species with a strong emphasis on the mechanisms of gametogenesis and gamete release (e.g.Bé, 1980;Bé et al., 1983;Be & Anderson, 1976;Caron et al., 1990;Spindler et al., 1978).Never a complete life cycle, from the zygote or offspring phase to its reproduction and to the production of another viable generation, had been described in planktonic foraminifera.
To unlock what is at the base of planktonic foraminifera success in the global ocean and understand their reproductive dynamics is necessary to constrain their entire life cycle further.One should continue to explore and document the potential different modes of reproduction that could co-exist in other species, for example in the spinose foraminifera for which fission has still not been reported.
Our observations of sexual reproduction and reproduction via multiple fission in both specimens of the nonspinose Globigerinita glutinata and Neogloboquadrina pachyderma collected (respectively) at the same sites suggest that their population dynamics do not rely entirely on one reproductive mode or the other.The combined use of sexual reproduction and reproduction via multiple fission could be the basis of the success of both species in the modern ocean (Meilland et al., 2023).Indeed, multiple fission produces a large number of offspring which could lead to quickly repopulating the environment when the conditions are favorable, for example after the ice break in polar environments, while gamete fusion (instead of schizogony, creating clones) would ensure gene recombination.The polar species N. pachyderma displays a very intense seasonal signal of abundance in high latitudes North and South (e.g.Jonkers et al., 2010;Mikis et al., 2019).This strong variability is also marked in specimens' morphology, with more or fewer specimens displaying clear signs of gametogenesis (e.g. the presence of a crust; Mikis et al., 2019), which could reflect a slight shift in the ratio between sexual reproduction and reproduction via multiple fission leading to sharp increases in N. pachyderma's abundances.
We should also take advantage of progress in cell fixation methodology and microscopy to assess the ploidy of specimens in natural populations, in specimens produced via multiple fission, at different stages of their life, and in specimens about to undergo gametogenesis to test if the "haplodiplont" life cycle hypothesis is applicable to planktonic foraminifera.
We need to keep constraining morphological traits in specimens reproducing via multiple fission in order to keep increasing its occurrence in the laboratory and we need to track the culture carefully to identify the schizonts and analyze their shell morphology in detail.

AC K NOW L E DGM E N T S
We would like to thank the captain and crew of the Maria S. Merian expedition MSM111 for their support in the sample collection and in maintaining the cold room at a stable temperature.We are grateful to Anjuly Janssen who supported us with the handling of the multinet, CTD-rosette, and water filtration during the expedition.We warmly thank Pushpak Martin and Jakob Salger who provided support in the laboratory to keep track of the culture.We are also thankful to Anna Biastoch and Tina Trautman who provided us with the strains of live diatoms used during this study.We would like to thank Sigal Abramovitch, Racheli Shaul, Gil Koplovitz, Emanuel Sestieri, and colleagues from the Inter-University Institute for Marine Sciences in Eilat, Israel, for their tremendous help and support in the collection of plankton in the Gulf of Aqaba and laboratory access.We also would like to thank the two anonymous reviewers who helped us improve our manuscript.Finally, J.
located in the deep and central part of the Gulf of Aqaba, in warm (21.8°C) and highly saline (40.7) waters.

F
Inverted light microscopy of (A) parental shell of Globigerinita glutinata, scale bar = 165 μm; (B) the cloud of offspring shortly after the reproductive event, scale bar = 100 μm; (C) a two-chambered offspring; (D) a three-chambered offspring feeding on a Nitzschia cell; (E, F) four-chambered offspring.Scale bar for (C), (D), (E), and (F) = 25 μm.T A B L E 2 Number of randomly and selectively picked right and left-coiling specimens (light gray), number of specimens that died before the end of the expedition (*), and number of specimens that reproduced (bold).

F
Light inverted microscopy images of 20 parents (M plus number, 135 to 230 μm) of the MSM111 expedition shortly after their reproduction and surrounded by some of their offspring (14 to 30 μm).
11th and 27th (of 2022); our first generation.Within this generation, multiple fission was again recorded for seven specimens from the middle of October to December 2022, i.e. between 40 and 70 days later.From two of the specimens of this second generation a third generation was spawned via multiple fission sometime between January and early February 2023, i.e. about 45 and 60 days later.This suggests, for the specimens reproducing via multiple fission, a generation time of about 2 months whereas specimens that died without apparent reproduction in our experiment could live up to 7 months.
Meilland and R. Morard were funded through the Cluster of Excellence-The Ocean Floor-Earth's Uncharted Interface-Receiver/Recorder group (EXC-2077, Project 390741603) funded by the German Research Foundation (DFG).Open Access funding enabled and organized by Projekt DEAL.ORC I D Julie Meilland https://orcid.org/0000-0001-8966-3115R E F E R E NC E S

T A B L E 1
List of samples (N.pachyderma specimens and water for the culture) location and time of sampling during the MSM111 cruise.

GeoB-Nr Device Date Longitude (°W) Latitude (°N) Sampling intervals & depth (m) Specimens picked Volume filtered (L)
1. Collection and filtration of in situ seawater.Once at the sampling station(s), a Niskin bottle (can be a CTD-rosette) and a Conductivity Temperature Density (CTD

Parent proloculus size (μm) Number of offspring analyzed Offspring proloculus median size (μm) SD
Note: The gray/white areas separate the lineages.