Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Ontogeny of juvenile freshwater pearl mussels, Margaritifera margaritifera (Bivalvia: Margaritiferidae)

  • Louise Lavictoire ,

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Freshwater Biological Association, The Ferry Landing, Far Sawrey, Ambleside, Cumbria, United Kingdom, Centre for Wildlife Conservation, University of Cumbria, Ambleside, Cumbria, United Kingdom

  • Andrew D. Ramsey,

    Roles Funding acquisition, Methodology, Resources, Supervision

    Affiliation College of Life and Natural Sciences, University of Derby, Derby, United Kingdom

  • Evelyn A. Moorkens,

    Roles Conceptualization, Methodology, Supervision, Writing – review & editing

    Affiliation University of Dublin, Trinity College, Dublin, Ireland

  • Graham Souch,

    Roles Resources

    Affiliation College of Life and Natural Sciences, University of Derby, Derby, United Kingdom

  • M. Christopher Barnhart

    Roles Writing – review & editing

    Affiliation Department of Biology, Missouri State University, Springfield, MO, United States of America

Ontogeny of juvenile freshwater pearl mussels, Margaritifera margaritifera (Bivalvia: Margaritiferidae)

  • Louise Lavictoire, 
  • Andrew D. Ramsey, 
  • Evelyn A. Moorkens, 
  • Graham Souch, 
  • M. Christopher Barnhart


The gills of juvenile freshwater bivalves undergo a complex morphogenesis that may correlate with changes in feeding ecology, but ontogenic studies on juvenile mussels are rare. Scanning electron microscopy was used to examine the ultrastructure and ontogeny of 117 juvenile freshwater pearl mussels (Margaritifera margaritifera) ranging in age from 1–44 months and length from 0.49–8.90 mm. Three stages of gill development are described. In Stage 1 (5–9 inner demibranch filaments), only unreflected inner demibranch filaments were present. In Stage 2 (9–17 inner demibranch filaments), inner demibranch filaments began to reflect when shell length exceeded 1.13 mm, at 13–16 months old. Reflection began in medial filaments and then proceeded anterior and posterior. In Stage 3 (28–94 inner demibranch filaments), outer demibranch filaments began developing at shell length > 3.1 mm and about 34 months of age. The oral groove on the inner demibranch was first observed in 34 month old specimens > 2.66 mm but was never observed on the outer demibranch. Shell length (R2 = 0.99) was a better predictor of developmental stage compared to age (R2 = 0.84). The full suite of gill ciliation was present on filaments in all stages. Interfilamentary distance averaged 31.3 μm and did not change with age (4–44 months) or with size (0.75–8.9 mm). Distance between laterofrontal cirri couplets averaged 1.54 μm and did not change significantly with size or age. Labial palp primordia were present in even the youngest individuals but ciliature became more diverse in more developed individuals. Information presented here is valuable to captive rearing programmes as it provides insight in to when juveniles may be particularly vulnerable to stressors due to specific ontogenic changes. The data are compared with two other recent studies of Margaritifera development.


Research and efforts to conserve freshwater mussels (Unionida) increased dramatically during the 20th Century in response to population losses and extinctions [1,2]. A species of particular concern is the freshwater pearl mussel (Margaritifera margaritifera) which is critically endangered [3] and has thus been the focus of captive rearing activities in Europe for over 30 years [49]. The life history of M. margaritifera is well documented [1013]. Parasitic glochidia larvae are released in summer and attach to the gills of salmonid host fish, where they remain over winter and metamorphose to the juvenile stage. The metamorphosed juveniles leave the host (excyst) the following spring and occupy interstitial spaces in the river substrate. Population declines have been attributed particularly to degradation of this interstitial habitat through siltation and pollution [1416].

As post-parasitic M. margaritifera grow from about 0.4 mm to several mm in length, the gills (ctenidia) develop from a few simple ciliated filaments to folded structures that transport water through internal spaces [1719]. It has been suggested that the simple gills of early juvenile mussels must be ineffective for suspension feeding, and that most particle capture instead involves the ciliated foot (pedal feeding [2025]). If a functional transition occurs as the gills develop, this transition could have implications for feeding ecology and captive culture methods [17,18]. However, the mechanisms involved in food capture by early juveniles are not well documented or understood. Direct internal observation and video imaging of particle capture by Unionids has been accomplished only for adults with fully formed gills [26,27].

For reference, the adult structures and terminology are provided in Fig 1. M. margaritifera is a eulamellibranch mussel displaying the homorhabdic gill condition. Each gill (ctenidium) consists of an inner and outer demibranch. The demibranchs are formed by rows of filaments that descend from the dorsal axis, and reflex upward to define an internal (suprabranchial) space. Each filament bears three sets of ciliary structures. Lateral cilia are responsible for water movement between the filaments, from the branchial (mantle) cavity to the suprabranchial cavity. Laterofrontal cirri are believed to be responsible for particle capture, although debate remains as to whether they act as ‘bats’ or ‘sieves’ moving particles to the frontal cilia [2832]. The frontal cilia transport particles ventrally towards the oral groove (ventral edge of the folded demibranch), in which particles are transported by cilia anteriorly to the labial palps for sorting into particles which are either transported to the mouth and consumed or egested as pseudofaeces.

Fig 1. Diagram of gill position and cross-section through a lamella showing ascending and descending limbs of a filament.

A: Gill position within a valve showing positioning of the anterior adductor (aa) and posterior adductor muscles (pa), budding zone (bz) and filaments (fi). B: A dorso-ventral section though a eulamellibranch showing the foot (ft), gill base (gb), inner demibranch (id), interlamellar junction (ilj), interlamellar space (ils), mantle (m) and outer demibranch (od). Used with permission from [38].

Juvenile mussels may have different biological and environmental requirements depending upon their mode of feeding [33] and mortality may increase when developmental changes occur [3436] due to inability to meet energetic demands during morphogenesis [37].

The aim of the present study was to improve the understanding of the morphological development of the gills of juvenile M. margaritifera and important additional features, such as the foot, labial palps, siphons, mouth and mantle, under a natural temperature and diet regime. This is important because it allows a better insight into potential high-risk periods for juveniles under natural conditions. The biological structure and ontogeny of juvenile gills and other pertinent structures at different ages were investigated using scanning electron microscopy (SEM). The study utilised captive-cultured individuals of known ages.

Materials and method

All experimental individuals were cultured at the Freshwater Biological Association, Windermere, UK and originated from a single mussel population (location details can be obtained from the author). Permission for use of juvenile Margaritifera margaritifera in this investigation was granted by Natural England and ethical approval was granted by the University of Cumbria. All work was carried out in the laboratories at the Freshwater Biological Association, Windermere, UK, and as such, no permission was required for use of land. Propagation methods are outlined in project reports [39,40] and in [5] but are briefly described here. Water is sourced from Windermere, a mesotrophic lake in Cumbria. Water is filtered to removed particles >20 μm for broodstock adults or 15 μm for juvenile mussels and delivered to the systems as outlined in the reports above. Juvenile mussels were collected in 200 μm plankton nets as they excysted from salmon in 2012, 2014 and 2015 and were transferred into substrate measuring 1–2 mm in either the tray system (2014 and 2015 juveniles) described in [40] or the aquarium system (2012 juveniles) outlined in [5]. Juveniles were maintained in these systems under a natural temperature regime until they were required. Experimental work took place between May 2015 and March 2016. Individuals were selected from three cohorts:

  1. 1–8 months old (excysted summer 2015).
  2. 13–20 months old (excysted summer 2014).
  3. 34–44 months old (excysted summer 2012).

Selection was made by disturbing sediment in the tray system described in [5] and siphoning water through a 0.3 mm sieve to retain mussels. Each individual was transferred to a 10 ml glass tube containing 1 mg ml-1 MS-222 [41] to induce the valves to gape and to expose soft tissues. Juveniles were processed with either both valves intact or one valve was removed completely to better observe the structures facing the shell. Juveniles were fixed in 2% glutaraldehyde in 0.1M Sorenson’s Phosphate Buffer (SPB) for <24 hours. They were then washed in SPB (x2) before being dehydrated through 25%, 50%, 75% and 2 x 100% ethanol washes. Two washes (30 minutes) of Hexamethyldisilazane (HMDS) were carried out in place of critical point drying. Samples were mounted onto SEM stubs and sputter coated with gold before being loaded into a Zeiss Leo 1450VP scanning electron microscope. The distance between filaments (interfilamentary space or distance) and distance between laterofrontal cirri couplets were measured from micrographs. Where pictures contained inter-filamentary junctions measurements were taken in the vicinity of junctions because, at this point, interfilamentary space is less variable. Effects of potential tissue shrinkage during sample preparation [42,43] were not quantified. All measurements of features on scanning electron micrographs were taken with ImageJ (version 1.48).

Data analysis

SPSS (v. 22) was used for data analysis. All data were tested for normality (Shapiro-Wilk) before performing parametric tests. ANOVA’s with post hoc Tukey’s HSD tests were performed to test for difference in interfilamentary space among individuals of different ages and to test the number of cilia per laterofrontal cirrus. Linear regression was performed for interfilamentary space against shell length and also for the number of inner demibranch ‘vs’ outer demibranch filaments on individuals where the outer demibranch was present.


Table 1 outlines summary information of specimens considered for this study. Three stages of juvenile development were observed. The age of specimens is provided in the top right of each micrograph. For some individuals, poor specimen quality meant that high quality micrographs were not always possible. In these cases, micrographs from a different cohort may be used to depict key features; where applicable, this has been outlined in the text.

Table 1. Summary information about the ranges of shell length (mm) and number of inner demibranch (ID) filaments for the different age cohorts.

The number of specimens considered is also provided (n).

Gills and labial palp morphogenesis

The ctenidia of 1 month old individuals consisted of only the inner demibranchs, each consisting of 3–4 simple filaments (Fig 2A). As the mussels grew, new filaments were added to the inner demibranch at the posterior budding zone. The number of filaments increased with shell length, with about 10.3 filaments added per mm shell length over the range of 1.0–8.9 mm (Table 1, Fig 3). Across all specimens, filament diameter averaged 28 μm (± 5 μm) and interfilamentary space 31 μm (± 10 μm). Filaments of 1 month old individuals were unconnected laterally to one another (Fig 2A). Attachment between the tips of adjacent filaments was observed at 3–4 months (> 0.8 mm; Fig 2B of a 14 month old individual). This area of attachment is called the ventral bend. Connections usually began posteriorly near the budding zone and the tissue was covered in simple cilia resembling frontal cilia (see below).

Fig 2. Main anatomical features of juvenile mussels.

a) Foot (FO), unreflected filaments (FI), gill axis (GA), left and right labial palps (LP). b) Distal tips of filaments are joined to each other by thin tissue connections (arrow heads); c) Gill reflection of the inner demibranch. Thin tissue connections join filaments at the ventral bend (VB) and the thicker fused dorsal bend (FDB) joins the terminal ends of the ascending arms. All three cilia types are present on the ascending limbs (AL); lateral cilia (LC), laterofrontal cirri (LFC) and frontal cilia (FC). The ascending limb is longer on medial filaments compared to those at either anterior or posterior ends (to the left and right of frame). Other features of note are the filament abfrontal surface (AS), descending limb (DL) and mantle (MA); d) Oral groove on inner demibranch (left) and absence of groove on outer demibranch (right); e) Ciliary junctions between approximately filaments 11–14 (*), after which tissue junctions were present (†); f) Right ID, OD and labial palps (LP). Inset: Labial palps are highly ciliated on the inner surface but devoid of cilia externally.

Fig 3. Scatter plots of the number of inner demibranch filaments against shell length (a) and age (b).

R2 values show shell length is a better predictor of number of inner demibranch filaments compared to age.

Inner demibranch filaments were added posteriorly and they began to reflect when individuals exceeded about 1.1 mm in length and when there were more than 9 filaments (Fig 2C). Reflection appeared to begin in the middle of the demibranch where the filaments were longer, so that the ascending limbs on the medial filaments were longer than on those and anterior and posterior (Fig 2C). In juveniles older than 34 months, new inner demibranch filaments were budding via cavitation extension. All filaments were reflected except the most anterior filament, which consisted of only a descending limb which was attached to the visceral mass along its entire length. In addition to connection at the ventral bend, filaments were also joined at the terminal end of the ascending limb by a continuous, thick piece of tissue, called the fused dorsal bend (Fig 2C). The fused dorsal bend was never observed attached to the visceral mass, even in the oldest/largest individuals.

No oral groove was observed on the inner demibranch at the size and age when gill reflection commenced (1.13–1.45 mm and 13–20 months old). However, the ventral bend appeared bulbous, which may be a precursor to oral groove development. The oral groove was present on the inner demibranch in older/larger specimens (>34 months and >2.6 mm). The groove was a deep, circular invagination on filaments that appeared anteriorly of approximately filaments 2–13 (numbering from posterior–anterior; Fig 2D).

Interfilamentary junctions were not observed in younger/smaller individuals but were observed from 34 months (2.66 mm). Filaments 1–11 had no interfilamentary junctions, filaments 11–15 had ciliary junctions, and more anterior filaments had tissue junctions (Fig 2E). Longer, more anterior filaments also exhibited additional interfilamentary junctions along their dorso-ventral length. In the largest individual considered during this study (shell length = 8.9 mm) the inner demibranch had 6 rows of tissue interfilamentary junctions in a dorsal-ventral direction. No ciliary interfilamentary junctions were observed in this individual.

Filaments of the outer demibranch first appeared in individuals larger than 3.1 mm. Outer demibranch filaments proliferated via cavitation extension (Fig 2F), giving rise to several filaments at once. The ventral bend on the outer demibranch was covered in simple cilia, and was flattened but not invaginated into an oral groove as observed on the inner demibranch (Fig 2F).

The full complement of filament ciliature (frontal and lateral cilia and laterofrontal cirri) was observed on the inner demibranch filaments in all age classes examined (Fig 4A). Laterofrontal cirri were complex, branching and had cirral plates orientated perpendicular to the filament. Each laterofrontal cirrus consisted of two parallel rows of cilia which were shortest towards the frontal surface and became progressively longer towards the lateral surface. Ciliation on the developing ascending limb was the same as on the descending limb with all types of cilia/cirri present although laterofrontal cirri were smaller and consisted of fewer cilia per cirrus. Laterofrontal cirri could reach over half way across interfilamentary spaces only in more developed individuals (as shown in Fig 4A). In all age groups, the abfrontal surface of filaments had only a very sparse, unorganised coverage of simple cilia.

Fig 4. Main anatomical features of juvenile mussels.

a) Ciliation of gill filaments; frontal cilia (FC), lateral cilia (LC) and laterofrontal cirri (LFC); b) Budding zone (BZ) and left inner (LID), left outer (LOD), right inner (RID) and right outer demibranchs (ROD). Inset box shows ciliary connection (CC) between left and right BZ. OG = oral groove; c) Labial palp primordia, lips and mouth; Foot (FO), left inner palp (LIP), left outer palp (LOP), lower lip (LL), mouth (MO), right inner palp (RIP), right outer palp (ROP), upper lip (UL); d) The labial palps were plicated internally but retained a flattened appearance on the outer surface. Inset shows simple cilia (†) and more complex cirri (*).

The filament budding zone was not very prominent in smaller individuals but in larger specimens it was covered in simple cilia. Gill buds were distinguished from ‘true’ filaments by the absence of laterofrontal cirri. Ciliary connections joined the left and right parts of the budding zone (Fig 4B). Each individual had approximately 3–5 buds before the onset of true filaments and the budding zone was not connected to the mantle. Laterofrontal cirri on newly-budded filaments consisted of fewer cilia per cirral plate (i.e. not as wide) as laterofrontal cirri on older filaments.

On younger individuals, labial palp primordia were observed as paired projections on either side of the mouth and were densely covered with simple cilia (Fig 4C). By 20 months old (~1.45 mm long), the labial palps were becoming plicate and a ciliary connection was also observed between the labial palp and the penultimate anterior filament in one individual. In older individuals the labial palps were plicate with both simple cilia and more complex cirri present on the inner but not the outer surfaces (Fig 4D).

Gill morphometry

Juvenile shell length was closely correlated with the number of inner demibranch filaments (n = 47; F(1,45) = 3520; P < 0.001; R2 = 0.99) and length was a better predictor than age (R2 = 0.84; Fig 3). The number of inner demibranch filaments also correlated with the number of outer demibranch filaments (Fig 5; F(1,4) = 483, P < 0.001, R2 = 0.99).

Fig 5. Scatter plot of the number of inner ‘vs’ outer demibranch filaments.

The number of inner demibranch filaments is able to predict the number of outer demibranch filaments and accounted for 99% of the explained variability in number of outer demibranch filaments.

The space between filaments differed significantly among individuals (F(4,74) = 11.40, P < 0.001) but did not show any significant trend with size or age (Table 2, Fig 6). The distance between filaments was more consistent when tissue junctions were present compared to ciliary junctions or no junctions (SD of interfilamentary distance was larger in individuals < 16 months old).

Fig 6. Interfilamentary space vs length (mm) and age (months).

There was no significant trend of interfilamentary space with length nor age.

Table 2. Mean interfilamentary space and filament diameter of individual specimens at given ages and shell lengths.

The number of measurements taken from each individual is also provided (n).

The number of cilia per laterofrontal cirrus increased as filaments developed with fewer cilia per cirrus on newly budded filaments and also on filaments of the ascending limb in juveniles undergoing gill reflection. However, comparing only cirri on the descending limb of reflecting filaments, there was no difference in the number of cilia per cirrus among 4 month olds (42 ± 2, n = 6), 16 month olds (43 (± 3), n = 3) and 34 month olds (39 (± 9), n = 2). Laterofrontal cirri couplets were spaced an average of 1.54 μm (± 0.40) apart (n = 21 from three 16 & 34 month old individuals).

The processes and timing of gill development were used to categorize three stages of development related to size and age. These stages are outlined in Table 3.

Table 3. Description of the three stages of juvenile gill development based upon observations during this study.

The age (months) at which individuals begin to display particular structures/developments is approximate and no attempt has been made to postulate when development of certain structures begins if they were not directly observed during this study. The number of inner demibranch (ID) and outer demibranch (OD) filaments are the number observed during this study and may differ depending upon population or other parameters.


Foot form and ciliation was consistent throughout all age classes. The foot appears to have two distinct regions (Fig 7A). The distal region is folded and has a dense coverage of simple cilia whilst the proximal region above the ‘heel’ is only very sparsely covered with patches of cilia. The largest 44 month old juvenile (8.9 mm), had a fine byssus thread attached. A thin hole was observed along the distal tip of the foot which was likely the byssus pit.

Fig 7. Notable foot and mantle features in juveniles of all age/size classes.

a) The foot showing two distinct regions. The distal portion, below and right of the ‘heel’ (dashed line) is heavily ciliated while the proximal region, above and left of the ‘heel’, bears only sparse ciliation. Inset shows ciliation in greater detail; b) Long simple cilia (*) and short cirral tracts (†) near the mantle margin. Ventral shell margin is to the bottom of the image; c) Mantle surface covered in rows of cilia (arrow heads).


In young individuals before ctenidia organogenesis, simple cilia were visible around the mantle margin with more dense aggregations near the posterior end. Ciliation was arranged into three rows. Firstly a band of short cirri composed of 6–10 individual cilia was present closest to the margin with two bands of long simple cilia above this (Fig 7B). Sparse aggregations of long simple cilia were distributed over the visible part of the mantle and occasional instances of long cilia in two parallel rows were observed. Ciliation continued to develop in individuals undergoing ctenidia organogenesis, becoming denser, especially posteriorly around the areas where siphons would eventually develop. By 44 months old the mantle surface was covered in organised rows of simple cilia (Fig 7C) with ciliation becoming denser close to the inhalant siphon. Tissue making up the inhalant siphon was plicated and ciliation extended to the shell-facing side of the mantle.


This paper is among the first attempts [1719] to describe the ontogeny of gill ultrastructure of juvenile freshwater pearl mussels Margaritifera margaritifera with scanning electron microscopy (SEM) and the first to investigate development under a natural temperature regime for this species. Earlier studies of the adult anatomy of freshwater mussels were directed mainly at taxonomic relationships [4446] and reproduction [47]. It is hoped that understanding juvenile gill development can improve understanding of feeding mechanisms and lead to evidence-based decisions about captive rearing practices. This study collected data from 117 juvenile mussels, focussing mainly on the ontogeny of ctenidia and associated structures. The ability of the ctenidia to efficiently capture particles is thought to depend on small interfilamentary space [48], development of the oral groove for particle transport to the labial palps [49], contact between anterior gill filaments and the labial palps [50] and the ability of ciliary mechanisms to drive a current of water through the mantle cavity to deliver suspended particles.

Margaritifera margaritifera undergoes substantial gill ontogeny during the first 44 months post-excystment (up to 8.90 mm). In Stage 1, ctenidia grow via simple proliferation of unreflected filaments until shell length > 0.8 mm (around 3–4 months old) when filaments begin to connect at the distal tips. The laterofrontal cirri are complex, branching and have a similar structure to those observed in other lamellibranch bivalves [32,49,51], but may not form an efficient sieve at this stage because the interfilamentary spaces are too wide. The labial palps at this stage are simple, flattened flaps which are not plicate at this stage but are heavily ciliated, as is the inner surface of the lips and the area around the mouth.

Stage 2 commences with the onset of gill reflection between 13–16 months old when shell length ≈1.2 mm, and when the inner demibranch consists of > 9 filaments. Shell length and the number of inner demibranch filaments attained before the onset of gill reflection is similar to those previously reported in other bivalve species [35,37,50,52]. However, M. margaritifera takes over a year, under the natural temperature regime in NW England, to commence gill reflection, the most delayed onset reported for any freshwater bivalve [19]. Medial inner demibranch filaments begin to reflect first, followed by anterior and posterior filaments. Once the most posterior filaments have reflected, proliferation from the budding zone is via cavitation extension. The number of inner demibranch filaments added per mm shell length in this study is consistent with results from Schartum et al. [17] (10.33 and 8.10 respectively). With further development during Stage 2, tissue at the ventral bend becomes thicker and more densely ciliated, probably as a precursor to oral groove development [50]. The outer demibranch is not present at this stage.

Stage 3 involves several important morphological changes. Firstly, the outer demibranch develops in individuals > 3.1 mm long and both the inner and outer demibranch filaments proliferate via cavitation extension, the same as reported for Crassostrea gigas [36]. The number of filaments on the inner and outer demibranch were highly correlated (R2 = 0.99). The age/size at which the outer demibranch develops is most likely species-specific.

Secondly, the oral groove begins to develop on the inner demibranch when individuals reach a shell length of between 1.45–2.60 mm. This result concurs with Araujo et al. [18] who found that the oral groove was present in 1.8 mm juveniles. The oral groove in M. margaritifera remains open, unlike the deep, enclosed oral grove of M. edulis [53] and is present anterior of approximately the 2 – 13th filament. The oral groove was never observed on the outer demibranch in the size-age range examined. Possibly the particles collected by the outer demibranch are passed to the frontal surface of the inner demibranch for transport to its oral groove [54,55]. Absence of an outer demibranch oral groove appears to be the norm in juvenile bivalves excepting M. edulis [53].

Thirdly, interfilamentary junctions on the elongating inner demibranch filaments were observed for the first time in Stage 3, beginning as ciliary junctions but quickly giving way to tissue junctions, similar to other eulamellibranchs [36,50,53]. Interfilamentary junctions developed in individuals >2.66 mm long (34 months old), it is likely that their development begins in slightly smaller and younger individuals. No ciliary junctions were observed in 44 month old specimens and it may be that only tissue junctions form after a certain size/age. Where several rows of interfilamentary junctions were present they appeared to be spaced approximately evenly along the dorso-ventral axis, likely to provide stability to elongating filaments. Addition of interfilamentary junctions along the dorso-ventral axis suggests that the site of elongation i.e. ventral growth of filaments, may be from the ventral portion of the filament rather than from the gill axis. Initial growth of the ascending lamella upon reflection is from the ventral bend region [38,53,56] but elongation may be from the terminal end of the ascending lamella i.e. near the fused dorsal bend. This area warrants further investigation. The current study concurs with the sequence of developmental stages [17] reported previously, at least until 44 months of age (shell length = 8.9 mm) (Table 3). Filament diameter (28 ±5 μm) and interfilamentary space (31 ±10 μm) in this study displayed isometry (Fig 6) and these dimensions are similar to those reported by Schartum et al. (28 μm and 29 μm respectively; [17]).

A substantial body of research exists on the topic of gill ciliation, particularly the role of laterofrontal cirri in particle capture [5761]. In the present study, the full suite of frontal and lateral cilia and laterofrontal cirri were present in even the youngest individuals and filaments developed cilia and cirri almost as soon as they were budded, similar to previous reports for other bivalves, both freshwater [50,52] and marine [36,53]. The structure of laterofrontal cirri in M. margaritifera is similar to previously described [32,48,51] and their function, whilst not observed directly in this study, is believed to be similar also. That is, cirri from opposing paired plates splay out in turn in opposing directions either acting as ‘nets’ to sieve particles or ‘paddles’ to create vortices that transfer particles to the frontal surfaces of the filaments [32,42,48,51,56,6062]. Whilst the number of cilia per laterofrontal cirrus is variable between species (S1 Table), the number in M. margaritifera was comparable to other marine and freshwater bivalve species.

Cirral plates in M. margaritifera are more closely aligned (1.54 μm ± 0.40) compared to other species (2.0–3.5 μm; [34,4547]). This type of laterofrontal cirrus is efficient at capturing small particles [63] and close alignment of cirral plates suggests that M. margaritifera may be capable of retaining particles < 2 μm, even smaller than previously suggested [64]. The relationships between gill structure, feeding mechanisms and ecological niche are important considerations for researchers and practitioners working with endangered mussel conservation and population recovery. If M. margaritifera is capable of retaining particles <2 μm and is also has a poor ability to select nutritious particles over non-nutritious particles [64], this species may be vulnerable to changes in the size and composition of particles in its interstitial environment. Sustainable recruitment is no longer observed in many populations. Historically oligotrophic conditions no longer exist in the majority of modern pearl mussel rivers. Silt loading in rivers introduces inorganic material not suitable for consumption that must be ejected in pseudofaeces, or lower the nutritional value of ingested material. Nutrient loading stimulates unnaturally high concentrations of relatively large algal cells, some of which may not be in the correct size range for mussels to consume. This again has energetic consequences of either egestion as pseudofaeces or restricting filtering to avoid the particles [6567]. Propagation facilities providing supplementary feeding to juvenile mussels might consider the implications of gill ciliation on the ability of juveniles to feed on the particles provided. Inter-cirral distance has been shown to increase with age in the marine bivalve Perna canaliculus [48] which may alter food particle size preference with increasing age. This was not observed in M. margaritifera in the size and age cohorts studied.

This study also made observations of other structures important for juvenile feeding. The labial palps developed significantly throughout this investigation. Labial palps began as small simple structures which grew and became plicate at approximately 20 months old (shell length = 1.45 mm). It is possible that this marks the beginning of more complex sorting capabilities of the labial palps although this hypothesis needs to be tested with further observations.

Foot ciliation showed a consistent pattern in all age classes studied. The distal portion was heavily invested with short simple cilia. These cilia are very active during locomotion and could resuspend deposited particles that then would be drawn into the pedal gape. This pattern of ciliation is similar to previous descriptions in other species [50,60,68] and supports the hypothesis that pedal cilia, along with mantle and gill cilia, have a function in feeding. However, there was no evidence of pedal ciliary tracts for direct particle transport.

At all developmental stages, mantle ciliation was denser around the posterior region where the apertures eventually develop. Ciliation was also present around the majority of the mantle margin, where cilia were arranged into areas of short, compound cirri nearest the margin and longer, simple cilia dorsal to this. Longer cilia may have a role in pseudofaeces transport [43]. The shorter cirri may have a role in creating water currents but direct observation in live specimens is required to confirm this. Ciliation elsewhere on the mantle was unorganised and patchily distributed in younger individuals, becoming more organised into tracts by 44 months old. It is possible that these cilia are involved in maintaining water currents through the branchial cavity. By 44 months old siphons were well-developed and the inhalant siphon had the characteristic papillose form and was highly ciliated.

Attainment of true siphonal filter feeding is likely to be a gradual process with individuals probably exploiting both pedal and siphonal filtering mechanisms concurrently for some time. In small juveniles, the entry of water into the mantle cavity is not limited to the posterior and also occurs along the anterior and ventral margins [20]. Limitation of water entry to the posterior incurrent aperture is likely to be a gradual process that culminates when individuals achieve a size and flow rate that requires access to the water column rather than the interstitial space.

Conclusions and implications for captive rearing programmes

Proliferation of freshwater mussel propagation programmes within at least the last 30 years [2,6,69] has necessitated increased understanding of the factors affecting juvenile growth and survival in captivity. Ontogenic studies are important to complement ecological and genetic studies and help improve understanding of endangered species in order to develop comprehensive management plans [70,71].

This work details information valuable to captive rearing programmes regarding the timing of key developments, how this may affect the survival and adaptation of M. margaritifera to oligotrophic conditions. A summary of key biological developments over the first four years post-excystment is provided in Table 4 including details of factors potentially important to juvenile survival in captivity, as well as providing insight into the most sensitive periods where pressures may impact efficient feeding and development in the wild. The current situation of enriched habitat conditions with unnaturally high suspended solids concentrations are assumed for this table.

Table 4. Summary of main periods during the first approximately 4 years post-excystment outlining whether the mortality risk is deemed to be low, medium or high and which factors may contribute to increased mortality during those periods.

Possible mitigation measures are also provided.

Gill reflection begins to occur around the middle/end of the second growth season (shell length >1.20 mm; 13-16 months old). Gill folding has the effect of increasing the surface area of the ctenidia relative to the dimensions of the mantle cavity. The effect of gill folding, in terms of the rates at which water transport and filtration occur, has not yet been directly tested. It would be useful and interesting to extend measurement of the allometric relationship between mass-specific filtration rate and size [61] into the size range of this developmental transition. Another area ripe for investigation is the relationship between filtration rate, interstitial space, and hydraulic conductivity of sediments, which together may determine when juvenile M. margaritifera must migrate to the substrate surface.

Whilst this investigation considered juveniles from only one population of M. margaritifera, some important conclusions can be drawn about how juvenile ontogeny may affect survival in captivity:

  1. Reported mortality during the first few months post-excystment [5,22,72] is not related to ctenidia organogenesis. This high mortality rate does not correlate with substantial ontogenic changes in gill morphology. Insufficient availability of appropriate food particles or feeding avoidance behaviour/over-production of pseudofaeces due to turbid habitat conditions may result in a net loss of energy, leading to mortality. Future work should focus on initial juvenile quality and optimising environmental conditions for the youngest/most vulnerable juveniles.
  2. Juveniles may be particularly sensitive to stress factors when gill reflection commences. Additional somatic reserves may be required to undergo significant morphological changes and the timing of reflection suggests that juveniles may have to store additional nutritional reserves over the second growth season to meet this proposed increased demand. Therefore, stress factors (Table 4) should be kept to a minimum during the second growth season and second winter. Future studies should consider the abundance of compounds such as lipids and polysaccharides for juvenile growth at all stages of development.
  3. The complex nature of laterofrontal cirri, the high numbers of cilia per laterofrontal cirrus and small inter-cirral distance implies that juvenile M. margaritifera are capable of filtering particles < 2 μm. This may limit the ecological niche of M. margaritifera to oligotrophic streams and may partially explain why recruitment has stalled in most surviving populations. This hypothesis is supported by the findings of ecological studies on the loss of juvenile function with increased catchment intensification and loss of oligotrophic conditions [73, 74].

These observations and initial measurements have increased our understanding of early Margaritifera development. Ontogenic studies are important to complement ecological and genetic studies and help improve understanding of endangered species such as the freshwater pearl mussel in order to develop comprehensive management plans.

Supporting information

S1 Table. Number of cilia per laterofrontal cirrus in various lentic and lotic freshwater and marine bivalves.



  1. 1. Lopes-Lima M, Teixeira A, Froufe E, Lopes A, Varandas S, Sousa R. Biology and conservation of freshwater bivalves: past, present and future perspectives. Hydrobiologia. 2014;735(1):1–13.
  2. 2. Barnhart MC. 100 years of mussel propagation: What have we learned and where are we going? In: 2nd international rearing of Unionoid mussels [Internet]. Clervaux, Luxembourg.; 2015.
  3. 3. Moorkens EA. Margaritifera margaritifera. The IUCN Red List of Threatened Species 2011: e.T12799A3382660. Downloaded on 26 February 2016.
  4. 4. Preston SJ, Keys A, Roberts D. Culturing freshwater pearl mussel Margaritifera margaritifera: a breakthrough in the conservation of an endangered species. Aquat Conserv Mar Freshw Ecosyst. 2007;17:539–49.
  5. 5. Lavictoire L, Moorkens E, Ramsey AD, Sinclair W, Sweeting RA. Effects of substrate size and cleaning regime on growth and survival of captive-bred juvenile freshwater pearl mussels, Margaritifera margaritifera (Linnaeus, 1758). Hydrobiologia. 2016;766(1):89–102.
  6. 6. Gum B, Lange M, Geist J. A critical reflection on the success of rearing and culturing juvenile freshwater mussels with a focus on the endangered freshwater pearl mussel (Margaritifera margaritifera L.). Aquat Conserv Mar Freshw Ecosyst. 2011;21(7):743–51.
  7. 7. Hruška J. Nahrungsansprüche der Flussperlmuschel und deren halbnatürliche Aufzucht in der Tschechischen Republik. Heldia. 1999;4(6):69–79.
  8. 8. Eybe T, Thielen F, Bohn T, Sures B. The first millimetre–rearing juvenile freshwater pearl mussels (Margaritifera margaritifera L.) in plastic boxes. Aquat Conserv Mar Freshw Ecosyst. 2013;23(6):964–75.
  9. 9. Lange M. Experiences with the rearing of freshwater pearl mussels within the Interreg III A Project, “Flussperlmuschel Drieländereck”. In: Vandré R, Schmidt C, editors. Proceedings from the Workshop “Pearl Mussel Conservation and River Restoration.” Bezirk Oberfranken & Anglerverband Südsachsen Mulde/Elster e.V.; 2005. p. 47–52.
  10. 10. Skinner A, Young M, Hastie L. Ecology of the freshwater pearl mussel. Conserving Natura 2000 Rivers Ecology Series No. 2 English Nature, Peterborough.
  11. 11. Hastie LC, Young MR. Freshwater pearl mussel (Margaritifera margaritifera L) glochidiosis in wild and farmed salmonid stocks in Scotland. Hydrobiologia. 2001;445(1–3):109–19.
  12. 12. Cunjak RA, McGladdery SE. The parasite-host relationship of glochidia (Mollusca: Margaritiferidae) on the gills of young-of-the-year salmon (Salmo salar). Can J Zool. 1991;69:353–358.
  13. 13. Murzina SA, Ieshko EP, Zotin AA. The freshwater pearl mussel Margaritifera margaritifera L.: Metamorphosis, growth, and development dynamics of encysted glochidia. Biol Bull. 2017;44(1):6–13.
  14. 14. Buddensiek V, Engel H, Fleischauer-Roessing S, Waechtler K. Studies on the chemistry of interstitial water taken from defined horizons in the fine sediments of bivalve habitats in several northern German lowland waters. 2. Microhabitats of Margaritifera margaritifera L., Unio crassus (Philipsson) and Unio tumidus. Arch für Hydrobiol Stuttgart. 1993;127(2):151–66.
  15. 15. Moorkens EA, Killeen IJ. Assessing near-bed velocity in a recruiting population of the endangered freshwater pearl mussel (Margaritifera margaritifera) in Ireland. Aquat Conserv Mar Freshw Ecosyst. 2014;24(6):853–62.
  16. 16. Santos RMB, Sanches Fernandes LF, Varandas SGP, Pereira MG, Sousa R, Teixeira A, et al. Impacts of climate change and land-use scenarios on Margaritifera margaritifera, an environmental indicator and endangered species. Sci Total Environ. 2015;511:477–88. pmid:25574975
  17. 17. Schartum E, Mortensen S, Pittman K, Jakobsen PJ. From pedal to filter feeding: Ctenidial organogenesis and implications for feeding in the postlarval freshwater pearl mussel Margaritifera margaritifera (Linnaeus, 1758). J Molluscan Stud. 2017;83(1):36–42.
  18. 18. Araujo R, Campos M, Feo C, Varela C, Soler J, Ondina P. Who wins in the weaning process? Juvenile feeding morphology of two freshwater mussel species. J Morphol. 2018; 279:4–16. pmid:28921628
  19. 19. Lavictoire L. Chapter 5: Investigations into the transformation from pedal to filter feeding in the freshwater pearl mussel Margaritifera margaritifera. In: Juvenile biology and captive rearing of the freshwater pearl mussel Margaritifera margaritifera. PhD thesis, Lancaster University; 2017. p. 115–190.
  20. 20. Reid RGB, McMahon RF, Ó Foighil D, Finnigan R. Anterior inhalant currents and pedal feeding in bivalves. The Veliger. 1992;35:93–104.
  21. 21. Yeager MM, Cherry DS, Neves RJ. Feeding and burrowing behaviours of juvenile rainbow mussels, Villosa iris (Bivalvia: Unionidae). J North Am Benthol Soc. 1994;13(2):217–22.
  22. 22. Gatenby CM, Parker BC, Neves RJ. Growth and survival of juvenile rainbow mussels, Villosa iris (Lea, 1829) (Bivalvia: Unionidae), reared on algal diets and sediment. Am Malacol Bull. 1997;14(1):57–66.
  23. 23. Gatenby CM, Neves RiJ, Parker BC. Influence of sediment and algal food on cultured juvenile freshwater mussels. J North Am Benthol Soc. 1996;15(4):597–609.
  24. 24. Hanlon SD. Release of juvenile mussels into a fish hatchery raceway: a comparison of techniques. MSc. Thesis, Viginia Polytechnic Institute and State University. 2000.
  25. 25. Kovitvadhi S, Kovitvadhi U, Sawangwong P, Thongpan A, Machado J. Optimization of diet and culture environment for larvae and juvenile freshwater pearl mussels, Hyriopsis (Limnoscapha) myersiana Lea, 1856. Invertebr Reprod Dev. 2006;49(1–2):61–70.
  26. 26. Ward J, Sanford L, Newell R, MacDonald B. The utility of in vivo observations for describing particle capture processes in suspension-feeding bivalve molluscs. Limnol. 2000;45(5):1203–10.
  27. 27. Ward JE, MacDonald BA, Thompson RJ, Beninger PG, Macdonald LBA, Thompson RJ, et al. Mechanisms of suspension feeding in bivalves: Resolution of current controversies by means of endoscopy. Limnol Ocean. 1993;38(2):265–72.
  28. 28. Riisgård HU, Larsen PS, Nielsen NF. Particle capture in the mussel Mytilus edulis: the role of latero-frontal cirri. Mar Biol. 1996;127:259–66.
  29. 29. Owen G, McCrae JM. Further studies on the latero-frontal tracts of bivalves. Proc R Soc London B Biol Sci. 1976;194(1117):527–44.
  30. 30. Jørgensen CB. Bivalve filter feeding revisited. Mar Ecol Prog Ser. 1996;142(1–3):287–302.
  31. 31. Riisgård HU, Funch P, Larsen PS. The mussel filter-pump—present understanding, with a re-examination of gill preparations. Acta Zool. 2015;96(3):273–82.
  32. 32. Silverman H, Lynn JW, Dietz TH. Particle capture by the gills of Dreissena polymorpha: structure and function of latero-frontal cirri. Biol Bull. 1996;191(1):42–54. pmid:29220214
  33. 33. Henley WF, Zimmerman LL, Neves RJ. Design and evaluation of recirculating water systems for maintenance and propagation of freshwater mussels. N Am J Aquac. 2001;63:144–55.
  34. 34. Fitt WK, Fisher CR, Trench RK. Larval biology of Tridacnid clams. Aquaculture. 1984;39:181–95.
  35. 35. Beninger PG, Dwiono SAP, Le Pennec M. Early development of the gill and implications for feeding in Pecten maximus (Bivalvia: Pectinidae). Mar Biol. 1994;119(3):405–12.
  36. 36. Cannuel R, Beninger PG. Gill development, functional and evolutionary implications in the Pacific oyster Crassostrea gigas (Bivalvia: Ostreidae). Mar Biol. 2006;149(3):547–63.
  37. 37. Veniot a, Bricelj VM, Beninger PG. Ontogenetic changes in gill morphology and potential significance for food acquisition in the scallop Placopecten magellanicus. Mar Biol. 2003;142:123–31.
  38. 38. Neumann D, Kappes H. On the growth of bivalve gills initiated from a lobule-producing budding zone. Biol Bull. 2003;205(1):73–82. pmid:12917224
  39. 39. Sweeting RA, Miles L. Freshwater pearl mussel ark project: Second year report (November 2008—October 2009). Freshwater Biological Association; 2010.
  40. 40. Sweeting RA, Miles L. Freshwater pearl mussel ark project: First year report (April 2007—October 2008). Freshwater Biological Association; 2008.
  41. 41. Galbraith HS, Frazier SE, Allison B, Vaughn CC. Comparison of gill surface morphology across a guild of suspension-feeding Unionid bivalves. J Molluscan Stud. 2009;75:103–7.
  42. 42. Silverman H, Achberger EC, Lynn JW, Dietz TH. Filtration and utilization of laboratory-cultured bacteria by Dreissena polymorpha, Corbicula fluminea, and Carunculina texasensis. Biol Bull. 1995;189(3):308–19. pmid:29244572
  43. 43. Beninger PG, Veniot A, Poussart Y. Principles of pseudofeces rejection on the bivalve mantle: Integration in particle processing. Mar Ecol Prog Ser. 1999;178:259–69.
  44. 44. Atkins D. On the ciliary mechanisms and interrelationships of lamellibranchs. II. Sorting devices on the gills. Q J Microsc Sci. 1937;79:339–73.
  45. 45. Ortmann AE. A monograph of the najades of Pennsylvania. Parts I and II. Mem Carnegie Museum. 1911;4(6):279–347.
  46. 46. Ortmann AE. The anatomical structure of certain exotic naiades compared with that of the North American forms. Nautilus. 1911;24(October):103–31.
  47. 47. Lefevre G, Curtis W. Reproduction and parasitism in the Unionidae. J Exp Zool. 1910;9:79–115.
  48. 48. Gui Y, Zamora LN, Dunphy B, Jeffs AG. Understanding the ontogenetic changes in particle processing of the greenshellTM mussel, Perna canaliculus, in order to improve hatchery feeding practices. Aquaculture. 2016;452:120–7.
  49. 49. Yonge CM. The pallial organs in the aspidobranch Gastropoda and their evolution throughout the Mollusca. Philos Trans R Soc Lond B Biol Sci. 1947;232:443–518. pmid:20296045
  50. 50. Trump MP. Feeding morphology of juvenile Utterbackia imbecillis (Bivalvia:Unionidae). MSc. Thesis, Wake Forest University; 2010.
  51. 51. Owen G. Studies on the gill of Mytilus edulis: The eu-latero-frontal cirri. Proc R Soc B Biol Sci. 1974;187(1086):83–91.
  52. 52. Kovitvadhi S, Kovitvadhi U, Sawangwong P, Machado J. Morphological development of the juvenile through to the adult in the freshwater pearl mussel, Hyriopsis (Limnoscapha) myersiana, under artificial culture. Invertebr Reprod Dev. 2007;50(4):207–18.
  53. 53. Cannuel R, Beninger PG, McCombie H, Boudry P. Gill Development and its functional and evolutionary implications in the blue mussel Mytilus edulis (Bivalvia: Mytilidae). Biol Bull. 2009;217(2):173–88. pmid:19875822
  54. 54. Tankersley RA. Multipurpose gills: effects of larval brooding on the feeding physiology of freshwater unionid mussels. Invertebr Biol. 1996;115(3):243–55.
  55. 55. Atkins D. On the ciliary mechanisms and interrelationships of lamellibranchs. Part III: Types of lamellibranch gills and their food currents. Q J Microsc Sci. 1937;79:375–421.
  56. 56. Ansell AD. The functional morphology of the larva, and the postlarval development of Venus striatula (Da Costa). J Mar Biol Assoc UK. 1962;42:419–43.
  57. 57. Atkins D. On the ciliary mechanisms and interrelationships of lamellibranchs. Part VII: Latero-frontal cilia of the gill filaments and their phylogenetic value. Q J Microsc Sci. 1938;80:345–435.
  58. 58. Riisgård HU, Funch P, Larsen PS. The mussel filter–pump–present understanding, with a re-examination of gill preparations. Acta Zool. 2015;96(3):273–82.
  59. 59. Ward JE, Shumway SE. Separating the grain from the chaff: Particle selection in suspension- and deposit-feeding bivalves. J Exp Mar Bio Ecol. 2004;300(1–2):83–130.
  60. 60. Lasee BA. Histological and ultrastructural studies of larval and juvenile Lampsilis (Bivalvia) from the upper Mississippi River. PhD Thesis, Iowa State University. 146 p. 1991.
  61. 61. Silverman H, Cherry JS, Lynn JW, Dietz TH, Nichols SJ, Achberger E. Clearance of laboratory-cultured bacteria by freshwater bivalves: differences between lentic and lotic unionids. Can J Zool. 1997;75(11):1857–66.
  62. 62. Riisgård HU. On measurement of filtration rate in bivalves-the stony road to reliable data: review and interpretation. Mar Ecol Prog Ser. 2001;211:275–91.
  63. 63. Riisgård HU. Efficiency of particle retention and filtration rate in 6 species of Northeast American bivalves. Mar Ecol Prog Ser. 1988;45:217–23.
  64. 64. Baker SM, Levinton JS. Selective feeding by three native North American freshwater mussels implies food competition with zebra mussels. Hydrobiologia. 2003;505(1–3):97–105.
  65. 65. Jørgensen CB. Bivalve Filter Feeding, Hydrodynamics, Bioenergetics, Physiology and Ecology. Olsen & Olsen, Denmark; 1990. 144p.
  66. 66. Ellis MM. Erosion silt as a factor in aquatic environments. Ecology. 1936;17:29–42.
  67. 67. Aldridge DW, Payne BS, Miller AC. The effects of intermittent exposure to suspended solids and turbulence on three species of freshwater mussels. Environ Pollut. 1987;45:17–28. pmid:15092759
  68. 68. Passos FD, Domaneschi O, Sartori AF. Biology and functional morphology of the pallial organs of the Antarctic bivalve Mysella charcoti (Lamy, 1906) (Galeommatoidea: Lasaeidae). Polar Biol. 2005;28:372–80.
  69. 69. Lopes-Lima M, Teixeira A, Froufe E, Lopes A, Varandas S, Sousa R. Biology and conservation of freshwater bivalves: past, present and future perspectives. Hydrobiologia. 2014;735:1–13.
  70. 70. Graf DL, Cummings KS. Palaeoheterodont diversity (Mollusca: Trigonioida plus Unionoida): what we know and what we wish we knew about freshwater mussel evolution. Zool J Linn Soc. 2006;148(3):343–94.
  71. 71. Bogan AE, Roe KJ. Freshwater bivalve (Unioniformes) diversity, systematics, and evolution: status and future directions. J North Am Benthol Soc. 2008;27:349–69.
  72. 72. Jones JW, Mair RA, Neves RJ. Factors affecting survival and growth of juvenile freshwater mussels cultured in recirculating aquaculture systems. N Am J Aquac. 2005;67(3):210–20.
  73. 73. Moorkens EA. Addressing the conservation and rehabilitation of Margaritifera margaritifera (L.) populations in the Republic of Ireland within the framework of the Habitats and Species Directive. J Conchol. 2010;40(3):339–50.
  74. 74. Österling ME, Arvidsson BL, Greenberg LA. Habitat degradation and the decline of the threatened mussel Margaritifera margaritifera: influence of turbidity and sedimentation on the mussel and its host. J Appl Ecol. 2010;47(4):759–68.