Elemental Composition and Morphometry of Rhyssoplax olivacea (Polyplacophora): Part I—Radula and Valves

Abstract

Rhyssoplax olivacea is the most common Mediterranean polyplacophoran species; however, no information exists regarding the functionality of its skeleton. The present study describes allometric trends related to its valves and radula and examines their chemical composition using specimens from the Aegean Sea, eastern Mediterranean Sea. Differences in valve width and thickness were found among all three valves; in particular, the intermediate valve had a significantly shorter length compared to the two terminal ones. The intermediate valve exhibited different trends for valve length to valve width and valve length to valve thickness compared to the terminal valves. However, all valve morphometrics to body length appeared to adhere to a shared trend. The radula to body length exhibited positive allometry. Regarding the elemental composition, all three valves appeared to have similar elemental compositions; however, the element concentrations in the radula differed with the tooth type. Iron was always the dominant element, with the highest values reported for the major lateral tooth (83% wt%). This study provides valuable insights into the different aspects of the skeleton of R. olivacea, enabling future research to focus on the skeletal functionality from evolutionary and ecological perspectives.

1. Introduction

Chitons are marine mollusks, belonging to the class Polyplacophora, which have existed since the Carboniferous period with little change in their morphology [1]. They can be locally abundant, found in midlittoral and shallow sublittoral hard substrata boulder fields around the world, and they exert top-down control on their habitats, making them important grazers [1,2,3,4]. Rhyssoplax olivacea is primarily a Mediterranean species, and is considered the most common chiton in the Aegean and Marmara Sea [5,6]. Its presence has been recorded from the eastern Mediterranean to southern Portugal [7]. It commonly inhabits boulders, though it has also been observed living on calcareous algae, sponges and anthozoans, at depths ranging from 0.5 to 62 m [6]. Recently, its population dynamics have been described in terms of population size and growth; however, not much is known about the evolution of its ecological parameters [8].

Different chiton species utilize different trophic sources correlated with radula morphology [9]. The chiton radula is composed of rows of teeth containing large concentrations of magnetite and apatite, which are considered some of the stiffest and hardest biological materials, allowing for continuous scraping of food from hard substrates [10,11]. Its four tooth types (central, lateral I, lateral II, marginal) exhibit different elemental compositions even at intraspecific level [12].

The chiton’s shell is composed of eight skeletal valves, which is an autapomorphy amongst mollusks [7]. These structures provide protection from wave action and predation, while their morphological features are affected by environmental conditions [13]. Apart from their primary role in protection, they also influence locomotion, enabling the animal to curve into a ball when dislodged from the surface it is attached to [13,14]. The valves are made of aragonite, and both their morphology and microstructure are considered to be taxonomic characteristics [15].

The present study is the first to describe the allometry and elemental composition of the valves and radula of R. olivacea in the Mediterranean Sea using pooled data from a geographic range along the Aegean Sea. Since polyplacophorans are relatively understudied in the Mediterranean, this will provide new insights regarding the ecological functionality of the skeletal elements of one of the most widely distributed chiton species.

2. Materials and Methods

2.1. Sample Collection and Study Sites

Overall, 98 individuals of R. olivacea were collected during summer 2023 and 2024 by scuba diving, at depths between 0.5 and 2 m, and transported to the laboratory in a cooler filled with ice for further processing. The minimum body length of the collected individuals was established at 14 mm to mitigate variations in mineralogy or allometry due to differences in ontogenic stages. The study included five distinct regions within the Aegean Sea (

Table 1. Sampled Areas, number of specimen, mean depth and mean body length.

Area	Coordinates	N	Depth (m)	Mean Length (mm)
Chalkidiki Peninsula	40°23′2.73″ N, 23°55′21.19″ E	31	2–2.5	14.4
Limnos Island	39°51′28.6″ N, 25°03′47.0″ E	9	0.5–1	15.01
Pagasitikos Gulf	39°20′50.09″ N, 22°58′22.84″ E	34	0.5–1	16.69
Evoia Island	38°37′31.35″ N, 24°7′16.09″ E	15	0.5–1	20.03
Paxos Island	39°11′51.47″ N, 20°11′10.74″ E	9	0.5	17.41

).

Paxos Island is located in the Ionian Sea, a transition area characterized by several small-scale oceanographic phenomena and great biodiversity, due to its special hydrodynamic and topographic features [16]. Evoia Island is situated in the central Aegean and presents relatively high tides, while anthropogenic pressure is also noticeable [17]. Pagasitikos Gulf is a shallow, semi-enclosed system located in the northwestern part of the Aegean Sea, with an average depth of 69 m and is significantly affected by anthropogenic activities [18]. Chalkidiki Peninsula is located in the north Aegean and is dominated by the presence of the low-salinity Black Sea Water (BSW) that enters into the Aegean through the Straits of Dardanelles, shaping the oceanographic characteristics of this region [19]. Limnos Island is also situated in the north Aegean and is mainly influenced by the complex interactions between the BSW and Aegean waters, which distinguish it from Chalkidiki [20].

2.2. Morphometric Measurements

To avoid measurement errors occurring due to body contraction, a relaxation protocol was used prior to the morphometric measurements, which involved gradually switching from seawater to fresh water (up to a 1:1 ratio) [21]. For each specimen, the total length (TL) was measured using a digital vernier caliper to the nearest 0.01 mm. The radula and valves (head I, intermediate IV and tail VIII) were carefully dissected and removed from the specimens. They were then rinsed with distilled water to remove any remaining tissue. The morphometric measurements consisted of the valve length, width, thickness and radular total length (RL). All measurements were taken using a digital vernier caliper (±0.01 mm), except for valve thickness, which was measured via scanning electron microscopy (Figure 1).

2.3. Energy-Dispersive Spectroscopy (EDS)

For the elemental composition of the radula and valves, forty samples of intact radulae and valves I, IV and VIII were taken, each from the same individuals. They were first rinsed with distilled water and subsequently critical point-dried. The processed samples were subsequently mounted on metal stabs with carbon-based tape and spatter coated with 6 nm carbon by a Q150R Plus-rotary pumped coater carbon thread evaporator and were then displayed using SEM JEOL JSM 6510 (JEOL Ltd., Tokyo, Japan). For the radula, the working zone was selected and radular teeth were categorized into four tooth types (central tooth, lateral I tooth, lateral II and marginal tooth) based on their morphology and position according to the relevant literature [12]. An EDS analysis was performed using a JEOL JSM 6510 scanning electron microscope outfitted with an Oxford Ling ISIS 300 system to examine the chemical composition of the valves and radulae teeth (Figure 1). Three measurements were taken from the upper surface of each tooth and the cross-section of each valve, lasting 240 s in a 0.102 mm area.

2.4. Statistical Analysis

To compare morphometric measurements across valve types, as well as elemental concentrations across valve and radular tooth types, a one-way analysis of variance (ANOVA) was employed, combined with Tukey’s post-hoc test with Bonferroni adjusted-p correction, accounting for multiple comparisons. Prior to the analyses, homogeneity and normality were tested using Levene’s and Kolmogorov-Smirnov tests, respectively. For the allometric relationships, the function SMA from the “smatr” package was used [22]. Variables were log transformed to linearize the allometric equation (y = ax^b) [23] and each one was then subjected each to Standard Major Axis (SMA) regression [24]. SMA is a Type II regression, which minimizes the distance between the regression line and data points along both axes, accounting for error in both variables. To determine whether the slope was equal to 1 (isometry), one-sample two-tailed t-tests were conducted for each set of variables.

3. Results

3.1. Morphometric Analysis

All relationships were significant (p < 0.05), with highest coefficient of determination (R²) observed for the valve length–valve width of the head valve, whereas the lowest R² was reported for the valve length–valve thickness of the tail valve (

Table 2. Allometric relationships for valves and radula.

Relationship	Skeletal Element	Equation	R2	Allometry
Length–Width	I	log(L) = 1.082log(W) − 0.365	0.65	Isometry
	IV	log(L) = 1.309log(W) − 0.723	0.46	Positive
	VIII	log(L) = 1.062log(W) − 0.286	0.61	Isometry
Width–Thickness	I	log(W) = 0.699log(T) + 1.011	0.38	Negative
	IV	log(W) = 0.66log(T) + 1.046	0.39	Negative
	VIII	log(W) = 0.705log(T) + 0.921	0.21	Negative
Length–Thickness	I	log(L) = 0.756log(T) + 0.729	0.33	Negative
	IV	log(L) = 0.864log(T) + 0.646	0.41	Isometry
	VIII	log(L) = 0.748log(T) + 0.692	0.16	Negative
Total Length–Valve Length	I	log(TL) = 1.094log(IL) + 0.646	0.41	Isometry
	IV	log(TL) = 0.874log(IVL) + 0.819	0.27	Isometry
	VIII	log(TL) = 1.009log(VIIIL) + 0.709	0.49	Isometry
Total Length–Valve Width	I	log(TL) = 1.183log(IW) + 0.0.257	0.55	Positive
	IV	log(TL) = 1.145log(IVW) + 0.186	0.49	Isometry
	VIII	log(TL) = 1.072log(VIIIW) + 0.421	0.47	Isometry
Total Length–Valve Thickness	I	log(TL) = 0.827log(IT) + 1.453	0.27	Negative
	IV	log(TL) = 0.756log(IVT) + 1.384	0.41	Negative
	VIII	log(TL) = 0.755log(VIIIT) + 1.41	0.27	Negative
Radula Length–Total Length	Radula	log(RL) = 2log(TL) − 1.6	0.5	Positive

). The relationship between radular length and total length exhibits positive allometry, while the mean ratio of radula to total length was 0.42 ± 0.08 with minimum and maximum values of 0.23 and 0.58, respectively (Table 2, Figure 2). Regarding the valve length to valve width allometry, valves I and VIII showed isometry, whereas positive allometry was observed for valve IV (Table 2, Figure 3A). All valves exhibited negative allometry for the valve width–valve thickness relationship, while for valve length–valve thickness both terminal plates showed negative allometry, and isometry was observed for the intermediate valve (Table 2, Figure 3B,C).

Allometric relationships between valves and the whole animal showed a uniform trend for all valves, with body length–valve length exhibiting isometry and body length–valve thickness presenting negative allometry for the three valves (Table 2, Figure 3D,F,G,I,J,L). The only exception was body length–valve width, for which valves IV and VIII showed isometry, whereas valve I exhibited positive allometry (Table 2, Figure 3E,H,K).

Comparing the standardized morphometrics among valves, the two terminal valves showed no statistical differences regarding length, exhibiting higher values compared to the intermediate valve (p < 0.001, F = 28.43, df = 270, MSE = 0.00052; Figure 4A). All three valves exhibited significantly different values for width and thickness (p < 0.001, F= 289.9, df = 270, MSE = 0.0017; p < 0.001, F= 21.14, df = 270, MSE = 0.000028, respectively), with the intermediate valve showing the highest values for both variables, while the tail valve presented lowest width and the head valve lowest thickness (Figure 4B,C).

3.2. Elemental Composition

Statistical differences were observed only for the elemental composition among tooth types, but not among valves (

Table 3. One-way ANOVA results for the elemental composition among tooth and valve types.

	Radula		Valves
	F	p	F	p
Ca	33.51	<0.001	0.266	0.766
Fe	137	<0.001	0.532	0.588
Mg	35.94	<0.001	0.886	0.413
K	61.34	<0.001	0.344	0.709
P	15.16	<0.001	3.058	0.482
Si	10.7	<0.001	0.169	0.844

).

Fe exhibited the highest wt% for all tooth types, followed by Ca, while the lowest values were observed for Si (Figure 5). All four tooth types appeared to have significantly different Fe content, with the major lateral tooth and marginal tooth exhibiting the highest and lowest values, respectively (Table 3, Figure 5). The Ca content did not differ among tooth types except for the lateral II, which exhibited significantly lower values compared to the other three types. The Mg concentration showed an inverse trend to that of Fe, with all tooth types having statistically different Mg content. The lowest values were observed for the major lateral tooth, whereas the highest values were observed for the marginal tooth. No statistical variations were reported between central and lateral I and lateral II and marginal tooth, with the former two presenting lower values. The lowest K values were observed for the major lateral tooth, followed by central and lateral I, which did not show any statistical variation, while the marginal tooth exhibited the highest values. Si followed the same trend as Ca, with the lateral II tooth showing significantly lower values compared to the other three tooth types (Figure 5).

Ca exhibited the highest values among the reported elements: 93.03 ± 15.12% for the head, 93.64 ± 11.97% for the intermediate and 94.28 ± 11.33% for the tail valve. The average valve Fe was reported to be 1.76 ± 7.49%, 1.33 ± 4.17% and 3.34 ± 4.17% for valves I, IV and VIII, respectively. The Mg, P, K and Si concentrations for the head valve were at 0.64 ± 2.07%, 0.69 ± 2.76%, 0.58 ± 1.64% and 0.47± 2.83%, respectively; for the intermediate valve, they were 0.39 ± 0.88%, 0.22 ± 0.56%, 0.58 ± 0.98% and 0.35 ± 1.32%, respectively; and for the tail valve, they were 0.47 ± 1.11%, 0.23 ± 0.53%, 0.47 ± 0.61% and 0.5 ± 1.8%, respectively.

4. Discussion

It was previously suggested that the relative length of the radula is indicative of its trophic position at species level [9]. The relative length of chiton radula varies with feeding preferences, while the average values reported here indicate that R. olivacea might not be herbivores, contrary to what was previously suggested [9,25]. Radula allometry changes with the life stage for many marine mollusks, with juvenile individuals exhibiting higher relative radular length compared to adults (e.g., [26]). In our study, positive allometry was observed, suggesting that the radula of adult individuals of R. olivacea continues growing at a fast rate. This might be related to this species’ short life cycle and high growth rate [8]; however, since this is the first study to examine the allometry of chiton radula, it is unclear whether this trend is characteristic of R. olivacea or whether it also occurs in other chitons.

Differences in valve allometry have been observed inter- and intra-specifically, while valve widths can differ among valves of the same animal [14,27,28]. Valve allometry was hypothesized to be linked to habitat preferences [29], while allometric differences among valves can be a result of different growth mechanisms [30]. Acanthopleura echinata, Chiton granosus, Enoplochiton niger and Tonicia elegans exhibited isometry regarding the valve length–total length relationship, which was in accordance with our findings for R. olivacea [30]. Since all five species belong to the family Chitonidae, this trend might be related to a shared evolutionary feature rather than to the effect of habitat, as they are found in different biogeographic zones. The allometric trends regarding the valve morphometrics of R. olivacea differ between terminal and intermediate valves, as was previously shown for other species [30,31]. These differences might be necessary for the functionality of the shell as a whole, providing maximal protection during the “passive” (stretched) and “defensive” (curled) phases (see [14] for details).

The comparison of chemical composition among chiton species across various families and genera reveals significant variability (

Table 4. Summarized data regarding the elemental composition of different polyplacophoran species.

Family	Species	Region	Method	Tooth	Chemical Elements (%)	Reference
Acanthochitonidae	Acanthochitona fascicularis	North Sea	EDX	C	(A) Ca (<1%), Mg (<1%), P (1%),	[32]
	Acanthochitona fascicularis	North Sea	EDX	L1	A Fe (<1%), Ca (1%), Mg (2%),P (2%), K (1%)	[32]
	Acanthochitona fascicularis	North Sea	EDX	L2	(A) Fe (29%), Ca (6%), Mg (4%), P (9%), K (1%)	[32]
	Acanthochitona fascicularis	North Sea	EDX	M	(A) Ca (6%), Mg (<1%)	[32]
	Cryptochiton stelleri		EMPA	L2	Fe (51.8%), Ca (3.8%), P (13.3%)	[35]
	Cryptochiton stelleri	Pacific Ocean	Raman; EDS	L2	(W) Fe (69%)	[33]
Chitonidae	Acanthopleura hirtosa	Indian Ocean	ICP-AES	Whole	(μg/g) Fe (59%),Ca (19%), Mg (3%), P (13%), K (<1%), Si (<1%)	[34]
	Acanthopleura echinata	Pacific Ocean	Raman; EDS	L2	(A) Fe (<3%), Ca (32%), P (16%)	[36]
	Acanthopleura echinata	Pacific Ocean	EDS	L2	(A) Fe (66%), Ca (<1%), P (<1%)	[37]
	Acanthopleura echinata	Pacific Ocean	Raman; EDS	L2	(A) Fe (1.3%), Ca (27.3%), Mg (0.4%), P (14.2%)	[37]
	Acanthopleura miles	Indian Ocean	Raman; EDS	L2	(A) Fe (7.1%), Ca (28.4%), Mg (0.6%), P (13.6%)	[37]
	Acanthopleura spinosa	Pacific Ocean	EDS	L2	(A) Fe (~62%), Ca (<1%), Mg (<1%), P (<1%)	[38]
	Chiton mamoratus	Atlantic	Raman; EDS	L2	(A) Fe (0.8%), Ca (33.2%), Mg (0.5%), P (16.7%)	[39]
	Clavarizona hirtosa	Indian Ocean	PIXE and PIGME	L2	Fe, Ca, K, P	[40]
	Onithochiton quercinus	Indian Ocean	EDS	L2	(A) Fe (66%), Ca (<1%), Mg (<1%), P (<1%)	[38]
	Onithochiton quercinus	Indian Ocean	Raman; EDS	L2	(A) Fe (0.2%), Ca (31.3%), Mg (0.4%), P (15.3%)	[39]
	Rhyssoplax olivacea	Mediterranean	EDS	C	(W) Fe (54%), Ca (15%), Mg (3%), P (4%), K (4%), Si (<1%)	Present study
	Rhyssoplax olivacea	Mediterranean	EDS	L1	(W) Fe (47%), Ca (15%), Mg (4%), P (6%), K (4%), Si (<1%)	Present study
	Rhyssoplax olivacea	Mediterranean	EDS	L2	(W) Fe (83%), Ca (7%), Mg (2%), P (4%), K (<1%), Si (<1%)	Present study
	Rhyssoplax olivacea	Mediterranean	EDS	M	(W) Fe (36%), Ca (17%), Mg (5%), P (7%), K (5%), Si (<1%)	Present study
	Sypharochtion pelliserpentis	Pacific Ocean	Raman; EDS	L2	(A) Fe (4.2%), Ca (30.5%), Mg (0.6%), P (16.3%)	[39]
Cryptoplacidae	Cryptoplax striata	Indian Ocean	EDS	L2	(A) Fe (~92%), Mg (5.5%), K (1%), Si (1%)	[41]
Ischnochitonidae	Ischnochiton australis	Indian Ocean	EDS	L2	(A) Fe (~62%), Ca (<1%), Mg (<1%), P (<1%)	[38]
	Ischnochiton australis	Indian Ocean	Raman; EDS	L2	(A) Fe (2.7%), Ca (16.2%), Mg (6.0%), P (13.6%)	[39]
Mopaliidae	Plaxiphora albida	Indian Ocean	ICP-AES	Whole	(μg/g) Fe (87%), Ca (2%), Mg (1%), P (5%), K (<1%), Si (<1%)	[34]
	Plaxiphora albida	Indian Ocean	EDS	L2	(A) Fe (~66%), Ca (<1%), Mg (<1%), P (<1%)	[38]
	Plaxiphora albida	Indian Ocean	Raman; EDS	L2	(A) Fe (26.7%), Ca (6.5%), Mg (2.4%), P (11.2%)	[39]
Tonicellidae	Lepidochitona cinerea	North Sea	EDX	C	(A) Fe (2%), Ca (4%), K (<1%)	[32]
	Lepidochitona cinerea	North Sea	EDX	L1	(A) Ca (4%), Si (<1%)	[32]
	Lepidochitona cinerea	North Sea	EDX	L2	(A) Fe (32%), Ca (8%), Mg (3%), P (7%), K (1%)	[32]
	Lepidochitona cinerea	North Sea	EDX	M	(A) Fe (1%), Ca (5%), K (<1%), Si (<1%)	[32]

(A) Atomic%; (W) weight%.

). Most research on radula mineralogy has primarily focused on the lateral II tooth (Table 4). High concentrations of iron have been reported in the major lateral tooth (lateral II) across various polyplacophoran species, including Acanthopleura hirtosa, Lepidochitona cinerea, Cryptochiton stelleri, and Plaxiphora albida (Table 4). However, comparing the specific percentages measured is difficult due to differing methodological approaches [32,33,34]. This study identified similar concentrations of Ca, Mg, P, K, and Si to those found in the polyplacophoran species in the major lateral tooth.

Regarding the central, lateral I, and marginal tooth types, the detection of high percentages of inorganic components contradicts the results of previous studies, which suggest that these tooth types are non-mineralized [12].

R. olivacea presents the highest values for Fe% within Chitonidae and the second highest among all species following Cryptoplax striata regarding the major lateral tooth. These species are also differentiated in terms of their overall elemental composition, with C. striata exhibiting no Ca content and a much higher Mg concentration (Table 4). High concentrations of iron and possibly Mg have been shown to confer superior mechanical properties (i.e., Young’s modulus and hardness) probably related to feeding preferences [42]. This suggests that C. striata might differ from R. olivacea regarding their dietary preferences, and this suggestion is further enhanced by substantial differences in relative radular length (0.24 and 0.42, respectively) [9].

5. Conclusions

The allometry and relative size of the radula of R. olivacea appear to be related to this species’ ecology, i.e., its feeding habits and resource utilization. Regarding the valve characteristics, there seems to be little variation among valves, with the intermediate valve showing some differentiation in relative size and allometric trends. All valves appear to exhibit similar elemental composition, whereas different tooth types present distinct elemental profiles. This is potentially related to different functional roles in food acquisition, while the high concentration of iron in the major lateral tooth might have a direct influence on feeding preferences. Given this species’ extended spatial distribution, geographic and bathymetric intraspecific differences related to environmental parameters might exist. Since this is the first study to focus on the allometry and elemental composition of R. olivacea, no intraspecific comparisons were possible; thus, future research should focus on other regions to elucidate how abiotic factors might influence the examined parameters.