ORIGINAL RESEARCH article

Front. Mar. Sci., 30 July 2026

Sec. Marine Conservation and Sustainability

Volume 13 - 2026 | https://doi.org/10.3389/fmars.2026.1875098

In search of resilient sponges as candidate habitat engineers for the renaturalization of polluted harbor environments

  • 1. Center for Advanced Studies of Blanes (CEAB-CSIC), Department of Marine Ecology, Blanes, Girona, Spain

  • 2. Ocean Ecostructures S. L., World Trade Centre, Barcelona, Spain

Abstract

Harbor infrastructures impose strong environmental stress on coastal marine communities, limiting the persistence of many native species and hindering ecological restoration. Identifying resilient habitat-engineering organisms is therefore key to developing effective and affordable renaturalization strategies. In this study, we evaluated the suitability of five Mediterranean sponge species to withstand the process of collection, laboratory manipulation, and transplantation to a recreational harbor, where they were monitored over 455 days using image-based morphometric analyses coupled with computational modeling. Among the species tested, only Chondrosia reniformis showed high resilience under harbor conditions, with 91.7% survival. In contrast, all individuals of Corticium candelabrum, Crambe crambe, Scalarispongia scalaris, and Ircinia oros died during laboratory preparation or shortly after deployment in the harbor. Individuals of C. reniformis initially underwent a moderate size reduction but subsequently stabilized their biomass. The sponges demonstrated remarkable morphological plasticity, including continuous substrate crawling (up to 12.7 cm of total displacement) and the capacity to recover from disease. Some individuals underwent body fission during warm months, generating clonal offspring that nearly doubled the size of the experimental population in the harbor. Collectively, the results indicate that the high survival of this species in the degraded harbor environment, together with its resistance to disease, its ability to proliferate clonally, and its capacity to reposition itself for improved environmental adaption, makes C. reniformis a highly suitable candidate as putative habitat engineer in future multitrophic approaches aimed at enhancing biodiversity and renaturalizing degraded harbor environments and other anthropogenically impacted coastal habitats.

1 Introduction

Human societies have traditionally settled and expanded along coastal areas where they often have greater access to basic resources. The increasing development of coastal human activities have caused important alteration or nearly complete loss of many natural, coastal habitats (Kennish, 2002; Pontes et al., 2021). Among the many human impacts on the coastal systems, commercial, fishing, and recreational harbors emerge as a major environmental pressure. Harbor construction often involves drastic habitat alterations for the species in adjacent areas (Palanques et al., 2022). Furthermore, the relatively restricted water circulation in most harbors favors accumulation of multiple microbiological, organic and chemical pollutants. Such a multifaceted impact is known to negatively affect the physiology, growth, health, behavior, and survival of marine species within harbors and in adjacent areas (Moreira et al., 2017; Agarwala and Saengsupavanich, 2023). Additionally, harbors are important entry points for allochthonous (i.e., non-native) species via ballast water (Carlton and Geller, 1993; Kurniawan et al., 2022) and as biofouling on ship hulls (Sylvester et al., 2011; Georgiades et al., 2021).

Because of their fundamental contribution to human societal development, historic harbors remain operational, and new ones continue to be constructed. In this context of ongoing harbor expansion and activity, the prospect of fully restoring the ecological function within harbors appears unrealistic. Yet, a growing number of initiatives worldwide are aimed at mitigating the ecological alterations that harbors impose on marine coastal ecosystems. A plausible strategy is to promote establishment of a variety of local species from different trophic levels but all capable of surviving the polluted harbor environment to subsequently act as pioneering habitat engineers. It is therefore necessary to identify organisms that can adapt well to harbor environments and help to directly or indirectly improving the environment and the availability of resources for other species. The underlying idea is that employment of “habitat engineers” operating at multiple trophic levels should accelerate the rehabilitation of at least part of the ecological function in the midterm (Ribeiro et al., 2026).

Among the pull of potential candidate benthic species to be used as habitat engineers are sponges (Phylum Porifera). These organisms represent a paradigmatic example of ecological success; they are not only ubiquitous in modern oceans but have also thrived over the last 600 million years, proliferating with notable accomplishment across diverse paleoceanic scenarios (Laubenfels and Moore, 1955; Wiedenmayer, 1994; Finks et al., 2004). Over the past decades, scientists have established that sponges play a crucial role in the ecological functioning of a large variety of marine benthic communities (reviewed in: Bell, 2008; Maldonado et al., 2012, 2017; de Goeij et al., 2017; Folkers and Rombouts, 2020). Their functional role often unfolds through complex mechanisms based on the interplay between biogeochemical processes and organismal interactions. It is long known that the three-dimensional body of many sponges increases the 3D structure of habitats, providing a multiplicity of physical refugia for small organisms that, in turn, attract a larger variety of others and their respective predators. This refuge effect of sponges is, for example, well known to foster local biodiversity of both invertebrates and fish (Tyler and Böhlke, 1972; Voultsiadou-Koukoura et al., 1987; Klitgaard, 1995; Pérez-Botello and Simões, 2021). In addition to favoring biodiversity through their role as refugia, sponges and their microbiomes interact with a variety of essential dissolved organic and inorganic nutrients, affecting the cycling of pivotal chemical elements, including carbon, silicon, nitrogen, phosphorus, and sulfur. Thus, sponges are habitat engineers not only because they add tridimensionality to benthic habitats, but also because they directly modulate the availability of nutrients that are crucial to the development of bacterioplankton and phytoplankton communities and that, in the end, regulate food and energy transfer from the seawater to the food webs (Corredor et al., 1988; Bell, 2008; Maldonado et al., 2012, 2021; de Goeij et al., 2013, 2017; López-Acosta et al., 2019; de Kluijver et al., 2021).

Because of their important functional roles, sponges should be considered in future multitrophic strategies for harbor renaturalization. However, these organisms may be sensitive to poor water quality and accumulation of contaminated harbor sediments, potentially limiting their suitability for deployment in degraded harbor environments (Johnston and Clark, 2007; Maldonado et al., 2008). Considering the high costs associated with restoration and renaturalization projects, selecting species with a strong likelihood of performing successfully under target environmental conditions is essential. Accordingly, the present study aims to identify sponge species suitable for incorporation into future renaturalization frameworks for degraded harbor environments. Specifically, we evaluated five common Mediterranean coastal sponge species after their translocation from their natural habitat to the degraded environment of a recreational port, assessing their tolerance to handling, attachment to experimental substrates, survival, growth, putative behavioral adaptations, and proliferative potential. It is important to note that this study was not designed to assess the renaturalization effects of these sponges. Rather, it represents a prior research step aimed at identifying pollution-tolerant sponge species that could be incorporated into future multitrophic strategies for the ecological recovery and renaturalization of degraded harbor environments.

2 Materials and methods

2.1 Study site and basic logistics

The study was conducted into recreational port “Marina Palamós”, located on the Palamós coast (Girona, Spain; Figures 1A, B). We built two 4 x 2 m, grided, metal structures that were coated by a 2 mm thick layer of calcium carbonate obtained by electrochemically enhanced deposition, a technique commonly used to build artificial reefs (Edvardsen et al., 2020; Robinson et al., 2024). These benthos-friendly structures, hereafter named as Life Boosting Units (LBUs), were fixed to a harbor seawall, at depths from 2 to 6 m, and about 10 m from the port fuel station. The LBUs were used to support the experimental plates bearing the sponges, which were positioned in vertical and overhanging orientations (Figures 1C–H) to mimic the preferred orientations exhibited by the sponges in their natural habitats.

Figure 1

2.2 Sponge species and attachment plates

For the initial assays, we selected five common sublittoral sponge species representing distinct phylogenetic lineages: Corticium candelabrum (Class Homoscleromorpha; Order Homosclerophorida), Chondrosia reniformis (Class Demospongiae; Order Verongiida), Crambe crambe (Class Demospongiae; Order Poecilosclerida), and Scalarispongia scalaris and Ircinia oros (Class Demospongiae; Order Dictyoceratida). Although these species belong to different phylogenetic groups, they coexist within the same habitat and are therefore exposed to nearly identical natural environmental conditions, including temperature, salinity, pH, UV radiation, turbidity, sediment deposition, food availability, and predation pressure. All five species are abundant on vertical and overhanging rocky walls, predominantly north-facing, at depths of 7–12 m at Santa Anna Point (41°40’22.4”N, 2°48’15.2”E; Blanes, Girona, Spain), where the specimens were collected.

These species were selected due to their high abundance at the collection site, an ecological success considered potentially advantageous for successful establishment and performance in degraded harbor environments. In addition, these species were selected to cover a broad range of contrasting anatomical and functional traits, allowing the identification of potential shared features associated with successful performance under polluted harbor conditions. Corticium candelabrum, the only representative of the relatively uncommon class Homoscleromorpha, is a high-microbial-abundance species (HMA) with a siliceous skeleton composed of minute spicules. This species can reattach to experimental substrata and survive for more than one year under aquarium conditions (Maldonado, unpublished information), although its performance in degraded harbor environments has not yet been evaluated. The remaining species belong to the class Demospongiae. Chondrosia reniformis is also a HMA species, but lacks both mineral and spongin skeletons. This sponge is known for its remarkable tolerance to handling and transplantation, as well as its ability to withstand various forms of coastal pollution (Wilkinson and Vacelet, 1979; Cebrian et al., 2006; Bertolino et al., 2016; Gökalp et al., 2019; Nemoy et al., 2021). Crambe crambe is arguably the most abundant coastal sponge species in the Mediterranean Sea. It possesses a siliceous skeleton whose degree of development varies substantially as a function of the environmental availability of dissolved silicic acid. Unlike the other species tested, it is a low-microbial-abundance (LMA) sponge, and is relatively difficult to manipulate and to maintain under laboratory conditions, showing only modest survival rates (Maldonado, unpublished information). Yet, it has been shown to persist for several months when translocated to polluted harbor environments (Cebrian et al., 2003). Scalarispongia scalaris and Ircinia oros are HMA species characterized by skeletons composed of irregular networks of spongin fibers and the absence of mineral skeletal elements. Both species exhibit relative resistance to disease outbreaks associated with heatwaves, which have severely affected other dictyoceratid sponges, including Ircinia fasciculatus (Maldonado et al., 2010; Garrabou et al., 2022) and Sarcotragus foetidus (Dinçtürk et al., 2023).

The sponge collection site is about 30 km south from “Marina Palamós” harbor (Figure 1A), where sponges were finally translocated. Upon collection, sponges were immediately transported alive to the Laboratory for Experimentation with Living Organisms (LEOV) in the Center for Advanced Studies of Blanes (CEAB), where they were transferred to 500 L aquarium tanks provided with unfiltered running seawater. Under these laboratory conditions, sponges were attached to 16.5 x 16.5 cm, experimental plates made from recycled ceramics. We initially tested three plate sizes (16.5 x 16.5 cm or 14.5 x 14.5 cm or 10.5 x 10.5 cm) before optimizing the logistics for live transport of sponge-bearing plates by car from the LEOV laboratory to Marina Palamós (Figure 1A).

To attach the sponges to the plates, we used two different approaches, depending on the species. Based on our preliminary knowledge of the performance of C. candelabrum (Figure 1E) and C. reniformis (Figure 1F) under laboratory conditions, we knew that if individuals of these two species are placed on an appropriate substrate and not disturbed for several days, they generate a basal spongin layer and reattach to the substrate by themselves. With C. crambe, S. scalaris and I. oros, this response does not happen. Thus, we used a two-component underwater epoxy resin, following a method previously used successfully by our team to attach sponges to experimental substrata (Maldonado and Young, 1998). Rock pieces on which thinly encrusting individuals of C. crambe were growing were glued with the epoxy putty to the ceramic plates (Figure 1G). Submassive individuals of S. scalaris and I. oros (Figure 1H) were directly glued to the plates using the putty, a method previously demonstrated to be successful for attachment of keratose demosponges (Maldonado and Young, 1998).

2.3 Sponge monitoring

Plates bearing one to three sponges were deployed on the LBUs in the harbor (Figure 1). Transplantation was conducted in batches (see Table 1). Initially, nine specimens of C. reniformis (Ch01 to Ch09) were deployed on November 19, 2024. One hundred days later (i.e., 2025/02/27), ten individuals of C. crambe were deployed (Cr01 to Cr10). A third deployment, 136 days after the initial one, involved eight additional individuals of C. reniformis (Ch10_Ch17). A fourth deployment was carried out on day 156, consisting on seven additional individuals of C. reniformis (Ch18-Ch24). A final, exploratory deployment (i.e., with low replication) was conducted on day 191 and included only dictyoceratid sponges, specifically five individuals of S. scalaris and three individuals of I. oros. Here, we may be criticized for the relatively small number of individuals used for these keratose species. This is because, as it will be explained in the Results, these species died shortly after deployment, following a mortality pattern that strongly discouraged further transplants of these species to the harbor, as they would inevitably result in mortality. Likewise, the species C. candelabrum was never transferred to the harbor, as all 15 individuals collected in two different trials failed to attach to the experimental plates offered in the LEOV and died in the laboratory. The remaining collected individuals successfully attached to the plates, either through natural settlement or artificial methods, and experienced no mortality during their stay in the wet laboratory.

Table 1

DateDayInterval
(days)
Individuals
deployed
19-11-2400Ch01 to Ch09
22-01-256464
04-02-257713
27-02-2510023Cr01-Cr10
04-04-2513636Ch10-Ch17
24-04-2515620Ch18-Ch24
29-05-2519135Ss01-Ss05; Io01-Io03
20-06-2521322
17-07-2524027
10-09-2529555
10-11-2535634
13-01-2642064
17-02-2645535

Time frame of the study summarizing deployment dates for the individuals of Chondrosia reniformis (Ch01 to Ch24), Crambe crambe (Cr01 to Cr10), Scalarispongia scalaris (Ss01 to Ss05), and Ircinia oros (Io01 to Io03), as well as the dates of additional photographic monitoring.

After deployment in the port, sponges were monitored at intervals ranging from 20 days to 64 days, over a period of 455 days (Table 1). Each monitoring action was based on digital underwater photography of each sponge individual using an Olympus Tough Stylus TG-4 camera, as well as detailed visual inspections of sponge health stage. Digital pictures obtained were subsequently analyzed through a Vision Transformer (ViT) pipeline to automatically identify the sponge bodies in the pictures, to generate binary masks of the sponge bodies, and to conduct subsequent morphological analyses. For the sake of simplicity, the information from each sponge individual was processed separately, even when up to three individuals coexisted on some plates (Figure 2). Plates placed outside the harbor that could serve as controls were not considered, as our objective was neither to compare survival between harbor and non-harbor environments nor to assess environmental differences that might drive site-specific mortality. Instead, the aim was to examine interspecific differences in survival under harbor conditions.

Figure 2

Image analysis required development of two custom-generated Python scripts (Script S1 and Scrip S2). The process is detailed in Supplementary Text Note S1. Briefly, image processing started manually by uploading each picture file in ImageJ software (Fiji is) to indicate Script S1 the sponge position into the framework of the image and, then, run Segment Anything Model (SAM2: sam2.1_1.pt; Ultralytics v8.3.103) to generate a standard binary image (Figure 2), where each pixel is either 1 (object) or 0 (background). These binary masks were rescaled and adjusted to 300 DPI resolution for being placed into a MsWord file as PNG images. Under this format, masks were provided to a second Python script (Script S2), developed to infer the morphological change and the spatial repositioning of the sponges between every two observation times in a nearly continuous mode, at a pace of 4 video frames per day. To infer these intermediate morphologies, we used an interpolation approach that combined two mathematical processes: 1) Signed Distance Field (SDF) Morphing, which transforms shapes into distance maps and physically moves the boundaries of the shapes (i.e., individual projected area) to describe growth or shrinkage; and 2) Linear Interpolation (LERP), which calculates the transition between two shapes at a given time. The interpolated SDF binary images are assembled as frames into a “mp4” video, which reconstructs in a continuous mode the most likely intermediate morphologies through which each individual has gone through during the monitoring period (Supplementary Video S1–S13).

Script S2 also calculated values of projected area (cm2) and perimeter (cm) for each individual in each video frame (Supplementary Dataset S1). By combining the information on area and perimeter, the circularity (C) of sponges was also calculated as:

Finally, Script S2 calculated the XY position of the geometric center (centroid) of each sponge individual (relative to the lower left corner of image) in every video frame. This information served to estimate, using Euclidean distance, the displacement (mm) of the sponges between frames or during other time periods of interest.

Through the monitoring period, some individuals of C. reniformis divided by fission, forming new clonal descendants. For this reason, Script S2 also incorporated a module for tagging the new individuals following their chronologic appearance. The script also infers the exact day in which each clonal individual was supposed to be “born” and tracks their subsequent displacement and morphological performance until their dead or the end of study.

2.4 Water temperature

To analyze the physical-chemical parameters characterizing the harbor environment and how they differ from those of the natural sponge habitat was beyond the scope of this study. Instead, our objective was to identify species capable of long-term survival in harbor conditions, irrespective of the specific environmental factors limiting survival. Yet, some environmental context is available. At the specific site where LBUs are placed in the Palamós Marina, which is located less than 10 m from the fuel station, seawater is relatively clear and well oxygenated. Sporadic measurements using a multiparameter probe indicate low turbidity (equivalent to 3 to 9 mg L-1) and high oxygen concentration (7.9-8.8 mg L-1). Pollution likely stems primarily from fuel spill derived from heavy recreational boating, which peaks dramatically during summer months.

Because temperature is a major modulator of both sexual and asexual reproduction in sponges (Riesgo and Maldonado, 2008; Fassini et al., 2012), as well as a main trigger of disease outbreaks (Perez et al., 2000; Maldonado et al., 2010; Dinçtürk et al., 2023; Mueller et al., 2023; Núñez-Pons et al., 2025), we recorded temperature during our dives using dive dataloggers. These harbor measurements are presented in the broader context of surface temperatures of the coastal water of Palamós (Figure 3A), obtained from the spectroradiometric readings collected by MODIS-Aqua satellite and accessed through https://www.seatemperature.org/.

Figure 3

3 Results

3.1 Survival

All 15 individuals of C. candelabrum taken to the laboratory for attachment on the settlement plates died after three to ten days. In contrast, all individuals of the remaining species successfully attached to the settlement plates by either natural or artificial means, experiencing no mortality during their laboratory period. However, once deployed in the harbor, all individuals of the species C. crambe, S. scalaris and I. oros died in the short term (Supplementary Figures 1, 2). In the case of C. crambe, 36 days after deployment, 3 individuals had died and 20 days later, all others showed substantial thinning and area reduction. One-hundred and thirty-one days after deployment, the five surviving individuals had lost about 50% of their initial extension, and disappeared 140 days after their deployment (Supplementary Figure 1). In the case of S. scalaris and I.oros (Supplementary Figure 2), the outcome was similar to that of C. crambe, with progressive body shrinkage that inexorably led to death 165 days after deployment. The unhealthy aspect observed in all these sponges shortly after their deployment in the harbor, together with their subsequent 100% mortality, discouraged us from carrying out the planned subsequent deployment batches, as all indicators suggested that additional transplants would inevitably result in mortality. Yet, given the limited sample size for these two keratose sponge species, the conclusion that they are unable to withstand harbor pollution should be considered preliminary.

In contrast, the individuals of C. reniformis transplanted into the harbor showed promising resilience. Of the 24 transplanted individuals, 22 (91.7%) survived on the plates through the entire monitoring period, which extended up to 455 days (Figure 3B; Supplementary Figure S3-S15; Supplementary Dataset S1). Importantly, the two individuals that did not persist (Ch01 and Ch09) exhibited a mortality pattern different from that observed in C. crambe, S. scalaris or I. oros: there was no progressive shrinkage. Rather, Ch09 was lost due to accidental detachment from the plate prior to its first monitoring, whereas Ch01 actively detached between days 191 and 213 (Figure 4). Prior to its intentional detachment, Ch01 underwent a pronounced body reorganization resulting in the formation of a “hanging” body portion. This morphological reshaping, well documented in previous studies as “creeping” behavior, has been regarded as a mechanism to generate body fragments for asexual propagation (Bonasoro et al., 2001; Fassini et al., 2012). Although other individuals also divided to produce clonal offspring during the study, none did so via creeping. Therefore, although we conservatively classified Ch01 and Ch09 as dead, their ultimate fate remains uncertain.

Figure 4

Importantly, three additional individuals escaped photographic monitoring during the study (Figure 3B). Individuals Ch02 and Ch11D (one of the four clonal individuals derived from the original individual Ch11) were lost from monitoring on days 295 and 360, respectively, after crawling to the edge of the plate and move onto its underside (Figure 5; Supplementary Figures S4, S9), where photographic imaging and script processing was no longer possible. Likewise, individual Ch13E (a small clone derived from original individual Ch13) squeezed through one drill hole on the plate on day 383 and migrated to the underside (Figures 4J–L; Supplementary Figure S10). As these three individuals that escaped photographic monitoring were still alive at the end of the study, they were not considered mortalities in the calculations of survival rates.

Figure 5

To emphasize the resilience and adaptability of C. reniformis, it is worth mentioning that individual Ch06 experienced a disease process immediately after its collection, when it was still attempting to attach to the settlement plate offered in the laboratory. The infection was so severe that divided the body into two portions, both of which completely recovered and successfully reattachment to their settlement plate (Figure 5). These two portions were subsequently deployed in the harbor as two separated individuals, designated Ch02 (the smaller clone) and Ch06 (the larger clone). Interestingly, Ch02 did not develop disease during the 295 days it was available for monitoring, before it migrated to the underside of the plate. However, Ch06 did develop disease (Figure 5), which first appeared 64 days after deployment. Although the infection resulted in partial tissue loss and left an evident scar, Ch06 successfully recovered by days 100-136 (Figure 5). Disease reappeared on day 240, but the sponge recovered again, being left with a second substantial scar still visible by day 295 (Figure 5). No other C. reniformis individual exhibited signs of disease during the study.

The low mortality rate among the transplanted individuals facilitated the subsequent growth of this experimental population. This occurred because seven individuals (i.e., Ch04, Ch 10, Ch11, Ch13, Ch17, Ch18, Ch20) underwent body fission (e.g., Figure 2, 6). It resulted in the asexual production of clonal offspring (see section on Clonal Propagation), which increased the number of individuals on the plates from 24 up to 40 (Figure 3B).

Figure 6

3.2 Growth and reshaping

We estimated changes in the body size of C. reniformis over time by measuring the projected area of the sponge body (Figure 7; Supplementary Dataset S1). In these thickly encrusting sponges, this metric provides a general indication of change in biomass through time. However, we are aware that projected area does not account for potential changes in body thickness, which may affect the accuracy of body area as a proxy for biomass. However, these limitations were relevant only to a single individual in the study, Ch01, which underwent substantial body reorganization as part of its “creeping behavior” (Figure 4). Such a reshaping led to a marked reduction in projected area during the first three months (Figure 7), although it reflected a substantial increase in body thickness rather than a true decrease in biomass.

Figure 7

Figure 7 shows that the majority of individuals that remained undivided experienced a reduction in their area during the first months of life in the harbor. However, after this initial period, the reduction process slowed down considerably, giving way to a period of nearly zero growth in most cases. In other words, their size became relatively stable (e.g., Ch02, Ch03, Ch05 to Ch08, Ch12, Ch21 to Ch24). A different pattern characterized all seven individuals that divided to produce clones (i.e., Ch04, Ch10, Ch11, Ch13, Ch17, Ch18). They showed a marked increase in projected area immediately prior to division, likely reflecting the extensive thinning and stretching of the body as different body regions migrated in opposite directions to produce body fissions (see Supplementary Videos S3, S7, S8, S10-S12). Newly formed clonal individuals generally exhibited a minimal decrease in projected area over time, except in the few cases where a clonal descendant subsequently divided to produce additional clones (e.g., Figure 7: Ch10, Ch11, etc.). Overall, these patterns suggest that a majority of individuals adapted well to survive in the harbor environment.

The net change in projected area for each individual over the course of the study was calculated by subtracting the area measured at the end of the monitoring period from the area recorded the first day of transplantation (Figure 8). For individuals that underwent division into clones, the final area was calculated as the sum of the final area of all resulting clones. The analysis of the data revealed mean initial and final area values of 13.2 ± 6.6 cm2 and 10.8 ± 5.7 cm2, respectively. That is, there was a small average decrease in area -2.4 ± 3.1 cm2 (i.e., negative growth), accounting for a 18.6 ± 22.1% area reduction. Despite this negative growth being of relatively modest magnitude, a t-paired test found it to be statistically significant (n= 23, t= 3.745, P = 0.001), since only 3 individuals (Ch4, Ch15, Ch17) out of the 24 initially deployed into the harbor showed net positive growth (Figure 8). It is worth noting that, in many individuals, the decrease in projected area was largely restricted to the first months after deployment. Thereafter, area values remained relatively stable over time (e.g., Figure 7: Ch02, Ch03, Ch05, Ch06, Ch07, Ch08, Ch12, Ch14, Ch22, Ch24).

Figure 8

We also examined whether the individuals that divided into clones (n=7) had experienced a greater decrease in area than those that did not divide (n=16). A Mann-Whitney rank sum test for non-normally distributed data indicated that although the median area decrease in the group of undivided individuals was -28.1% and that of the individuals that divided into clones was -19.7%, the difference is not great enough to exclude the possibility that it is produced by random variability (U = 34.000, P = 0.142).

In the calculations of growth, we did not consider the individuals that left the monitoring process after migrating to the underside of plates. However, as previously described in the literature (see Discussion), these individuals, once installed on the underside, grew at much faster rates than their counterparts remaining on the exposed side. A remarkable example is provided by the small clonal individual Ch13E, which after squeezing through a drill hole to reach the underside, it expanded its projected area from 0.50 to 2.93 cm2 in about three months, representing a 486% increase (Figures 4J-L).

Inspection of the circularity index (Figure 9) indicated that most sponges initially exhibited a high value, that is, a preferred pseudocircular shape. Yet, frequent short-term fluctuations in circularity occurred through time, reflecting a highly dynamic body reshaping as the sponges crawled onto the substratum (Figure 9; Supplementary Video S1-S13). It is worth noting that, in those sponges that generated clones (Ch04, Ch10, Ch11,Ch13,Ch17, Ch18, Ch20), circularity values decreased markedly prior to fission, as body form became increasingly elongated (Figure 9).

Figure 9

3.3 Movement

While most sponge species do not crawl on the substrate to which they are attached, C. reniformis does it. This crawling ability was long noticed (Wilkinson and Vacelet, 1979) but, to our knowledge, it has seldom been analyzed morphometrically. Herein, we have reconstructed sponge crawling in a continuous mode using SDF interpolation (Supplementary Video S1–S13).

Exploration of the cumulative displacement attained by the centroid of the individuals during their respective monitoring periods reveals that all individuals crawl on the sustrata (Figure 10), irrespective of whether it is individuals transferred from the wild to the harbor conditions or clonal individuals generated later in the harbor conditions by fission of the former ones. Individual Ch11A crawled the longest distance (12.7 cm), but several other sponges moved between 6 and 8 cm. The magnitude of these distances suggest that C. reniformis, if given enough time, can achieve substantial displacements. This repositioning may have important ecological implications, as it has been demostrated by the migration of individual Ch13E through a drill hole in the plate to reach the underside of the plate, where it grew at much faster rate than its counterpart that remained in the upper side of the plate (see Discussion).

Figure 10

The total distance crawled by the different individuals cannot be directly compared, since different individuals were monitored for different time periods. Dividing the total distance crawled (mm) by the time elapsed (days) yields the average speed (mm day-1) over the monitoring period, a metric that is comparable across individuals. Inspection of these velocity data (Figure 11; Supplementary Dataset S2) shows that the fastest individual crawled at an average speed of 424 micron day-1. This record corresponds to individual Ch01, whose persistence in the experiment (i.e., longevity) was the shortest (less than 191 days) because it detached soon after deployment as a result of creeping behavior (Figure 4). While this observation might suggest that mean crawling velocity depends on the duration of the monitoring period (i.e., longevity), no correlation was found between those two variables (Figure 11). Likewise, we examined whether the sponge size (area) was related to the average crawling speed. No such a relationship emerged from the data, despite the wide size range present, from about 0.1 cm2 (resulting from fission events) to about 23 cm2. Inspection of Figure 11 also indicates that newly formed clonal descendants move at a range of velocities comparable to that of their progenitors and similat to that of the individuals that did not undergo division. Mean crawling velocity showed only a weak relationship with total distance crawled (adjusted r2 = 0.559; Figure 11). Such a relationship was somehow expected, as higher average velocities inevitably result in greater distances when sufficient time is allowed (Figure 10).

Figure 11

3.4 Clonal propagation

Seven of the 24 individuals introduced in the harbor divided their body at least once to produce from one to five clonal descendants. See as examples Figure 2, 3, 6, S5, S9, S10, S12, S13, S14, as well as Supplementary Video S3, S7, S8, S10-S12. It is worth noting that the process of body division mostly occurred from late July to mid November, throughout the warm season (Figures 3A, B, 6). Nevertheless, it is also worth noting that this warm period is coincidental with the peak of boat usage in Marina Palamós and its associated hydrocarbon pollution.

In most cases, clonal offspring were directly derived from the body of the individuals initially deployed in the harbor. However, the interpolation suggests that, in at least certain cases, clonal individuals “born” in the harbor were able to undergo further division to generate additional clonal individuals (e.g., Supplementary Video S7, S8, S10). This global production of clonal offspring caused the size of the experimental population to rise from 24 to 40 individuals.

4 Discussion

4.1 Survival

Harbors represent polluted environments, in which many of the most locally abundant coastal species cannot survive, as it has also been the case of the sponges Ircinia oros, Scalerospongia scalaris, and Crambe, crambe. Some previous attempts to transplant small explants of the species Scopalina lophyropoda into the Blanes harbor also reported slow regression of the individuals, which ended in high mortality (Maldonado et al., 2008). In contrast, a separate study monitoring the performance of adult individuals of C. crambe deployed in Blanes harbor reported high survival (90%), despite important accumulation of heavy metals and reduction of the projected area and the reproductive output (Cebrian et al., 2003). This between-study discrepancy regarding C. crambe survival might derive from the fact that Cebrian et al. (2003) monitored the sponges for only 4 months (from February to June) while we did it for more than 15 months. In contrast, C. reniformis exhibited not only high survival and relative maintenance of body size, but also considerable adaptive potential, partially derived from its ability to reposition the body on the substrate and produce clonal offspring through body fission.

Our findings regarding the high survival of C. reniformis are in agreement with previous literature. For instance, deployment of this species in the harbor of Blanes, followed by a four-month monitoring, resulted in 100% survival (Cebrian et al., 2006). This species appears capable of tolerating environments with elevated concentrations of heavy metals, such as copper —which the sponge does not accumulate (Cebrian et al., 2006)— and mercury—which is sub-lethally accumulated in its cells to the extent that the species has been proposed as a bioindicator of mercury pollution (Roveta et al., 2022). In addition, C. reniformis has been shown to tolerate high levels of organic pollution, including conditions found near urban wastewater outfalls and fish farms (Gökalp et al., 2019, 2020). As a high-microbial-abundance sponge, it typically incorporates ammonium from seawater and releases nitrate (Jiménez and Ribes, 2007; Nemoy et al., 2021; Maldonado et al., 2023). Therefore, it is likely that its ability to cope with high organic loads is mediated by its rich and diverse microbiome, as well as by shifts in microbial community composition that may occur when the holobiont is translocated between habitats.

Although in our study C. reniformis showed remarkable resistance to infections of unidentified nature (Figure 5), it has been demonstrated that this sponge can be fatally attacked by the fungus Aspergillus tubingensis, which kills the individuals in about two weeks (Greco et al., 2019). It has also been reported that explants of this species deployed in both the polluted surroundings of fish farms and in pristine sites developed frequent infections that resulted in high mortality (Gökalp et al., 2019). These infections occurred whenever the explants were prepared in mid-july, under rising seawater temperatures (Gökalp et al., 2019). Nevertheless, to our knowledge, events of mass mortality which occur during Mediterranean heat waves in the genera Spongia, Ircinia, Sarcotragus and Petrosia have not been reported in the natural populations of C. reniformis (Osorio-Tafall and Cardenas, 1945; Perez et al., 2000; Cerrano et al., 2001; Maldonado et al., 2010; Cebrian et al., 2011; Garrabou et al., 2022; Dinçtürk et al., 2023).

4.2 Growth and clonality

Most individuals of C. reniformis exhibited a highly dynamic morphology through the monitoring period (Videos S1–S13), as it had also been reported in previous studies (Wilkinson and Vacelet, 1979; Bavestrello et al., 1998). The individuals underwent a net mean reduction in projected area of about 18.6%, which is regarded as a relatively successful response by comparison with the rapid negative growth leading to death in all other species translocated to the harbor. Earlier studies monitoring C. reniformis in the surroundings of fish farms and urban water discharges have reported positive growth and, in some case, even higher growth in the polluted sites compared to pristine sites (Gökalp et al., 2019, 2022). These contrasting results probably derive from the organic pollution characterizing aquaculture outputs being less detrimental (or even beneficial, as bacteria are food for sponges) to the sponges and their microbiomes than the mix of organic, heavy-metal, and hydrocarbon pollution characterizing harbor environments. Negative growth of C. reniformis transplants has also been reported in pristine conditions, but it appears to be associated with the sponges being exposed to direct sunlight and sediment deposition (Gökalp et al., 2019, 2022). Many sponge species are known to grow better on vertical than horizontal substrates (Knott, 1999), likely because they are less exposed to light (i.e, UV radiation) and sediment accumulation. Intense sediment deposition has been shown to negatively affect larval settlement and early juvenile survival (Maldonado et al., 2008; Abdul Wahab et al., 2019). Thus, reduced exposure to light and sediment might explain why three individual in our study (Ch02, Ch11D, Ch13E) migrated to the underside of plates, a preferential movement pattern that has also been reported in previous studies (Wilkinson and Vacelet, 1979).

Regarding body fragmentation, the ecological triggers of this process remain unclear. Unlike in our study, translocation of C. reniformis to Blanes harbor triggered no instances of fission (Cebrian et al., 2006). The absence of fissions in the latter study might be due to the relative short duration of observations and their timing, which spanned from February to June and, importantly, it did not include the warm season. High temperatures have been proposed to stimulate clone production (Fassini et al., 2012). Nevertheless, a separated two-year monitoring of wild C. reniformis individuals in a relatively pristine natural habitat reported no significant changes in biomass or body shape, as well as no instances of fission (Garrabou and Zabala, 2001). Therefore, on the basis of these two previous studies, it cannot be ruled out that the processes of body fission observed in our study were induced by specific environmental conditions in the harbor, which, in addition to organic, heavy-metal, and hydrocarbon pollution, is characterized by high seawater temperature values during summer (Figure 3A). In this regard, it has also been reported that the proportion of individuals of C. reniformis engaged in creeping behavior to produce clonal fragments in the natural communities closely mirrors changes in seawater temperature over the year, with creeping behavior peaking during the warmest months (Fassini et al., 2012).

In modular animals, such as sponges and corals, body fragmentation into functional clonal subunits may provide important ecological advantages. It may enhance survival of genotypes by transforming “mortality and genotype extinction” into just “partial mortality”, increasing the chances of species persistence in harsh or disturbed environments (Huitzil and Huepe, 2024). Likewise, clonal propagation facilitates rapid population expansion and colonization of new habitats (Oborny, 2019). Clonality may also buffer disease spread, by fragmenting and isolating the risk of infections, as exemplify in Figure 5. Formation of clones may also facilitate evolutionary innovation, since independent mutations and variation will likely arise in each clone (Huitzil and Huepe, 2024).

In addition to the asexual reproduction by body fission, C. reniformis reproduces sexually and water temperature also appears to modulate this process (Lévi and Lévi, 1976; Riesgo and Maldonado, 2008). The species is gonochoric, with no recognizable sexual dimorphism between males and females. The individuals reproduce broadcasting gametes for external fertilization and embryo development until formation of a free-swimming larva (Lévi and Lévi, 1976; Maldonado and Abdul Wahab, 2025). It has been shown that shallow-water populations of C. reniformis exhibit marked seasonality in oocyte production, with elevated density of oocytes in the mesohyl and high fecundity, whereas their mesophotic counterparts are characterized by a nearly continuous production of larger oocytes but in lower number, leading to lower reproductive output and a greater contribution of asexual reproduction to the population (Idan et al., 2020). In our study, we decided not to collect sponge tissue to assess whether the individuals transplanted to the harbor were forming gametes, because it involves repetitive wounding, which could have had unpredictable effects on survival and/or body dynamics of the individuals. Our local population of Chondrosia reniformis is known to release gametes in July (Maldonado, unpublished observations). Therefore, we carefully inspected the plates in September 2025, searching for new settlers that could derive from sexual reproduction among the transplanted individuals, but none were detected.

Finally, behaviors such as crawling and body fission, while interpreted here as flexible responses to environmental heterogeneity, may also have practical implications for restoration design. Mobility and clonal propagation could facilitate spatial self-organization, persistence, and local population expansion in engineered substrates. Understanding how these processes operate in artificial environments may therefore contribute to the development of more resilient and self-sustaining biodiversity enhancement strategies in coastal infrastructure.

4.3 Movement

The ability of C. reniformis to re-locate its position by crawling on the substrata represents an unparalleled ecological advantage relative to most other sponges, since it enables the individuals to readjust their position in relation to neighbors and environmental factors. In our study, this advantage is paradigmatically exemplified by the relatively short migration of clonal individual Ch13E, which moved through a drill hole in the plate to reach its underside. This relocation drastically modified its exposure to UV radiation, sediment deposition, food availability, spatial competitors, predators, hydrodynamics, and other factors. Coinciding with this repositioning, individual Ch13E experienced a record growth of about 486% in less than three months. In agreement with these results, Wilkinson and Vacelet (1979) showed that two individuals initially transplanted on the illuminated side of experimental plates crawled distances of 7 and 10 cm to move to the underside of the plates. In that more suitable, shaded microhabitat, the two individuals grew much more than the specimens that remained exposed to light. The ability to crawl away from neighboring organisms, whether conspecific or not, may provide a major advantage by reducing competition for both particulate food in the water column and space for growth (Bond and Harris, 1988; Maldonado and Uriz, 1999; Morganti et al., 2021). Yet, the reasons why C. reniformis crawl on the substrata remain unclear. Our interpolation videos show that all monitored individuals exhibited crawling behavior (Figure 8; Video S1-S13). This common response could be interpreted as an attempt by transplanted sponges to find locations with lower exposure to detrimental factors, such as contaminants, light and UV radiation, sediment deposition, and/or predators. However, this interpretation is not straightforward. Explants of C. reniformis deployed around fish farms under high organic loads (i.e., high food availability) showed enhanced growth but also significant migration and body fission (Gökalp et al., 2019), despite inhabiting a favorable environment in terms of food supply. Therefore, crawling and body fragmentation, though being mechanisms that enhance the chances of the sponge to better exploit available ecological niches, are not necessarily responses to environmental stress. Rather displacement may represent a routine exploratory behavior (i.e., random search) aimed at maximizing exposure to the most favorable environmental conditions, irrespective of the overall suitability of the habitat.

4.4 Implications for ecological restoration

Although this study was not designed as a restoration experiment, our results have broader implications for ecological restoration and the eco-engineering of artificial coastal habitats. In particular, the ability of Chondrosia reniformis to survive, grow, reproduce asexually, and actively reposition itself under degraded harbor conditions suggests that additional sponge species may function as promising biological components in nature-based solutions aimed at enhancing biodiversity and ecological functioning in highly modified coastal systems. Traditional ecological restoration in marine environments has focused primarily on habitat recovery in natural or semi-natural settings, whereas harbors and other artificial infrastructures have often been considered too disturbed to support meaningful ecological enhancement. However, the emerging field of eco-engineering of coastal infrastructure has challenged this view by proposing that built environments can be designed or managed to promote biodiversity and restore at least part of ecosystem functioning (i.e., renaturalization), even under persistent anthropogenic pressures (Ido and Shimrit, 2015; Firth et al., 2024; Gad et al., 2026). Within this framework, habitat-forming organisms that tolerate pollution and physical disturbance may act as foundational species capable of initiating local biogenic complexity and supporting associated communities.

Sponges are particularly relevant in this context because of their dual role as ecosystem engineers: they increase habitat structural complexity and actively mediate nutrient and energy fluxes processing bacterial loads and several forms of suspended and dissolved organic matter through their enormous filtering activity (Pham et al., 2019). Our findings indicate that at least some sponge species can persist under harbor conditions while maintaining growth and exhibiting strong behavioral plasticity. Such traits may be advantageous in engineered settings, where persistence under stress is a prerequisite for long-term functional contribution. Nevertheless, the strong species-specific differences observed in survival highlight an essential constraint for restoration-oriented applications: tolerance cannot be generalized across taxa, even within ecologically similar sponge assemblages. This reinforces the need for careful species selection when designing biodiversity enhancement or renaturalization interventions in artificial marine habitats. In this sense, our results provide a first step toward identifying candidate species that may be suitable for incorporation into future eco-engineering strategies, rather than demonstrating renaturalization outcomes per se.

5 Conclusion

After evaluating the performance of five coastal sponge species (C. candelabrum, C. crambe, S. scalaris, I. oros, and C. reniformis) as potential components of future multitrophic strategies for harbor renaturalization, only C. reniformis emerged as a suitable candidate. This conclusion is supported by the demonstrated ability of this species to self-attach to experimental substrates, tolerate harbor pollution, recover from disease, readjust its spatial position for maximizing ecological niche exploitation, and increase population size through clonal offspring.

Statements

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

The manuscript presents research on animals that do not require ethical approval for their study.

Author contributions

MM: Methodology, Writing – original draft, Data curation, Resources, Visualization, Investigation, Conceptualization, Validation, Project administration, Funding acquisition, Formal Analysis, Supervision, Writing – review & editing, Software. CE: Writing – review & editing, Investigation, Data curation, Visualization. AL: Writing – review & editing, Funding acquisition, Resources.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work has been supported by grants from three different funding programs: 1) Spanish Ministry of Sciences, Innovation, and Universities (MICIU) Program (PRO-Si, PID2023-151426NB-I00), awarded to MM. 2) Biodiversa+, the European Biodiversity Partnership under the 2021-2022 BiodivProtect joint call, co-funded by the European Comission (GA No. 101052342) and with national cofunding organizations Spanish National Research Council (CSIC) and Fundación Biodiversidad, awarded to MM. 3) ECOFOSS grant, framed in the Primary Projects (PNA-020100-2023-30) of the INNCODIS tractor project (PNA-010000-2023-1) funded by the European Union – NextGenerationEU through the PERTE Naval call of the Spanish Ministry of Industry and Tourism within the Recovery, Transformation and Resilience Plan (PRTR), awarded to AL. We acknowledge support of the publication fee by the CSIC Open Access Publication Support Initiative (PROA) through its Unit of Information Resources for Research (URICI).

Acknowledgments

The authors thank Manel Bolivar for support during scuba dives, Francesc X. Coloma for help to develop script S1 and Laia Lería for critical reading of the manuscript. Jan Tapiolas, Edgard Ballesta, Alessia Pons, and Ocean Ecostructures’ team and Port Marina Palamós’ team are also thanked for operational support during the LBUs installation and through the monitoring sessions.

Conflict of interest

Author AL was employed by Ocean Ecostructures S. L.

The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author MM declared that they were an editorial board member of Frontiers at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. A free version of ChatGPT was used to revise the English language for clarity and readability through different sections of the text. A free version of Google Gemini was used to assists in the development of scripts S1 and S2.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmars.2026.1875098/full#supplementary-material

References

Summary

Keywords

benthic communities, ecological restoration, functional ecology, marine pollution, Porifera

Citation

Maldonado M, Escarré C and Lloveras A (2026) In search of resilient sponges as candidate habitat engineers for the renaturalization of polluted harbor environments. Front. Mar. Sci. 13:1875098. doi: 10.3389/fmars.2026.1875098

Received

07 May 2026

Revised

12 June 2026

Accepted

19 June 2026

Published

30 July 2026

Volume

13 - 2026

Edited by

Nathan Knott, NSW Government, Australia

Reviewed by

Andy R. Davis, University of Wollongong, Australia

Dondy Arafat, IPB University, Indonesia

Senem Tarantini, University of Salento, Italy

Updates

Copyright

*Correspondence: Manuel Maldonado,

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics