The lava flows are located on the west coast of the island in the Tazacorte municipality (28.6075°N, 17.9252°W) ( Figure 1 ). The lava reached the coast four times between September 29^th and December 1^st of 2021 (Carracedo et al., 2022). Before the eruption, the impacted zone consisted of a boulder field, gravel, and sand shoreline in the intertidal zone, whereas sandy bottoms predominated in the subtidal zone up to a depth of 50 m with some rocky enclaves (Martin-Garcia et al., 2013). The new coastal sections formed by the lavas included one in the old Playa de Los Guirres, which is approximately 1,900 m long, and the other in Playa de La Vina, which is approximately 650 m long. The new and heterogeneous coastline is primarily composed of cliffs that are a few meters high, along with boulders, gravel, and some platforms, although it has experienced significant erosion by waves. In the subtidal area, consolidated rocks are mixed with rocky fragments resulting from rapid cooling of the lava. The penetration of lava into the seabed was uneven. In the area of Playa de La Vina, the lava flows reaches a depth of 10 m, whereas in that of Playa de Los Guirres, the lava front reaches depths varying between 20 and 50 m, although at some points, the lava flowed down the insular slope to depths exceeding 125 m. The new rocky bottoms are highly abrupt, heterogeneous, and feature a notable slope. The predominant basaltic lavas appear stony and rough, including ‘type-aa, ‘type-pahoehoe, and ‘type-pillows. Consolidated flows of large rock blocks alternate with fragments of loose rock, varying in size and displaying irregular shapes (Sangil et al., 2023).

To compare the colonisation processes of the lava flows, we selected a nearby closed zone to the south ( Figure 1 ). This choice was made to identify a zone that was as similar as possible to the new habitats in terms of geomorphology and oceanographic conditions. In so doing, we aimed to avoid or minimise site-specific effects. The sampling sites in the control zone are located on the lava flows formed after the San Juan and Tajuya eruptions in 1949 and 1585, respectively.

Benthic rocky communities in unaffected intertidal areas consist of cespitose macroalgae, such as non-crustose coralline (geniculate) and Rhodomelaceae, and crustose coralline algae (CCA) (Sangil et al., 2005). Mollusks are common, with some species such as Phorcus sauciatus and Patella spp. that are exploited for human consumption (Lopez et al., 2012). Meanwhile, non-crustose and crustose corallines thrive along with Dictyotaceae algae and some filamentous red algae in shallow subtidal areas. Wave action prevents the establishment of macroscopic organisms on the sandy bottoms of this part of the island. Populations of the garden eel Heteroconger longissimus and some macroinvertebrates develop only in deeper (3040 m) waters (Martin-Garcia et al., 2013).

During the active eruption, anomalous data, such as pH, temperature, salinity, and CO₂ inorganic system, were registered at the interface between the lava and seawater (Gonzalez-Santana et al., 2022). When we started monitoring the benthic communities, we also began collecting data on temperature, salinity, and pH at each sampling site using a portable multiparameter sonde (HANNA) and dataloggers (HOBO^®). However, we did not observe anomalies at any of the monitored sites ( Supplementary Table S1 ), although some persisted at other locations.

Sampling commenced immediately upon receiving authorization from the authorities responsible for the emergency, and it took place in February, April-May, and July 2022, corresponding to 2, 4.5, and 7 months after the eruption had concluded, respectively. In each zone, for both the lava flows and control zone, we selected three sites for sampling in the subtidal area and three sites in the intertidal area ( Figure 1 ). In the subtidal area, sampling was performed on rocky reefs, while in the intertidal area, sampling was performed on rocky platforms. In the intertidal area, we studied all the initially formed coastal platforms, whereas in the subtidal area, sampling locations were chosen randomly, without following any specific criteria, except for ensuring a minimum distance between sites. The locations in the control zone were previously known. Fishes, macroinvertebrates, and algae were studied in the subtidal area, whereas macroinvertebrates and algae were studied in the intertidal area. Subtidal samples were obtained by scuba diving at depths ranging from 5 to 20 m. Within this depth range, we used a stratified sampling approach with three levels: 5-10 meters, 10-15 meters, and 15-20 meters, to ensure an equal distribution of samples. Intertidal sampling was conducted during the periods of maximum low tide each month. Samples were equally distributed across three intertidal levels defined for desiccation tolerance (i.e., low, medium, or high) of the different organisms collected (Lewis, 1964). The lower level ranged from 0-0.4 m above sea level during low tide, the medium level from 0.4-0.8 m, and the upper level from 0.8-1.2 m, approximately, coinciding with the regional mean tidal range. This design aimed to encompass all the spatial variability within these areas, establishing a robust method for long-term monitoring.

To estimate fish abundance, we used an in-situ point-count method (Bortone et al., 1989), in which the observer takes a position at the centre of a circle (100 m²) and records the number (abundance) and approximate size of individuals of each species for 5 min. At each depth range, three visual censuses (nine per site), minimum 25 m apart, were performed. To estimate visually in-situ the macroinvertebrate abundance (number of individuals of each species), we used the belt transect method (10 × 2 m). The minimal body size of individuals sampled was about 0.5 cm. During sampling, we checked crevices and holes but did not turn any rocks over to search for species underneath. At each level, we laid out a 50-m long tape to measure on the rocky bottom and counted the macroinvertebrates within the following intervals: 0-10 m, 20-30 m, and 40-50 m along this distance, leaving a separation of 10 m between samples. Sponges were excluded from the analysis due to the difficulty in identification. Algal cover was estimated from photographs of 25 × 25 cm quadrats divided into 5 × 5 cm. High-resolution photographs were obtained with a Nikon D700 camera housed in a Subal D700 housing and equipped with two Subtronic strobes. At each depth level, photographs were taken every 5 m along the tape measure used to estimate the abundance of macroinvertebrates, with 10 photoquadrats per level and 30 photoquadrants per site. To confirm the identity of small taxa sampled in the field, samples were collected for later confirmation in the laboratory at species or genus level. CCA, cyanophytes, were not identified at species level; instead, they were grouped for later analyses.

Intertidal sampling was conducted during the low-tide periods. Macroinvertebrate abundance was estimated also visually using 10 m × 1 m transects. We also checked crevices and holes but did not turn any rocks over, in these areas, we were able to count some gastropods that were only a few millimeters long. Three transects were established at each level. Similarly, we used a tape measure, but owing to the small area of the intertidal platforms, the transects were separated horizontally by 5 m only.

Algal cover was estimated using quadrats (25 × 25 cm). Sampling was conducted 5 m apart on the tape measure used for invertebrate estimation. At each site, seven quadrats were tested per tidal level, with 21 quadrats per site. We focused solely on the emerged surfaces; pools were not present on any of the platforms studied. The percentage cover of each taxa was estimated in-situ from its perpendicular projection onto the substrate. The cover percentage by epiphytes and other small species was also estimated in the field but species identification was conducted in the laboratory afterwards using samples taken in the field. Similarly to subtidal, some species were not identified at a specific level in the intertidal area. Many of these taxa require molecular studies for definitive species delimitation. In both the intertidal and subtidal, the algae belonging to the Order Ectocarpales and the Superclass Bacillariophyceae.s.l. grew forming tangled masses where it was impossible to separate one taxon from another, Consequently, in the analyses, both have been considered together as ectocarpaleans/diatoms.

All taxa data records for each study zone and period are shown in Supplementary Material ( Supplementary Tables S2–S6 ).

The mean species per sample of subtidal invertebrates and intertidal algae in the lava flows was lower and significantly different from that in the control zone, but there has not been a significant increase in the number of species in these groups over time ( Table 1 ; Figures 2C, D ). Indeed, little more than half of the species present in the control zone colonised the lava flows. Further, the mean species density of fishes, intertidal invertebrates and subtidal algae was also lower in lava flows in the first two sampling periods ( Figures 2A, B, E ). After seven months, species density was similar in both zones at least for fishes and subtidal algae. However, the ‘zone × time interaction effects detected by statistical analysis for these groups showed significant differences between zones at 2 and 4.5 months for the a posteriori pairwise comparisons. Further, the a posteriori analysis indicated significant differences in fishes and subtidal algae within each zone over time ( Table 1 ; Figures 2A, E ). For all groups, the species shared between both areas are common and dominant in the undisturbed benthic communities of the island ( Supplementary Tables S2–S6 ). Particularly in algae, species observed in lava flows, some of them dominant, are considered rare in typical communities.

Total species abundance also showed different group-dependent temporal patterns. In the intertidal area of the lava flows, algae rapidly covered the rocky substrate, and no differences between zones were observed from the beginning of the monitoring ( Table 1 ; Figure 2D ). Fishes and subtidal algae abundance was significant different between zones. Subtidal algae showed lower total-abundance values in the lava flows in all periods, fishes showed lower values in the two first period and higher in the last sampling but the test did not find the interaction between time and zone factors ( Figures 2A, E ). The number of intertidal macroinvertebrates in the lava flow zone was much lower than that in the control zone. Similar to the trend observed in fish populations, there was an increasing in the number of individuals in the lava flows, but this was not supported by the analysis ( Figure 2B ). On the other hand, subtidal invertebrates constituted a different case altogether; i.e., the ‘zone × time interaction was significant, and, while the mean number of individuals was similar between zones at 2 and 4.5 months, it significantly differed at seven months, as shown by the a posteriori analysis, with the number of individuals in the lava flows more than doubling that in the control zone at the latter sampling time point ( Table 1 ; Figure 2C ).

Putting all the recognized species into context for each time and area, we observed that the number of species sampled in the first two samplings was similar but increased in the last sampling ( Figure 3 ), and after 7 months, the total number of species recorded in the lava flows was 70, representing 64% of those found in the control zone. Overall, the number of species found in the lava flows for all groups was greater in the last sampling.

The analysis of the community structure found significant statistical differences between zones for all groups but following two different trend patterns. On one hand, fish and intertidal and subtidal invertebrate communities differed between lava flows and the control zone throughout the duration of the study, as shown by the statistical differences for the factor ‘zone ( Table 1 ). Additionally, differences were found for fish in relation to the factor ‘time. On the other hand, intertidal and subtidal algae communities differed between zones but interacted with the factor ‘time. The a posteriori analysis for this interaction detected differences between lava flows and the control zone at all three sampling time points, as well as within lava flows, in the ‘time factor ( Table 1 ). In most of the comparisons of the levels of the ‘T x Z terms, the level of significance obtained was lower than the Bonferroni-corrected significance level, supporting the robustness of the results ( Supplementary Table S7 ). In other words, as supported by the results of the nMDS analyses ( Figure 4 ), the intertidal and subtidal invertebrate samples within each zone ( Figures 4B, C ) were mixed and did not show significant differences over time. Alternatively, potential differences were not large enough to be considered significant. In the case of fish communities, the samples tended to be distinct across different times and zones whereas, intertidal and subtidal algal communities were distinct only in the lava flows ( Figures 4A, D, E ).

Graphically, these distinctions between zones and temporal shifts were evident when the abundances of the six dominant taxa within each community were plotted ( Figure 5 ). These six most abundant groups of taxa that contributed to total abundance as follows: fishes, 91%; intertidal invertebrates, 93%; subtidal invertebrates, 68%; intertidal algae, 81%; and subtidal algae, 65%. The SIMPER analyses ( Table 2 ) enabled us to identify the main taxa contributing to these differences. In the case of fish communities, the predominant species remained consistent across both areas under study. However, all species exhibited a higher abundance towards the control area, except for Canthigaster capistrata. Some intertidal invertebrate species were not recorded in the lava flows even after seven months. Moreover, those sampled from lava flows were consistently more abundant and closely related to the control zone. With respect to subtidal invertebrates, we observed two distinct communities: one dominated by Percnon gibbesi and Antedon bifida in the lava flows, and the other dominated by Hermodice carunculata or Holoturia sanctorii in the control zone. As for algae, in both the intertidal and subtidal zones, differences between the zones were driven by distinct taxa compositions and abundances in the communities ( Table 3 ). For instance, in the intertidal lava flows, ectocarpaleans/diatoms, Ulva spp., and Bangia atropurpurea were predominant, whereas Corallina berteroi, Jania virgata, and CCA were the most abundant taxa in the intertidal control zone. Similarly, in the subtidal lava flows, ectocarpaleans/diatoms dominated, and the CCA (notably after seven months) was significant. In contrast, Taonia atomaria, Lophocladia trichoclados, Canistrocarpus cervicornis, and CCA (the latter in the first two samples) were the most abundant taxa in the subtidal control zone. Temporal differences in the communities of lava flows were primarily attributed to changes in the abundance of taxa present throughout the study period.

All studied communities followed a similar temporal trajectory, and the differences between the lava flows and the control zone decreased over time. However, a regression was observed in invertebrate communities of the sublittoral area between months 2 and 4.5 ( Figure 6 ). Further, the composition and structure of the communities in both zones became increasingly similar at each sampling time points, with the samples growing closer to each other over time. However, when comparing different communities of organisms, distinctions were observed from the establishment of the communities, which persisted for seven months. For example, the initial fish communities that colonised the lava flows closely resembled those in the control zone. This contrasted with the subtidal algae and invertebrate communities, in which case, the differences between zones after seven months were more pronounced than the initial differences between fish. Additionally, different rates of evolution were observed among the communities. Specifically, changes in algal communities were smaller than those in fish or intertidal invertebrates.

Species-specific responses of the main taxa revealed different colonisation patterns on lava flows, even within the same group of organisms ( Table 4 ). Among fish species, Chromis limbata and Boops boops quickly colonised lava flows ( Figures 7B, C ), with no detectable differences, as early as within the first two months after the eruption. On the other hand, Thalassoma pavo and Similiparma lurida showed significant differences between zones, with populations of these species being higher in the control zone throughout the entire study ( Figures 7A, E ). As for Canthigaster capistrata and Sparisoma cretense populations, an interaction was observed between the ‘zone and ‘time factors. Thus, populations of C. capistrata were similar at two and four and a half months in both zones but significantly higher in lava flows after seven months. Meanwhile, populations of S. cretense were significantly higher at two and four and a half months in the control zone but were higher and statistically different in the lava flows at seven months ( Figures 7C, D ).

In the intertidal area of the lava flows, some characteristic species of these habitats, such as the hermit crab Clibanarius aequabilis, and gastropods Tectarius strictus and Patella piperata, were not recorded during the first seven months ( Figures 8B, C, F ). However, other gastropods were present, such as Phorcus sauciatus, Patella ordinaria, and P. aspera. Populations of P. sauciatus were significantly lower in lava flows and incipient to observe any clear trends over time, and although P. ordinaria and P. aspera were less abundant in the lava flows, no significant differences were detected in these species ( Table 4 ; Figures 8A, D, E ). A similar pattern emerged for certain sublittoral macroinvertebrates and echinoderms, such as Holoturia sanctori and Diadema africanum, which were common in the control zone but were not observed in the lava flows ( Figures 8J, K ). However, unlike in the intertidal area, populations of some species, such as Percnon gibbesi, Antedon bifida and Ascidia mentula were more abundant in the lava flows, although for A. bifida and A. mentula no significant differences were found between areas ( Table 4 ; Figures 8G, I, L ). The proliferation of P. gibbesi was extraordinary, presenting a high density of individuals in the lava flows from the beginning of monitoring. Also notably, despite being one of the most abundant species in the lava flows, Hermodice carunculata showed significantly more numerous in the control zone through the time ( Table 4 ; Figure 8H ).

Moreover, perennial algal species, such as Corallina berteroi and CCA were more abundant in the control zone. They colonised the flows slowly from two months, but their cover was significantly lower, and without very large changes between samplings to cause the interaction between the time and area factors. Jania virgata was also more abundant in the control zone. Although it was only detected in the third sampling timing, no significant differences were found ( Table 4 ; Figures 9B, E, F ). In contrast, fast-growing algae such as Ulva spp., ectocarpaleans/diatoms, and Bangia atropurpurea dominated lava flows. These taxa also showed significant temporal variations in the lava flows ( Table 4 ; Figures 9A, C, D ). Meanwhile, in the subtidal zone, algal species such as Lophocladia trichoclados, Canistrocarpus cervicornis, and Taonia atomaria established themselves in the lava flows from the beginning, but their cover was significantly lower than that of the control zone without proliferating excessively during the study ( Table 4 ; Figures 9I, J, L ). Conversely, Cotoniella filamentosa proliferated quickly and, although its cover in the flow zone was less than that in the control zone, the difference was not significant ( Table 4 ; Figure 9K ). Ectocarpaleans/diatoms were the most abundant taxa in the lava flows with limited cover in the control zone. However, these algae showed a significant decrease in cover over time under the lava flows. In contrast, CCA increased their cover in the lava flows reaching higher values by the seventh month, but were always less abundant than in the control zone ( Table 4 ; Figures 9G, H ).

The capacity of a species for dispersal is a key factor governing the connectivity between zones and the initial stages of primary succession (Cowen and Sponaugle, 2009). There are various dispersal strategies in the marine environment (Cowen et al., 2007) and, depending on their effectiveness, certain species establish themselves before others. In lava flows, species arrived through different mechanisms, including active displacement of adults or juveniles or passive dispersion aided by currents and waves via larval, germinative, or vegetative propagules.

Fishes most rapidly reached and established new populations in the lava flows, as observed in similar lava flow environments (Godwin and Kosaki, 1989). With greater mobility, adult fishes have the capacity to quickly colonize new territories (Leitao et al., 2008). Fishes encountered minimal spatial or physical barriers such as sandbanks or great depths in the study area. They were able to migrate from nearby areas without losing contact with the rocky bottoms. Furthermore, the areas surrounding the lava flows served as the primary source of species because they did not experience significant impacts. Even populations of sessile organisms survived only a few meters from the advancing incandescent lava. After seven months, the species density of fish communities in the lava flows resembled that of the control zone, although the number of individuals in the lava flows was still lower. From a trophic perspective, it seems that there was no limitation for the fishes, despite the coexistence of different feeding strategies. The lava flows were not an inhospitable habitat; some species are planktivores, others are herbivores or omnivores that could benefit from the presence of highly palatable primocolonizing algae, and for the microcarnivorous species, there was also a high number of preys, such as Percnon gibbesi. The recovery of the fish community was also positively influenced by the establishment of a marine exclusion zone surrounding the lava flows where no activities, including fishing, were permitted. For the remaining organisms, species density was significantly lower after seven months. Only subtidal macroalgae showed a similar number in the lava flows to that in the control zone. With a few exceptions, such as the macroinvertebrates Octopus vulgaris, Sepia officinalis, and Cronius ruber (of which we observed some adults), all species colonised lava flows via larvae or propagules, which slowed the colonisation process. The incorporation of individuals through larvae or propagules depends on the characteristics of the life cycle of the species. The life cycle plays a fundamental role in determining new individuals integrate the populations (Levin, 1984; Coma et al., 2000), as well as their availability at specific times. Depending on the foundational events, the trajectories of community assemblages can take different directions (Underwood and Chapman, 2006). In this context, it is important to note that many dominant invertebrate species may not have been in their reproductive period immediately after the formation of lava flows. As it has been extensively documented, the abundance of larvae from benthic invertebrate can vary significantly through the year (Wong et al., 2022; Menendez, 2023). In this study, we observed distinct patterns in the number of species and population sizes in intertidal and subtidal habitats. Specifically, in the subtidal area, many species settled in lava flows from the beginning, soon after the eruption of the volcano. Conversely, in the intertidal area, both the number of species and their population sizes were significantly reduced and were notably lower than those in the control zone. Apparently, there were no larval limitations for species colonising subtidal habitats; for instance, Percnon gibbesi larvae are present in canary waters year-round (Landeira and Lozano-Soldevilla, 2018). However, it seems that this limitation existed for intertidal species. Intertidal invertebrate communities on La Palma Island are primarily dominated by limpets (Patella spp.) and marine snails (Phorcus sauciatus). Both species exhibited multiple breeding periods throughout the year, but gamete production occurred mainly in the summer. Furthermore, veliger larvae settle rapidly after a few days (Sousa et al., 2018). Consequently, their settlements are concentrated in summer and early autumn, which did not coincide with the formation of lava flows.

Although only a few species arrived at the lava flows during the first few months, algae proliferated exponentially within a short time. Many taxa colonizing the lava flows were ephemeral species capable of producing new generations throughout the year, both through sexual and asexual reproduction, such as the ectocarpaleans. In both the intertidal and subtidal habitats, they exhibited high cover values for the lava flows, similar to those in the control zone from the beginning. The early stages of primary succession are also characterised by low interactions between organisms, and the exclusion of herbivores can alter the trajectory of algal communities (Mayakun et al., 2020; Srednick et al., 2023). The limited presence of herbivore populations in lava flows, particularly in the intertidal zone, undoubtedly contributed to the abundance of algae (Williams et al., 2000). In subtidal habitats, the number of algal taxa increased at a faster rate. It is worth noting that habitat interactions also play a role in the primary succession of algae, as previously demonstrated (Gulliksen et al., 1980). The interaction with the habitat can be as important, or even more important, than the development and feeding modes of the species (Defeo and McLachlan, 2011). Although this aspect was not observed due to the limited time duration of the study, in subsequent samplings, we observed how the remodeling of lava flows by waves influences the longer-term trajectory of benthic communities in lava flows.

The disparities in colonisation became more pronounced when we examined the specific composition of each community. The fish species that dominated the lava flows during this period were essentially identical to those that dominated the control zone. The differences between the zones stemmed primarily from the varying abundances of the dominant species in each area. Species that were sparsely abundant in the control zone and not yet present in the lava flows had minimal influence on the statistical analysis. This contrasts with the previously described primary succession in similar studies, where either early in succession (a few months) (Godwin and Kosaki, 1989) or a few years later (Pinault et al., 2013), the fish community in lava flows differs from that in other areas in terms of specific composition. Although these examples pertain to coral reef communities, where fish species are strongly linked to other benthic organisms, such as corals, and their proliferation depends on prior colonization. Studies on the colonisation of new volcanic habitats by invertebrates have also noted that the early stages of succession are dominated by a few pioneer species that are later replaced (Gulliksen et al., 2004; Bollard et al., 2013). However, in our study of invertebrates, a similar pattern emerged as with fishes and lava flows, and the control zone shared many dominant species. Generally, these species were more abundant in the control zone, although some species such as Percnon gibbesi and Antedon bifida proliferated successfully in the lava flows. Again, the weak interactions between species in these early stages, particularly in terms of carnivory, may explain the proliferation of certain populations (McCook and Chapman, 1993).

While changes in the communities of fishes and invertebrates were driven solely by a progressive increase in the number of species and their population size, which resulted in primary succession without species turnover (Mori et al., 2008), this was not the case for algae. In both the intertidal and subtidal communities of the lava flows, the transformations were more profound and involved shifts in specific composition ( Supplementary Figure S1 ), aligning with typical seaweed succession on new substrates (Serisawa et al., 1998; Jung and Choi, 2022). Further, as observed in previous studies on lava flows (Doty, 1967; Jonsson, 1968; Zubia et al., 2018), initially, a community partially dominated by primary colonising species proliferated, but many of these taxa were gradually replaced by the dominant species in the climax communities. While the structure of the algal communities was similar to that of the other animal communities, eventually converging towards the communities in the control zone, following a similar trajectory, the greatest changes in the algal communities resulted in a delay in their evolution compared to the other communities.

By focusing on the abundance of the main species, which collectively represented over 90% of the total abundance in all groups, we successfully revealed additional events that occurred in this primary succession, as well as specific responses within each studied group. Specifically, with respect to fishes, species such as Boops boops and Chromis lumbata have few relationships with other benthic organisms, because they are bentho-pelagic and plankton-feeding species (El-Maremie and El-Mor, 2015; Anderson et al., 2017), and have established themselves in the lava flows with an abundance comparable to that in the control zone. On the other hand, Thalassoma pavo and Similiparma lurida, slowly and progressively colonised lava flows, primarily through adult migration from densely populated areas. Conversely, species such as Canthigaster capistrata and Sparisoma cretense proliferated in the lava flows through juvenile settlements, with most size records in month seven belonging to small individuals. This massive colonisation by juvenile fishes is a common phenomenon in lava flows (Godwin and Kosaki, 1989), and is likely attributable to low predation, as piscivorous fishes play a key role in primary succession processes by controlling recruitment populations (Herrera et al., 2002). Consistently, Aulostomus spp. are the primary piscivores in tropical and subtropical ecosystems (Aronson, 1983), including the Canary Islands. Indeed, we observed that the population density of Aulostomus strigosus was 20 times lower in lava flows than in the control zone.

Low predation pressure on certain subtidal invertebrate species may also have contributed to their initial success. Percnon gibbesi, a fast-growing invertebrate whose populations were observed early, serves as a common prey for many demersal fishes (Guerra-Marrero et al., 2023). Facilitative interactions, a mechanism of ecological succession (Connell and Slatyer, 1977; Bruno et al., 2003), suggests that the arrival of a species depends on the arrival of a preceding species. In this context, the absence of such interactions between organisms can slow down changes in communities. Although mollusk populations in the intertidal area were not very large, it is worth noting that they developed in the absence of their main predator, gastropod Stramonita haemastoma (Watanabe and Young, 2006). Without the prior mortality of Phorcus sauciatus, the hermit crab Clibanarius aequabilis, which is abundant in the control zone, has no opportunity for shell-selection behaviour, even with the large availability of this crab larvae in the surrounding waters.

As previously demonstrated, algae on lava flows followed different dynamics. Fast-growing species with a single-thallus morphology have competitive advantages in the early stages of succession (Sousa, 1979; Littler and Littler, 1980). These ephemeral species, including Ulva spp., Bangia atropurpurea, ectocarpaleans, and diatoms, not only grow rapidly, but invest significantly in reproduction as well. A substantial portion of their thalli, if not the entirety, can serve as propagative cells, enabling their populations to grow exponentially (De Wreede and Klinger, 1988; Santelices, 1990). In the intertidal area, Corallina berteroi and Jania virgata, as well as CCA in both intertidal and subtidal areas, have established themselves as the main perennial taxa in lava flows. Although they exhibit comparatively slower growth, they are highly competitive in primary succession (Mariath et al., 2013) and have begun to displace ephemeral species in specific environments, such as the lower intertidal and upper subtidal zones. The interplay between low competitiveness and high demographic dynamics is a common feature of all primary successions (Huston and Smith, 1987; McCook, 1994). This allowed for the proliferation of not only certain species, but also many other taxa that are uncommon in mature communities. In fact, 46 taxa (25% of those recorded in the study), primarily algae, displayed higher abundances or were exclusively detected in lava flows.