Non-indigenous species are considered as a major driver of change for marine biodiversity, affecting the functions and services provided by native species. Macroalgae constitute a significant component of non-indigenous marine species which, in combination with other processes such as eutrophication or habitat degradation, represent a major threat to marine biodiversity (Schaffelke et al., 2006). Understanding the invasion process of those species, their ecology and their impacts is an important challenge for conservation and management of local communities in the context of climate change (Andreakis and Schaffelke, 2012).

The red seaweed Asparagopsis armata Harvey, 1855 is a clear example of non-indigenous seaweed in the Atlantic European coasts. Being endemic to the Indo-Pacific region, in the Southern Hemisphere (Dijoux et al., 2014), this species was firstly described outside its natural boundaries in France in 1923 (Svedelius, 1933). Currently, A. armata is spread in the north-eastern Atlantic from the British Isles to Senegal, including Canary Islands, Azores and the Mediterranean Sea (Guiry and Dawes, 1992). Marine transport through the Suez Canal seemed the most probable vector of introduction, though no direct evidences are available (Zanolla et al., 2022). In non-native locations, negative effects of this species on the abundance and taxonomic composition of native seaweed communities and their epifaunal assemblages have been reported (Rubal et al., 2018; Silva et al., 2021).

Life history of A. armata consisted of a triphasic life cycle with a gametophytic and a tetrasporophytic phase morphologically distinct (Feldmann and Feldmann, 1942). In fact, both phases were previously recognized as two different species, the tetrasporophyte being named Falkenbergia rufolanosa and the gametophyte being named A. armata. Phase alternation has been shown to be temperature, nutrient and photoperiod dependent. Indeed, tetrasporogenesis occurs within a narrow range of daylength and temperature conditions, which explains why sexual reproduction is limited in areas close to its biogeographical limits (Bonin and Hawkes, 1987; Guiry and Dawes, 1992; Zanolla et al., 2022). The absence of sexual reproduction indicators in such places and the observed differences in the distribution patterns of gametophytes and tetrasporophytes led to the hypothesis that both phases can reproduce independently through vegetative mechanisms (Dixon, 1965). For example, A. armata gametophytes produce specialized vegetative structures (morphologically similar to hooks) with the capacity to attach to new surfaces and to regenerate into new individuals once dislodged from the original thalli (Codomier et al., 1979). On the other hand, tetrasporophytes can also propagate through vegetative means by breaking their filamentous aggregations (Chihara, 1960).

Vegetative propagation is, actually, an important reproductive mechanism for many seaweeds within the Bonemaisoniales order (e.g. Asparagopsis taxiformis or Bonnemaisonia hamifera) (Breeman et al., 1988; Zanolla et al., 2018b) and others (e.g. Gelidium corneum, Gracilaria sp.) (Santelices and Doty, 1989; Juanes and Puente, 1993), being in some cases the main form of population maintenance. Environmental variables such as temperature, nutrients, photoperiod or irradiance may influence this process. The response to these variables seems to be different for each phase and may help to explain why gametophytes are characterized by seasonal patterns in some regions while tetrasporophytes can be found in the field all year round (Zanolla et al., 2022). However, this effects on the vegetative propagation process are still scarcely understood. For A. armata gametophytes, only the effects of medium composition (i.e. nutrients) (Codomier et al., 1981; Haslin and Pellegrini, 2001) and the type of substrata haven been explored (Wright et al., 2022). These studies showed how the production and growth of vegetative structures (plantlets) was promoted by substances such us iron sulfate and potassium bromide. Those plantlets are the result from the development of hook-like structures’ apical cells, opposite to the spines (Haslin and Pellegrini, 2001). On the other hand, mussel ropes were the preferred substrata to attach from the different materials and structures tested (polypropylene ropes and nylon mono and multifilament nets). More research has been done for tetrasporophytes, being temperature, nutrients and culture density among the most explored factors (e.g. Ní Chualáin et al., 2004; Mata et al., 2006; Mata et al., 2007; Zanolla et al., 2014).

Given that climate change is modifying the current ranges of those environmental variables, it is necessary to understand how key processes (e.g. vegetative propagation) will be affected by those changes in order to predict future impacts and adopt effective monitoring and management strategies. Indeed, predictive models on A. armata have already anticipated distributional changes for this species with contractions on the Atlantic coasts of Europe and shifts in the Mediterranean (Zanolla et al., 2018a). Considering all this information, the objective of the present study is to analyze the interactive effects of temperature and irradiance on the vegetative propagation process of gametophytes and tetrasporophytes of A. armata. Temperature and irradiance conditions were selected individually for each phase according to their ecology. Different response variables were also measured for each life cycle stage. The survival rates of hook-shaped specialized structures and the production of plantlets were analyzed for gametophytes. In the case of tetrasporophytes, the vegetative propagation process was characterized by the growth rates of tufts and the phycobiliprotein content.

Vegetative propagation responses for gametophytes and tetrasporophytes of A. armata were influenced by temperature and irradiance. For gametophytes, the survival rates of the hooks after 30 days of culture were lower (though not always significant) under increasing levels of temperature and decreasing levels of irradiance. However, patterns of appearance of plantlets at different temperature and irradiance conditions were not clear. In the case of tetrasporophytes, only positive growth rates were registered at 15°C and 55-60 µmol/m²/s after 30 days of culture. Higher concentration of phycobiliproteins was detected at higher temperatures during the first days of culture, while not clear patterns were detected at the end. The results for this phase indicated that temperature is the main factor controlling the survival of broken pompon-like structures when irradiance is not a limiting factor.

Study findings showed how the development of new fronds from single hooks collected in northern Spain (1-3cm length) is possible. These findings differ to those found for A. armata in New Zealand and Tasmania where regrowth from single hooks was not possible (Bonin and Hawkes, 1987; Wright et al., 2022), but are in close agreement to those found in France where regrowth occurred (Codomier et al., 1979; Haslin and Pellegrini, 2001). The lack of regrowth in New Zealand and South Australia has been suggested to be attributed to seasonal effects or the existence of genetically different strains (Wright et al., 2022). In fact, there are genetic evidences of different clades of A. armata clearly linked to its geographic distribution (Dijoux et al., 2014; Preuss et al., 2022) and differences among these strains have been previously reported for other phenological responses (e.g. Ní Chualáin et al., 2004).

In our study, vegetative propagation through hooks was conditioned by the low survival rates of those structures, being always below 50% after 60 days of culture. The low survival of those specialized structures and the weakness of the attachment bundles produced by a single hook (Wright et al., 2022) may explain why they are produced in high numbers, occurring every 3-4cm along the thallus (Wright et al., 2022). Both temperature and irradiance seemed to have influenced the survival rates, being the rates higher at increasing levels of irradiance and decreasing temperatures. In fact, temperatures over 18°C seemed to be lethal for excised hooks. In a preliminary study, these structures were cultured also at 20 and 25°C resulting in 100% mortality rates within the first ten days of culture (unpublished data). These results were congruent with the annual summer biomass decline observed in the field for this species at the collection area. According to Ní Chualáin et al. (2004), A. armata can grow at temperatures between 9 and 21-23°C, with a wider range of survival (up to 25-28°C). Presented results suggest that the vegetative propagation through hooks may be limited to a narrower range as a consequence of its survival capacity, being 18°C the upper limit. Temperatures below 15°C were not tested, so the lower limit for this process is still to be determined. Regarding light availability, lower irradiances resulted in lower survival rates. For A. armata gametophytes the light compensation point (I_c) was set at a mean value of 20 µmol photons/m²/s (Flores-Moya et al., 1996). Being continuously under this compensation point was critical for hooks growing at 18°C, as only a 3% of the initial hooks survived after 60 days of culture, but not for those growing at 15°C. According to this, it seems that I_c may be influenced by temperature as it has been shown for tetrasporophytes (Zanolla et al., 2014). Those results contrast to that found for other red seaweeds such as Gelidium sp., where more vegetative structures were produced under low levels of irradiance (Fonck et al., 2007; Otaíza et al., 2018; Sainz-Villegas et al., 2023). In fact, Lüning et al. (1990) stated that germlings of red algae growing under canopy achieve its maximum growth rates at very low levels of incident light (approx. 4-5 µmol/m²/s). However, as shown by the results, it seems that the under-canopy development of A. armata may be scarce. In fact, germlings grow mainly on top of other species (as an epiphyte) or over complex surfaces where the specialized structures (hooks) get entangled (personal observation) in the field.

Regarding regrowth, hooks developing new plantlets were registered in every treatment after 60 days of culture but a clear response pattern to temperature and/or irradiance was not detected. The ratio recruit size/hook size was always between 0-2. These results differ from those reported by Haslin and Pellegrini (2001) were ratios scored values between 0-3.5 after one month of culture. In fact, the first vegetative structures in our study did not show until the 30^th day of culture, while in Haslin and Pellegrini (2001) appeared after 10 days. Temperature and irradiance conditions in that study were close to the ranges explored in our study, suggesting nutrients to be the main factor influencing those differences.

It is well known that many red algae can produce structures that allow attachment through asexual reproduction (Cecere et al., 2011). Here, after 60 days of culture, attachment structures were not identified for any of the cultured hooks, despite contact between tissues and substrata was assured with a plastic mesh. Instead, a complex system of axes through the different threads of the net was observed, preventing the original thalli from moving. In fact, it was impossible to remove the hooks from the mesh without breaking any biological tissue. Nevertheless, no solid points/discs of attachment nor a rhizomatous system were found. These results suggest that it is not necessary the attachment of the hook to another object to induce vegetative propagation as it has been previously suggested (Svedelius, 1933). Actually, it seems that the entanglement, and thus, the attachment is promoted by the growing vegetative axis of the new recruits. In contrast, some field collected individuals presented morphologically different hooks with bigger and rounded spines not seen in the laboratory experiments. These hooks seemed to be attached to other species, but it was difficult to discern if there was through attachment structures or not due to the entanglement degree. In some cases, those modified spines seemed to grow and trap the other species tissue (personal observation), but this require a more detailed analysis to be confirmed.

A. armata tetrasporophytes are also capable of reproducing through vegetative propagation by the breakage of the pompom-like structures into different tufts (Zanolla et al., 2022). Previous studies on the life cycle of this species showed how phase alternation is temperature, light and nutrient dependent. In fact, sporogenesis occurs only at a narrow range of temperatures ( ± 4-6°C) and daylengths below 9 hours (Oza, 1977; Guiry and Dawes, 1992; Ní Chualáin et al., 2004). Temperature ranges vary among strains, being 13-17° C for Australian individuals, 15-21°C for the Irish strain and 17-21°C for Italian seaweeds (Guiry and Dawes, 1992). In this study, we have shown that the vegetative propagation process is also influenced by temperature and light conditions. Actually, both the growth and the phycobiliprotein content were affected by temperature and/or irradiance at some point during the 30 days of culture.

Regarding tetrasporophyte growth, gaining of biomass was only registered at 15°C and 55-60 µmol/m²/s after 30 days of culture. At 20°C, a weak increase of biomass was observed during the first 10 days of culture followed by biomass losses after 20 and 30 days. At 25°C biomass losses were registered at every time interval. Previous studies on A. armata tetrasporophytes established a survival range from 3 to 25°C and a growing range from 7°C to 23°, being wider or narrower depending on the strain considered (Ní Chualáin et al., 2004; Mata et al., 2006). In Ní Chualáin et al. (2004), the survival range of the strain from the site closest to our collection area was 5-25°C and the growth range was 9-21°C. In our research, once divided into tufts, A. armata tetrasporophytes were not able to grow at 20°C after 30 days of culture. However, when the light levels were between 55-60 µmol/m²/s, biomass losses were around 25% after 20 days and remained stable after 30 days. In addition, survival at 25°C was either not possible, as almost 100% of the initial weight was lost after 30 days. Many possible factors could explain those differences. However, we have to consider that 20°C and 25°C are close or at the top limits of growing (21°C) and survival (25°C) respectively. In this sense, differences in the environmental conditions previous to the collection dates (e.g. hotter summer stressful conditions, lower nutrient levels, etc.) or even the stress induced during the sample processing may have affected the results.

Considering light, it has been shown how the lowest irradiance level (5-10 µmol/m²/s) induced higher biomass losses. Previous studies on the photosynthetic plasticity of A. armata tetrasporophytes found low light compensation points (I_c) values which were temperature dependent (e.g. Mata et al., 2006; Zanolla et al., 2014). For example, I_c values at 15°C were around ~5-6 µmol/m²/s. This means that one of our treatments was close to the light compensation point during the whole experiment and may explain why a stable biomass loss (15-25%) was observed at each time interval. At 20°C, direct estimations of I_c were not carried out, but considering the closest temperatures, the I_c should be between 7.5 and 12.5 µmol/m²/s. At 25°C, I_c values should be in the range 12.5-15 µmol/m²/s. According to this, being continuously under the compensation point at 25°C led to biomass losses clearly visible during the first days of culture (50% of the biomass was lost within the first 10 days). In this sense, short extreme marine heatwaves may be lethal for those tetrasporophytes growing under canopy, epiphytically at a certain depth or floating in high density biomass aggregations. On the other hand, photoinhibition was not a problem in our study as it is induced at an irradiance level of 150 µmol/m²/s, regardless of the temperature considered (Zanolla et al., 2014).

The analysis of phycobiliproteins was carried out in order to quantitatively evaluate the bleaching or depigmentation process. Although chlorophyll-a is considered an essential pigment for photosynthesis due to its role in the capture of light and its function as an electron donor in the electron transport chain (Raven et al., 2005), we focused the analysis on phycobiliproteins, considered accessory pigments (Häder and Figueroa, 1997), because they are directly linked to that bleaching process. This process has been previously considered as a morphological “health” indicator for red seaweeds growing under stressful conditions (e.g. Quintano et al., 2017; Quintano et al., 2018). Bleaching is the consequence of a physiological response to stress which result in a complete degradation of photosynthetic pigments. In our study, frond bleaching did not show until the second half of the experiment. In fact, after 10 days of culture no depigmentation was registered as shown by the phycobiliprotein analysis. Even at 25°C and low irradiance, where 50% of the biomass was lost, depigmentation was not visible. Actually, phycobiliprotein content was higher at lower irradiances at that time. The increasing of photosynthetic pigments to compensate low light availability is a common strategy of seaweeds and it has been previously reported for other red macroalgae (Ramus et al., 1976; Xu and Gao, 2008). However, the following ten days were critical for tetrasporophytes growing at 25°C. During this time, some metabolic thresholds were probably exceeded leading to frond bleaching in almost every thallus. At this moment, depigmentation was also observed for some tetrasporophytes growing at 20°C and 5-10 µmol/m²/s, as reflected by the slight variance increase at the pigment analysis. After 30 days, depigmentation was generalized over the replicates. This loss of pigments is usually linked to a reduction in growth, thallus weakening and death (Martone et al., 2010). However, this was not entirely supported by our results. At 25°C, depigmentation and weight losses were clearly linked and evidenced tetrasporophytes death. Although depigmentation was clearly visible, it was not confirmed by the pigment analysis as it was impossible to perform it due to the lack of biomass at this temperature. The same pattern was observed at 20°C and low irradiance, but bleaching was not detected at 55-60 µmol/m²/s despite registering weight losses. At 15°C, depigmentation was not observed though the pigment analysis reflected a clear reduction in the content of red pigments. However, conclusions derived from the red pigment analysis should be taken carefully as tests were clearly limited by biomass losses. More research is still needed on the photosynthetic responses of vegetatively grown tetrasporophytes, from the analysis of a wider pool of pigments to its relationship with photosynthetic parameters.

Vegetative propagation has been previously identified as the major form of population maintenance for other species in the genus Asparagopsis (Zanolla et al., 2018b). Considering this, it is easy to find synergies between the studies predicting changes in A. armata distributions under projected temperature scenarios (e.g. Zanolla et al., 2018a) and our study, where temperature was the main factor limiting propagation. But not only at biogeographical scales, it might be also linked to changes in local processes such as seasonality. Cantabrian Sea populations were characterized by four peaks of abundance and biomass throughout the year, corresponding to four different cohorts (Aranda et al., 1984). According to our results, the number and duration of cohorts and the peak values may be affected by temperature changes that limit the vegetative propagation process particularly in late spring and summer. In fact, A. armata gametophytes seemed to have already changed at the collection area, as only one peak of biomass in spring has been observed in the last years (personal observation). Similar patterns have been previously reported for other populations (e.g. Azores populations, Neto, 2000). Understanding this patterns of change is crucial from a management perspective, as the negative impacts of this invasive seaweed on native ecosystems are well documented. For example, it has been associated to reductions in macroalgal species richness and abundance (e.g. Silva et al., 2021) or negative impacts on native epifauna as a consequence of the production of toxic exudates (Guerra-García et al., 2012; Rubal et al., 2018; Silva et al., 2020). In this sense, the results presented here have important implications for the analysis of invasion risk and the detection of potentially threatened areas as it has been shown that the spread of the species by vegetative means is dependant on temperature and irradiance. A possible way to include this information on risk analysis is through the use of Hybrid Species Distribution Models (SDMs), which integrate the vegetative propagation responses in correlative models enhancing a better delimitation of vulnerable areas and conservation/protection zones.

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

SS-V: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. BS-A: Formal analysis, Investigation, Writing – review & editing. AP: Conceptualization, Formal analysis, Funding acquisition, Project administration, Writing – review & editing. JJ: Conceptualization, Formal analysis, Funding acquisition, Project administration, Writing – review & editing.