1. Introduction
Biological invasions, i.e., the successful establishment and spread of species outside their native range [
1,
2], are a major component of global change [
3,
4,
5]; they reshuffle the planet’s biota and represent a distinctive mark of the Anthropocene [
6,
7]. Biological invasions have been dubbed “biological pollution” [
8,
9] that causes biodiversity loss and alters the composition of communities and ecosystem functioning [
10,
11], thus compromising ecosystem services [
12]. In terrestrial habitats, biological invasions have been the concern of biogeographers and ecologists for decades [
13,
14], whereas in marine environments they have remained less known until recently [
15,
16]. Today, meta-analysis and reviews are available that highlight the impact of marine invasions [
12,
17,
18,
19]. Studies in the Mediterranean Sea indicate that biological invasions represent a serious ecological and economical menace leading to fishery and tourism impairment [
20].
In terrestrial environments, invasive alien species have been considered important agents of biotic homogenization at different spatial scales [
21,
22,
23]. In the present paper, biotic homogenization means the process that diminishes floral and faunal differences among previously distinct communities within a specific region [
24,
25]. The rationale beyond the idea that biological invasions may cause homogenization in the recipient biota is simple—invasive species typically show great adaptability to a wide range of environmental conditions and may thus enter different habitats, often becoming dominant [
26,
27]. If an abundant invader replaces or simply adds to the native species that used to characterize the recipient communities, then the dissimilarity among communities is expectedly reduced (
Figure 1). This rather simplistic scheme just aims at illustrating the rationale of our study, while more complete schemes illustrating the many ways invaders can impact recipient communities can be found in review papers [
24,
28,
29].
We tested the assumption above by analyzing the change that may be observed in the zonation of communities along a sharp ecological gradient (i.e., depth) in the presence or absence of an alien invader. We used a comprehensive database collected in the marine protected area of Portofino (NW Italy), where
Caulerpa cylindracea—one of the worst invaders in the Mediterranean Sea [
30,
31]—has become abundant in recent years [
32]. After a descriptive analysis of the comprehensive database, we generated two distinct datasets to go through multivariate analysis: the first contained observed data, and the second was a “simulated” one where
C. cylindracea was excluded from the dataset, under the hypothesis that the latter dataset will exhibit greater dissimilarity than the former one. We used the term “simulated” to make it clear that we did not perform a real, physical removal experiment in the field. Our procedure ties in with the so-called “inclusion versus exclusion” approach, which has been said to be not only relevant to invasion biology, but to any field where the component of interest is an integrated part of the response [
18,
33].
4. Discussion
Caulerpa cylindracea was abundant in the rocky reefs of Portofino, as illustrated by field data. Relative abundance (here expressed as % cover) is a recommended index to gauge the invasion level of an alien species [
70]. With an average cover of about 10% (with peaks to 25%),
C. cylindracea can be defined a strong invader [
54]; nevertheless, its impact to the recipient communities has been shown significant already with low cover [
71].
C. cylindracea spread through the whole Marine Protected Area, including the “no entry, no take” zone. MPAs are mostly designed to manage fisheries [
72], and may even help restore ecosystems through enhanced top-down control [
73], but cannot prevent the introduction of alien species [
74,
75]. In Portofino, constancy of
C. cylindracea cover values for six years (2012 to 2017) suggests that it has reached the so-called phase of persistence—the last of the four phases identified in the process of naturalization of introduced species [
76,
77].
As already observed in other Mediterranean sites [
41],
C. cylindracea showed a wide depth range at Portofino, with a preference for about 20 m depth, and invaded all the communities existing there, although with different intensities. Our study is the first to make a direct comparison of the quantitative incidence (in term of % cover) in as many as 14 communities (defined following EUNIS habitat classification). Notwithstanding significant differences of
C. cylindracea cover among the 14 individual habitats, no clear pattern (aside from depth) was discernible according to the nature of the habitats, with macroalgal associations, seagrass beds, coral facies, and zooxanthellate coral facies all being colonized by
C. cylindracea. It has been recognized that the impacts of invasive species are often context-dependent, both on land [
78,
79] and in the sea [
29], and may depend on the idiosyncrasy of the eco-evolutionary history of the recipient communities [
80]. The relationship between species richness and the invasions of alien species may be either positive or negative, mostly depending on scale [
81]. At Portofino, increased cover values of
C. cylindracea were associated with reduced species richness, as reported in many studies [
82,
83]. It has been shown that reduced species richness can in turn further accelerate invasion [
81].
The depth gradient in the presence of
C. cylindracea showed no discontinuity, suggesting that the benthic communities replaced each other gradually. On the contrary, when
C. cylindracea was excluded from the analysis three sharply distinguished situations were recognizable, separated by discontinuities that correspond well with known ecological thresholds in light intensity and water movement [
84]. Such a result indicates that the high environmental adaptability of
C. cylindracea [
40] overcomes the severity of the ecological gradients that typically dictate the depth zonation of native communities in the infralittoral and circalittoral zones in Mediterranean rocky reefs [
84]. This effect and other indicators and indices (multivariate dispersion, average dissimilarity, standard deviation, and coefficient of variation) consistently point to the homogenization caused by
C. cylindracea invasion.
The problem of homogenization in terrestrial ecosystems is better known than in marine ecosystems, which affects our ability to make ecological assessments of change [
85]. We used a simulation-like approach to evidence that
C. cylindracea is cause of homogenization in the recipient ecosystem. In particular, the depth gradient was better defined, with a greater multivariate dispersion, when the invader was excluded from the analysis. Multivariate dispersion is considered a measure of β diversity [
86], i.e., the variability in species composition among communities within an area. Consistently, it has been demonstrated experimentally that
C. cylindracea depresses β diversity in invaded ecosystems [
87].
Apparently
C. cylindracea did not replace any native species, most of them having already disappeared or got rarer due to climatic and local human impacts during the 1980–90s ecosystem shift at Portofino reefs [
32]. Thus,
C. cylindracea acted more as a “passenger”, i.e., a passive opportunist taking advantage of environmental change, than a “driver”, i.e., a genuine actor of ecosystem degradation [
88]. There is presently a debate about whether invasive species act as passengers or drivers of global change [
89,
90]. Removal experiments led to the conclusion that
C. cylindracea acted as a passenger at the beginning of the invasion to become a driver once established [
91]. As drivers, invasive species exert an impact on Mediterranean marine ecosystems that may exceed that of sea water warming [
92]. The analysis of many correlative and experimental studies led to the conclusion that the invasion by
C. cylindracea is one of the most threatening in the Mediterranean Sea [
93].