Plant Species for Sustainable Bioretention Systems’ Implementation in Mediterranean Italian Regions: A Review

Abstract

Vegetation is a key component of bioretention systems, significantly influencing their functionality and overall performance. The sustainability of these systems largely depends on the plant species’ ability to withstand the primary hydrological stresses induced by the infrastructure’s design, in combination with local climatic conditions. This study focuses on plant selection for bioretention applications in Sub-Mediterranean and Mediterranean regions of Italy, which are increasingly affected by extreme weather events, including intense rainfall, heat waves, and prolonged droughts. A review of scientific literature and stormwater management manuals developed for Mediterranean climates was conducted to evaluate both non-native species and Italian native species already used in international applications. The findings reveal a limited range of drought-tolerant non-native species in scientific literature, whereas such species are more prominently featured in California-based manuals. Moreover, among the relatively few native species identified, only a small number are truly suited to fully Mediterranean conditions, based on evaluations of Ellenberg index values. These results highlight a significant research gap, emphasizing the need for further studies, particularly focused on the target ecosystems of native flora in central and southern Italian regions.

1. Introduction

Driven by increasingly intense rainfall linked to climate change, together with widespread soil sealing in rapidly urbanizing areas, sustainable stormwater management has become a priority for municipalities worldwide [1,2]. In this context, Green Stormwater Infrastructures (GSI)—referred to by different terms depending on the region, such as Sustainable Urban Drainage Systems (SuDS) or Low Impact Development (LID)—have gained growing attention over the past three decades [3], emerging as priority Nature Based Solutions (NbS) for stormwater management [4]. Among GSI practices, bioretention systems, commonly known as rain gardens (RGs) and often implemented as bioswales [5], represent one of the most widely adopted approaches. These systems not only promote the infiltration and filtration of stormwater but also deliver a range of ecosystem services [6,7], including support for biodiversity, enhancement of urban aesthetics, and various social benefits [8,9].

Climate change makes stormwater management infrastructures particularly relevant for the Italian peninsula, not only in Northern Italy but also in its Central and Southern regions. While mostly Northern Mediterranean areas—such as South France, North Italy, and the Balkans—are projected to experience increasingly extreme precipitation events [10,11,12], Central Italy is also showing an increase in the frequency and intensity of such events, particularly during the summer, with a shift towards shorter and more intense convective rainfall [13,14,15]. Furthermore, although the southern Mediterranean is generally expected to experience a decrease in both average and extreme precipitation [10,11,12], local studies nevertheless highlight a trend toward more intense short-duration rainfall events [16,17,18].

Consequently, GSI are increasingly recognized as a crucial strategy for climate change adaptation within Italian urban areas [19,20]. The broader ongoing SuDS implementation initiative is represented by the Milan Sponge City Project [21], while several pilot RG projects have been developed across the country [19], although they remain largely undocumented in the scientific literature. To date, the only experimental RG described in the literature was established in Padova in 2011 [22,23,24]. In Central and Southern Italy, the municipality of Rome is currently promoting the implementation of stormwater management measures such as RGs [25], while in Sicily, the intention to adapt coastal areas to extreme climatic events through the use of NbS is reflected by the completed Gifluid project [26] and participation in the ongoing Horizon project Cardimed [27]. A first RG was recently implemented in the city of Catania [28].

Mediterranean climates occupy a large portion of the Italian peninsula across its Central and Southern regions, according to both the Köppen climate classification [29] and, more importantly for vegetation analysis, the Rivas-Martinez bioclimatic classification [29,30]. Furthermore, a large portion of northern Central Italy, currently classified as Sub-Mediterranean macrobioclimate according to Rivas-Martínez [29], is projected to shift toward Mediterranean climatic conditions from the 2040s onward [31]. The magnitude of this climate-driven southward shift toward warmer climatic conditions is also well documented by Bastin et al. [32]. E.g., in fifty years, Madrid is projected to experience the climate of Marrakesh, and London that of Istanbul.

Since the overall performance and long-term functionality of RGs strongly depend on vegetation, which plays a central role in regulating hydrological processes, supporting microbial activity, and ensuring system resilience over time [33,34,35], selecting climate-adapted species is crucial to the infrastructure’s sustainability. Under Mediterranean conditions, plants in RGs are not only required to tolerate a more or less prolonged and pronounced temporary waterlogging [8,36], depending on the infrastructure’s design and their positioning within the RG [37], but also severe heat and drought stress [38]. Although a limited number of studies have addressed bioretention design in xeric climates [39,40,41], applications in Mediterranean-climate regions remain comparatively underrepresented in the literature [5], highlighting a significant research gap regarding the identification of suitable plant species.

Therefore, this review aims to address the current knowledge gap regarding vegetation selection for bioretention systems in Mediterranean environments in Italy. Specifically, the objectives of the study are:

• to identify the botanical families, genera, and growth forms most frequently associated with bioretention systems in scientific literature and Mediterranean stormwater manuals;
• to evaluate the ecological compatibility of recurrent non-native taxa with Sub-Mediterranean and Mediterranean Italian conditions, with particular reference to tolerance to alternating flooding and drought stress;
• to identify Italian native taxa potentially suitable for bioretention applications through the integration of ecological indicators, habitat affinity, and flooding-tolerance information;
• to develop a structured preliminary reference framework for future experimental testing and plant selection in Mediterranean bioretention systems.

2. Materials and Methods

2.1. Studies’ Research

First, two of the most comprehensive books on bioretention (rain gardens) were selected as primary sources to identify potentially suitable taxa in both British [37] and North American [9] literature. Subsequently, the research conducted in April 2025 focused on identifying the most recent and comprehensive review papers in the scientific literature on bioretention, with particular emphasis on vegetation, or, at a minimum, on including the names of the plant species reported in the analyzed studies [2,5,42]. The construction of the main database was based on expanding the taxa list proposed by Corduan and Kühn [5]. Within the bibliographies of the selected reviews, only studies directly associated with cited plant species were considered, while duplicate articles were counted only once. Consequently, only species explicitly mentioned within the databases and texts of the selected review papers were included in the initial species list.

The database was subsequently expanded to include both the most recent studies on bioretention (2024–2026) and the species experimentally tested under cyclic flooding conditions. To achieve this, two additional searches were conducted in March 2026, referred to as “search 1” (S1) and “search 2” (S2), respectively (Figure 1). Both searches were performed in Web of Science across all available databases, including only English-language publications.

For S1, the following search criteria were applied: Topic = “bioretention” OR “bioswale” OR “rain garden” OR “rain garden” OR “stormwater biofilter”. The results were first filtered by publication year (2024, 2025, and 2026), after which two additional filters—“vegetation” and “plant”—were applied separately, generating two result sets. Duplicate records between the two sets were subsequently removed. An initial screening based on titles and abstracts excluded studies unrelated to bioretention, and articles already included in the review papers were considered as primary references. A second screening, based on abstracts and full texts, applied the following exclusion criteria: unavailable full texts, absence of scientific plant names, studies specifically focused on tropical environments, and review papers to avoid duplication of sources.

For S2, the following search criteria were applied: Topic = (“bioretention” OR “bioswale” OR “rain garden”) AND (“plant selection” OR “cyclic flooding” OR “cyclical flooding”). Results were refined using the combined filters “vegetation” and “plant”. Articles unrelated to bioretention or already identified in the previous search were first excluded, followed by records lacking full-text availability or species names. Overall, the final database included 169 scientific publications: 2 books and 167 articles.

Grey literature on stormwater management was also surveyed, with a specific focus on the Mediterranean climate. An initial online search was performed in Google Chrome using the keywords “stormwater management manual plant list” and “stormwater management LID manual plant list”, yielding 29 results from various North American locations. Based on the Köppen climate classification of the corresponding cities or regions, three manuals from Csa climate regions (Mediterranean climate with dry, hot summers) were selected [43,44,45], along with three additional manuals representative of Csb climates (Mediterranean climate with dry, warm summers) [46,47,48]. A secondary practice-oriented database was compiled from plant species lists reported in the selected stormwater management manuals representative of Mediterranean-type climates. These manuals were considered as technical grey literature intended to complement, rather than replace, peer-reviewed scientific evidence.

2.2. Species’ Analysis

All the species found in the aforementioned scientific literature and stormwater manuals were listed in a first Excel database, available as Table S1a,b in Supplementary Materials S1. The database was designed as a structured reference tool for preliminary species screening, rather than as a definitive plant-selection algorithm. Taxa were included according to a transparent source hierarchy: first, scientific literature on bioretention and cyclic flooding experiments; second, stormwater management manuals from Mediterranean-type climates; and third, botanical and ecological databases to standardize each taxon’s attributes. For each record, accepted taxonomic name, botanical family, growth form, native range, and Italian native/non-native status were compiled, based on the information in the Kew Gardens’ Plants of the World online database [49].

Species were then separated into herbaceous dicotyledons, herbaceous monocotyledons, shrubs, and trees/shrub-trees, because these groups differ in ecological function, available information, and potential use within bioretention systems. For each of these groups, the most recurring families and genera were assessed. For further analysis of the single species, Italian native and non-native taxa were considered separately. For non-native taxa, drought tolerance, moisture use, anaerobic tolerance, minimum precipitation, and wetland status were extracted when available. For Italian native taxa, Ellenberg temperature and humidity values were combined to derive an environmental compatibility class, while tolerance to temporary waterlogging was classified using evidence from cyclic flooding studies, wetland status, or habitat affinity.

The screening procedure did not apply numerical weights, as the available sources provide heterogeneous information and rarely report comparable quantitative performance metrics. Instead, the database applies a reproducible qualitative classification based on explicitly defined criteria. Therefore, the database should be interpreted as a preliminary decision-support framework for identifying promising candidate species, not as a substitute for local experimental validation.

2.2.1. Non-Native Taxa: Scientific Literature

For non-native taxa identified through the scientific literature survey, only species with at least two citations were considered in order to narrow down the pool of candidate taxa. Given the predominance of North American species, the following attributes, available in the USDA Plants Database [50], were primarily assessed: anaerobic tolerance, drought tolerance, moisture use, minimum precipitation, and wetland status. When drought-tolerance information was unavailable, and especially for particularly promising taxa, the USDA data were integrated and cross-checked with information from the Missouri Botanical Garden online database [51] and the University of Texas database [52].

Table 1. Drought-tolerance assessment.

		Drought Tolerance
Source	Description	Low	Low/Med.	Medium	Med./High	High
Missouri Botanical Garden [51]	dry moisture; drought tolerant				x
	dry to medium moisture; some tolerance for drought			x
	medium moisture; occasional drought		x
	medium to wet moisture	x
	high moisture	x
Lady Bird Johnson Wildflower Center [52]	drought tolerance: high				x	x
	soil moisture: dry				x
	soil moisture: dry, moist			x
	soil moisture: moist	x

summarizes the conversion of textual information into the drought-tolerance ranges derived from the USDA Plant Database.

Additional information from species descriptions was also used to assign drought-tolerance ranges when not available in the aforementioned literature and to list the species’ preferred habitats. Multiple databases, floras, and nursery sources were consulted for species native to North America [53,54,55,56,57,58], Oceania [59,60,61,62], and other geographical regions [63,64,65]. Whenever available, this information was further compared with Ellenberg indicator values [66,67], which were interpreted according to the criteria described in Section 2.2.2. The aim of the analysis was to identify the most suitable species for a potential use in the drought-prone Mediterranean environments. The resulting candidate taxa are reported in Section 3, along with information on their current occurrence within the Italian territory, to provide a preliminary assessment of potential invasiveness risk.

2.2.2. Non-Native Taxa: Stormwater Management Manuals

Besides performing the same analysis described in Section 2.2.1 for the recurrent taxa, all species reported in the Csa-climate manuals were additionally considered in order to further characterize the non-native taxa. Cultivars and subspecies were reported to species rank and therefore counted only once. Due to the availability of additional ecological information within the manuals, tolerance to submersion, water regime, and shading were also assessed. Csb-climate manuals were likewise consulted when the same species were included in those sources. Regarding submersion tolerance, species were classified as “Y” (tolerant) when indicated for the bottom zone of the basin or described as having broad ecological tolerance to submersion, whereas “(Y)” indicated limited or uncertain tolerance, corresponding to species recommended for the slope zone of the basin or more generally classified as flood-tolerant. Water-regime values ranged from L (Low) to M+ (High). Values reported as “VL” (Very Low) in the manuals were conservatively reclassified as L. Water-regime classifications were subsequently verified by comparison with the WUCOLS database, when available for the species considered [68].

For the analysis of plant communities, habitats, and environmental conditions, the Calflora database was used as the reference source [57], given the predominance of California-native taxa. Information regarding maximum temperatures during the hottest month (“July High”), minimum annual precipitation, plant communities, and habitat types—including wetland occurrence—was extracted. The complete analysis is provided in Supplementary Materials S6.

2.2.3. Italian Native Taxa

For the analysis of Italian native species, the following attributes were considered: Raunkiaer life-forms, chorotype, Ellenberg indicator values, USDA wetland status, distribution within the Italian peninsula, and main habitats. The first three categories were primarily derived from Pignatti [69]. For Monocotyledon indicator values, Guarino et al. [66] were additionally consulted, while species assigned an “X” value were further integrated using the corresponding values reported by Tichy et al. [70], which were included in brackets. Whenever available, USDA wetland status classifications were also extracted from the USDA Plants Database [50]. Species’ distribution within Italy was assessed in terms of both geographical and altitudinal range using online floristic databases [71,72] and Pignatti [73]. The same sources were used to identify and synthesize the preferred habitats of the most promising taxa, which were compiled into a final summary table for each plant group (herbaceous Dicotyledons, Monocotyledons, shrubs, and trees).

To synthesize the ecological information most relevant to environmental compatibility under Sub-Mediterranean and Mediterranean conditions, namely temperature (T) and humidity (H), a qualitative index based on these two Ellenberg indicators was developed. The index was intended as a preliminary screening and classification tool and is reported in

Table 2. Environmental compatibility (“−−” = very poor; “−” = poor; “+/−” = mediocre; “+” = moderate; “++” = good; “+++” = very good).

	T < 4	4 ≤ T < 6	6 ≤ T < 8	8 ≤ T < 10	10 ≤ T	T = X
H ≤ 2	+/−	+	++	+++	+++	t.b.e.
2 < H ≤ 4	−	+/−	+	++	+++
4 < H ≤ 6	−−	−	+/−	+	++
6 < H ≤ 8	−−	−−	t.b.e.	t.b.e.	t.b.e.
8 < H	−−	−−	−−	t.b.e.	t.b.e.
H = X	t.b.e.

.

Based on the ecological significance attributed to Ellenberg temperature and humidity values in Pignatti’s interpretation, the index might indicate the following gradient of environmental compatibility:

• the value “+/−” may correspond to conditions requiring average irrigation in Sub-Mediterranean contexts or frequent irrigation under Mediterranean conditions;
• the value “+” may correspond to moderately irrigated Sub-Mediterranean conditions or average irrigation under Mediterranean conditions;
• the value “++” may correspond to moderately irrigated Mediterranean conditions;
• the value “+++” may correspond to Mediterranean conditions requiring only infrequent or no irrigation, apart from establishment and emergency interventions.

The category “t.b.e.” (“to be evaluated”) was assigned both to species with incomplete ecological information (e.g., “X” values for T and/or H) and to more or less thermophilous taxa associated with elevated moisture requirements. This category was intended to highlight the need for additional ecological or technical evaluation of drought and heat tolerance, or for further inference based on habitat affinity [71,73,74,75], particularly in cases where species also exhibited high flooding tolerance associated with elevated humidity values.

A second index was developed to assess the species’ tolerance to waterlogging. Due to the frequent occurrence of taxa derived from cyclic flooding studies, the most recurrent limit flooding duration—72 h—was adopted as a reference threshold. The classification system was primarily based on the study of Eben et al. [76] concerning experimentally tested herbaceous species, and is reported in

Table 3. Flooding tolerance (“−−” = very poor; “−” = poor; “+/−” = mediocre; “+” = moderate; “++” = good).

		Visual Score
		Minimum	Moderate	Good	Maximum
mortality	0–25%	−	+/−	+	++
	25–50%	−−	−	+/−	+/−
	50–75%	−−	−−	−−	−−
	75–100%	−−	−−	−−	−−

.

The results of comparable cyclic flooding studies [77,78,79] were subsequently standardized using this reference framework, incorporating mortality and visual and/or growth assessments. In other cases, flooding tolerance was inferred from the experimental conditions described in the corresponding studies. When direct information was unavailable, species associated with highly moist or flood-prone ecological niches [71,73,74,75] were assumed to have elevated tolerance to flooding. For woody taxa, the “Try Plant Trait Database” was also consulted [80], particularly regarding the trait “Species tolerance to waterlogging”, mainly derived from the study of Niinemets and Valladares [81]. In ambiguous cases, or when flooding tolerance could not be inferred from the available literature, the value was reported as “?”.

3. Results

3.1. Overall Results

Scientific literature. The scientific literature survey identified a total of 37 taxa reported at the genus level, 704 distinct species, 67 cultivars, and 9 botanical varieties or subspecies. Of the recorded taxa, 67% were herbaceous, whereas 33% were woody (see Table S1a in Supplementary Materials S1).

Among herbaceous dicotyledon species, classified primarily as forbs (296 taxa), the vast majority were Hemicryptophytes, while only a limited number of species were Therophytes or Biennials. Monocotyledons were also predominantly represented by Hemicryptophytes, mainly consisting of grasses (94 taxa) and graminoid species, including sedges (44 taxa) and rushes (12 taxa). The remaining taxa are Geophytes, mostly rhizomatous, whereas bulbous species were comparatively rare. The few Chamaephytes were identified as “small shrubs” and partially included within the herbaceous Dicotyledons due to their reduced dimensions and predominantly herbaceous habit. Among Phanerophytes (260 taxa), shrubs represented the dominant life form (126 taxa), followed by trees (73 taxa) and taxa exhibiting both shrub and tree habits (53 taxa), while only a few species were woody climbers.

Figure 2 illustrates the distribution of the taxa across the different native geographical ranges. Most taxa were native to North America (370 taxa), of which 84% were exclusively North American. Only 184 taxa, being 23% of the censused total, were native to Italy.

Stormwater management manuals. The total number of records identified in manuals from both Csa and Csb climates amounted to 533 taxa, including 15 genera, 450 species, and 117 cultivars or subspecies. Of these, 172 taxa had already been identified in the scientific literature survey (see Table S1b in Supplementary Materials S1). Compared with the scientific literature survey, stormwater manuals showed a greater prevalence of woody taxa (58%), whereas herbaceous taxa accounted for 42% of the total records (37% considering only Csa manuals). However, the distribution of growth forms generally reflected the same ecological patterns observed in the scientific literature. Among herbaceous dicotyledons (130 taxa), most species were Hemicryptophytes, while only a few taxa (6 species) were Therophytes. Monocotyledons were also predominantly Hemicryptophytes, mainly including grasses (38 taxa), sedges (20 taxa), and rushes (7 taxa), whereas the remaining taxa were mostly rhizomatous Geophytes. Among woody taxa, 45 species were classified as “small shrubs” (Chamaephytes or Nano-Phanerophytes), while Phanerophytes constituted the dominant group, including 128 shrubs, 94 trees, 33 taxa exhibiting both shrub and tree habits, and 5 woody climbers.

Across all vegetation groups, the surveyed literature revealed a strong predominance of North American taxa associated with temperate-climate bioretention systems (Figure 3). Conversely, Mediterranean-native species were comparatively underrepresented, particularly among taxa experimentally validated as tolerant of combined drought and flooding. This imbalance highlights the need for future region-specific testing focused on Mediterranean ecological conditions and native plant communities.

3.2. Herbaceous Dicotyledons

Herbaceous dicotyledons represented the most taxonomically diverse component of the surveyed vegetation, with a strong predominance of perennial forbs associated with wetland environments, temperate grasslands, wet to dry meadows, and disturbance-adapted habitats. The results reveal a marked prevalence of North American taxa recurrently used in bioretention systems, while Mediterranean and Italian native species remain comparatively underrepresented. This pattern reflects both the historical development of bioretention research in temperate North American regions and the limited availability of experimental studies specifically focused on Mediterranean environments.

In the survey of the scientific literature, a total of 312 records (including genera, species, cultivars, and subspecies) across 46 botanical families were identified. The most represented families are Asteraceae (31.4% of the taxa), followed by Lamiaceae (7.4%), Plantaginaceae (5.1%), Fabaceae (5.8%), and Rosaceae (4.2%). Overall, the recorded taxa belong to 168 genera, of which 61% are represented in Italy by native species. The most recurrent genera were all members of the family Asteraceae, namely Rudbeckia, Solidago, Symphyotrichum, Echinacea, and Eutrochium. Among genera including Italian native species, the most frequently cited were Solidago, Filipendula, Achillea, Lythrum, and Geranium. Complete lists of botanical families and genera are provided in Tables S2 and S3 (Supplementary Materials S2).

The predominance of Asteraceae taxa, particularly perennial prairie-associated genera such as Rudbeckia, Solidago, and Echinacea, likely reflects their broad ecological plasticity and tolerance to alternating wet and dry conditions. Similarly, the recurrent occurrence of genera such as Achillea, Lythrum, and Geranium among Italian native taxa suggests that ecotonal meadow and semi-ruderal communities may be relevant reference ecosystems for designing Mediterranean bioretention plantings.

Within stormwater manuals, in addition to Asteraceae (25.4% of the taxa), Lamiaceae, and Rosaceae, the most represented families are Onagraceae and Phrymaceae. The most frequently cited genera were Achillea, Epilobium, and Erigeron, all of which are also present in Italy with native species.

3.2.1. Non-Native Taxa

Scientific literature. Although drought-tolerant taxa were identified (36%), most recurrent species still belonged to intermediate moisture-use categories, suggesting that many currently adopted bioretention species may remain more suitable for temperate or Sub-Mediterranean conditions than for fully Mediterranean climates characterized by prolonged summer drought. This finding further highlights the limited availability of experimentally validated taxa specifically adapted to combined flooding and severe drought stress. These taxa are reported in

Table 4. Non-native drought-tolerant herbaceous Dicotyledons (“Anaer. Resist.” = Anaerobic resistance; “Drought Tol.” = Drought tolerance; “Prec. Min.” = Minimum precipitation; “Cas.” = Casual; “Nat.” = Naturalized; “N” = Northern Italy, “C” = Central Italy, “S” = Southern Italy).

-		USDA Plant Database					Other Ref.
Most Cited Species	n. of cit.	Wetland Status	Anaer. Resist.	Drought Tol.	Moisture Use	Prec. Min. (mm)	Drought Tol.	Presence in Italy
Baptisia australis	8	FACU/UPL	-	High	Medium	-	Medium	no
Rudbeckia hirta	7	FACU	None	Medium	Medium	711.2	Medium	Cas./Nat. (N/C)
Asclepias tuberosa	5	-	None	High	Low	711.2	High	Cas. (N)
Penstemon digitalis	5	FACW/FAC	Low	High	Low	762.0	Medium	no
Echinacea pallida	4	-	None	Medium	Medium	355.6	Medium	no
Heliopsis helianthoides	4	FACU/UPL	-	Medium	Medium	-	Medium	Cas. (N)
Rudbeckia laciniata	4	FACW/FAC	Low	High	Low	381.0	Low/Medium
Solidago rigida	4	FACU/UPL	None	High	Medium	355.6	Low/Medium	no
Aquilegia canadensis	3	FAC/FACU	-	High	Medium	-	Medium	no
Armeria maritima	3	FAC/FACU	None	Low	Medium	812.8	Medium	no
Baptisia bracteata	3	-	-	-	-	-	High	no
Coreopsis verticillata	3	-	-	-	-	-	Medium/High	no
Ratibida pinnata	3	-	None	Medium	Medium	457.2	Medium	Cas. (N)
Amsonia orientalis	2	-	-	-	-	-	Medium	no
Geum triflorum	2	FACU	None	High	Medium	406.4	Medium	no
Liatris aspera	2	-	-	-	-	-	Medium/High	no
Parthenium integrifolium	2	-	-	-	-	-	Medium/High	no
Silphium laciniatum	2	-	-	-	-	-	Medium/High	no
Solidago californica	2	-	-	-	-	-	High	no
Solidago missouriensis	2	-	Low	High	Low	304.8	Medium/High	no
Solidago ptarmicoides	2	-	-	-	-	-	Medium/High	no
Solidago speciosa	2	UPL	None	Medium	Low	-	Medium	no
Symphyotrichum oolentangiense	2	-	-	-	-	508.0	Medium/High	no

, while the complete list of species and their characteristics is reported in Table S10a (Supplementary Materials S3).

Manuals. The analysis of the most recurrent non-native species showed that 61% of the taxa have a medium to high drought tolerance, although most fell within the “medium” tolerance range. The drought-tolerance assessment is reported in Table S10b (Supplementary Materials S3).

3.2.2. Italian Native Taxa

Scientific literature. Among the surveyed taxa, 90 herbaceous dicotyledon species were native to the Italian peninsula. Of these, 86 occur in Northern Italy, 77 in Central Italy, and 67 in Southern Italy, with 71, 77, 64, and 33 taxa, respectively, distributed within the lowland, hilly, montane, and subalpine belts. The complete species list is provided in Table S11a (Supplementary Materials S3).

The predominant chorotype was Eurasian (16 taxa), followed by Paleotemperate and South European–South Siberian chorotypes (10 taxa each). In contrast, only 9 taxa exhibited a Eurimediterranean distribution, while only 2 taxa belonged to the Mediterranean montane chorotype. The limited representation of Mediterranean chorotypes among recurrent taxa further emphasizes the current mismatch between the climatic contexts in which bioretention systems have been most extensively investigated and the Mediterranean environments where climate-adapted vegetation selection is increasingly needed.

This pattern is partially reflected in the Ellenberg indicator values analysis, although the trend is more pronounced for the other plant groups (Monocotyledons, shrubs, and trees). In this case, average Ellenberg values were 7 for light, 5.7 for temperature, 4.7 for humidity, 6.6 for reaction, and 3.9 for nutrients. Despite their broad occurrence in Southern Italy, most taxa (46%) fell within the temperate to Mediterranean-montane temperature range (6 ≤ T < 8), predominantly with the value 6, while a comparable proportion of species was associated with cooler montane environments (4 ≤ T < 6). Regarding humidity, most taxa (54%) were characteristic of habitats ranging from relatively dry to fluctuating moisture conditions (2 < H ≤ 4), followed by species associated with humid to infrequently submerged environments (6 < H ≤ 8). Overall, these findings suggest that many of the drought-tolerant taxa identified are nevertheless primarily associated with relatively cool temperate, montane, or Mediterranean-montane environments.

Table 5. Environmentally compatible Italian native herbaceous Dicotyledons from surveyed literature: a = scientific literature; b = stormwater manuals (“x” = presence; “(x)” = limited presence).

(a)
Family	Species	Biological Form	Chorotype	N. of References	Env. Compatibility	Cyclic Flooding	Flood Tolerance	Northern Italy	Central Italy	Southern Italy
Asteraceae	Achillea millefolium	H Scap	Eurosib.	5	t.b.e. [+]	Y	(++)/(+/−)	x	x	x
Asteraceae	Artemisia absinthium	Ch Suffr	E-Medit.-Mont.	1	(+)	Y	(++)	x	x	x
Apiaceae	Bupleurum falcatum	H Scap	Eurasian	1	(+)	Y	(−)	x	x	x
Campanulaceae	Campanula glomerata	H Scap	Eurasian	1	(+)		?	x	x	x
Campanulaceae	Campanula rapunculoides	H Scap	Europ.-Caucas.	1	(+)	Y	(++)	x		x
Caryophyllaceae	Cerastium tomentosum	Ch Suffr	Endem.	1	(+)		(+/−)	x	x	x
Asteraceae	Cichorium intybus	H Scap	Paleotemp.	1	(+)	Y	(+)	x	x	x
Asteraceae	Cota tinctoria	H Scap	Central-Europ.	1	(+)	Y	(+/−)	x	x	x
Apiaceae	Daucus carota	H Bienn	Paleotemp.	1	(+)		(+)	x	x	x
Boraginaceae	Echium vulgare	H Bienn	Europ.	1	(+)		?	x	x	x
Euphorbiaceae	Euphorbia cyparissias	H Scap	Central-Europ.	1	(+)	Y	(++)	x	x	x
Asteraceae	Galatella linosyris	H Scap	S-Europ.-S-Sib.	1	(+)	Y	(++)	x	x	x
Rubiaceae	Galium verum	T Scap	Europ.-Caucas.	1	(+)	Y	(++)	x	x	x
Geraniaceae	Geranium sanguineum	H Scap	Europ.-Caucas.	1	(+)	Y	(++)	x	x	x
Hypericaceae	Hypericum perforatum	H Scap	Paleotemp.	1	(++)	Y	(++)	x	x	x
Asteraceae	Lactuca perennis	H Scap	W-Eurimedit.	1	(+)	Y	(+/−)	x	x	x
Linaceae	Linum flavum	H Scap	S-Europ.-S-Sib.	1	(+)	Y	(++)	(x)
Lamiaceae	Marrubium vulgare	H Scap	S-Europ.-S-Sib.	1	(++)	Y	(+/−)	x	x	x
Brassicaceae	Mutarda nigra	T Scap	Eurimedit.	1	(+)		?	x	x	x
Lamiaceae	Nepeta cataria	H Scap	E-Medit.-Turan.	2	(+)	Y	(+/−)	x	x	x
Lamiaceae	Origanum vulgare	H Scap	Eurasian	1	(+)		(−)	x	x	x
Asteraceae	Pentanema ensifolium	H Scap	SE-Europ.-Pont.	2	(++)	Y	(+)	(x)
Asteraceae	Pentanema hirtum	H Scap	S-Europ.-S-Sib.	1	(+)	Y	(++)	x	x	x
Caryophyllaceae	Petrorhagia saxifraga	H Caesp	Eurimedit.	1	(+++)	Y	(+)	x	x	x
Apiaceae	Peucedanum cervaria	H Scap	Eurosib.	1	(+)	Y	(+/−)	x	x
Crassulaceae	Petrosedum rupestre	Ch Succ	W- E C-Europ.	1	(+)	Y	(++)	x	x	x
Lamiaceae	Salvia pratensis	H Scap	Eurimedit.	1	(+)	Y	(+)	x	x
Caprifoliaceae	Scabiosa ochroleuca	T Scap	S-Europ.-S-Sib.	1	(+)		(+)	(x)
Lamiaceae	Stachys recta	H Scap	N-Medit.-Mont.	1	(+)	Y	(++)	x	x	x
Asteraceae	Tanacetum corymbosum	H Scap	Eurimedit.	1	(+)	Y	(+/−)	x	x	x
Lamiaceae	Teucrium chamaedrys	Ch Suffr	Eurimedit.	1	(++)	Y	(−)	x	x	x
Fabaceae	Trifolium rubens	H Scap	Central-Europ.	1	(+)	Y	(+)	x	x
Plantaginaceae	Veronica teucrium	H Scap	SE-Europ.-Pont.	1	(+)	Y	(+)	x	x	x
(b)
Family	Species	Biological Form	Chorotype	N. of Ref.	Env. Compatibility	Cyclic Flooding	Flood. Tolerance	Northern Italy	Central Italy	Southern Italy
Caryophyllaceae	Cerastium arvense	H Scap	Paleotemp.	1	t.b.e. (+/−)		(+)	x	x	(x)
Rosaceae	Fragaria vesca	Ch Rept	Eurosib.	1	t.b.e. (+/−)		(+/−)	x	x	x
Crassulaceae	Hylotelephium telephium	H Scap	Eurosib.	1	(+)		?	x
Rosaceae	Potentilla pusilla	H Scap	Central-Europ.	1	(++)		?	x	(x)
Lamiaceae	Prunella vulgaris	H Scap	Circumbor.	2	(+/−)		(+)	x	x	x

a reports all the species selected for environmental compatibility, based on the compatibility classes (from + to +++) assigned according to the criteria described in Section 2.2.2. Species showing unsuccessful waterlogging tolerance (−−) were excluded from the table. Most taxa included in Table 5a were identified exclusively through cyclic flooding studies, particularly those conducted by Eben et al. [76], which specifically targeted species characterized by drought tolerance or ecological adaptability.

Table S12 (Supplementary Materials S3) also reports, for the species with adequate environmental compatibility (from +/− to ++), the habitats of occurrence (Figure 4). The prevalence of dry meadows, steppic grasslands, rocky slopes, roadside margins, and other disturbance-prone habitats suggests that species adapted to fluctuating resource availability and shallow or skeletal soils may represent promising candidates for Mediterranean bioretention systems. These habitats are characterized by periodic water limitation, high solar exposure, recurrent disturbance regimes, and often compacted, occasionally saturated soils, conditions partially analogous to those experienced in engineered stormwater infrastructures.

Manuals. An additional six Italian native taxa were identified; the complete list is provided in Table S11b (Supplementary Materials S3). The average Ellenberg indicator values were 7.0 for light, 6.0 for temperature, 4.3 for humidity, 6.3 for reaction, and 3.8 for nutrients. No Mediterranean chorotypes were recorded among these taxa. Table 5b reports the species selected for environmental compatibility.

Overall, the results suggest that herbaceous Dicotyledons potentially suitable for Mediterranean bioretention systems are primarily associated with stress-tolerant grassland and ecotonal communities rather than wetland-specialist vegetation. However, the relatively limited number of Mediterranean native taxa identified in the literature indicates the need for future experimental validation specifically targeting southern European climates.

3.3. Herbaceous Monocotyledons

Herbaceous monocotyledons were primarily represented by graminoid species associated with wetlands, meadows, and disturbance-adapted grassland ecosystems. Compared with Dicotyledons, this group showed a stronger prevalence of taxa functionally associated with hydrological regulation, soil stabilization, and tolerance to fluctuating moisture conditions, particularly among Poaceae, Cyperaceae, and Juncaceae. However, Mediterranean-native monocotyledons remained poorly represented, further highlighting the limited availability of experimentally tested taxa adapted to hot, dry Mediterranean climates.

In the scientific literature survey, a total of 212 records (genera, species, and cultivars/subspecies) belonging to 16 botanical families were identified. The most represented families were Poaceae (45.3% of the taxa), Cyperaceae (20.8%), Iridaceae (4.7%), Juncaceae (6.1%), and Asphodelaceae (5.2%). Overall, the recorded taxa belonged to 85 genera, of which 58% are represented in Italy by native species. The most recurrent genera were Carex, Juncus, Iris, Panicum, and Dianella; among genera represented in Italy by native taxa, the most frequently cited were Poa, Festuca, and Calamagrostis, in addition to the first four genera mentioned above. Complete lists of botanical families and genera are reported in Tables S4 and S5 (Supplementary Materials S2). The predominance of graminoid genera such as Carex, Juncus, Panicum, and Poa reflects the functional importance of Monocotyledons in bioretention systems, particularly regarding erosion control, root-mediated soil stabilization, and tolerance to hydrological fluctuations. At the same time, the recurrent use of North American prairie and wetland-associated taxa suggests that most experimentally validated Monocotyledons remain linked to temperate climatic conditions rather than fully Mediterranean environments.

The analysis of stormwater manuals revealed a similar distribution among the dominant families, whereas the most recurrent genera were Festuca and, to a lesser extent, Nassella. Deschampsia and Melica were the other most frequently cited genera represented in Italy by native species.

3.3.1. Non-Native Taxa

Compared with herbaceous Dicotyledons, Monocotyledons showed a generally higher representation of drought-tolerant taxa, particularly among warm-season grasses, often characterized by C4 metabolism, and rhizomatous species. Taxa sharing these morphophysiological traits may offer functional advantages under Mediterranean conditions due to their efficient water use and extensive underground organs, which contribute to resilience under alternating dry and temporarily flooded conditions.

Scientific literature. The analysis of the non-native species with at least two citations showed that 54% of the taxa exhibited medium to high drought tolerance. These taxa are reported in

Table 6. Non-native drought-tolerant Monocotyledons (“Anaer. Resist.” = Anaerobic resistance; “Drought Tol.” = Drought tolerance; “Prec. Min.” = Minimum precipitation; “Cas.” = Casual; “Nat.” = Naturalized; “N” = Northern Italy, “C” = Central Italy, “S” = Southern Italy).

		USDA Plant Database					Other Ref.
	n. of cit.	Wetland Status	Anaer. Resist.	Drought Tol.	Moisture Use	Prec. Min. (mm)	Drought Tol.	Presence in Italy
Panicum virgatum	21	FACW/FAC	Medium	Medium	Medium	304.8	Medium	Cas. (N/C)
Dianella revoluta	8	-	-	-	-	-	Medium/High	no
Ficinia nodosa	7	FACU	-	-	-	-	Medium/High	no
Miscanthus sinensis	7	FAC/FACU/UPL	Low	Medium	Medium	609.6	Low/Medium	Cas./Nat. (N/C)
Poa labillardierei	7	-	-	-	-	-	Medium/High	no
Iris tectorum	6	-	-	-	-	-	Medium	Nat. (C)
Schizachyrium scoparium	6	FACU/UPL	None	High	Low	254.0	High	no
Andropogon gerardii	5	FAC/FACU	Medium	High	Low	304.8	High	no
Bouteloua gracilis	4	-	None	High	Medium	203.2	High	no
Cenchrus purpurascens	4	-	None	Medium	Low	812.8	Low/Medium	Cas. (N/C/S)
Juncus patens	4	FACW	-	-	-	-	Medium	no
Lomandra longifolia	4	-	-	-	-	-	Medium	no
Sorghastrum nutans	4	FACU	Low	Medium	Medium	279.4	Medium	no
Sporobolus heterolepis	4	FACU/UPL	-	-	-	-	Medium/High	no
Carex pansa	3	FAC/FACU	-	-	-	-	Medium	no
Chrysopogon zizanioides	3	FACW/FAC	High	High	Low	304.8	High	no
Iris lactea	3	-	-	-	-	-	Medium	no
Sporobolus virginicus	3	FACW/FAC	Medium	High	Low	508.0	High	no
Tradescantia ohiensis	3	FACW/FAC/FACU	-	-	-	-	Medium	no
Agapanthus	2	-	-	-	-	-	Medium/High	no
Cenchrus americanus	2	FACU/UPL	None	Medium	Low	762.0	-	no
Dianella brevipedunculata	2	-	-	-	-	-	Medium	no
Festuca glauca	2	-	-	-	-	-	Medium	no
Juncus kraussii	2	-	-	-	-	-	Medium	no
Tradescantia bracteata	2	FACU/UPL	-	-	-	-	Medium	no

(for the complete list of species, see Table S13a in Supplementary Materials S4).

Manuals. The analysis showed that 67% of the newly most-cited taxa have medium to high drought tolerance. The drought-tolerance assessment of the species is reported in Table S13b (Supplementary Materials S4).

3.3.2. Italian Native Monocotyledons

Scientific literature. Among the surveyed taxa, 40 monocotyledon species were native to the Italian peninsula, while one species (Iris germanica) was identified as a naturalized Archaeophyte. Of these taxa, 40 occur in Northern Italy, 34 in Central Italy, and 29 in Southern Italy, while 33, 29, 30, and 22 taxa are, respectively, distributed within the lowland, hilly, montane, and subalpine belts. The complete list is provided in Table S14a (Supplementary Materials S4).

The predominant chorotype was Circumboreal (9 taxa), followed by the Eurasian and Eurosiberian chorotypes (6 taxa each), and 7 taxa classified as Cosmopolitan or Subcosmopolitan. Only one taxon exhibited a Eurimediterranean distribution. Compared with Dicotyledons, this pattern indicates an even more pronounced underrepresentation of Mediterranean taxa, likely reflecting both the predominance of European applications in cooler and moister environments and the availability of drought-tolerant North American taxa better adapted to harsher climatic conditions. Average Ellenberg indicator values further reflect this trend, being 7.1 for light, 5.9 for temperature, 6.4 for humidity, 6.1 for reaction, and 4.6 for nutrients. Similar to Dicotyledons, most taxa (39%) fell within the temperate to Mediterranean-montane temperature range, predominantly with the value 6. The second largest group was associated with cooler montane environments, while the third corresponded to species characterized by a variable temperature range (“X”). Only one species was associated with not-too-hot Mediterranean environments (8 ≤ T< 10), and another with hot Mediterranean conditions (10 ≤ T). Regarding humidity, most taxa (27%) were characteristic of habitats ranging from fluctuating to moist soil (4 < H ≤ 6), followed by species associated with consistently moist to infrequently inundated soils (6 < H ≤ 8), and, finally, taxa typical of frequently or permanently submerged conditions (8 < H).

Table 7. Environmentally compatible Italian native Monoctyledons from surveyed literature: a = scientific literature; b = stormwater manuals (“x”= presence; “(x)”= limited presence).

Family	Species	Biological Form	Corotype	N. of Ref.	Envi. Compatibility	Cyclic Flooding	Flood. Tolerance	Northern Italy	Central Italy	Southern Italy
Poaceae	Agrostis stolonifera	Ch Rept	Circumbor.	1	t.b.e. [+/−]		(++)	x	x	x
Poaceae	Avenella flexuosa	H Caesp	Subcosmop.	1	t.b.e. [+/−]	Y	(++)	x	x	x
Poaceae	Brachypodium pinnatum	H Caesp	Eurasian	1	(+/−)	Y	(-)	x		(x)
Poaceae	Briza media	H Caesp	Eurosib.	2	t.b.e. [+/−]	Y	(+)	x	x	x
Poaceae	Dactylis glomerata	H Caesp	Paleotemp.	1	(+)		(+)	x	x	x
Poaceae	Deschampsia caespitosa	H Caesp	Paleotemp.	9	t.b.e. [+/−]	Y	(++)	x	x	x
Poaceae	Festuca ovina	H Caesp	Eurasian	3	t.b.e. [+/−]	Y	(++)	x
Poaceae	Festuca rubra	H Caesp	Circumbor.	1	(+/−)		(+)	x	x	x
Iridaceae	Gladiolus palustris	G Bulb	Central-Europ.	1	(+/−)		(++)	x	(x)
Poaceae	Helictotrichon sempervirens	H Caesp	Endem.	1	(+/−)		?	x
Poaceae	Imperata cylindrica	G Rhiz	Thermcosmop.	1	(++)		(+)		x	x
Poaceae	Iris germanica	G Rhiz	Natur. Archeo.	1	(+)	Y	(++)	x	x	x
Iridaceae	Iris sibirica	G Rhiz	Eurosib.	2	(+/−)	Y	(++)	x
Poaceae	Koeleria macrantha	H Caesp	Circumbor.	1	(+)		?	x	x	x
Poaceae	Koeleria pyramidata	H Caesp	Oroph. Europ.	1	(+/−)		?	x	x	x
Poaceae	Melica ciliata	H Caesp	Eurimedit.	2	(++)	Y	(−)/(++)	x	x	x
Poaceae	Molinia caerulea	H Caesp	Circumbor.	4	t.b.e. [+/−]	Y	(++)	x	x
Poaceae	Stipa capillata	H Caesp	Eurasian	1	(++)	Y	(+)	x	x	x
Poaceae	Stipa pennata	H Caesp	Subatlant.	1	(++)	Y	(++)	x

reports all species selected for environmental compatibility, based on compatibility classes (from +/− to ++; due to the limited number of species in the higher compatibility ranges), assigned according to the criteria described in Section 2.2.2.

Table S15a (Supplementary Materials S4) also reports the habitats of occurrence for the species with adequate environmental compatibility (from +/− to ++). The analysis revealed a broad distribution across various meadow and pasture types, along with a notable occurrence on dry-slope habitats, often characterized by rocky or gravelly substrates, as shown in Figure 5.

Overall, the results suggest that Monocotyledons may be among the most functionally suitable vegetation groups for Mediterranean bioretention systems, particularly with respect to hydrological resilience and soil stabilization. However, as with Dicotyledons, the limited representation of Mediterranean-native taxa highlights the need for future experimental studies focused on southern European climates and native grassland communities.

Manuals. An additional four Italian native taxa were identified; the complete list is provided in Table S14b (Supplementary Materials S4). The average Ellenberg indicator values were 7.8 for light, 6.0 for temperature, 8.3 for humidity, 5.0 for reaction, and 3.8 for nutrients. No Mediterranean chorotypes were recorded among these taxa, although Iris pallida exhibited a SE-European distribution. Notably, I. pallida was also the only species classified as environmentally compatible (+) with Mediterranean stress conditions.

3.4. Woody Species

Woody species were an important component of both the scientific literature and stormwater manuals, particularly in practice-oriented applications in Mediterranean-climate regions. Compared with herbaceous taxa, woody vegetation shows a stronger association with long-term structural functions, including shading, evapotranspiration enhancement, slope stabilization, and resilience under low-irrigation conditions.

In the scientific literature survey, a total of 121 shrub records from 39 botanical families were identified. The most represented families were Rosaceae (17.6% of the taxa), Viburnaceae (6.4%), Aquifoliaceae (5.6%), Cornaceae (5.6%), Cyrillaceae (1.8%), and Myrtaceae (5.6%). For trees and species capable of developing as either shrubs or trees, 136 records from 39 botanical families were identified. The most represented families were Myrtaceae (15.6% of the taxa), Betulaceae (8.9%), Sapindaceae (7.4%), Fagaceae (5.9%), and Rosaceae (8.9%).

The shrub taxa belonged to 76 genera, of which 38% are represented in Italy by native species. The most recurrent genera were Viburnum, Ilex, Cornus, Itea, and Salix; among genera represented in Italy by native taxa, the most frequently cited were Spiraea, Betula, and Vaccinium, in addition to Viburnum, Ilex, and Salix. Tree and shrub/tree taxa belonged to 72 genera, of which 39% are represented in Italy by native species. The most recurrent genera were Melaleuca, Acer, Quercus, Betula, and Magnolia; among genera represented in Italy by native taxa, the most frequently cited were Ligustrum, Salix, and Alnus, in addition to Acer, Quercus, and Betula. Complete lists of botanical families and genera are reported in Tables S6–S9 (Supplementary Materials S2). Overall, the recurrent occurrence of genera such as Quercus, Acer, Melaleuca, and Salix reflects the coexistence of two main ecological strategies in bioretention design: species adapted to periodically flooded environments and drought-tolerant woody taxa capable of persisting under prolonged summer water deficit.

In the stormwater manuals survey, the most represented families for shrubs and subshrubs, besides Rosaceae (14.8% of taxa), were Asteraceae, Ericaceae, Rhamnaceae, and Lamiaceae. The most recurring genera were Arctostaphylos, Berberis, and Ribes, followed, among genera represented in Italy by native species, by Salvia, Artemisia, and Frangula. Among shrubs/trees, aside from Fagaceae, Sapindaceae, and Myrtaceae, the most represented families were Cupressaceae and Fabaceae, while the most common genera were Quercus, Acer, and Platanus.

3.4.1. Non-Native Taxa

The analysis of the non-native shrub species showed that 50% of the recurring taxa exhibited medium to high drought tolerance, although most fell within the medium tolerance range. At the same time, an equivalent proportion of species was classified as obligate or facultative wetland taxa. Drought-tolerant species are reported in

Table 8. Non-native drought-tolerant shrubs (“Anaer. Resist.” = Anaerobic resistance; “Drought Tol.” = Drought tolerance; “Prec. Min.” = Minimum precipitation; “Cas.” = Casual; “Nat.” = Naturalized; “N” = Northern Italy, “C” = Central Italy, “S” = Southern Italy).

		USDA Plant Database					Other Ref.
Most Cited Species	n. of cit.	Wetland Status	Anaer. Resist.	Drought Toler.	Moisture Use	Prec. Min. (mm)	Drought Toler.	Presence in Italy
Aronia melanocarpa	5	OBL/FACW/FAC	Medium	Medium	-	609.6		no
Leucophyta brownii	5	-	-	-	-	-	High	no
Leptospermum continentale	4	-	-	-	-	-	High	no
Correa alba	3	-	-	-	-	-	Medium	no
Hibbertia scandens	3	-	-	-	-	-	Medium	no
Photinia × fraseri	3	-	None	Medium	Medium	889.0	Low/Medium	no
Physocarpus opulifolius	3	FACW/FAC/FACU	None	High	Low	889.0	Medium	Nat (N)
Amelanchier utahensis	2	FAC/UPL	None	High	Medium	304.8	High	no
Aucuba japonica	2	-	-	-	-	-	-	Nat (N/C)
Berberis repens	2	-	None	High	Medium	381.0	Medium	no
Cornus racemosa	2	FAC	Low	Medium	Medium	609.6	Low/Medium	no
Hypocalymma angustifolium	2	-	-	-	-	-	Medium	no
Ilex decidua	2	FAC/FACW	Medium	Medium	Medium	762.0	Low/Medium	no
Ilex vomitoria	2	FAC	None	High	Low	914.4	Medium	no
Myrica pensylvanica	2	FAC	Low	High	Medium	812.8	Medium/High	no
Rhus aromatica	2	-	None	High	Low	838.2	High	no
Westringia fruticosa	2	-	-	-	-	-	High	no

(see Table S16a in Supplementary Materials S5 for the complete species list).

The analysis of non-native shrubs/trees and tree species showed that 62% of the recurrent taxa exhibited medium to high drought tolerance, with most falling within the medium/high tolerance range. These drought-tolerant taxa are reported in

Table 9. Non-native drought-tolerant shrubs/trees and trees (“Anaer. Resist.” = Anaerobic resistance; “Drought Tol.” = Drought tolerance; “Prec. Min.” = Minimum precipitation; “Cas.” = Casual; “Nat.” = Naturalized; “N” = Northern Italy, “C” = Central Italy, “S” = Southern Italy).

		USDA Plant Database					Other Ref.
Most Cited Species	n. of cit.	Wetland Status	Anaer. Resist.	Drought Tol.	Moisture Use	Prec. Min. (mm)	Drought Tol.	Presence in Italy
Melaleuca ericifolia	6	-	-	-	-	-	Medium	no
Cercis canadensis	5	UPL/FACU	None	High	Low	508	Medium	no
Acer rubrum	4	FAC	Medium	Medium	High	635.0	Low	no
Ligustrum japonicum	4	FAC/FACU/UPL	None	Medium	Low	762.0	High	Cas. Al.
Celtis occidentalis	3	FAC/FACU	Medium	High	Low	355.6	Medium	no
Ulmus americana	3	FAC/FACW	Low	Medium	High	381.0	Medium	no
Allocasuarina littoralis	2	-	-	-	-	-	Medium/High	no
Banksia integrifolia	2	-	-	-	-	-	Medium/High	no
Cercocarpus ledifolius	2	-	None	High	Low	254.0		no
Cercocarpus montanus	2	-	None	High	Low	228.6		no
Dodonaea viscosa	2	UPL/FACU	None	High	Low	203.2		no
Hakea laurina	2	-	-	-	-	-	Medium/High	no
Kunzea ericoides	2	FACU	-	-	-	-	High	no
Lophostemon confertus	2		-	-	-	-	Medium/High	no
Melaleuca fulgens	2	-	-	-	-	-	Medium/High	no
Melaleuca incana	2	-	-	-	-	-	Medium/High	no
Melaleuca linariifolia	2	-	-	-	-	-	Medium	no
Melaleuca pachyphylla	2	-	-	-	-	-	Medium	no
Osmanthus fragrans	2	-	-	-	-	-	Medium	no
Pittosporum tobira	2	-	None	High	Medium	609.6	Medium/High	Nat. Al.
Quercus alba	2	FACU	None	Medium	Medium	762.0	Medium/High	no
Quercus macdanielii	2	-	-	-	-	-	Medium	no
Quercus macrocarpa	2	FAC/FACU	None	High	Medium	381.0	Medium/High	no
Triadica sebifera	2	FAC	-	-	-	-	Medium	(invasive)

(see Table S17a in Supplementary Materials S5 for the complete species list).

Manuals. Compared with the scientific literature, Mediterranean-climate stormwater manuals showed a stronger representation of drought-tolerant woody species and xeric-adapted taxa, particularly from California floras. This likely reflects the practical need, in operational design guidelines, to prioritize long-term survivability under low-irrigation conditions. The analysis of the non-native shrub species showed that 87% of the recurrent taxa exhibited medium to high drought tolerance, although most fell within the medium tolerance range. The drought-tolerance assessment is reported in Table S16b (Supplementary Materials S5). In the case of non-native shrub/tree and tree species, 62% of the recurring taxa exhibited medium to high drought tolerance, although, also in this case, most fell within the “medium” tolerance range. The drought-tolerance assessment is reported in Table S17b (Supplementary Materials S5).

3.4.2. Italian Native Taxa

Scientific literature. Among the surveyed taxa, 26 woody species were native to the Italian peninsula, while 2 species were identified as naturalized Archaeophytes. Of these 28 taxa, 22 occur in Northern Italy, 21 in Central Italy, and 20 in Southern Italy, with 21, 22, 17, and 10 taxa, respectively, distributed within the lowland, hilly, montane, and subalpine belts. The complete species list is provided in Table S18a (Supplementary Materials S5).

The predominant chorotypes were European-Caucasian, Eurasian, and Eurosiberian (5 taxa each), while only one taxon exhibited a North-Mediterranean montane distribution. Average Ellenberg indicator values were 6.9 for light, 5.5 for temperature, 6.0 for humidity, 6.2 for reaction, and 4.6 for nutrients. Regarding temperature, most species (50%) were primarily associated with cooler montane environments, followed by taxa belonging to the temperate to Mediterranean-montane range, while two species were characteristic of upper-montane or subalpine conditions. In terms of humidity, most taxa (36%) were associated with consistently moist to infrequently inundated soils, followed by species characteristic of habitats ranging from fluctuating to moist soil conditions.

Table 10. Environmentally compatible Italian native woody species from surveyed literature: a = scientific literature; b = stormwater manuals (“x”= presence; “(x)”= limited presence).

(a)
Family	Species	Biological Form	Corotype	N. of Ref.	Env. Compatibility	Flood. Tolerance	Northern Italy	Central Italy	Southern Italy
Sapindaceae	Acer campestre	P Scap	Europ.-Caucas.	3	(+/−)	(++)	x	x	x
Betulaceae	Carpinus betulus	P Scap	Europ.-Caucas.	1	(+/−)	(++)	x	x	x
Cornaceae	Cornus sanguinea	P Caesp	Eurasian	2	(+/−)	(++)	x	x	x
Elaeagnaceae	Hippophae rhamnoides	P Caesp/P Scap	Eurasian	1	t.b.e. [+/−]	(++)	x	x
Oleaceae	Ligustrum vulgare	Np	Europ.-Caucas.	1	t.b.e. [+]	(++)	x	x	x
Platanaceae	Platanus orientalis	P Scap	SE-Europ.	1	(+/−)	(++)			x
Rosaceae	Prunus cerasifera	P Caesp/P Scap	Nat. Archeoph.	1	(+/−)	(++)	x	(x)
Lythraceae	Punica granatum	P Caesp/P Scap	Nat. Archeoph.	1	(++)	?	x	x	x
Salicaceae	Salix alba	P Scap	Paleotemp.	1	t.b.e. [+/−]	(++)	x	x	x
Lamiaceae	Salvia officinalis	Ch Suffr	N-Medit.-Mont.	1	(++)	(++)			x
Rosaceae	Torminalis glaberrima	P Caesp/P Scap	Paleotemp.	1	(+)	(+/−)	x	x	x
(b)
Family	Species	Biological Form	Corotype	N. of Ref.	Env. compatibility	Flood. tolerance	Northern Italy	Central Italy	Southern Italy
Ericaceae	Arbutus unedo	P Caesp	Stenomedit.	1	(++)	?	x	x	x
Anacardiaceae	Cotinus coggygria	Np/P caesp/P scap	S-Europ.-Sudsib.	2	(+)	?	x	x
Rosaceae	Dasiphora fruticosa	Np	Circumbor.	1	(+/−)	(++)	x
Cistaceae	Helianthemum nummularium	Ch Suffr	Europ.-Caucas.	1	t.b.e. [+]	(+/−)	x	x	x
Cupressaceae	Juniperus communis	NP/P scap	Circumbor.	1	t.b.e. [+]	(++)	x	x	x
Lauraceae	Laurus nobilis	P Caesp/P scap	Stenomedit.	2	t.b.e. [+/−]	(+)	x	x	x
Apocynaceae	Nerium oleander	P Caesp	S-Stenomedit.	1	t.b.e. [++]	(+)	x	x	x
Hydrangeaceae	Philadelphus coronarius	Np	Endem.	1	(+)	?	x	x
Fagaceae	Quercus frainetto	P Scap	SE-Europ.	1	(+/−)	?		x	x
Fagaceae	Quercus ilex	P Caesp/P scap	Stenomedit.	2	(++)	(+/−)	x	x	x
Fagaceae	Quercus robur	P Scap	Europ.-Caucas.	1	(+/−)	(++)	x	x	x
Fagaceae	Quercus suber	P Scap	W-Eurimedit.	1	(++)	?		x	x
Lamiaceae	Salvia rosmarinus	Np	Stenomedit.	1	(+++)	?	x	x	x
Styracaceae	Styrax officinalis	P Caesp	NE-Stenomedit.	1	(++)	?		(x)	(x)
Malvaceae	Tilia cordata	P Scap	Europ.-Caucas.	1	(+/−)	(++)	x	x	x

a reports all the native species selected for environmental compatibility, based on the compatibility classes (from +/− to ++) assigned according to the criteria described in Section 2.2.2.

Table S19a (Supplementary Materials S5) also reports the habitats of occurrence for the species included in Table 10a. The analysis revealed a predominance of mesophilous or meso-thermophilous woodlands and riparian habitats.

Manuals. An additional 19 Italian native woody species were identified through the manual survey. Of these, 16 occur in Northern Italy, 18 in Central Italy, and 16 in Southern Italy, while 17, 9, 7, and 4 taxa are, respectively, distributed within the lowland, hilly, montane, and subalpine belts. The complete species list is available as Table S18b (Supplementary Materials S5). Besides the European-Caucasian chorotype (6 taxa), the second most represented group consisted of Mediterranean chorotipes (5 taxa). Average Ellenberg indicator values were 6.6 for light, 6.8 for temperature, 4.7 for humidity, 5.5 for reaction, and 4.3 for nutrients. Regarding temperature, most taxa (40%) belonged to the temperate to Mediterranean-montane range, while the second-largest group included species associated with climates ranging from Mediterranean-montane to warm Mediterranean conditions (8 ≤ T < 10). One species was characteristic of hot Mediterranean environments (10 ≤ T). In terms of humidity, most species (40%) were associated with habitats ranging from dry to fluctuating moisture conditions, followed by taxa characteristic of fluctuating to moist environments. Table 10b reports all native species selected for environmental compatibility.

Table S19b (Supplementary Materials S5) also reports the habitats of occurrence for the species included in Table 10b. The analysis revealed a predominance of scrublands, rocky habitats, and dry thermophilous woodlands.

Overall, woody taxa appeared particularly relevant for Mediterranean bioretention applications due to their structural and ecological resilience under stressful urban conditions. Nevertheless, most recurrent taxa originated from North American or Australian floras, whereas experimentally validated Mediterranean-native woody species remained comparatively scarce. In particular, most native species identified through the scientific literature survey were associated with cooler and moister environments, potentially limiting their sustainable applicability under hotter and drier Mediterranean conditions. In this respect, stormwater manuals offered a broader palette of environmentally compatible species, including properly Mediterranean taxa.

3.5. Csa-Climate Stormwater Management Manuals Analysis: Non Native Taxa

The complete list of non-native taxa from the Csa manuals is reported in Table S20 (Supplementary Materials S6). A total of 85 herbaceous species were recorded, including 50 Dicotyledons and 35 Monocotyledons, of which 64 were native to California. Overall, most species exhibited either “Low” (31%) or “Low to medium” (18%) water requirements. For California native taxa (including two non-native species in the Calflora database), the predominant maximum temperature range was 30–35 °C (43%), followed by 35–40 °C (35%). The most represented minimum precipitation range was 200–400 mm/year (52%), followed by 400–600 mm/year (25%). The most recurrent plant communities were the Yellow Pine Forest (54%) and chaparral (51%), followed by wetland-riparian, Foothill Woodland, and Red Fir Forest communities. 31 California native taxa (48%) were reported to occur in wetlands.

A total of 164 woody species were recorded, including 116 shrubs, 18 shrub/tree taxa, and 30 trees, of which 133 were native to California. Most species exhibited either “Low” (32%) or “Very Low” (31%) water requirements. Among California native taxa (together with six non-native species included in the Calflora database), the predominant maximum temperature range is 30–35 °C (41%), followed by 35–40 °C (39%). The most represented minimum precipitation range was 200–400 mm/year (59%), followed by 400–600 mm/year (21%). The most recurrent plant communities were Chaparral (55%), Foothill Woodlands (34%), Yellow Pine Forest (28%), Coastal Sage Scrub (27%), and wetland-riparian (21%). Furthermore, 34 California native taxa (26%) were reported to occur in wetlands.

4. Discussion

The present review integrates worldwide peer-reviewed literature and Mediterranean-climate technical manuals to assess the potential suitability of plant taxa proposed for bioretention systems in Mediterranean-climate regions of Italy. Both the recurrent non-native species and the Italian native taxa were assessed, with particular attention to the combined effects of drought, heat, and temporary flooding. Flood tolerance, especially for native species, was also considered to evaluate the potential applicability of taxa across the full moisture gradient typically present within bioretention facilities.

The main survey featured horticultural books and peer-reviewed studies, while Mediterranean-climate stormwater manuals were included to broaden the range of taxa currently adopted in operational practice, and to identify Mediterranean species already used in stormwater infrastructures. Although technical manuals do not constitute peer-reviewed scientific evidence, they provide relevant insight into species currently employed by practitioners, particularly in Mediterranean-climate regions where experimental studies remain scarce. Nevertheless, the criteria underlying plant selection in these manuals often remain unclear, and the transferability of horticultural recommendations to different environmental contexts may be uncertain, as already highlighted by several authors [8,82,83,84]. This is particularly true for stormwater manuals, which, on the other hand, are supposed to rely on empirical evidence and horticultural knowledge.

The reference scales developed for drought tolerance, environmental compatibility (i.e., the combination of heat and drought), and flooding tolerance should therefore be interpreted as pragmatic tools for a preliminary screening and comparisons within and across plant groups, rather than as definitive predictive models. Indeed, the review revealed both a lack of scientifically grounded information and multiple, often divergent, and scarcely comparable sources. Guzzon et al. [83] reported similar difficulties in selecting ornamental shrubs, highlighting the contrast between qualitative, experience-based horticultural information—which is often inconsistent—and highly specialized technical literature that focuses only on specific morphophysiological responses.

The predominance of North American taxa in the surveyed literature is likely a consequence not only of ecological suitability but also of the historical development of bioretention research itself. Since modern bioretention systems were primarily developed and experimentally tested in North American urban contexts [5,85], especially in temperate regions, the species most frequently investigated tend to reflect locally available flora and prairie-associated vegetation models. E.g., within the first survey of herbaceous Dicotyledons, all the main genera have a North American distribution. This is also reflected in North American stormwater manuals, which generally encourage the use of native species due to their environmental adaptations and ecological value, as shown in the plant lists proposed for bioretention [84]. This geographical research bias has probably contributed to the limited experimental evaluation of Mediterranean-native species, despite their potential relevance under increasingly drought-prone climatic conditions.

The limited representation of Mediterranean-native taxa may also derive from intrinsic ecological and operational constraints. Many Mediterranean plant communities are adapted to prolonged summer drought but not necessarily to repeated short-term waterlogging, which represents a key functional requirement in bioretention systems. Furthermore, Mediterranean grasslands and xeric communities remain comparatively understudied in stormwater infrastructure research, while commercially available plant palettes used by practitioners are often derived from established North American or ornamental horticultural selections.

Growth-form analysis revealed contrasting patterns between the worldwide literature and Mediterranean-climate manuals. Herbaceous taxa dominate the scientific literature, whereas woody species—particularly shrubs—prevail in Mediterranean manuals, especially those from Csa-climate regions. This pattern is consistent with Mediterranean landscaping literature [86], which extensively relies on drought-tolerant shrubs and sub-shrubs, capable of coping with aridity through morphophysiological adaptations such as deep rooting, sclerophylly, and woody tissues. Similarly, Houdeshel et al. [39] recommend using shrubs and bunchgrasses in bioretention design for xeric climates.

Among the dominant herbaceous genera, many taxa are associated with medium to high moisture availability (e.g., Eutrochium, Symphorythricum, Epilobium, Lythrum, Carex, Juncus, and Calamagrostis), whereas a limited part has a broader ecological variability (e.g., Solidago, Geranium, Poa), and only a few are also characteristic of drought-prone habitats (e.g., Achillea, Festuca) [50,53,70,73]. Although this partially reflects the family’s amplitude and distribution, the strong predominance of Asteraceae among herbaceous Dicotyledons remains particularly remarkable, and suggests considering this taxon for further research within the native flora. For shrubs, besides the variability of the family Rosaceae, the main taxa identified in the scientific literature are associated with moist or riparian environments (Ilex, Viburnum, Cornus, and particularly Salix, Betula, and Vaccinium), whereas the trees’ genera are more variable (e.g., Quercus, Melaleuca), but still encompass high moisture-loving taxa (e.g., Betula, Salix, Alnus) [50,53,59,70,73]. Overall, only a relatively limited number of taxa appear potentially suitable for prolonged drought-prone Mediterranean urban environments.

The results also suggest a frequent functional trade-off between drought tolerance and flooding tolerance. Species associated with wetlands or permanently moist environments often exhibit high waterlogging resistance but limited drought resilience, whereas most Mediterranean xerophytic taxa may tolerate prolonged summer drought while remaining poorly adapted to repeated inundation. This dual hydrological stress likely represents one of the major challenges for vegetation selection in Mediterranean bioretention systems. For this reason, the numerous taxa reported in the literature from wetland or riparian habitats—often strictly hygrophilous species associated with cooler environments—are unlikely to represent a sustainable choice under Mediterranean conditions. Only a limited number of specialized ecological niches within southern Italian wetland systems may undergo a seasonal transition from winter waterlogging to complete summer desiccation, potentially providing valuable ecological models for Mediterranean bioretention design. Nevertheless, none of the species reported in the literature occur in such habitats, and only a few species (e.g., E. cannabinum, E. hirsutum) might be considered for use if provided adequate irrigation.

Among herbaceous taxa, Monocotyledons—particularly graminoids—generally showed greater adaptability to drought stress than Dicotyledons, both in terms species richness and of tolerance intensity. Several North American grasses, including S. scoparium, A. gerardii, B. gracilis, S. viriginicus, S. heterolepis, and M. rigens, stand out for their combined tolerance to drought and temporary flooding [45,50,53]. Interestingly, the largely used P. virgatum is absent from the Mediterranean-climate manuals, suggesting, as a tallgrass prairie species, a limited suitability under fully Mediterranean conditions. Nevertheless, the adoption of the above-mentioned species for Italy—even if convenient in terms of plant resilience—may involve ecological risks, particularly considering the high invasive potential of Poaceae [87]. Although species from more arid habitats and from North America currently appear less problematic at the global scale [87], Poaceae already represent the second most abundant family among naturalized species in Italy, and some Sporobolus species are considered already invasive [88].

Among woody species, Mediterranean manuals include a broader range of drought-tolerant species than the scientific literature, although this may partly reflect the inclusion of taxa suitable for general LID applications rather than specifically for bioretention systems. Nevertheless, several drought-tolerant woody taxa are also indicated for flood tolerance [45], suggesting the need for further experimental investigation of their compatibility—and that of similar species—with temporary waterlogging. Within the scientific literature survey, trees generally exhibited greater adaptability than shrubs, largely due to the occurrence of drought-tolerant Australian Myrtaceae species [59,60,61].

Several Italian native herbaceous Dicotyledons were identified as potentially compatible with Sub-Mediterranean or Mediterranean conditions. However, most taxa fell within the moderate compatibility range (+), suggesting limited tolerance to harsher urban Mediterranean conditions without irrigation support. This interpretation is further supported by the absence of the Stenomediterranean chorotype, and by the fact that the most promising species (e.g., P. saxifraga, H. perforatum, M. vulgare) appear to be more strongly adapted to aridity only, rather than to the combination of drought and high temperatures [69]. Furthermore, most of the species come from waterlogging [78] or cyclic flooding experiments [76,77] and not from bioretention field implementations. Of these, the main experiment was conducted in a temperate climate, despite the focus on drought-tolerant species [76]. Indeed, for this reason, most species come from dry meadows or pastures [71,72,73], yet not from the hotter Mediterranean stations, even when present in Southern Italy [69].

Selected Monocotyledons appeared even less adapted to heat and drought stress, with most taxa showing only limited environmental compatibility (+/−). This pattern is reflected in the prevalence of moist meadow habitats among recurrent species. Remarkable exceptions are M. ciliata and Stipa spp., which were again tested only for cyclic flooding, with contrasting [76,77] and promising results [76], respectively. These findings may indicate, unexpectedly, the potential relevance of the respective plant communities of these taxa, i.e., dry slopes (M. ciliata, S. capillata) and steppic meadows (S. pennata) [73], and of other Southern Stipa species and communities. It is uncertain to what extent species such as A. flexuosa, D. caespitosa, and M. coerulea can tolerate heat and drought, even though they are included in the WUCOLS species list and have moderate water requirements for non-desertic regions [68]. This suggests a certain compatibility with Sub-Mediterranean conditions.

Among the woody species derived from the first survey, only a few show minimum environmental compatibility, and most belong to the “limited” range (+/−). Conversely, nearly all species appeared capable of tolerating short-term flooding and, in some cases, longer inundation periods [81], consistently with their generally high moisture affinity and frequent occurrence in riparian habitats [71,73]. In this respect, stormwater manuals provide a broader selection of drought-adapted taxa, including a significant proportion of Mediterranean species; however, their tolerance to temporary flooding often remains insufficiently documented or uncertain. In scientific literature, only a few garigue species, including Salvia officinalis, have been experimentally evaluated for waterlogging tolerance, with promising results [89].

Although the scientific literature survey was conducted globally, the technical manuals considered in this study were primarily derived from North American Mediterranean-climate regions, particularly California. Future research should extend the comparative analysis to additional Mediterranean-climate regions, primarily Australia, and potentially other regions such as South Africa or Chile, to further assess the transferability of plant-selection strategies across different ecological and urban contexts.

California stormwater manuals provide a particularly broad range of drought- and heat-adapted taxa, a significant portion of which is reportedly also flood-tolerant [45]. Many of these species occur under minimum-precipitation regimes, comparable to the driest Mediterranean regions of Italy, and tolerate high summer temperatures. However, the transferability of ecological information derived from natural habitats [57] to highly altered urban environments should be evaluated cautiously [83]. The recurrent occurrence of “chaparral”—sensu lato as part of the division “Californian Chaparral, Coastal Scrub & Grassland”, or sensu stricto as a macro-group of more or less xeric shrubland communities [90]—within California manuals may provide a useful ecological parallel for the identification of Mediterranean-compatible species. In particular, the coexistence of taxa adapted to both dry shrublands and periodically moist habitats suggests potentially valuable adaptive strategies for bioretention systems. Nevertheless, these ecological combinations may also characterize aggressive species with invasive potential, as demonstrated by Amorpha fruticosa, which is invasive across several European regions [91].

The use of non-native species in NbS remains controversial, particularly in biodiversity-oriented urban ecological planning [92]. On the one hand, some non-native taxa recurrently identified in the bioretention literature exhibit functional traits potentially advantageous under Mediterranean urban conditions, including tolerance to alternating flooding and drought stress, rapid establishment, and resilience in compacted or nutrient-poor soils (e.g., Panicum virgatum). These characteristics may partially explain their frequent adoption in operational stormwater infrastructures. On the other hand, the introduction of non-native species may involve ecological risks, including potential invasiveness, ecological homogenization, and reduced support for local biodiversity. Moreover, when species are not adequately adapted to local ecological conditions, long-term resilience under climate change may remain uncertain. In Mediterranean regions, where urban ecosystems are already subjected to significant ecological pressure [93,94], the extensive transfer of plant palettes developed for temperate North American systems may therefore conflict with broader ecological concerns.

Given the complexity of the framework and the trade-offs involved, however, the question of whether to adopt non-native species in the urban environment remains open. The main concern about non-natives is their potential invasiveness, which, however, has been shown to depend on a combination of biological traits that natives might also possess [95], especially in response to a rapidly evolving climate change scenario [96]. Conversely, several non-native non-invasive species have been shown to provide ecosystem services, including support for biodiversity and biological interactions, as well as for natives [84]. Climate adaptation is again a trait that might be shared by natives and non-natives from analogous environmental conditions; therefore, planting sustainability should rely on the species’ compatibility with soil and competition conditions, together with establishing and management choices [95]. Furthermore, Rhami et al. [84] highlight the progressive shift in the literature toward plant functional traits—e.g., drought tolerance, root system type—as a guiding criterion for plant selection in bioretention.

Nevertheless, invasive species remain a major threat to Italian ecosystems [97], as highlighted in the previous paragraph, particularly for the Poaceae group. E.g., in recent decades, Sicilian grasslands have been disrupted by the rapidly spreading grass species Cenchrus setaceus [98], which, on the other hand, might represent an ideal candidate for its resilience to the extreme stresses of urban Mediterranean conditions. Viciani et al. [99] provided a list of the most invasive-impacted vegetation types in Italy, in which ruderal and hygrophilous herbaceous communities were the most affected, as were hydrophytic habitats and alluvial, marshy, and riparian communities. Competitive bioretention plants with high water affinity might therefore further contribute to this trend.

Overall, species selection for Mediterranean bioretention systems should not rely exclusively on species origin, but rather on a balanced evaluation of ecological compatibility, invasiveness risk (based on critical biological traits), functional performance, and contribution to urban biodiversity. Consequently, while carefully selected non-native taxa may still represent useful short-term functional options in specific urban contexts, future research should increasingly prioritize Mediterranean-native species and stress-adapted local plant communities that can simultaneously tolerate drought, heat stress, and temporary waterlogging.

5. Conclusions

This study highlights significant limitations in current knowledge and application of plant selection for bioretention systems in Mediterranean and Sub-Mediterranean contexts in Italy. Existing scientific literature and stormwater management manuals are strongly biased toward North American species, while European ones, particularly Italian-native taxa, remain underrepresented. As a result, plant palettes commonly used in bioretention design do not adequately reflect the ecological and climatic conditions of central and southern Italy.

A key finding is the scarcity of species simultaneously tolerant to drought, heat, and periodic flooding, conditions that are increasingly relevant under Mediterranean climate change scenarios. Herbaceous species documented in scientific literature are often associated with moist environments, whereas Mediterranean manuals emphasize drought-adapted woody taxa, especially shrubs. However, flood tolerance for many of these species remains insufficiently investigated.

Italian native species show some potential for use in Mediterranean bioretention systems; however, their overall adaptability to harsh urban Mediterranean conditions often appears limited, frequently requiring supplemental irrigation. Among herbaceous taxa, several species associated with dry meadows and pasture communities appear to combine relatively high environmental compatibility with unexpectedly good flooding tolerance, suggesting the need for further investigation of similar plant communities and for expanding flooding-tolerance testing. At the same time, habitats occasionally subjected to temporary flooding should also be considered, particularly where soils undergo substantial summer desiccation. Furthermore, future research should progressively adopt a plant-community-oriented approach, which may improve the understanding of long-term species interactions, ecological dynamics, and overall system resilience.

In this sense, integrating non-native species into such communities presents both opportunities—such as greater stress tolerance and complementary or extended phenological phases- and risks, particularly regarding invasiveness. The potential aggressiveness of alien species should be singularly evaluated based on their biological traits and the resulting expected behavior, across both anthropically affected and natural ecosystems.

Overall, these findings highlight a substantial research gap and emphasize the need for targeted theoretical and field-based studies focused primarily on native Mediterranean flora from Central and Southern Italy. Based on this, future experimental research and testing should rely on integrating ecological-context analysis with targeted investigations of plant functional traits, including both morphological characteristics (e.g., rooting depth and growth form) and physiological responses to stresses.