As Bioswale Revetments

sustainability

Article Performance of Rolled Erosion Control Products (RECPs) as Bioswale Revetments

Arnaldo T. Coelho 1, Gustavo B. Menezes 2,*, Terezinha C. de Brito Galvão 3 and Joaquim F. T. Coelho 1

1 Ingá Engenharia, Belo Horizonte 30320-130, MG, Brazil; [email protected] (A.T.C.); [email protected] (J.F.T.C.) 2 Department of Civil Engineering, California State University, Los Angeles, CA 90032, USA 3 Independent Researcher, Atlanta, GA 30033, USA; [email protected] * Correspondence: [email protected]

Abstract: Vegetated swales, or bioswales, are among the most commonly used type of green infrastructure (GI) for managing stormwater in temperate climate regions. However, performance data on bioswale drainage technology applied to highly weathered soils (low fertility, high acidity, and erosion prone) in tropical and subtropical climates are still limited. Aimed at closing this gap, this research investigated the performance—assessed in terms of vegetation biomass, biodiversity and coverage of swale, the structural integrity of revetments, and erosion control potential—and cost effectiveness of ﬁve rolled erosion control products (RECPs) currently available on the market, in combination with herbaceous vegetation as the revetment of drainage swales, in tropical soils. Additionally, the research project evaluated the performance of a new preseeded RECP, the Pre- seeded Reinforcement Mat, for drainage in areas that are difﬁcult to access. The performances of

all six RECPs were generally adequate as bioswale revetments in the conditions investigated, with performance index values ranging from 6 to 10 in a 0 to 10 scale. At the same time, some RECPs

Citation: Coelho, A.T.; Menezes, were more conducive to the growth of regional herbaceous vegetation species, measured in terms of G.B.; de Brito Galvão, T.C.; Coelho, biodiversity, which ranged from 2 to 14 species in the different bioswales, and some were more cost J.F.T. Performance of Rolled Erosion effective than others, with costs ranging from 19% to 106% of the cost of concrete lined swales. Control Products (RECPs) as Bioswale Revetments. Sustainability Keywords: bioswales; erosion control; RECPs; stormwater management 2021, 13, 7731. https://doi.org/ 10.3390/su13147731

Academic Editor: Teodor Rusu 1. Introduction In recent years, green infrastructure (GI) practices, which appeared in the 1990s, Received: 18 May 2021 have become a very important tool in stormwater management [1] and erosion control [2] Accepted: 5 July 2021 Published: 11 July 2021 systems, in both rural and urban areas. According to Wise [3], GI strategies can be deﬁned as an interconnected network of open spaces and natural areas—greenways, wetlands,

Publisher’s Note: MDPI stays neutral parks, forest preserves, and native plant vegetation—that naturally manages stormwater, with regard to jurisdictional claims in reduces the risk of floods, captures pollution, and improves water quality [4]. GI solutions published maps and institutional affil- such as green roofs, porous pavements, bioretention cells, and bioswales have become iations. important bioengineering approaches due to their inherently lower cost in the phases of implementation and operation, and the fact that they are environmentally appropriate and generally aesthetically pleasing. Performance studies have been carried out to compare GI with gray infrastructure made of concrete and steel, and the combination of both green and gray infrastructure, which in some cases has led to better results [5]. For instance, Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. the results of a life–cost analysis performed by Cohen [6] indicate that hybrid solutions This article is an open access article involving concrete (gray) and green infrastructure were more cost effective than solutions distributed under the terms and with concrete only. conditions of the Creative Commons Bioswales, which are among the most commonly used GI, are vegetated water con- Attribution (CC BY) license (https:// veyance systems used to manage stormwater runoff [7]. The benefits of bioswales include creativecommons.org/licenses/by/ the potential for retention and subsequent infiltration into the subsurface, and evapotran- 4.0/). spiration, which when combined can reduce stormwater runoff by 90% for precipitations of

Sustainability 2021, 13, 7731. https://doi.org/10.3390/su13147731 https://www.mdpi.com/journal/sustainability Sustainability 2021, 13, x FOR PEER REVIEW 2 of 16

the potential for retention and subsequent infiltration into the subsurface, and evapotran- spiration, which when combined can reduce stormwater runoff by 90% for precipitations of up to 25 mm accumulated over 24 h [8]. Considering the high concentration of pollutants in the “first-flush” of precipitation [9–11], this potential for retention also leads to a reduction in water pollution [12]. In recent years, a robust amount of research has been carried out on the effectiveness of bioswales as a tool for the management of stormwater as well as carbon emission control [13–18], including large scale, citywide projects [19], and, therefore, data on stormwater runoff performance and long term hydrological simulations that integrate bioswales to other GI solutions have become widely available [20–22]. However, data on veg- Sustainability 2021, 13, 7731 etation growth, erosion control potential, and guidelines for bioswales are not2 of widely 18 disseminated, and studies under tropical conditions are limited. This is important because vegetation cover in bioswales is directly connected to their hydrological performance, and uppoor to 25 vegetation mm accumulated establishment over 24 h has [8]. been Considering often associated the high concentration with the reduced of pollutants performance in of thebioswales “first-flush” [23]. of In precipitation Brazil, for instance, [9–11], this there potential are no for technical retention alsostandards/technical leads to a reduction guidelines inthat water could pollution guide [ 12the]. selection of materials and the type of vegetation to be used in the implementationIn recent years, of abioswales, robust amount for instance, of research what has beentype carriedof seeds out would on the better effective- develop in nessacidic, of bioswaleshighly weathered, as a tool for and the low management fertility soils of stormwaterunder tropical as well conditions. as carbon emission controlThis research [13–18], includingwork focuses large on scale, researching citywide projectsthe implementation [19], and, therefore, and operation data on of veg- stormwateretated swales runoff with performance rolled erosion and longcontrol term pr hydrologicaloducts (RECPs) simulations revetment, that in integrate rural areas un- bioswales to other GI solutions have become widely available [20–22]. However, data on der tropical conditions. More specifically, it targets study areas that are erosion prone, vegetation growth, erosion control potential, and guidelines for bioswales are not widely disseminated,with multiple and erosional studies under ramifications, tropical conditions and that are are limited. situated This isclose important to the because foundations of vegetationpower transmission cover in bioswales towers, is directlysuch as connected that depict to theired in hydrological Figure 1. The performance, removal of and protective poorvegetation vegetation cover establishment and soil disturbance has been often during associated the installation with the reduced of transmission performance towers of have bioswalesbeen reported [23]. In as Brazil, causing for instance,serious erosion there are and no technical affecting standards/technical the natural hydrology guidelines of disturbed thatsites could [24,25]. guide the selection of materials and the type of vegetation to be used in the implementationThis work of aims bioswales, to contribute for instance, to the what curren typetly of limited seeds would scientific better literature develop inon the use acidic,of RECPs highly as weathered, revetments and for low bioswales fertility soils that under could tropical potentially conditions. lead to the development of This research work focuses on researching the implementation and operation of technical standards/guidelines on the choice of materials and vegetation for revetments to vegetated swales with rolled erosion control products (RECPs) revetment, in rural areas underbe used tropical under conditions. tropical conditions. More specifically, In this it targetssense, studythis work areas carried that are erosionout a field prone, scale com- withparative multiple evaluation erosional of ramifications, six erosion control and that mats are situated available close on to the the market, foundations as well of as a re- powercently transmission developed product towers, suchaimed as thatfor the depicted revetment in Figure of swales,1. The removal the preseeded of protective reinforcement vegetationmat. The coveroverarching and soil hypothesis disturbance is during that the installationadded structural of transmission protection towers of bioswale have beds beenwith reported RECP will as causing fortify seriousthe upper erosion soil and layer, affecting enhancing the natural the chances hydrology of ofvegetation disturbed to thrive sitesin the [24 low,25]. soil fertility conditions described above.

FigureFigure 1. 1.Gully Gully erosion erosion due due to poor to poor land land management management practices practices near a transmissionnear a transmission tower in thetower in the studystudy area. area.

This work aims to contribute to the currently limited scientiﬁc literature on the use of RECPs as revetments for bioswales that could potentially lead to the development of technical standards/guidelines on the choice of materials and vegetation for revetments to be used under tropical conditions. In this sense, this work carried out a ﬁeld scale comparative evaluation of six erosion control mats available on the market, as well as a recently developed product aimed for the revetment of swales, the preseeded reinforcement mat. The overarching hypothesis is that the added structural protection of bioswale beds with RECP will fortify the upper soil layer, enhancing the chances of vegetation to thrive in the low soil fertility conditions described above. Sustainability 2021, 13, x FOR PEER REVIEW 3 of 16

Sustainability 2021, 13, 7731 3 of 18

2. Materials and Methods 2. Materials and Methods 2.1. Study Area 2.1. Study Area The study areas, Thewith study a total areas, length with of a total200 m, length are located of 200 m, in are the located cities inof theSão cities Gonçalo of São Gonçalo do Abaeté—Site 1—anddo Abaet Córregoé—Site 1—and Danta—Site Córrego 2—both Danta—Site in the 2—both state of in Minas the state Gerais, of Minas Brazil Gerais, Brazil (Figure 2). Both sites(Figure are2 located). Both in sites the areSão located Francisco in the basin S ã whicho Francisco drains basin an area which of 641.000 drains an area of km2. 641.000 km2.

FigureFigure 2. Locations 2. Locations of study area:of study São Gonçaloarea: São do Gonçalo Abaeté—Site do Abaeté—Site 1 (18.3415◦ S, 45.83281 (18.3415°◦ W)—and S, 45.8328° Córrego W)—and Danta—Site 2 (19.8238Córrego◦ S, 45.9043 Danta—Site◦ W). 2 (19.8238° S, 45.9043° W).

2.2. Climate 2.2. Climate The climate of Site 1 is, according to Köppen [26,27], characterized as Aw, which is The climate typicalof Site of1 is, a tropical according savannah, to Köppen with a dry[26,27], winter characterized with an average as Aw, precipitation which is of less than typical of a tropical60 savannah, mm. The rainy with period a dry starts winter inNovember with an average and lasts precipitation until April (summer of less inthan the Southern 60 mm. The rainy Hemisphere).period starts Thein November average annual and lasts precipitation until April is 1317 (summer mm. The in predominantthe Southern vegetation Hemisphere). Thetype average is grasses, annual while precipitation bushes and undergrowthis 1317 mm. vegetationThe predominant are also observed.vegetation type is grasses, while Thebushes climate and for undergrowth Site 2 is also characterized vegetation are as Aw, also with observed. an average annual precipitation The climate forof 1566 Site mm2 is andalso atcharacterized least 70% or moreas Aw, of annualwith an precipitation average annual occurring precipita- in the warmest tion of 1566 mm andsix monthsat least [70%28]. or more of annual precipitation occurring in the warmest six months [28]. 2.3. Soils Soils of the study area, “Cerrado Biome”, originated from thick layers of sediments 2.3. Soils dating back to the Tertiary. They are characterized by a soil profile of a depth of around Soils of the study10–25 m,area, a red “Cerrado or yellowish Biome”, red color, originated porous, from permeable, thick welllayers drained, of sediments and highly weath- dating back to theered Tertiary. [29,30]. They are characterized by a soil profile of a depth of around 10–25 m, a red or yellowishThe soil red color color, is porous imparted, permeable, by coatings well ofvery drained, fine particlesand highly of humifiedweath- organic ered [29,30]. matter, iron hydroxides, iron oxides, and sometimes manganese oxides. Manganese oxide minerals, although less prevalent than iron oxides in soils, commonly have an even stronger The soil color is imparted by coatings of very fine particles of humified organic mat- pigmenting effect upon soil. Iron oxides impart a characteristic yellowish (goethite) or red ter, iron hydroxides,color iron (hematite). oxides, The and color sometimes can be quantified manganese by visual oxides. comparison Manganese to color oxide charts, such minerals, althoughas theless Munsell prevalent soil colorthan charts iron [oxides31]. By usingin soils, the Munsellcommonly chart, have the presencean even of goethite stronger pigmentingand hematiteeffect upon is inferred soil. Iron from oxides the hue; impart soils with a characteristic hues ranging fromyellowish 2.5 Y to (goe- 7.5 YR usually thite) or red color (hematite). The color can be quantified by visual comparison to color charts, such as the Munsell soil color charts [31]. By using the Munsell chart, the presence of goethite and hematite is inferred from the hue; soils with hues ranging from 2.5 Y to 7.5 Sustainability 2021, 13, x FOR PEER REVIEW 4 of 16

YR usually contain only goethite; if associated with hematite, the hue becomes redder than 7.5 YR. Pure hematite tends to give the reddest hues (5 R to 2.5 YR) [29,32]. From the point of view of granulometric size, its texture is dominated, in general, by the sand fraction, followed by clay and finally silt. Therefore, the soils can be classified as predominantly sandy, sandy-clay, clayey-sandy, and, as such, erosion prone. In addition, Sustainabilitytheir2021, 13 water, 7731 holding capacity is relatively low. 4 of 18 The organic matter content of these soils is usually between 3 and 5%, which is very small. As the climate is seasonal, with a long dry period, humus decomposition is very little known [29].contain only goethite; if associated with hematite, the hue becomes redder than 7.5 YR. As for theirPure electrochemical hematite tends characteristics, to give the reddest they hues (5are R tovery 2.5 YR)acidic, [29,32 with]. a pH that can From the point of view of granulometric size, its texture is dominated, in general, by vary from less thanthe sand4 to fraction,a little followedmore than by clay5 (Figure and finally 3). This silt. Therefore, strong acidity the soils is can in be large classified part as due to the high levelspredominantly of Al3+, which sandy, sandy-clay,makes them clayey-sandy, alumino and,toxic as to such, most erosion plants. prone. High In addition,levels of Fe and Mn ionstheir also water contribute holding capacityto their is toxicity. relatively Low low. cation exchange capacity, low base sum, and high Al3+ saturationThe organic mattercharacterize content ofthese these soil soilss, iswhich usually are between profoundly 3 and 5%, dystrophic which is very small. As the climate is seasonal, with a long dry period, humus decomposition is very and, therefore, unsuitablelittle known for [29 ].agriculture. In their electrochemical behavior, the soils from the study area differ Asfundamentally for their electrochemical from the characteristics, soils of temperate they are zones. very acidic, Soils with from a pHtemper- that can ate regions generallyvary from contain less than large 4 to amounts a little more of than2:1 layer 5 (Figure silicate3). This clay strong minerals acidity iswith in large a per- part 3+ manent negativedue surface to the charge high levels resulting of Al from, which isomorphous makes them aluminosubstitution toxic toin mostthe clay plants. struc- High levels of Fe and Mn ions also contribute to their toxicity. Low cation exchange capacity, ture. This type of charge is known as constant charge. On the other hand, soils from trop- low base sum, and high Al3+ saturation characterize these soils, which are profoundly ical regions are generallydystrophic dominated and, therefore, by unsuitable 1:1 layer for silicates agriculture. along In with their electrochemicaliron and aluminum behavior, oxide minerals, allthe of soils which from thehave study a surface area differ charge fundamentally dependent from on the soilsthe pH of temperate value. This zones. sur- Soils face charge can befrom positive temperate or regionsnegative. generally This containtype of large charge amounts is known of 2:1 layer as constant silicate clay surface minerals potential. Some ofwith the a differences permanent negative between surface trop chargeical and resulting temperate from isomorphous zone soils substitutionare attributed in the clay structure. This type of charge is known as constant charge. On the other hand, soils to the existence offrom positive tropical electric regions charges are generally and dominatedto the extreme by 1:1 layerease silicateswith which along cations with iron are and hydrolyzed in tropicalaluminum soils. oxide Thus, minerals, on the all ofba whichsis of have electrochemical a surface charge behavior, dependent soils on the may pH value. be classified accordingThis surfaceto whether charge they can be have positive a constant or negative. surface This type charge of charge or constant is known assurface constant potential. In general,surface tropical potential. soils Some are of thecharacterized differences between as constant tropical surface and temperate potential, zone while soils are attributed to the existence of positive electric charges and to the extreme ease with which temperate zone soilscations are are constant hydrolyzed charge in tropical [29,30,32]. soils. Thus, on the basis of electrochemical behavior, soils Figure 3 showsmay a be typical classified point according of zero to whethercharge theyfor the have soils a constant of “Cerrado surface chargeBiome” or constant(pH = 5.2). Above the surfacesurface potential.charge is In negative general, tropical and below soils are it, characterized the surface as charge constant becomes surface potential, positive [29]. while temperate zone soils are constant charge [29,30,32].

7.5 -7.5

5 -5 ) - -1

2.5 -2.5 ) OH ) -1 0 0 CaCl2 2M (cmol kg(cmol -2.5 2.5 0.2M + H -5 0.02M 5 kg (cmol Charge Surface 0.004M -7.5 7.5 44.555.566.577.588.59 pH

Figure 3. Point of zeroFigure charge 3. Point for ofsoils zero located charge forin soilsstudy located area. in study area.

Soils from the study area can be classified from an engineering point of view as residual soils, as they were developed in situ by the weathering of igneous, sedimentary, Sustainability 2021, 13, 7731 5 of 18

Figure3 shows a typical point of zero charge for the soils of “Cerrado Biome” (pH = 5.2). Above the surface charge is negative and below it, the surface charge becomes positive [29]. Soils from the study area can be classified from an engineering point of view as residual soils, as they were developed in situ by the weathering of igneous, sedimentary, and metamorphic rocks. The mineral grains of residual soils should remain in the same position or at least have little movement relative to the position in which there were in the parent material during the weathering process. Residual soils are more prevalent in the tropics than elsewhere. In the tropical environment, chemical and biochemical weathering of the underlying material have much more influence on the residual soil formation process than in temperate and cold climates, where physical weathering prevails. Thus, under the name residual soils, an engineer might expect a wide range of different types of soils derived from different parent materials (igneous, sedimentary, and metamorphic rocks) presenting different conditions of soil mineralogy, morphology, climate, and degrees of weathering. Depending on the region, residual soils also receive different names, such as lateritic soils, laterites, latosols, saprolitic soils, black cotton soils, tropical soils, etc. The situation is further complicated by the fact that various investigators have given to the terms laterite, latosol, and lateritic soils so many connotations that they have become almost meaningless [29,32]. Table1 presents the results of the granulometric analysis of the soil of Sites 1 and 2. The soil in the area was classified as a sandy soil, which is erosion prone, and erosion control measures needed to be put in place before the implementation of any additional engineering works.

Table 1. Results of the granulometric analysis.

Site Location Coarse Sand Fine Sand Silt Clay 61% 22% 15% 2% 1 São Gonçalo do Abaeté, MG Soil type: sandy soil, erosion prone 58% 23% 13% 6% 2 Córrego Danta, MG Soil type: sandy soil, erosion prone

2.4. Materials Drainage bioswales were installed in 20 experimental parcels with different types of revetments. Each bioswale had a trapezoidal cross section with a top width, bottom width and height of 1.5 m, 0.6 m and 0.5 m, respectively, and a length of 10 m. Eleven experimental parcels were installed in Site 1 (São Gonçalo do Abaeté), and nine in Site 2 (Córrego Dantas). Each experimental parcel was planted with seeds of the following species: Urochloa umidicola, Urochloa decumbens, Hypharrenia rufa, Crotalaria juncea, Melinis minutiflora, and Avena strigose. A total of six commercially available rolled erosion control products (RECPs) from three different manufacturers (Deflor [33], Maccaferri [34], and Tensar [35]) and one RECP that was specifically developed for this study—the preseeded reinforcement mat—were used as swale revetments in the experimental parcels: (a) Preseeded reinforcement mat (PSRM)—this revetment was developed specifically for this project. Its fabric is made with coconut fiber intertwined with two polypropylene networks by means of a longitudinal seam, and included seeds, organic adhesives, and a granulated chemical fertilizer. This geotextile, when moistened, promotes the germination and establishment of vegetation that will constitute the lining of the swale (Figure4). It has a weight/area ratio of 500 g/m 2. Its lifespan is 12 months. (b) High performance turf reinforcement mat (HPTRM) W3000 (Tensar)—this product is made with polypropylene monofilaments stabilized against UV radiation, forming a corrugated fabric in a three dimensional structure (Figure5). It has a weight/area ratio of 500 g/m2 and a lifespan of 15 years. Sustainability 2021, 13, 7731 6 of 18

(c) Erosion control rolled reinforcement mat, Fibrax 400 BF (Deflor)—this product’s fabric Sustainability 2021, 13, x FOR PEER REVIEW 6 of 16 Sustainability 2021, 13, x FOR PEER REVIEW is made with coconut fiber intertwined with 2 layers of polypropylene by means6 of 16 of longitudinal seams (Figure6). It has a weight/area ratio of 400 g/m 2 and a lifespan of 18 months. d)(d) ErosionErosion control control rolled rolled geomat, geomat, Macmat Macmat (Maccaferri)—this (Maccaferri)—this geotextile geotextile is is flexible flexible as as it d) Erosionhashas thick thick control polypropylene polypropylene rolled geomat, filamentsfilaments Macmat whichwhich (Maccaferri)—this areare fused fused and and spaced spaced geotextile throughout throughout is flexible the the contactas con- it hastactpoints thick points polypropylene of theof the fabric fabric (Figure filaments(Figure7). 7). It which consistsIt consists are of fused of about about and 90% 90% spaced voids. voids. throughout It It has has a a weight/area weight/area the con- tactratioratio points of of 500 500of theg/m g/m fabric2 and2 and (Figurea lifespan a lifespan 7). ofIt of consists10 10 years. years. of about 90% voids. It has a weight/area e)(e) ratioErosionErosion of 500 control controlg/m2 and rolled rolled a lifespan geomat, geomat, of MacMatR310 MacMatR3 years. (Maccaferri)—in (Maccaferri)—in addition addition to to what what is is de- de- e) Erosionscribedscribed control in in item item rolled d), d), this this geomat, product product MacMatR3 has has an an extra extra (Maccaferri)—in layer layer of of polyester polyester addition fabric fabric to(Figure (Figure what 8 8).).is ItItde- hashas scribedaa weight/area weight/area in item d), ratio ratio this of product of 500 500 g/m g/m has2 and2 anand a extra lifespan a lifespan layer of of of10 polyester 10years. years. fabric (Figure 8). It has f)(f) a weight/areaErosionErosion control control ratio rolled of 500 reinforcementreinforcement g/m2 and a lifespan mat,mat, MacMat MacMat of 10 years. R R (Maccaferri)—this (Maccaferri)—this RECP RECP geomat geo- f) Erosionmatconsists consists control of about of rolled about 90% reinforcement 90% voids voids and and is fabricatedmat, is fabricated MacMat with with R thick(Maccaferri)—this thick filaments filaments of polypropylene.of RECP polypropyl- geo- matene.In consists addition, In addition, of itabout has it ahas90% metallic a voids metallic reinforcement and reinforc is fabricatedement in the with in form the thick form of afilaments double of a double twisted of polypropyl- twisted hexagonal hex- ene.agonalmesh In addition, (Figuremesh (Figure9 it). has Its lifespan a9). metallic Its lifespan is 20reinforc years. is 20ement years. in the form of a double twisted hex- g)(g) agonalGeogridGeogrid mesh Triton Triton (Figure 15001500 9). BX BX Its (Tensar)—this (Tensar)—this lifespan is 20 geogrid years.geogrid is madeis made with with monofilaments monofilaments of polypropy- of poly- g) Geogridpropylenelene stabilized Triton stabilized 1500 against BX against UV(Tensar)—this radiation. UV radiation. It geogrid has a weight/areaIt hasis made a weight/area with ratio monofilaments of 500 ratio g/m of2 500and of g/m a poly- lifespan2 and propyleneaof lifespan 15 years. stabilized of 15 years. against UV radiation. It has a weight/area ratio of 500 g/m2 and a lifespan of 15 years.

(a) (b) (a) (b) FigureFigure 4. 4. SwaleSwale with with preseeded preseeded mat—Site mat—Site 1: 1: ( (aa)) immediately immediately after after implementation implementation and and ( (bb)) 90 90 days days after after implementation. implementation. Figure 4. Swale with preseeded mat—Site 1: (a) immediately after implementation and (b) 90 days after implementation.

(a) (b) (a) (b) Figure 5. Swale with Tensar reinforcement mat W3000 revetment—Site 1: (a) immediately after implementation and (b) Figure90Figure days 5. Swale 5.afterSwale implementation. with with Tensar Tensar reinforcement reinforcement mat matW3000 W3000 revetment—Site revetment—Site 1: (a) 1: immediately (a) immediately after implementation after implementation and (b and) 90 (daysb) 90 after days implementation. after implementation. Sustainability 2021, 13, 7731 7 of 18 Sustainability 2021, 13, x FOR PEER REVIEW 7 of 16

Sustainability 2021, 13, x FOR PEER REVIEW 7 of 16

(a) (b) (a) (b) FigureFigure 6. 6.SwaleSwale with with Deflor Deﬂor mat FibraxFibrax 400 400 BF BF revetment—Site revetment—Site 1: ( a1:) immediately(a) immediately after implementationafter implementation and (b) and 90 days (b) 90 days Figure 6. Swale with Deflor mat Fibrax 400 BF revetment—Site 1: (a) immediately after implementation and (b) 90 days afterafter implementation. implementation. after implementation.

(a) (b)

FigureFigure 7. 7. SwaleSwale with with geomat geomat Macmat—Site Macmat—Site(a) 1: 1: ( (aa)) immediately immediately after after implementation implementation and and ( (bb)) 90 90 days days(b after) after implementation. implementation. Figure 7. Swale with geomat Macmat—Site 1: (a) immediately after implementation and (b) 90 days after implementation. SustainabilitySustainability 20212021, 13,,13 x ,FOR 7731 PEER REVIEW 8 of8 18 of 16 Sustainability 2021, 13, x FOR PEER REVIEW 8 of 16

(a) (b) Figure 8. Swale with geomat Macmat R3—Site 2: (a) immediately after implementation and (b) 90 days after implementa- FigureFigure 8. Swale 8. Swale with with geomat geomat Macmat Macmat R3—Site 2:2: ((aa)) immediately immediately after after implementation implementation and and (b) 90 (b days) 90 days after implementation.after implementation. tion.

(a) (b) Figure 9. Swale with Maccaferri geomat MacMat R revetment—Site 2: (a) immediately after implementation and (b) 90 FigureFigure 9. 9.SwaleSwale with with Maccaferri Maccaferri geomat MacMatMacMat R R revetment—Site revetment—Site 2: ( a2:) immediately(a) immediately after implementationafter implementation and (b) and90 days (b) 90 days after implementation. daysafter after implementation. implementation. In addition to the individual performances of the ﬁrst six RCEPs listed above (a–f), the In addition to the individual performances of the first six RCEPs listed above (a–f), thecombined combined performances performances of the of Geogridthe Geogrid Triton Triton 1500BX 1500BX plus theplus preseeded the preseeded reinforcement reinforce- themat combined (Figure 10performances), and the Geogrid of the Geogrid Triton 1500BX Triton plus 1500BX the Deﬂor’splus the Matpreseeded Fibrax reinforce- 400 BF ment mat (Figure 10), and the Geogrid Triton 1500BX plus the Deflor’s Mat Fibrax 400 BF (Figure 11) were also evaluated. In these two cases, the Geogrid Triton was added as a turf reinforcement to increase the revetment’s shear strength. All RCEPs were anchored Sustainability 2021, 13, 7731 9 of 18

Sustainability 2021, 13, x FOR PEER REVIEW 9 of 16 Sustainability 2021, 13, x FOR PEER REVIEW 9 of 16 (Figure 11) were also evaluated. In these two cases, the Geogrid Triton was added as a turf reinforcement to increase the revetment’s shear strength. All RCEPs were anchored to the swaleto the lining swale by lining using by organic using organic adhesives adhesive and steels and clamps, steel clamps, CA-60, inCA-60, an inverted in an inverted “U” shape “U” withtoshape the dimensions swale with liningdimensions of by 4.2 using mm of 4.2organic× 200 mm mm ×adhesive 200 (diameter mms (diameterand× steellength). clamps, × length). CA-60, in an inverted “U” shape with dimensions of 4.2 mm × 200 mm (diameter × length).

(a) (b) (a) (b) FigureFigure 10. 10. Swale with with preseeded preseeded mat mat and and Tensar Tensar Geog Geogridrid Triton Triton 1500 1500 BX BXrevetment—Site revetment—Site 1: (a 1:) immediately (a) immediately after imple- after Figurementation 10. Swale and ( bwith) 90 preseededdays after implementation.mat and Tensar Geog rid Triton 1500 BX revetment—Site 1: (a) immediately after imple- implementation and (b) 90 days after implementation. mentation and (b) 90 days after implementation.

(a) (b) (a) (b) Figure 11. Swale with Deflor Mat Fibrax 400 BF + Tensar Geogrid Triton 1500 BX revetment—Site 2: (a) immediately after Figureimplementation 11. Swale withand ( DeﬂorDeflorb) 90 days Mat Fibraxafter implementation. 400 BF + Tensar Geogrid Triton 1500 BX revetment—Site 2: (a) immediately after implementation and (b) 90 days after implementation. implementation and (b) 90 days after implementation. 2.5. Methods 2.5. Methods To evaluate the performance of the experimental parcels, a performance index is pro- posedTo basedevaluate on the five performance performance of indicators: the experime (1)ntal the parcels,number a of performance plant species index that is devel- pro- posedoped basedin the swale;on five (2) performance the dry biomass indicators: of the (1vegetation) the number developed of plant on species each swale that bed;devel- (3) oped in the swale; (2) the dry biomass of the vegetation developed on each swale bed; (3) Sustainability 2021, 13, 7731 10 of 18

2.5. Methods To evaluate the performance of the experimental parcels, a performance index is proposed based on five performance indicators: (1) the number of plant species that developed in the swale; (2) the dry biomass of the vegetation developed on each swale bed; (3) the structural integrity of each swale; (4) the vegetative cover level of the swale; and (5) the occurrence of erosive processes in the swale surfaces, around them, and immediately downstream of these drainage systems. For each of these performance indicators, a weight varying from 0 to 2 was assigned, and then the performance index of each one of the bioswale revetments investigated was defined as the sum of all performance indicator weights. Table2, below, describes each parameter in detail, where the higher the weight (indicator of performance), the better the overall performance. For cost effectiveness analysis, the implementation cost of each of the parcels was compared to the implementation cost of a concrete lined swale, which is defined as the cost ratio (implementation unit area cost of swales with RECPs, divided by implementation cost of a concrete lined swale).

Table 2. Performance indicators.

Performance Indicators Condition Weight

• Up to 3 species of vegetation 0 Amount of vegetation species—A • Between 4 and 5 species of vegetation 1 • Above 5 species of vegetation 2

• Up to 300 g/m2 0 Dry biomass of the vegetation that was • 2 1 developed inside the swale—B From 300 to 500 g/m • Above 500 g/m2 2

• Less than 30% 0 Vegetative cover of the swale—C • From 30 to 70% 1 • Above 70% 2

• Damages in more than 40% of the parcel’s length 0 Structural integrity of revetment—D • Damages in 10 to 40% of the parcel’s length 1 • Damages in 10% or less of the parcel’s length 2

• Erosional Process: occurrence of erosional Occurrence of advanced erosional processes 0 • processes on the swale, laterals and Initial phase of soil mobilization due to erosional 1 downstream—E processes 2 • Absence of erosional processes

Performance Index = A + B + C + D + E

Vegetation sampling for the biomass was obtained by randomly selecting from areas of about 40 x 60 cm2, oriented toward the flow direction of the swale. From these randomly selected areas, all exposed (above ground) vegetation was sampled for biomass calculations. Immediately after sampling, the aerial biomass was labeled, weighed and bagged. After the drying process at the laboratory, the biomass was weighed again to estimate the biomass of each parcel according to procedures defined by Quincozes et al. [36]. The identification of the vegetation species present on each experimental parcel was carried out over the entire length of the swale (10 m), according to the databank from the International Plant Name Index [37], Missouri Botanical Garden [38], and Brazilian Flora [39], and it adopted the circumscription of the botanical families as proposed by the angiosperm phylogeny group classification—APG IV [40]. The performance evaluation was completed 48 months after the installation, to evaluate long term integrity and vegetation growth in the bioswale. Sustainability 2021, 13, x FOR PEER REVIEW 11 of 16

Sustainability 2021, 13, 7731 11 of 18

parcels 5 (preseeded mat + Geogrid Triton 1500 BX), 9 (Deflor Mat Fibrax 400 BF), 13 (De- flor Mat Fibrax 400 BF + Tensar Triton 1500 BX), and 14 (Geomat Macmat), with at least 3. Results 10 different species, which might have been brought to the swales by wind and/or ani- Table3, below, shows the performance of each revetment of the swales in Sites 1 and 2, mals, and were not48 part months of afterthe initial implementation. experiment. The timeframeIt is important for performance to note assessmentthat from wasthe selected originally planted tospecies, allow forthe the species bioswales Urochloa to be exposedumidicola to fourand dryAvena and strigose four wet were seasons, not basedpre- on the sent after 48 monthsauthors’ of implantation, 20 years of experiencewhile Melinis working minutiflora with the was bioengineering present in 16 of soils.out of The the greatest 20 parcels. The othervegetation three biodiversityspecies, Urochloa developed decubens, within theHypharrenia swales was rufa observed and forCroatalaria the experimental juncea, were presentparcels in 8 5out (preseeded of the 20 mat parcels. + Geogrid Similarly Triton 1500 notably, BX), 9 (Deflor there Matwere Fibrax four 400 species BF), 13 (Deflor (Setaria sp, EchinolaenaMat Fibraxinflexa, 400 Megathyrsu BF + Tensars Tritonmaximus, 1500 BX),and andStylosanthes 14 (Geomat macrocephala) Macmat), with that at least 10 different species, which might have been brought to the swales by wind and/or animals, were not planted butand were were notpresent part ofin the at initialleast 25 experiment.% of the parcels. It is important Finally, to note it seems that from that the the originally organic decay of theplanted geomat species, materials the species mayUrochloa have also umidicola helpedand theAvena growth strigose of thesewere notspecies present after inside the bioswales,48 monthssince theof implantation,soils are acidic while andMelinis have minutiflora low fertility.was present in 16 out of the 20 parcels. In terms of overallThe other performance, three species, parcelsUrochloa 2 (preseeded decubens, Hypharrenia mat), 10 (Deflor rufa and FibraxCroatalaria 400 BF), juncea , were 13 (Deflor Mat Fibraxpresent 400 inBF 8 + out Tensar of the Triton 20 parcels. 1500 BX), Similarly 14 (Geomat notably, Macmat), there were 17 four (Geomat species (Setaria sp., Echinolaena inflexa, Megathyrsus maximus, and Stylosanthes macrocephala) that were not MacMat R3), 18 (Geomat MacMat R3), 19 (Geomat MacMat R) and 20 (Geomat MacMat planted but were present in at least 25% of the parcels. Finally, it seems that the organic R) had the highest performancedecay of the geomat index, materials with values may have of 9, also 9, 10, helped 10, the9, 9, growth 9 and of9, theserespectively. species inside the Parcels 2 (preseededbioswales, mat) and since 10 the(Deflor soils are Fibr acidicax 400 and BF) have also low had fertility. the lowest implementation cost (materials andIn labor), terms of with overall a cost performance, ratio (defined parcels as 2 (preseeded the ratio mat),of RECP 10 (Deflor cost Fibraxto the 400 BF), concrete revetment13 cost (Deflor per Mat unit Fibrax area) 400 of BF0.19 + Tensarand 0.22, Triton respectively, 1500 BX), 14 (Geomatas shown Macmat), in Figure 17 (Geomat 12. However, the damageMacMat R3),level 18 was (Geomat higher MacMat in these R3), two 19 (Geomat parcels; MacMat therefore, R) and long-term 20 (Geomat per- MacMat R) had the highest performance index, with values of 9, 9, 10, 10, 9, 9, 9 and 9, respectively. formance may be compromised.Parcels 2 (preseeded mat) and 10 (Deflor Fibrax 400 BF) also had the lowest implementation Figure 13 showscost the (materials correlations and labor), between with a costthe ratiotype (defined of vegetative as the ratio cover, of RECP structural cost to thein- concrete tegrity and dry biomassrevetment of the cost 20 per experimental unit area) of 0.19 parcels and 0.22, located respectively, in Study as shown Sites in 1 Figureand 2. 12 As. However, expected, the structuralthe damage integr levelity (measured was higher in as these percent two parcels;damage) therefore, of swales long-term was inversely performance may proportional to thebe vegetative compromised. cover percentage, with a Pearson correlation factor of −0.51.

10 1.2 9 1 8 7 0.8 6 5 0.6 4 0.4 3 2 Green to Gray Cost Ratio Performance Index [0–10] Index Performance 0.2 1 0 0 1234567891011121314151617181920 Parcel ID

Performance Index Cost Ratio

FigureFigure 12. 12.Comparison Comparison of performance of performance index and index green and to gray green cost to ratio gray for cost all experimental ratio for all parcels experimental in Sites 1 andpar- 2. cels in Sites 1 and 2. Sustainability 2021, 13, 7731 12 of 18

Table 3. Performance results after 48 months.

Swale Dry Biomass Vegetative Cover Structural Integrity Observed Erosional Performance Parcel ID Revetment/ Vegetation Species/Count Cost Ratio (g/cm2) (%) (as %damage) Processes Index Site Pennisetum glaucum, Setaria sp., Hancornia speciosa, Melinis Pre-seeded mat/ minutiflora, Echinolaena inflexa, 1 Site 1 0.55 40 30 Absent 8 0.19 Urochloa decumbens, Melinis repens, (Figure3) Calliandra cf. dysantha, Andira humilis/9 species Urochloa decumbens, Melinis Pre-seeded mat/ minutiflora, Megathyrsus maximus, 2 0.81 90 10 Absent 9 0.19 Site 2 Mellinis retusa/ 4 species Melinis minutiflora, Urochloa sp., Pre-seeded mat/ 3 Crotalaria sp., S. 0.31 40 30 Absent 7 0.19 Site 2 adistringens/4 species Urochloa decumbens, Melinis Pre-seeded mat + Geogrid minutiflora, Megathyrsus maximus, 4 Triton 1500 BX 0.71 70 20 Absent 8 0.50 Crotalaria retusa/4 Site 1 Species Melinis minutiflora, Microlicia sp., Trachypogon spicatus, Stylosanthes Pre-seeded mat + Geogrid macrocephala, Myrcia variabilis, 5 Triton 1500 BX Erythroxylum sp., Bauhinia sp., 0.45 45 25 Absent 8 0.50 Site 1 Pennisetum Setaria sp., Euphobia cf. Hyssopifolia, Diplusodon sp., Anemopaegma arvense/11 species Melinis minutiflora, Eugenia Tensar HPTRM W3000/ 6 dysenterica, Setaria sp., 0.1 20 0 Absent 7 1.06 Site 1 (Figure4) Megathyrsus maximus/4 species Melinis minutiflora, Eugenia Tensar HPTRM W3000/ dysenterica, Setaria sp., 7 0.1 25 0 Absent 7 1.06 Site 2 Megathyrsus maximus/ 4 species Sustainability 2021, 13, 7731 13 of 18

Table 3. Cont.

Swale Dry Biomass Vegetative Cover Structural Integrity Observed Erosional Performance Parcel ID Revetment/ Vegetation Species/Count Cost Ratio (g/cm2) (%) (as %damage) Processes Index Site Stylosanthes macrocephala, Melinis Deflor mat Fibrax 400 BF/ minutiflora, Chamaecrista cathartica, 8 Site 1 0.21 25 40 Absent 7 0.22 Calopogonium mucunoides, Clitoria (Figure5) guianensis, Setaria sp/6 species Melinis minutiflora, Setaria sp., Borreria capitata, Stylosanthes macrocephala, Echinolaena inflexa, Urochloa decumbens, Melinis repens, Deflor mat Fibrax 400 BF/ 9 Calliandra cf. dysantha, Andira 0.35 50 45 Absent 7 0.22 Site 1 humilis, Davilla elliptica, Chamaecrista cathartica, Trachypogon spicatus, Oxalis sp., Antonia ovata/14 species Melinis minutiflora, Echinolaena inflexa, Davilla elliptica, Calliandra cf. dysantha, Urochloa decumbens, Myrcia variabilis, Pennisetum Deflor mat Fibrax 400 BF/ glaucum, Euphobia cf. hyssopifolia, 10 0.44 70 40 Absent 9 0.22 Site 1 Oxalis sp., Microlicia sp., Chamaecrista cathartica, Calolisianthus speciosus, Clitoria guianensis, Trachypogon spicatus/ 14 species Melinis minutiflora, Echinolaena Deflor mat Fibrax 400 BF/ 11 inflexa, Baccharis sp. 0.67 50 40 Absent 6 0.22 Site 2 /3 species Deflor mat Fibrax 400 BF + Melinis minutiflora, Urochloa 12 Tensar Triton 1500 BX/ decumbens, Echinolaena 0.68 100 0 Absent 8 0.52 Site 2 inflexa/3 species Sustainability 2021, 13, 7731 14 of 18

Table 3. Cont.

Swale Dry Biomass Vegetative Cover Structural Integrity Observed Erosional Performance Parcel ID Revetment/ Vegetation Species/Count Cost Ratio (g/cm2) (%) (as %damage) Processes Index Site Urochloa decumbens, Melinis minutiflora, Oxalis sp., Crotalaria retusa, Stylosanthes macrocephala, Deflor mat Fibrax 400 BF + Calopogonium mucunoides, 13 Tensar Triton 1500 BX/ 0.67 80 10 Absent 10 0.52 Pennisetum glaucum, Microlicia sp., Site 2 Calliandra cf. dysantha, Andira humilis, Macroptilium lathyroides, Borreria capitata/12 species Urochloa decumbens, Pennisetum glaucum, Megathyrsus maximus, Andira humilis, Oxalis sp., Chamaecrista cathartica, Melinis Geomat Macmat/ 14 minutiflora, Echinolaena inflexa, 0.43 70 0 Absent 10 0.34 Site 1 Calopogonium mucunoides, Stylosanthes macrocephala, Neea theifera, Melinis repens, Machaerium villosum/13 species Geomat Macmat/ M. minutiflora, M. albicans, M. 15 0.89 90 0 Absent 8 0.34 Site 2 maximus/3 species Geomat Macmat/ Urochloa sp., Melinis minutiflora, 16 0.64 100 0 Absent 8 0.34 Site 2 Crotalaria sp./3 species M. minutiflora, Urochloa sp., 17 Geomat MacMat R3/Site 2 Crotalaria sp., Baccharis sp./ 0.92 90 0 Absent 9 0.40 4 species Melinis minutiflora, Urochloa sp., 18 Geomat MacMat R3/Site 1 E. inflexa, Crotalaria sp./ 0.71 85 0 Absent 9 0.40 4 species M. minutiflora, E. inflexa, Bauhinia 19 Geomat MacMat R/Site 2 sp., Caryocar brasiliens, A. 0.57 85 0 Absent 9 0.96 fastigiatum/5 species Melinis minutiflora, Urochloa sp., 20 Geomat MacMat R/Site 2 E. inflexa, Crotalaria sp./ 0.83 95 0 Absent 9 0.96 4 species Sustainability 2021, 13, 7731 15 of 18

Figure 13 shows the correlations between the type of vegetative cover, structural integrity and dry biomass of the 20 experimental parcels located in Study Sites 1 and 2. As Sustainability 2021, 13, x FOR PEER REVIEW 12 of 16 expected, the structural integrity (measured as percent damage) of swales was inversely proportional to the vegetative cover percentage, with a Pearson correlation factor of −0.51.

100 1 )

80 0.8 2

60 0.6

40 0.4 Percentage, %

20 0.2 Dry Biomass (g/m

0 0 1234567891011121314151617181920 Parcel ID

Vegetative Cover (%) Structural Integrity (% of damages) Dry Biomass

FigureFigure 13. Correlation 13. Correlation between between vegetative vegetative cover, structural cover, struct integrityural (measured integrity as(measured % damage) as and % damage) dry biomass and for the treatmentsdry biomass of all experimental for the treatments parcels in Sitesof all 1 ex andperimental 2. parcels in Sites 1 and 2.

Table 3. Performance4. Discussion results and Conclusions after 48 months. The performances of the seven RECPs used as erosion control revetments for drainage swales were evaluated using five performance indicators: (1) biodiversity; (2) dry biomass of the vegetation; (3) structural integrity of swale revetment; (4) vegetative cover; and (5) the occurrence of erosive processes. Structur In terms of biodiversity, the swale revetments with polypropylenePerfo materials presented a low performance in termsDry of biodiversityVegetati insideal the swales, in particular those treated to Parce Swale Observed rman protect against UV radiation.Bioma On theve otherIntegrit hand, it was observed that parcelsCost 5 (preseeded l ID Revetment/ Vegetation Species/Count Erosional ce mat + Geogrid Triton 1500ss BX),Cover 9 (Fibrax mat),y 13 (Fibrax mat + GeogridRatio Triton), and Site 14 (Macmat mat) presented the highest number of species.Processes Out Inde of the originally planted (g/cm2) (%) (as %da species, the species Urochloa umidicola and Avena strigose were not presentx after 48 months of implantation, while Melinis minutiflora wasmage) present in 16 out of the 20 parcels. The other three species, Urochloa decubens, Hypharrenia rufa and Croatalaria juncea, were present in 8 out of the 20 parcels. Similarly notably, there were four species (Setaria sp., Echinolaena inflexa, Megathyrsus maximus, and Stylosanthes macrocephala) that were not planted but were present in at least 25% of the parcels. In general, RECPs with a high opening in the mat had Pennisetum glaucum, Setaria sp, Hancornia Pre-seeded better performances in terms of biomass and vegetative growth, likely because openings speciosa, Melinis minutiflora, Echinolaena mat/ made it easier for vegetation to grow. 1 inflexa, Urochloa decumbens, Melinis repens, 0.55 40 30 Absent 8 0.19 Site 1 Parcels 13 (Deflor Mat Fibrax 400 BF + Tensar Triton 1500 BX) and 14 (Geomat Macmat) Calliandra cf. dysantha,had Andira the highesthumilis/9 performance index, with minimal to no damage observed, and with an (Figure 3) speciesimplementation cost ratio of 0.32 and 0.54, respectively. Pre-seeded Urochloa decumbens, MelinisWhen minutiflora, both performance index and cost ratio are considered, Parcels 2 (preseeded mat) and 10 (Deflor Fibrax 400 BF) had the best cost benefit, which means a lower cost ratio and 2 mat/ Megathyrsus maximus,higher Mellinis performance retusa/ index.0.81 The innovative90 preseeded 10 Absent mat developed 9 during0.19 this research Site 2 4 speciespresented great potential for utilization in the revetment of swales, with the advantage that Pre-seeded it is already preseeded by the manufacturer. In general, RECPs were less expensive than Melinis minutiflora, Urochloa sp., Crotalaria 3 mat/ concrete lining, while providing0.31 the40 inherent 30 environmental Absent benefits7 of0.19 bioswales, such sp., S. adistringens/4 species Site 2 Pre-seeded mat Urochloa decumbens, Melinis minutiflora, + Geogrid 4 Megathyrsus maximus, Crotalaria retusa/4 0.71 70 20 Absent 8 0.50 Triton 1500 BX Species Site 1 Pre-seeded mat Melinis minutiflora, Microlicia sp., + Geogrid Trachypogon spicatus, Stylosanthes 5 0.45 45 25 Absent 8 0.50 Triton 1500 BX macrocephala, Myrcia variabilis, Erythroxylum Site 1 sp., Bauhinia sp., Pennisetum Setaria sp, Sustainability 2021, 13, 7731 16 of 18

as increased infiltration, reduced stormwater runoff flow, increased carbon sequestration, and increased groundwater recharge. Additionally, transport and installation of RECPs is usually simpler when compared to concrete, especially in areas with difficult access such as the remote sites where transmission line towers are usually installed. Furthermore, the preseeded reinforcement mat presented some advantages over conventional RECPs, since it requires fewer materials during installation and a simpler implementation procedure, since the seeds and fertilizers needed for the growth and development of vegetation are included as part of the mat matrix. In addition, the adhesive layer used to fix the seeds to the mat creates a semi-impermeable layer that protects the structural integrity of the bioswale bed, allowing for the initial development and growth of the vegetation. In conclusion, results have indicated that the implementation of bioswales was a successful and economical solution as a stormwater management system, and RECPs demonstrated great potential for use as a swale revetment under tropical conditions. The proper choice of seeds and type of RECPs can provide organic matter for the top soil horizon, consequently increasing its soil water capacity retention (which is very important in the long dry season of this study area), improve soil fertility and diminish the inherent soil acidity by increasing its pH, thus providing the desirable conditions for vegetation growth not only of the planted seeds, but also the ones transported by wind, humans and animals, as evidenced in this project.

Author Contributions: Conceptualization, A.T.C. and J.F.T.C.; methodology, A.T.C., T.C.d.B.G., G.B.M.; formal analysis, G.B.M., T.C.d.B.G., A.T.C.; investigation, A.T.C., T.C.d.B.G.; writing—original draft preparation, A.T.C., T.C.d.B.G., G.B.M.; writing—review and editing, T.C.d.B.G., G.B.M.; super- vision, A.T.C. and J.F.T.C.; project administration, A.T.C. and J.F.T.C.; funding acquisition, A.T.C. and J.F.T.C. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Agência Nacional de Energia Elétrica (ANEEL; GT-0581) and performed in collaboration with CEMIG GT—Geração e Trasmissão. Institutional Review Board Statement: Not Applicable. Informed Consent Statement: Not Applicable. Acknowledgments: The authors would like to thank the CEMIG GT—Geração e Transmissão for being a proponent of the research project; ANEEL for providing the financial support, and NSF CREST Center for Energy and Sustainability (NSF #1547723) at California State University Los Angeles; and all of the CEMIG GT community for the support provided to the research activities reported in this paper. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funding agencies. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References 1. Pennino, M.J.; McDonald, R.I.; Jaffé, P. Watershed-scale impacts of stormwater green infrastructure on hydrology, nutrient fluxes, and combined sewer overflows in the mid-Atlantic region. Sci. Total Environ. 2016, 565, 1044–1053. [CrossRef] 2. Harbor, J. Engineering geomorphology at the cutting edge of land disturbance: Erosion and sediment control on con-struction sites. Geomorphology 1999, 31, 247–263. [CrossRef] 3. Wise, S. Green infrastructure rising. Planning 2008, 74, 14–19. 4. Tota-Maharaj, K.; Scholz, M. Efficiency of permeable pavement systems for the removal of urban runoff pollutants under varying environmental conditions. Environ. Prog. Sustain. Energy 2010, 29, 358–369. [CrossRef] 5. Casal-Campos, A.; Fu, G.; Butler, D.; Moore, A. An integrated environmental assessment of green and gray infra-structure strategies for robust decision making. Environ. Sci. Technol. 2015, 49, 8307–8314. [CrossRef][PubMed] 6. Cohen, J.P.; Field, R.; Tafuri, A.N.; Ports, M.A. Cost comparison of conventional gray combined sewer overflow control infrastructure versus a green/gray combination. J. Irrig. Drain. Eng. 2012, 138, 534–540. [CrossRef] Sustainability 2021, 13, 7731 17 of 18

7. USEPA. What Is Green Infrastructure? US Environmental Protection Agency 2017. Available online: https://www.epa.gov/ green-infrastructure/what-green-infrastructure (accessed on 1 August 2017). 8. Bloorchian, A.A.; Ahiablame, L.; Zhou, J.; Osouli, A. Performance Evaluation of Combined Linear BMPs for Reducing Runoff from Highways in an Urban Area. In World Environmental and Water Resources Congress 2016; American Society of Civil Engineers (ASCE): Reston, VA, USA, 2016; pp. 20–29. 9. Bertrand-Krajewski, J.-L.; Chebbo, G.; Saget, A. Distribution of pollutant mass vs volume in stormwater discharges and the first flush phenomenon. Water Res. 1998, 32, 2341–2356. [CrossRef] 10. Lee, H.; Lau, S.-L.; Kayhanian, M.; Stenstrom, M. Seasonal first flush phenomenon of urban stormwater discharges. Water Res. 2004, 38, 4153–4163. [CrossRef] 11. Taebi, A.; Droste, R.L. First flush pollution load of urban stormwater runoff. J. Environ. Eng. Sci. 2004, 3, 301–309. [CrossRef] 12. Anderson, B.S.; Phillips, B.M.; Voorhees, J.P.; Siegler, K.; Tjeerdema, R. Bioswales reduce contaminants associated with toxicity in urban storm water. Environ. Toxicol. Chem. 2016, 35, 3124–3134. [CrossRef][PubMed] 13. Xiao, Q.; McPherson, E.G.; Zhang, Q.; Ge, X.; Dahlgren, R. Performance of two bioswales on urban runoff man-agement. Infrastructures 2017, 2, 12. [CrossRef] 14. Shetty, N.H.; Hu, R.; Mailloux, B.J.; Hsueh, D.Y.; McGillis, W.R.; Wang, M.; Chandran, K.; Culligan, P.J. Studying the effect of bioswales on nutrient pollution in urban combined sewer systems. Sci. Total Environ. 2019, 665, 944–958. [CrossRef] 15. Yang, H.; Spoonamore, S. Stormwater runoff treatment using Bioswales augmented with advanced nanoengineered materials. In Aquananotechnology: Global Prospects; CRC Press: Boca Raton, FL, USA, 2012; pp. 661–680. 16. Lucas, S.; Clar, M.L.; Gracie, J. A Green Street Retrofit in a Chesapeake Bay Community Using Bioswales. In Low Impact Development Technology: Implementation and Economics; American Society of Civil Engineers (ASCE): Reston, VA, USA, 2015; pp. 81–90. [CrossRef] 17. Vogel, J.R.; Moore, T.L.; Coffman, R.R.; Rodie, S.N.; Hutchinson, S.; McDonough, K.; McLemore, A.J.; McMaine, J.T. Critical Review of Technical Questions Facing Low Impact Development and Green Infrastructure: A Perspective from the Great Plains. Water Environ. Res. 2015, 87, 849–862. [CrossRef] 18. Kavehei, E.; Jenkins, G.A.; Adame, M.F.; Lemckert, C. Carbon sequestration potential for mitigating the carbon footprint of green stormwater infrastructure. Renew. Sustain. Energy Rev. 2018, 94, 1179–1191. [CrossRef] 19. Nickel, D.; Schoenfelder, W.; Medearis, D.; Dolowitz, D.; Keeley, M.; Shuster, W. German experience in managing stormwater with green infrastructure. J. Environ. Plan. Manag. 2013, 57, 403–423. [CrossRef] 20. Kirby, J.T.; Durrans, S.R.; Pitt, R.; Johnson, P.D. Hydraulic Resistance in Grass Swales Designed for Small Flow Conveyance. J. Hydraul. Eng. 2005, 131, 65–68. [CrossRef] 21. Sartor, J.; Mobilia, M.; Longobardi, A. Results and findings from 15 years of sustainable urban storm water man-agement. Int. J. Saf. Secur. Eng. 2018, 8, 505–514. 22. McFarland, A.R.; Larsen, L.; Yeshitela, K.; Engida, A.N.; Love, N.G. Guide for using green infrastructure in urban environments for stormwater management. Environ. Sci. Water Res. Technol. 2019, 5, 643–659. [CrossRef] 23. Mazer, G.; Booth, D.; Ewing, K. Limitations to vegetation establishment and growth in biofiltration swales. Ecol. Eng. 2001, 17, 429–443. [CrossRef] 24. McCollough, S.; Ring, T. Electrical Transmission Lines in Montana: Mitigation of Impacts to Soil and Biological Sources. In Proceedings of the Billings Symposium on Disturbed Land Rehabilitation, Billings, MT, USA, 25–30 March 1990; Available online: https://www.osti.gov/biblio/56925 (accessed on 1 January 2021). 25. Iqbal, M.F.; Khan, I.A. Spatiotemporal Land Use Land Cover change analysis and erosion risk mapping of Azad Jammu and Kashmir, Pakistan. Egypt. J. Remote Sens. Space Sci. 2014, 17, 209–229. [CrossRef] 26. Köppen, W.; Geiger, R. Klimate der Erde; Verlag justus perthes: Gotha, Germany, 1928; pp. 91–102. 27. Júnior, A.D.S.; Carvalho, L.G.; Da Silva, F.F.; Alves, M.D.C. Application of the Köppen classification for climatic zoning in the state of Minas Gerais, Brazil. Theor. Appl. Clim. 2011, 108, 1–7. [CrossRef] 28. Climate Data. 2020. Available online: http://climate-data.org (accessed on 15 July 2020). 29. De Brito Galvão, T.C. Mineralogical and Geotecnical Properties of a Latosol from Minas Gerais, Brazil. Ph.D. Dissertation, Purdue University, West Lafayette, Indiana, (Diss. Abstr. 94-20816), 1993. 30. Galvão, T.C.D.B.; Schulze, D.G. Mineralogical Properties of a Collapsible Lateritic Soil from Minas Gerais, Brazil. Soil Sci. Soc. Am. J. 1996, 60, 1969–1978. [CrossRef] 31. Munsell Color (Firm). Munsell Soil Color Charts: with Genuine Munsell Color Chips; Munsell Color: Grand Rapids, MI, USA, 2010. 32. De Brito Galvão, T.C.; Drnevich, V.P.; Schulze, D.C. Chemical, Mineralogical, and compressibility characteristics of a collapsible lateritic soil from Minas Gerais, Brazil. In Proceedings of the Unsaturated soils/sols non saturés—Proceedings of the 1st International Conference on Unsaturated Soils, Unsat 95, Balkema, Rotterdam, 6–8 September 1995; Alonso, E.E., Delage, P., Eds.; pp. 39–44. 33. Deflor, Biomantas Antierosivas. Available online: http://deflor.com.br/biomanta-fibrax/ (accessed on 14 April 2021). 34. Maccaferri, Geomantas–MacMat. 2021. Available online: https://www.maccaferri.com/br/produtos/geomantas-macmat (accessed on 14 April 2021). 35. Tensar: Erosion Control Products Brochure. 2012. Available online: https://pdf.archiexpo.com/pdf/tensar-international/tensar- erosion-control-brochure/63203-234206.html (accessed on 14 April 2021). Sustainability 2021, 13, 7731 18 of 18

36. da Quincozes, E.R.F.; Vargas, L.M.; Scivittaro, W.B.; De Mello, J.M.; Bandeira, D.L. Embrapa (in Portuguese)–Foliar Analysis System–User Manual. 2007. Available online: https://www.embrapa.br/busca-de-publicacoes/-/publicacao/745893/sistema- de-analise-foliar-manual-do-usuarioEMBRAPA (accessed on 17 June 2019). 37. Garden, M.B. Tropicos.org. 2011. Available online: Http://www.Tropicos.Org (accessed on 20 August 2019). 38. MBGPF. Missouri Botanical Garden. 2019. Available online: http://www.missouribotanicalgarden.org/plantfinder/ plantfindersearch.aspx (accessed on 1 December 2019). 39. Flora do Brasil. Rio de Janeiro Botanic Garden (in Portuguese). 2020. Available online: http://floradobrasil.jbrj.gov.br/ (accessed on 17 June 2020). 40. Chase, M.W.; Christenhusz, M.; Fay, M.; Byng, J.; Judd, W.S.; Soltis, D.; Mabberley, A.N.; Sennikov, P.S.; Stevens, P.F. An update of the angiosperm phylogeny group classification for the orders and families of flowering plants: APG IV. Bot. J. Linn. Soc. 2016, 181, 1–20.