Introduction

The Mediterranean region shows varying fluctuations not only in temperature but also in rainfall patterns, due to its particular climate conditions characterized by a distinct mild and wet period followed by a warm and dry period. Additionally, this region is influenced by a sequence of regular and often extreme flooding and drying periods (Bonada & Resh, 2013; Gasith & Resh, 1999).

As a consequence, freshwater habitats are among the most endangered ecosystems by physicochemical characteristics of water, anthropogenic pressures, and hydrological factors which would change the common patterns of habitat diversity and connectivity by altering habitat dynamics and giving rise to new conditions that would affect the phenology of species, the distribution range, the composition and dynamics of communities and their adaptation capacity (Alba-Tercedor et al., 2017; Dudgeon et al., 2006). Thus, aquatic species are expected to respond differentially to climate change and environmental characteristics that affect their structure and function depending on their sensitivity according to each bioclimatic zone (Kalogianni et al., 2017; Keçi et al., 2012; Lytle & Poff, 2004; Mabrouki et al., 2019).

Benthic macroinvertebrates are important components of freshwater communities. Therefore, any change in water bodies can strongly influence their spatial and temporal distribution making them good indicators of water quality and a useful tool for hydrobiological analysis (Romero et al., 2013; Selvanayagam & Abril, 2015; Silviera et al., 2006). Indeed, some aquatic insects have different habitat preferences and pollution tolerance making them suitable for assessing specific impacts (Ab Hamid et al., 2016). Among these insects, Ephemeroptera, Plecoptera, and Trichoptera (EPT) are important elements in the ecological dynamics of streams and are considered as crucial biological indicators that bring some fruitful insights into many aspects of stream ecology. In addition, ecological data on these insects provide valuable information for building and improving multimetric indices for biomonitoring and predicting the state of the river ecosystem in response to various environmental disturbances (Karr, 1991; Potikha, 2015; Righi-Cavallaro et al., 2010; Romero et al., 2013).

Moreover, the composition and biotypology of these three orders are determined by their high sensitivity to a wide array of environmental variables. Indeed, EPT species have limited levels of tolerance to changes in the watercourses and require high-quality water for their existence (Ab Hamid et al., 2016; Stoyanova et al., 2014). For this reason, many ecological studies have been dedicated to that specific assemblage (Melo, 2005; Righi-Cavallaro et al., 2010).

Within a biogeographical context, the Mediterranean side of Morocco can be considered as an island for aquatic fauna that is isolated by natural boundaries from the Mediterranean Sea to the north and arid lands to the south (Beauchard et al., 2003; Bonada et al., 2007). In this context, the River Martil, which occupies a strategic geographical position in northwestern Morocco on the southern side of the Mediterranean basin, is an excellent example of a studied site of the combined effects of seasonal variation, environmental factors, and human pressures on aquatic macroinvertebrates (Guellaf & Kettani, 2021a).

In this framework, the main goals of this study are (1), to analyze the composition, variation, and distribution of Ephemeroptera, Plecoptera, and Trichoptera (EPT) to improve our knowledge of those insects in the Martil basin. (2), to evaluate the relationship between the selected environmental drivers and EPT assemblages through a biotypological classification. (3), to examine responses of EPT taxa according to altitudinal gradient, land use, and hydrogeological characteristics of the prospected sites.

Material and methods

Study area and sampling sites

The study area corresponds to the Martil catchment, located in the mountainous domain of the Rif and belongs to the Tangier-Tetouan-Al hoceima Region in the northwestern of Morocco and covers 1259 km2 (Oualad Mansour et al., 2009).

Altitudes vary between 0 m on the coasts and 1782 m with an average altitude of about 424 m. The hydrological regime of this basin is characterized by a strong seasonal contrast presenting the maximum flows during winter and spring and minimum flows in summer and autumn. Indeed, the annual rainfall is of a great variability characterized by the rainy season during October–April and a dry season during May–September, and varies between 500 and 750 mm/year, while the climate is mainly Mediterranean with a mean annual temperature ranging from 15 to 19 °C (Karrouchi et al., 2016).

The hydrogeology of our study area is highly variable, as is its composition and land use pattern. The Oued Martil hydrographic network originates in forests and natural areas in the mountainous part of the limestone ridges in the south-east of the basin, and from Jbel Bouhachem in the south-west of the basin, passing then through areas of more moderate slope mainly associated with agricultural and rural activities and where the Oueds Mhajrat, Khemis, and Chekkoûr join to supply the Oued Martil river. This latter watercourse, which shares the same name as the watershed, is mainly associated with agricultural, urban and industrial areas before reaching the Mediterranean sea (Guellaf & Kettani, 2021a; Karrouchi et al., 2016). Moreover, the hydrological regime in the last decade was significantly altered due to the construction of dams and landscaping project on the edge of the Martil River.

In respect to the composition land, the stream bed of the southeastern watercourses (Oueds Nakhla, Zarka and Mhajrat) are dominated by calcareous, sandstone and clary substratum, while those located at southwestern streams (Oueds Tkaraa, Taida, Hamma, El Kebir, Chekkoûr) are flowing through substrates composed mainly by sandstone and clay. As for the lowlands streams in the northern half of the basin (Oueds Khemis and Martil), they are dominated by marly and sandy deposits (Karrouchi et al., 2016).

The study was carried out at 19 sampling sites (Fig. 1) between an interval of 5 m and 1062 m of altitude belonging to the most important streams of Martil basin through eight seasonal campaigns which took place from spring 2016 to summer 2018 and distributed along upstream–downstream gradient through the three major hydrogeological zones that characterize the Martil basin.

Fig. 1
figure 1

Location of the Martil Basin, land use and the location of sampling sites

Full size image

Sampling and identification

Benthic macroinvertebrates were sampled at each station with two sampling techniques, Surber net (20 × 20 cm) which was positioned firmly embedded into the substrate with its opening faced against the water flow, and used to sample riffles by dislodging, removing all organisms of each rocky substrate for holding them subsequently in the net of the sampler. The other meaning was a standard hand net (25 × 25 cm) used to get sufficient samples representative from all types of habitats.

Afterward, collected organisms were placed in a plastic box in a 95% alcohol. The material was immediately separated from any debris, rocks and leaves to reduce damage to the specimen, and preserved in jars with 90% of alcohol.

In the laboratory, the EPT larvae were sorted and identified to family level under a binocular microscope using specific keys of Sansoni (1992) and Tachet et al. (2002), and subsequently identified to species level with the help of the co-authors.

Physicochemical and habitat analysis

Water samples were carried out seasonally from winter 2016 to spring 2018 in the Martil basin. Water temperature (T), pH, electrical conductivity (CE), and dissolved oxygen (DO) were measured in the field immediately before the sampling of EPT taxa with the specialized portable device (EUTECH CyberScan PCD 650). Whereas BOD5 and COD were analyzed in the laboratory of the Loukkos Hydraulic Basin Agency (ABHL, Tetouan) using Pastel UV portable UV analyzer. The physicochemical parameters were determined as recommended by (Rodier 2009). Additionally, habitat characteristics were estimated through the average of depth (m), width (m) and current speed (m/s) of streams obtained from three measurements for each station.

Statistical analysis

Descriptive statistical analyses were performed to analyze the environmental variables and biotic data, including the calculation of mean, standard deviation and the coefficient of variation (CV) at the level of sampled sites and seasons. Kolmogorov–Smirnov test was applied to test variables normality. Since normality was not achieved, Kruskal–Wallis and Dunn’s post hoc tests were applied to test the differences between the studied sites, streams and seasons.

Spearman correlation coefficient was used to test the significant correlation between environmental variables, EPT species richness and abundance.

In order to quantify variables variance across the studied ecological levels (streams, sites, years and seasons) a variance decomposition procedure was performed (Kassout et al., 2019). The “lme” function in the “nlme” package (Pinheiro et al., 2019) was used to fit a general linear model using the restricted maximum likelihood method (RMEL) across the studied levels, and then the “varcomp” function in the “ape” package (Paradis et al., 2004) was used to extract variance expressed at each level.

Species composition and structure of the studied sites were analysed using correspondence analysis (CA). To study the effect of the environmental variables on the composition and structure of EPT species, a correspondence canonical analysis (CCA) was performed.

Hierarchical Cluster Analysis (HCA) was investigated with the objective to group the sampling stations under the effects of physicochemical and hydrological based on their level of associations. Word agglomeration method and jaccard similarity index were used for HCA analysis.

All statistical analysis were performed using the R software (R 3.3.3; R Development Core Team, 2015).

Results

Macroinvertebrate assemblages

In total, 10,370 specimens of EPT taxa were sampled from the Martil basin and identified. They belong to 22 families, 36 genera and 48 species (Annex 1).

Ephemeroptera was the most widespread, diverse in terms of species richness and abundant group, being present in all localities and contributed with 6 families, 13 genera and 19 species, followed by Trichoptera with 10 families, 13 genera and 18 species, and finally Plecoptera with 7 families, 10 genera and 11 species (Fig. 2).

Fig. 2
figure 2

Box–plots (median, min–max) showing the variations of taxa richness and abundance of EPT in the Martil basin during the study period

Full size image

Baetidae (Ephemeroptera), Perlodidae (Plecoptera) and Hydropsychidae (Trichoptera) were the most abundant families with 80%, 40% and 92%, respectively of the total catch for each group. In the same way, the genus Baetis Leach comprised the highest diversity with 9 species, when the most frequent species was Baetis rhodani, registered at all sampled sites with 20% of the total EPT catch.

Concerning the spatiotemporal variations, EPT taxa were more abundant and diverse in winter and spring compared to summer and autumn. Thus, spring 2017 showed the highest taxa richness (37 species). Moreover, only 10 species (Baetis maurus, Baetis punicus, Baetis pavidus, Baetis rhodani, Cloeon dipterum, Ecdyonurus rothschildi, Isoperla cf. kir, Hydropsyche iberomaroccana, Hydropsyche lobata and Hydropsyche pellucidula) were captured in all sampling seasons. The variation of species richness observed between stations during the overall period of study is shown in Fig. 3.

Fig. 3
figure 3

Seasonal variation of taxa richness (Mean, min–max) during our study period in the Martil Basin

Full size image

With regard to the stations, Oued Zarka (ZR1) was the station with the greatest abundance (1803 individuals), when Oued El Kebir (EK2) was the richest one in the terms of species (27 species).

Analysis of the EPT composition, based on the chorological categories according to their distribution area, reveals that this community occurring in the Martil basin can be divided into three main categories. The first one shows a clear predominance of the Mediterranean chorotype which constitutes 60% of the reported species, followed by the Palaearctic ones (32%) and finally the elements with wide distribution (8%). The Mediterranean chorological category is subdivided into seven groups of species, three of them are endemic in the broadest sense which constitutes 77% of the whole Mediterranean elements. Within the endemic species, Ibero-Maghrebian elements (10 species) represent 59%, followed by the Maghrebian (4 species) with 23% and lastly, 3 typically Moroccan species (Choroterpes volubilis, Isoperla cf. kir and Siphonoperla lepineyi) represent 18% of the total captured taxa in the Martil basin during our investigation (Appendix 1).

Correspondence analysis (CA) was used to search for possible sites grouping according to their species assemblage and composition. The first two axes (CA1 and CA2) explained 19.3% and 12.7% of the total inertia, respectively. Therefore, from the CA analysis, three EPT assemblage groups were discriminated during our study period in the Martil basin (Fig. 4).

Fig. 4
figure 4

Results of the Correspondence Analysis (CA), a distribution of EPT species on the CA1–CA2 factorial plans, b distribution of the sampling sites on the CA1–CA2 factorial plans

Full size image

Group A; It is composed of Epeorus sylvicola, Rhithrogena sp, Isoperla cf. kir, Siphonoperla lepineyi, Nemoura sp, Brachyptera auberti, Brachyptera algirica, Brachyptera sp, Capnioneura sp, Capnopsis schilleri, Leuctra sp, Rhyacophila fonticola, Agapetus sp, Wormaldia sp, Polycentropus kingi, Stenophylax sp and Calamoceras sp, situated in the right side of the biplot and mainly reported during flowing conditions and rocky/stony watercourses of the upper mountainous section (TK1, TA1, HM1, EK1 and EK2).

Group B; this group is constituted by Acentrella almohades, Baetis fuscatus, Baetis maurus, Baetis punicus, Ecdyonurus rothschildi, Serratella ignita, Eoperla ochracea, Protonemura sp, Rhyacophila munda, Hydroptila vectis, Cheumatopsyche atlantis, Hydropsyche fezana, Hydropsyche iberomaroccana, Hydropsyche lobata, Hydropsyche maroccana, and Hydropsyche pellucidula, which are thermophilic and eurytherms species of medium altitude and streams with a preference for calcareous substrates. These rhithral species tended to colonize lotic systems with high/moderate current velocity in the natural and agricultural areas of the Martil basin (ZR1, NA1, NA2, NA3 and MH1) positioned mostly in the upper left quadrant. When Alainites muticus, Caenis pusilla, Eoperla ochracea, Cheumatopsyche atlantis, Ecnomus deceptor, which also belong to this group, are only encountered in streams with limestone characteristic.

Group C; this group is composed by Baetis pavidus, Baetis rhodani, Cloeon dipterum, Choroterpes atlas, Choroterpes volubilis, Ephoron virgo, Caenis luctuosa, Hemimelaena flaviventris, Capnioneura sp and Mesophylax asperses, considered as thermophilic and eurytope species, able to colonize a wide variety of habitats. These species, mostly grouped in the lower left quadrant of the CA biplot, appear to be common throughout our region, looking to their considerable dispersal capacity under different altitudinal and climatic intervals. Thus, they inhabiting moderate and slow flowing rivers situated downstream mainly associated with agricultural (CH1, KA1, KA2, KA3) and urban (MA1, MA2, MA3, MA4 and MA5) sites.

Along the first CA axis, nineteen species (CVO, Hsp, RFO, Ssp, Casp, Wsp, BAU, Nsp, BAL, BPA, CSC, ESY, EDE, EVI, SLE, CPU, PKI, HLO, Bsp) have a strong and important contribution. On the second CA axis, the highest contribution was due to fourteen species (CPU, HMA, CVO, CDI, CMA, BMA, PPU, Asp, HVE, HFL, Csp, BFU, PCO, Lsp).

Physicochemical, hydrological and biotic variables

The Kruskal–Wallis and Dunn’s post hoc tests showed significant differences between the mean values of physicochemical, hydrological features, species richness and abundance through the prospected sites and streams in the Martil basin except for temperature (Tables 1 and 2). However, when considering inter-season variability, all variables show significant differences, except for water pH and EC (Table 3).

Table 1 Mean ± SD, and Kruskal–Wallis Test showing spatiotemporal changes of physicochemical, hydrological and biotic variables measured at the sampling sites of Martil basin during our study period
Full size table
Table 2 Mean ± SD, and Kruskal–Wallis Test showing spatiotemporal changes of physicochemical, hydrological and biotic variables measured at the different prospected streams of Martil basin during our study period
Full size table
Table 3 Mean ± SD, and Kruskal–Wallis Test showing seasonal variations of physicochemical, hydrological and biotic variables measured at the sampling sites of Martil basin during our study period
Full size table

The values of temperature ranged from 8.5 °C observed at Oued Taida (TA1) in autumn 2017 to 32.6 °C at Oued Martil (MA3) in summer 2016. Water temperature fluctuated significantly between seasons from highlands to lowlands. For all sites, the pH levels showed a slightly alkaline tendency ranged from 6.54 at Oued Chekkoûr (CH1) in summer 2017 to 8.71 during summer 2016 at Oued Mhajrat (MH1). Electrical conductivity was maximum (2900 μS/cm) at Oued Martil (MA5) in spring 2017 and minimum (26.5 μS/cm) in Oued Tkaraa (S1) during the same period. However, the highest mean of electrical conductivity levels usually corresponded to localities linked to agricultural areas. Regarding to BOD5 and COD, their levels followed the same trend and showed higher mean values especially at urban and industrial localities than the rest of the sites. These last parameters varied from 3,8 and 10.5 mg/L, respectively at Oued Nakhla (NA3) during autumn 2017 to 120 and 340 mg/L, respectively at Oued Martil (MA4) in summer 2017. Contrariwise, the values of dissolved oxygen being significantly higher in natural localities than the other sites associated with human activities. Thus, the highest value (19,9 mg/L) was reported at Oued Nakhla (NA3) in autumn 2017, whereas the lowest (0.2 mg/L) was enregistred at Oued Martil (MA4) in summer 2017.

Furthermore, the values of the three morpho-hydrological factors varied depending on upstream/downstream position, flow regime, altitude and intermittency level between seasons. The current speed (CS) varied from 1 m/s during winter 2018 at Oued Tkaraa (TK1) to 11.5 m/s in summer 2016 at Oued Martil (MA3). Likewise, water depth (WD) was significantly lower (0.14 m) at Oued Taida (TA1) in summer 2017 compared to the value of 0.70 m observed at Oued Martil (MA5) during spring 2016, while stream width (SW) accounted for a rising trend ranging from 2.10 m in summer 2016 at Oued Tkaraa (TK1) to 65 m during winter 2018 at Oued Martil (MA4). It should be pointed out that stream channels were sometimes dry or constituted by isolated pools in some temporary streams during drought periods.

Regarding the biotic metrics, the average values of taxa richness ranged from 1.75 recorded at Roumana (MA4) to 8.75 registered at Oued Taida (TA1). In the same way, abundance showed average values ranging from 2.62 to 215.12 recorded at Roumana (MA4) and Oued Zarka (ZR1), respectively.

The results show an intra- and inter-seasonal variation of the environmental factors, species richness and abundance among seasons, especially in dry periods. However, temperature, DO and CS varied significantly with seasons (Kruskal–Wallis test, p ˂ 0.0001), (Table 2).

Winter 2017 was the season of the highest stability of temperature (SD = 1.47), pH (SD = 0.17), EC (SD = 243.6), CS (SD = 2.01), SW (SD = 2.3) and abundance (SD = 44.72). The lowest values of SD in BOD5 (SD = 6.41) and COD (SD = 12.63) were observed in winter 2018. DO remained more stable in spring 2017 (SD = 1.85), species richness (SD = 2.3) in summer 2016. Whereas, the highest stability of water deep (SD = 0.3) was observed in autumn 2017.

Spearman correlation analyses were used to explore the relationship among physicochemical, hydrological and biotic factors (Table 4). In general, altitude, DO, species richness and abundance were positively correlated with each other and negatively correlated with most of the other physicochemical hydrological and biological factors considered.

Table 4 Spearman's correlation coefficients among different physicochemical, hydrological and biotic variables measured at the sampling sites of Martil basin during our study period
Full size table

Furthermore, significant and strong correlations were observed between BOD5 and COD (r = 0.947, p = 0.001), species richness and abundance (r = 0.931, p = 0.001), WD and CS (r = 0.918, p = 0.001), WD and temperature (r = 0.883, p = 0.001), altitude and species richness (r = 0.869, p = 0.001), DO and species richness (r = 0.863, p = 0.001), altitude and DO (r = 0.857, p = 0.001). Conversely, negative significant and strong correlations were observed between CS and species richness (r = -0.958, p = 0.001), WD and species richness (r = −0.913, p = 0.001), CS and abundance (r = −0.908, p = 0.001), WD and altitude (r = −0.907, p = 0.001), temperature and altitude (r = -0.889, p = 0.001), temperature and DO (-0,882, p = 0.001), temperature and species richness (r = −0.858, p = 0.001). However, no significant correlations were found between pH and the other measured factors.

The coefficient of variation showed the lower values for pH (CV = 2.31) in winter 2018 corresponding to very narrow variation ranges compared with the studied variables (Table 5). On the other hand, abundance (CV = 157.48) and stream width (CV = 139.12) show important variation in autumn 2017, indicating high uniformity along the studied sites considering the sampling seasons (Table 5). However, DO show important variation according to the studied seasons ranging from CV = 128.39 (autumn 2017) to CV = 26.24 (spring 2017).

Table 5 The coefficient of variation showing the amplitude of seasonal variability of physicochemical, hydrological and biotic variables measured at the sampling sites of Martil basin during our study period
Full size table

Hierarchical decomposition of variance related to the physicochemical, hydrological and biotic variables was applied, in order to assess the amount of variance of each variable across the three studied spatiotemporal scales (streams, sites, years and seasons) in Martil basin.

The percentage variance of CS was higher between-sites level. Besides, stream-level showed high EC, DO and WD percentage variance (41–64%). The greatest variability expressed between-streams was observed in DO with 63.75%, and the lowest was observed for pH with 9.58% of the total variance.

However, temperature, pH, DO, BOD5 and COD were the variables with the highest variability at the ‘between seasons’ level ranging from 49 to 84% (Table 6). In addition, temperature explained the main proportion of the variance between season with 83.36%, whereas SW only explained 6.78% of total variance.

Table 6 Estimated percentage variance across the studied hierarchical levels (streams, sites, years and seasons) among different physicochemical, hydrological and biotic variables measured at Martil basin during our study period
Full size table

Taxa richness (TR) expressed, respectively, 50.11% and 49.81% of total variance at the streams and season levels. Thus, species abundance (AB) shows important variability at the stream and season level with 47.16% and 44.20% of total variance, respectively. Conversely, SW accounted, respectively, for 32.38%, 27.48%, 30.43% of the total variance at the sites, streams and years levels.

Correspondence correlation analysis

The CCA analysis was performed to highlight the biotypological patterns of EPT community structure among sampling sites under the effects of environmental factors. The first two CCA axes explained 61.7% of the variation in EPT species and environmental relationships in the data set (axis 1:46.96% and axis 2:14.74%).

In the positive direction along axis-1, altitude and DO showed a close relationship with the most sensitive EPT taxa found in the Martil basin such as Baetis punicus, Procloean concinnum, Epeorus sylvicola, Rhithrogena sp, Habrophlebia sp, Serratella ignita, Isoperla cf. kir, Siphonoperla lepineyi, Nemoura sp, Protonemura sp, Brachyptera auberti, Brachyptera algirica, Capnioneura sp, Capnopsis schilleri, Leuctra sp, Rhyacophila fonticola, Agapetus sp, Wormaldia sp, Hydropsyche fezana, Polycentropus kingi, Stenophylax sp and Calamoceras sp. Those taxa seemed to be associated with high current speed locations in highlands areas, whereas the other environmental parameters were negatively correlated with axis-1 (Fig. 5).

Fig. 5
figure 5

Canonical correspondence analysis showing biplots of (a) taxa in relation to physicochemical, hydrological features, b sampling sites in relation to physicochemical, hydrological features

Full size image

The lower right quadrant of the plot includes species related to midstream localities such as Acentrella almohades, Alainites muticus, Baetis fuscatus, Baetis maurus, Ecdyonurus rothschildi, Ephoron virgo, Caenis pusilla, Eoperla ochracea, Rhyacophila munda, Hydroptila vectis, Cheumatopsyche atlantis, Hydropsyche iberomaroccana, Hydropsyche lobata, Hydropsyche maroccana, Hydropsyche pellucidula, Ecnomus deceptor and Psychomyia pusilla. The majority of these species are notably abundant at sampling sites influenced by a significant percentage of limestone substratum and slighty alkaline waters.

The bottom of the left part of the plot revealed that the widespread and/or tolerant taxa are associated with lowland stations of Martil basin such as Baetis pavidus, Baetis rhodani, Choroterpes atlas, Choroterpes volubilis, Caenis luctuosa, Hemimelaena flaviventris and Chimarra marginata, and are influenced by the variation of T, EC, BOD5, COD, SW, WD and CS levels. However, the upper part of the same quadrant contains Cloeon dipterum and Mesophylax asperses which are also positively associated with relatively wide and deep lowland localities and being characterized by high levels of temperature and pollutant concentrations.

Hierarchical cluster analysis

The Hierarchical Cluster Analysis revealed two clusters on the basis of to the physicochemical and hydrological variables as shown in Fig. 6.

Fig. 6
figure 6

Dendrogram shows the cluster analysis of the physicochemical, hydrological and biotic variables recorded during the study period at 19 sampling sites in Martil basin

Full size image

Group I contains 11 stations which represents the lower watercourses situated in the northern side of the basin. This group could subdivided into 2 subgroups:

Subgroup I.a includes stations (ZR1, MH1 and CH1) located mainly in the foothills reaches and subjected to intensive agricultural land use practices and/or flow through rural areas. These stations recorded slightly higher EC values.

Subgroup I.b correspond to stations (KA1, KA2, KA3, MA1, MA2, MA3, MA4 and MA5) located at lowland streams and covered by the agricultural, industrial and urban areas. They are characterized by slow-moving water flowing through a wide and deep watercourses and expressed high levels of temperature, EC and organic loads.

Group II contains 8 stations which represents mountainous streams situated in the southern side of the basin. This group is subdivided into 2 subgroups (Fig. 2c):

Subgroup II.a counts the stations (TK1 and TA1) with headwater stream characteristics occupied by relatively dense forest cover. Generally, these well oxygenated locations had the lowest values of physicochemical properties.

Subgroup II.b include stations (NA1, NA2, NA3, HM1, EK1 and EK2) associated with natural/agricultural areas of middle mountain reaches and exposed to low and/or moderate levels of mineral and organic pollution.

Discussion

It is known that the hydro-climatical, physicochemical variables and habitat characteristics including the geological nature, altitude and land use appear as the most meaningful drivers of variation in EPT composition and dynamics in the Mediterranean area (Mebarki et al., 2017; Tall et al., 2015; Vuori et al., 1999).

Our results demonstrated that upstream–downstream gradient and seasonality had strong impacts on physicochemical, hydrological characteristics and subsequently EPT richness and abundance. In the present study, the results obtained highlighted a strong spatial structure of EPT species according to their degrees of sensitivity to the physicochemical and hydrological variables measured during our investigation. However, species richness and abundance were positively correlated with DO, pH and altitude and negatively correlated with temperature and the other parameters measured. In this sense, our results were similar to what was known in other regions of the Ibero-Maghrebian areas (Bispo et al., 2006; García et al., 2014; Monbertrand et al., 2019; Narangarvuu et al., 2014; Pastuchová et al., 2008).

In this context, correspondence analysis (CA) was used to define groupings sites and species according to composition and structure. Thus, three EPT assemblage groups were discriminated during our study period in the Martil basin: a group of rheophilic taxa typical of mountainous streams exhibiting higher affinities for lotic habitats with rapidly flowing waters found mainly in natural stations under low levels of land use intensity, a group of thermophilic rheophilic species of midstream reaches exposed to agricultural activities which had more affinities for calcareous substrates with moderately to rapidly flowing waters influenced mainly by agricultural runoffs, and lastly a group of lentic, thermophilic and eurytope species with a widespread distribution in the Martil river basin and being located mostly in downstream areas after the dams and exposed to agricultural, industrial and urban activities.

With reference to the upstream–downstream gradient and hydro-climatic variations which acting as an important factors in the spatiotemporal fluctuation in aquatic macroinvertebrate in Mediterranean streams (Bauernfeind & Moog, 2000; Kalogianni et al., 2017; Lounaci et al., 2000; Pinheiro et al., 2020), several researches have already highlighted the effect of these variables throughout Moroccan watersheds: Sebou (Dakki, 1987), Oum Rbia (Bouzidi, 1989), Bou Regreg (El Agbani et al., 1992), Moulouya (Berrahou et al., 2001; Mabrouki et al., 2019) Oued Laou (El Alami, 2002; El Alami & Dakki, 1998; Errochdi, 2015), Bni Bousera region (Khadri, 2018) and Loukkos (Slimani, 2018). Their findings were similar to those observed in our study area showing that the taxonomic richness and abundance are also linked to stream types and land use.

In general, Baetis rhodani and Baetis pavidus were the most abundant and common species widespread throughout the various prospected localities. This result is in accordance with diverse researches addressing the distribution of EPT species conducted in the Maghreb region (Bebba et al., 2015; Khadri et al., 2017; Mabrouki et al., 2017; Mebarki et al., 2017). Concomitantly, Alainites muticus, Protonemura sp, Eoperla ochracea, Stenophylax sp and Ecnomus deceptor were collected with only a single specimen during our investigation in the Martil basin.

Thus, it is worth noting that the highland/midland permanent stations (TK1, TA1, EK1, ZR1, NA1 and NA3) contained 6036 individuals belonging to 43 species due to the wetted habitats which likely allowing for the development of mosaics of microhabitats and providing available refuges for different EPT taxa (Bonada et al., 2007), when 2178 individuals grouped in 34 species were observed in midland intermittent streams (HM1, EK2, NA2, KA1). As for midland/lowland ephemeral watercourses (MH1, CH1, KA2, KA3, MA1), 1760 individuals belonging to 29 species were encountered, and finally, only 249 individuals grouped in 12 species were found in the highly impacted permanent stations of the downstream part (MA2, MA3, MA4, MA5).

The most interesting faunistic findings consist in the first citation of Ephoron virgo, Eoperla ochracea, Protonemura sp, Brachyptera algirica, Capnioneura sp, Psychomia pusilla, Stenophylax sp and Mesophylax aspersus for the Martil watershed. In addition, Cloeon dipterum, Ecdyonurus rothschildi, Choroterpes atlas, Agapetus sp, Wormaldia sp, Ecnomus deceptor and Calamoceras sp seem to expand their range of distribution in the north of Morocco.

The biogeographical analysis of EPT species based on the chorological categories attributed to each taxon according to its area of occupancy shows that they are essentially composed of Mediterranean elements. The predominance of Ibero-Maghrebian species is explained by the great genetic exchanges between Iberian Peninsula and North African fauna across to the Betico-Rif massif which acted as a land bridge permitting the inter-continental migration of species and thus offers unique biogeographical settings for the Rif region (Lavergne et al., 2012). Our findings showed a strong similarity to those of El Alami (2002), Errochdi and El Alami (2008), Hajji et al., (2012, 2013), Errochdi et al. (2014a), El Bazi et al. (2017) and Slimani (2018) observed for EPT taxa in other northwestern localities in Morocco.

The comparison of the biogeographical patterns of EPT species richness in the Rif with other regions of Morocco (Navás, 1934; Aubert, 1961; Meinander, 1967; Miron, 1972; Giudicelli & Dakki, 1984; El Agbani, 1984; Qninba, 1986; Dakki, 1987; Ajakane, 1988; Bouzidi, 1989; Chergui et al., 1990; El Agbani et al., 1992; Ouahsine, 1993; Sánchez-Ortega & Azzouz, 1997; Vinçon & Sánchez-Ortega, 1999; Berrahou et al., 2001; El Alami, 2002; Touabay et al., 2002; Chillasse & Dakki, 2004; Maamri et al., 2005; Errochdi & El Alami, 2008; Dakki, 2009; Errochdi et al., 2014a; Errochdi et al., 2014b; Vinçon et al., 2014; Errochdi, 2015; Lamri et al., 2016; Mabrouki et al., 2016; Mabrouki et al., 2017; Hajji, 2017; Khadri et al., 2017; Khadri, 2018; Mabrouki et al., 2020) allowed us to conclude that in addition to the position of the study area as a bridgehead between Europe and Africa, the high number of EPT richness could be explained by the wide hydro-climatic interval as well as the geological nature of its lands (Khadri, 2018). On the other hand, this is due to the fact that this region has benefited from previous studies based on an integrated approach for identifying maximum species richness (El Alami, 2002; Errochdi, 2015 et Hajji, 2017; Khadri, 2018) (Table 7).

Table 7 Comparison of the specific richness of EPT families found through our survey in the Martil basin between the major biogeographic areas of Morocco (Rif, Middle Atlas, High Atlas, Central Plateau, and Eastern Morocco)
Full size table

Indeed, the specific richness (48 species) of the Ephemeroptera, Plecoptera and Trichoptera found in the Martil basin following our study constitutes about 42,85% and 30% of total species richness found in the Rif region and the whole of Morocco, respectively. However, the EPT taxa richness of our study area is comparable to other nearby regions with similar characteristics such as Tahaddart basin (36 species) (Errochdi, 2015), Bni Bousera region (49 species) (Khadri, 2018) and Loukkos basin (43 species) (Slimani, 2018).

Because of the lack of information on biological traits and the morphological/ecological adaptations to seasonal variability of EPT species, more research is clearly needed for assessing the concordance between ecotypes streams, environmental conditions and macroinvertebrates communities.

Undoubtedly, macroinvertebrate biodiversity information is a key tool to help managers to determine and improve appropriate actions for long-term conservation strategies for species and their ecosystems.

Conclusions

The results of this investigation reveal that:

  • Physicochemical parameters varied according to the anthropogenic pressures between natural, agricultural and urban/industrial areas, while hydrological factors are mainly affected by seasonal variations and upstream–downstream gradients.

  • The spatiotemporal occurrence of EPT taxa has shown different responses according to the combination of the parameters measured. Thus, the biotypological pattern observed supports the view that EPT taxa are distributed depending on their ecological requirements and degree of sensitivity to environmental and seasonal factors.

  • Ephemeroptera was the most diverse group during all sampling periods showing some stability toward seasonal variations, Plecoptera was occurred mainly during the wet period, while the highest taxa diversity of Trichoptera was especially enregistred at the beginning of the dry period. However, some taxa of Baetidae (Betis pavidus and Baetis rhodani) were the most abundant and widespread during the whole period of investigation.

  • Finally, given the geographical situation of the Martil basin in the Mediterranean Inter-continental Biosphere Reserve and the large percentage of endemic species that occurred, our study has considerably increased the knowledge about these insects in this region of Morocco and which can support afterward the conservation and management plans at regional and local scales.