Introduction

Ecosystems have been changed by humans over the years (Millenium Ecosystem Assessment 2005). They provide benefits for the well-being of humans which is termed as ecosystem services (Costanza et al. 1997). The most important factor that influences ecosystem services apart from climate change is land use and land cover (LULC) change (Lang et al. 2017). Recent studies around the world have proven that LULC change is the main driving force of change in the provision of ecosystem services (Kertész et al. 2019; Milheiras and Mace 2019; Schröter et al. 2019). Urbanization and agricultural expansion are among the major anthropogenic drivers of LULC change that have direct consequences on ecosystem services (Tolessa et al. 2017). Numerous studies have shown that changes in LULC can either increase or decrease ecosystem services such as carbon storage, soil conservation, water yield, and water purification (Geneletti 2013; Stockmann et al. 2015) and eventually affect human well-being (Quintas-Soriano et al. 2016). Ecosystem services are derived from different LULCs, and managing these locations while understanding their changes is critical for their sustainability. A better way to manage ecosystem services is to have adequate knowledge on the dynamic patterns of these services and to understand the ecosystem functioning and relationships (Leh et al. 2013). Studying the relationship between LULC and ecosystem services provides evidence of the current status of these services, and policies could be established to protect the lands and manage the ecosystems for the well-being of the society (Peng et al. 2015). The ecosystem services framework is useful for land use management and reshaping national development policy and planning (Goldstein et al. 2012).

Mediterranean countries have been experiencing anthropogenic LULC change over the past decades (Grove and Rackham 2003). The Mediterranean area has been intensely affected by human activity for millennia which caused only 4.7% of its primary vegetation to remain untouched (Geri et al. 2010). Agricultural expansion, urbanization and industrialization, population growth, development of tourism activities, and rural exodus are some of the main reasons leading to the alterations of the Mediterranean ecosystems (Serra et al. 2008). Rapid urban growth due to rural urban migration is putting enormous pressures on natural resources in the southern Mediterranean particularly Morocco (García-Ruiz et al. 2011). Marraccini et al. (2015) stated that transformation of natural vegetation into agriculture is prevalent in the southern Mediterranean region. Morocco’s significant urban growth is creating LULC competition especially between housing and agricultural demands (Debolini et al. 2015). Agricultural activities within the country are also expanding to other LULCs which is affecting other vegetation and water resources (Johannsen et al. 2016). These changes are having impacts on the provision of ecosystem services in the country (Schilling et al. 2012).

Most of the studies on ecosystem services in Morocco have targeted one of the services (e.g., carbon storage by Maanan et al. (2019), water yield by Rochdane et al. (2012), and sediment export by Simonneaux et al. (2015)) whereas the rest with multiple ecosystem services focus on single watershed or specific locations within the country (e.g., Bouahim et al. (2015), Ghazi et al. (2018), Kusi et al. (2020)). Multiple ecosystem services are not sufficiently documented in Morocco especially at the national scale. Out of the 52 ecosystem services assessment studies conducted by Wangai et al. (2016) in Africa, only 11 of them were carried out at national scales and none was from the northern part of the Sahara Desert. The few ecosystem services assessment studies conducted at the national scale rarely incorporate the different watersheds of the country, and few of them provide enough information on areas of intervention at the watershed scale. Since the impacts of LULC change and management on ecosystem services are strongly reflected at the watershed scale, it is crucial to incorporate studies at the watershed scale into national ecosystem services assessment in order to better understand and improve human well-being (Jiang et al. 2016). This would also facilitate implementation of policies and strengthen sustainable governance (Colvin et al. 2016). Numerous studies concluded that ecosystem services assessments need to be conducted at various spatial and temporal scales (de Groot et al. 2010). A study done by Willcock et al. (2016) after surveying stakeholders in sub-Saharan Africa also concluded that there is a need for more information on ecosystem services at national scales in Africa. Information provided from multiple ecosystem services assessment at the national scale improves policies and decision-making on LULC management in developing countries like Morocco (McKenzie et al. 2011). The analysis at the local stage within the national assessment of ecosystem services is of great importance since the local people are highly dependent on ecosystem services for their livelihoods.

In this study, we assessed the impacts of LULC change and management activities on multiple ecosystem services in Morocco, North Africa. During this process, Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) (Tallis and Polasky 2009) was used to quantify and map five ecosystem services: carbon sequestration, sediment export, water yield, nitrogen export, and phosphorus export for the years 1992, 2003, and 2015. InVEST is a spatially explicit modeling tool which uses maps and biophysical tables of land use together with other ecological factors such as climate, soil, and topography as data inputs to quantify, map, and value multiple ecosystem services at different scales (Polasky et al. 2011). It is more convenient for assessing the impacts of the changes of LULC on multiple ecosystem services (Bagstad et al. 2013) since it provides a large documentation that is adequate, simple, and sufficient enough to help anyone to use while using different approaches to address the scarcity of data. InVEST has been used in different parts of the world by several authors with much success (Goldstein et al. 2012; Hamel et al. 2015). Another tool, Resource Investment Optimization System (RIOS) (Vogl et al. 2015), was also used to select appropriate areas at the watershed scale where activities focused on restoration and conservation could be implemented in order to improve ecosystem services. RIOS uses information from biophysical conditions such as land use, climate, soil, and topography, and social information from a given location to help with interventions and management (Vogl et al. 2017). RIOS generates land use scenarios from the implemented land management activities and investment alternatives. InVEST and RIOS models are often used together in ecosystem services assessment (Lüke and Hack 2018).

This paper aims to assess the effects of the principal LULC changes and management activities on five selected ecosystem services. The objectives of this study were (1) to analyze changes in LULC in 1992, 2003, and 2015; (2) to quantify and map changes in the five ecosystem services (carbon sequestration, sediment export, water yield, nitrogen export, and phosphorus export) at both watershed and national scales; (3) to analyze the overall ecosystem services combined at the national and watershed scales; and (4) to improve ecosystem services at the national and watershed scales through the implementation of land management activities.

Materials and method

Study area

Morocco is located in the northwestern part of Africa ranging between 21 N and 36 N latitude and between 1 and 17 W longitude (Fig. 1). It is bordered by the Mediterranean Sea on the north, the Atlantic Ocean on the west, Algeria on the east, and Mauritania on the south. The total surface area is 710,820 km2 with 62% occupied by the pre-Saharan or Saharan area. Morocco has a mountainous landscape with the highest altitude of 4414 m at Toubkal and the lowest altitude of − 55 m at Sebkha Tah. The climate which is diverse varies from humid and sub-humid to Saharan and desert, also including arid, semiarid, and high mountain regions of the Rif, Middle Atlas, Anti Atlas, and High Atlas. The country has two major seasons: a hot and dry summer and a short winter with concentrated precipitation which are characteristics of a Mediterranean climate. The average maximum and minimum temperatures reach 50 °C in the Sahara and − 4 °C in the mountainous regions with a yearly average that can reach 10 °C. Precipitation is high in the north (more than 500 mm), medium in the center (between 100 and 500 mm), and low in the south (less than 100 mm). Precipitation falls between the periods of September and May, with November and December having the wettest months, but falls in the High and Middle Atlas as snow (Knippertz et al. 2003; Weisrock et al. 2006). Morocco has a bioecological diversity and important natural habitats which are divided into three main zones (Benabid 1985). The Mediterranean vegetation zone which consists of forests, bushland and thickets, shrublands, and matorral; the Mediterranean–Saharan transition zone which includes the argan scrub forest and steppes; and the Saharan zone which consists of sparse desert vegetation with Chenopodiaceae, grasses, and acacia trees. The grazing lands which are found in the south are the most dominant land use occupying more than 54 M ha (Sobrino and Raissouni 2000). Arable lands are about 22% of the land use in Morocco (Maliha et al. 2008).

Fig. 1
figure 1

Location of Morocco and the different watersheds in the country

Full size image

Ecosystem services modeling

The Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) tool was used in this study to model, map, and quantify ecosystem services in the study area (Tallis 2011). This study selected five ecosystem services to model: water yield, sediment export, carbon storage, nitrogen export, and phosphorus export. These services were selected because they are the most studied services in Morocco and data to model these services are easily accessible. All the services are related to water which is crucial to Morocco except carbon storage which would help with climate regulation because climate change will strongly affect Morocco (Simonneaux et al. 2015). All the data inputs necessary for this tool and their source are shown in Appendix 2. LULC maps for 1992, 2003, and 2015 were obtained from the ESA/CCI Landcover Project database (https://www.esa-landcover-cci.org) at 300 m resolution. This study aimed to analyze changes in LULC of Morocco for over 20 years, and the ESA/CCI LC maps cover a period of 24 years from 1992 to 2015 which allows us to assess annual LULC change dynamics (Kirches et al. 2017). Although the full time series of ESA/CCI LC maps provide a better temporal coverage for changes in LULC compared to other global land cover maps, its overall accuracy is 71.5% (Bontemps et al. 2011) which is lower than Global Land Cover (Globeland30) (83.5%) (Chen et al. 2015) and Moderate Resolution Imaging Spectroradiometer (MODIS) (74.8%) (Friedl et al. 2010). The maps were reclassified into eight LULC (agriculture, forest, grassland, wetland, urban areas, sparse vegetation, bare areas, and water bodies) to facilitate the calculation of the tool (Appendix 1). The MODIS land cover type (MCD12Q1; Friedl et al. 2002) datasets from 2001 to 2010 were also analyzed to compare with the results from the ESA/CCI LC maps. The 30 m digital elevation model (DEM) was downloaded using the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Model (GDEM) from the Earth Observing System Data and Information System of NASA. The Department of Water and Forests of Morocco has divided the country into different watersheds for a proper management of natural resources while the larger basins are managed by the regional basin authorities. This study used the same division but divided the largest watershed, Saquia El Hamra, into 9 sub-watersheds (watersheds 20–28) for easy comparison and analysis using the DEM (Fig. 1). Other authors also used the same watersheds in their studies (FAO 2015; Schyns and Hoekstra 2014). All data was resampled to a 30-m spatial resolution and transformed into the same Universal Transverse Mercator (UTM) coordinate system.

Carbon storage

The InVEST Carbon model was used to quantify the carbon storage in Morocco. This model aggregates the amount of carbon stored in four different pools: aboveground biomass (all living plant material above the soil such as trunks and branches), belowground biomass (living root systems), soil organic matter (soil), and dead organic matter (litter and dead wood) (Sharp et al. 2018). The model demands a raster dataset for the current LULC map and biophysical data comprising four carbon pools for each LULC. The carbon values were estimated through the InVEST guide and a set of average literature values for each LULC category shown in Appendix 2 (Maanan et al. 2019; Oubrahim et al. 2015; Zaher et al. 2020). Aboveground biomass, belowground biomass, and soil organic carbon were the pools considered for this study. Lack of data and limited literature for dead wood carbon values partly due to restrictions from the authorities were the reasons why it was not considered for this study.

Sediment retention

Sediment retention refers to the ability of the watershed to keep the soil, and when soil loss is controlled, water quality is improved. The universal soil loss equation (Wischmeier and Smith 1978) is used for this model in predicting the potential soil loss for each pixel on the study area.

$$ \mathrm{USLE}=R\times K\times \mathrm{LS}\times C\times P $$
(1)

where USLE is the average annual soil loss, R is the erosivity factor, K is the soil erodibility factor, LS is the field topography factor, C is the cropping and management factor, and P is the factor for supporting conservation practices. The values for cover management (C) and the practice factor (P) for each land use category were obtained from literature (Chadli 2016; Gaubi et al. 2017; Gourfi et al. 2018). Soil export was used as the main output for this model.

Water yield

In InVEST, water yield refers to the amount of water that runs off the landscape which will be useful for humans. The annual water yield which the model calculates uses mean annual precipitation, annual reference evapotranspiration, and correction factors for vegetation type, soil depth, and plant-available water content as data inputs (Sharp et al. 2018). All data source can be found in Appendix 2. The model estimates the total annual water yield (Y) for each pixel (x) of the watershed as total annual precipitation (P) minus total annual actual evapotranspiration (AET) (Eq. 2).

The water yield is determined for each pixel x by using this equation:

$$ Yx=\left\{1-\mathrm{AET}x/ Px\right\}\ Px $$
(2)

where Yx is the water yield, AETx is the annual actual evapotranspiration, and Px is the annual precipitation. Using the methodology developed by Budyko (1974) and Zhang et al. (2004), the model relates AET to potential evapotranspiration (PET) since the measurement of AET is difficult at the watershed scale (Eq. 3). PET is estimated as the product of the reference evapotranspiration and the plant evapotranspiration coefficient for each pixel.

$$ \mathrm{AET}x/ Px=1+\mathrm{PET}x/ Px\hbox{--} \left\{1+\left[\mathrm{PET}x/ Px\right]\ \omega \right\}\ 1/\omega $$
(3)
$$ \omega =Z\ \left\{\mathrm{AWC}x/ Px\right\}+1.25 $$
(4)

where ω is an empirical parameter which is related to the plant available water content (AWC), precipitation, and the constant Z (Eq. 4) (Sharp et al. 2018). The evapotranspiration coefficients (KC) were estimated based on literature and the land use category (Hou et al. 2016; Redhead et al. 2016; Sharp et al. 2015).

Nitrogen retention

The model which has interest in water quality is used to estimate the quantity of nutrients retained by different land uses for water purification. The model uses a three-step procedure to estimate the amount of nutrients exported from one landscape to the streams (Leh et al. 2013). Firstly, it calculates the annual average water yield based on the method used for the water yield model. Then, the nitrogen and phosphorus export coefficients supplied by the user for each land use category are used to calculate the average annual amount of nutrients exported from each pixel. The loading value of nutrients at each pixel is adjusted by the inclusion of a hydrological sensitivity score.

$$ \mathrm{ALV}x=\mathrm{HSS}x\cdotp \mathrm{pol}x $$
(5)

where ALVx is the adjusted loading value of the nitrogen and phosphorus at pixel x, polx is the nitrogen and phosphorus export coefficient at pixel x, and HSSx is the hydrologic sensitivity score at pixel x which is calculated as

$$ \mathrm{HSS}x=\lambda x/\lambda w $$
(6)

where λw is the mean runoff index for the watershed and λx is the runoff index at pixel x calculated using the formula

$$ \lambda x=\log\ \left(\sum Y\right) $$
(7)

where Y is the sum of water yield pixels along the flow path from pixel x and above. Finally, the quantity of nutrient load retained by the landscape is calculated using the retention capacity of nitrogen and phosphorus of each LULC class as

$$ \mathrm{retention}=\mathrm{ALV}x\times \mathrm{filtration} $$
(8)

where filtration is the nitrogen and phosphorus retention capacity of a given LULC class. The data on nutrient export coefficients for Morocco is rare, so we used data from literature outside of the country (Berg et al. 2016; Gunaratne et al. 2017).

Total ecosystem services analysis at different scales

The study did an integrated assessment of ecosystem services at the watershed scale in order to find specific watersheds that need appropriate management activities to help improve the overall ecosystems in the country. The data outputs from the InVEST model were processed and analyzed to determine the performance of ecosystem services at the watershed scale.

Data processing

A hotspot analysis was performed using the Getis-Ord Gi* hotspot analysis method in ArcGIS 10.2 (ESRI, Redlands, CA) in order to identify hotspots for each ecosystem services for the study area. The Getis-Ord Gi* hotspot analysis determines where the features with high and low z scores and p values tend to form a cluster in the study area. z score output represents the statistical significance of clustering for a specified distance, whereas the p value indicates the probability that the observed spatial pattern was created by some random process. The z score and the p value calculated by the analysis tool for each ecosystem service help to indicate cold and hot spots in the study area. The study also combined the five services for the year 2015 where each service was reclassified into a unitless value in order to visualize the current state of ecosystem services at the watershed scale in Morocco using ArcGIS 10.2. All ecosystem services were reclassified into five classes (very low, low, moderate, high, and very high) based on the study of Arunyawat and Shrestha (2016). Carbon storage and water yield were reclassified from 1 (very low ecosystem service), 2 (low), 3 (medium), 4 (high), and 5 (very high) whereas sediment export and nitrogen and phosphorus exports were reclassified the opposite way in ArcGIS. Finally, all the reclassified maps were put together to determine the overall performance of the ecosystem services using the weighted sum tool in ArcGIS. The zonal statistics tool in ArcGIS was used to determine the degree of the overall ecosystem services for each watershed at the national scale. Moreover, a percentage change of ecosystem services for each watershed at the national scale was also performed to analyze the watersheds that had the highest change between 1992 and 2015.

Land management activities in a selected watershed

One watershed was selected in order to perform specific land management activities to help improve ecosystem services at this location and Morocco. The selection was based on the watersheds with the lowest performance of the overall ecosystem services and higher percentage decreases from 1992 to 2015 of the total ecosystem services (Figs. 7 and 8). The study selected the RIOS tool which is used for making investments with the aim of attaining the greatest ecosystem service returns regarding multiple objectives (Vogl et al. 2015). The tool is made up of two modules known as the Investment Portfolio Adviser and the Portfolio Translator, which offers the opportunity to select cost-effective locations for management activities with the aim of protecting, maintaining, or restoring ecosystem system services for human well-being at low costs. The tool permits the user to select different activities linked to seven transitions supported by RIOS (keeping native vegetation, assisted or unassisted revegetation, agricultural vegetation management, ditching, fertilizer management, and pasture management) (Vogl et al. 2015). This study selected three main activities—protection (keeping native vegetation), reforestation (assisted or unassisted revegetation), and silvo-pasture (pasture management)—and all other inputs were selected based on the management and investment program of the Department of Water and Forests aimed at conserving water and soil within the selected watershed shown in Appendix 3 (HCEFLCD 2013). The RIOS tool gives three different maps as output, but the transitioned map was selected for this study. The transitioned map is the LULC map that was generated as a result of implemented activities showing the new LULC combinations and protected areas. This tool has been used by several authors, and it is mostly used in conjunction with the InVEST model (Esmail and Geneletti 2017; Lüke and Hack 2018; Vogl et al. 2015). The transitioned map was used as the main LULC input for the InVEST model in order to compare and contrast the differences in the five ecosystem services to the base map of 2015.

Results

Land use/land cover change between 1992 and 2015 in Morocco

The major changes in LULC over the years are shown spatially at the national scale and the watershed scale in Figs. 2 and 3. The most predominant land use in Morocco is the bare area which occupies more than half of the total surface area, followed by sparse vegetation and agriculture (Fig. 2). From 1992 to 2015, there were changes in all the LULC categories except wetlands. There were increases in forests by 1858.7 km2 (0.25%), sparse vegetation by 9420.7 km2 (1.3%), agriculture by 2250.6 km2 (0.3%), and urban areas by 1076.9 km2 (0.14%) while grasslands and bare areas decreased by 2853.6 km2 (0.4%) and 11812.7 km2 (1.6%) respectively (Appendix 1). The increase in forests came mostly from agriculture and sparse vegetation whereas that of agriculture came from sparse vegetation, bare areas, forests, and grasslands respectively. The increase in sparse vegetation was as a result of a decline in bare areas, agriculture, and grasslands while the decline of agriculture, sparse vegetation, and bare areas provoked the increase in urban areas respectively. At the watershed scale, all the watersheds experienced changes with the highest coming from the northeastern part of the country and the least coming from the southern part. The watersheds that were highly altered were Moulouya, Souss, Ziz Rheris, and Guir, whereas Saquia El Hamra, Loukous, and Drader Souier were slightly altered (Fig. 3).

Fig. 2
figure 2

Land use maps for 1992, 2003, and 2015

Full size image
Fig. 3
figure 3

Specific changes in the land use maps from 1992 to 2015 at the national and watershed levels

Full size image

The expansion of forests happened in Sebou (493.7 km2), Oum Er Rabia (159.1 km2), and Loukous (117.8 km2) while the increase in agriculture mainly took place in Moulouya (1004.5 km2), Souss (712.7 km2), and Isly (649.5 km2). The increase in sparse vegetation occurred in Moulouya (5205.7 km2), Guir (2086.9 km2), and Ziz Rheris (1722.6 km2) whereas the increase in urban areas occurred in Sebou (108.5 km2), Casablanca Atlantic Coast (68 km2), and Tangerois (20.4 km2). The decline in grassland happened in Oum Er Rabia, Draa, and Tensift whereas the decrease in bare areas took place in Moulouya, Guir, and Ziz Rheris. The results of LULC change using the ESA/CCI LC datasets were compared with the MODIS datasets from 2001 to 2010 (Appendix 1). The results from MODIS datasets showed increases in forests (227.6 km2), sparse vegetation made up of shrublands and woodlands (4327.4 km2), agriculture (9217.2 km2), and urban areas (45.1 km2) but a decrease in bare areas (14,942 km2). The major difference between these two datasets was the grasslands which increased using the MODIS datasets but decreased when the ESA/CCI LC datasets were used.

Changes in ecosystem services

The changes in LULC from 1992 to 2015 affected all the ecosystem services in Morocco. The study showed decreases in all the ecosystem services except carbon storage in 2003 and 2015. Carbon storage increased at 1.66% in 2003 whereas water yield, sediment export, nitrogen exports, and phosphorus exports decreased at 6.2%, 0.21%, 0.88%, and 0.78% respectively. Between 1992 and 2015, there was an increase in carbon storage at 1.69% but there were decreases in water yield at 15.5%, sediment export at 7.4%, nitrogen export at 0.43%, and phosphorus export at 0.4% (Table 1). The highest change in carbon storage, water yield, and sediment export at the national scale occurred in the year 2015 whereas the year 2003 saw the highest decrease in nitrogen and phosphorus exports.

Table 1 Total values of ecosystem services estimated by the InVEST model in 1992, 2003, and 2015
Full size table

Figure 4 shows the spatial presentation of the ecosystem services at the national and watershed scales for the different years used in this study. Carbon storage was high in the northwestern parts of the country but low in the southern watersheds for all the years. Water yield was very high at the northeastern watersheds and very low at the southern parts of the country. Sediment export was high at the central parts toward the south whereas nitrogen and phosphorus exports were higher at the southern parts and low within the northern watersheds. The hotspot analysis shows that the watersheds in the red zone are the hotspots of a given ecosystem service which has high values of z score whereas the watersheds in blue are the cold spots of an ecosystem service with low values of z score (Fig. 5). The watersheds in red show statistically significant clustering or dispersion, an evidence of underlying spatial processes at work.

Fig. 4
figure 4

Ecosystem services in the various watersheds at different years in Morocco

Full size image
Fig. 5
figure 5

The distribution of hotspots for the ecosystem services in the different watersheds in Morocco

Full size image

Changes in ecosystem services at the watershed scale of 2003 and 2015 are presented in Fig. 6 in the form of percentage change from 1992. Carbon storage saw the highest increase in the northern watersheds, with the Guir watershed showing the highest increase at 4% in 2003 and 25% in 2015. The highest decrease for carbon storage was found in Isly which is also found in the northeastern part of Morocco. The highest decline in water yield was found in Souss at 22% in 2003 and Moulouya at 28% in 2015, with other watersheds showing similar decreases such as watersheds Sebou, Guir, Mediterranean Coast, and Ziz Rheris. Safi El Jadida Coast, Saquia El Hamra, and Loukous were the only watersheds that showed increases in water yield for both years, with the highest increase coming from Ziz Rheris in 2003 and Safi El Jadida Coast in 2015. The highest increase in sediment export was found in Isly whereas the highest decrease was located in Ziz Rheris in 2015 (Fig. 7). The nitrogen and phosphorus exports presented higher increases in Isly for both years and higher decreases in Guir for 2015 and Saquia El Hamra for 2003.

Fig. 6
figure 6figure 6

Percentage change of ecosystem services from 1992 within the main watersheds of Morocco in 2003 and 2015. a Carbon storage, b water yield, c sediment export, d nitrogen export, e phosphorus export

Full size image
Fig. 7
figure 7

Percentage change of the total ecosystem services for the different watersheds of Morocco for the latter year 2015

Full size image

Integrated ecosystem services assessment and management synergies

Figure 8 shows the performance of the overall ecosystem services at the watershed scale. The overall ecosystem services assessment shows that the northwestern and western watersheds showed higher levels of ecosystem services whereas the northeastern, the eastern, and the southern watersheds displayed moderate to lower levels of ecosystem services. More than half of the country (over 60%) has medium level of ecosystem services in 2015 whereas about 30% has high ecosystem services (Fig. 8). The total percentage changes in the overall ecosystem services from 1992 to 2015 for each watershed are shown in Fig. 7. There were higher decreases in Moulouya, Ziz Rheris, Souss, Guir, and Tamri watersheds whereas watersheds such as Massa, Essaouira Coast, Safi El Jadida Coast, and Saquia El Hamra presented the only increase for the overall ecosystem services in Morocco. Since almost all the watersheds located in the southern part are bare areas, Guir watershed which is located in the northeast was selected as the watershed that requires management activities in order to improve the state of the overall ecosystem services in the future.

Fig. 8
figure 8

Overall ecosystem services in Morocco for the latter year 2015

Full size image

Implementation of land management activities

The selected output from the RIOS model which shows the implementation of the selected activities on the base map is shown in Fig. 9. The transitioned map shows the conversion of sparse vegetation to forests based on the reforestation activity through assisted revegetation and the conversion of sparse vegetation to agriculture from the silvo-pasture activity through pasture management (Fig. 9). This caused an increase in carbon storage at 7.6% and decreases in sediment export, nitrogen export, and phosphorus export at 3.6%, 4.5%, and 2.6% but a negligible decrease in water yield for the transitioned scenario at the watershed scale (Table 2). At the national scale, the management activities implemented in the Guir watershed caused an increase in carbon storage at 0.2% (6,762,668 t) whereas water yield, sediment export, nitrogen export, and phosphorus export decreased at 0.07% (4,780,879 m3), 0.09% (4,014,833 t), 0.17% (227,382 kg), and 0.09% (10,124 kg) respectively. The highest change occurred in carbon storage whereas the least change occurred in water yield.

Fig. 9
figure 9

Land use maps of the Guir watershed including the base and the transitioned maps

Full size image
Table 2 Total values of ecosystem services for the Guir watershed
Full size table

Discussion

Principal LULC changes in Morocco

The LULC changes in Morocco revealed increases in forests and agriculture and urban areas between 1992 and 2015. This corresponds with the other studies on land use change in Morocco (e.g., Maanan et al. 2019; Simonneaux et al. 2015; Sobrino and Raissouni 2000). The results from the MODIS datasets also showed increases in forests and agriculture and urban areas which validates the LULC changes from the ESA/CCC LC results. Morocco is a country where rangeland, pasture, and forest occupy over half of the land area and agriculture or livestock raising is the source of livelihood of almost half of the working population (Barrow and Hicham 2000). The country is experiencing significant urbanization and urban growth like other southern Mediterranean countries where two thirds of the inhabitants already live in urban areas with an annual growth rate of 4.3% (Debolini et al. 2015). The urban population increased from 12.6 million in 1992 to 20.8 million in 2015 compared to the rural population which increased from 12.5 million to 13.4 million. The annual growth rate for the urban population was 3.27% and that of the rural population was 0.5% in 1992; it was 1.96% for the urban population and − 0.33% for the rural population in 2015 (HCP 2015). The population increase in urban areas had five regions (Casablanca-Settat, Rabat-Sale-Kenitra, Tanger-Tetouan-Al Hoceima, Sous-Massa, Marrakech-Safi) contributing to over 70% of the urban population. This is putting a lot of pressure on the land uses and creating land use competition between housing and other land uses. It is estimated that the southern regions of the country will have an annual growth rate at 1.4%, higher than the national average at 0.96% in 2030 (de Miras et al. 2005). Our results revealed that urban areas had the highest change rate between 1992 and 2015 at 93%, and these changes were observed mostly in the aforementioned five main regions.

Agricultural lands keep increasing because agriculture accounts for about 17% of the country’s gross domestic product and employs about 40% of the country’s labor force (Diao et al. 2008). Schilling et al. (2012) compared the importance of the agricultural sector to both the GDP and employment of the North African countries and Morocco ranked very high with Egypt. Agricultural lands have been increasing at the expense of grassland (McGregor et al. 2009). Agriculture expansion is driven by the need of raw material for national and international agro-processing industries. Transformation from traditional agriculture to huge agro-industrial plantations of vegetables, citrus fruits, and bananas is common in the southern part of Morocco (Peter et al. 2014). The government’s aim to increase agricultural productivity by up to 59% till 2020 through intensification of production and extension of cropland also caused the increase in agricultural lands in the country (Schilling et al. 2012).

The increase in forests is as a result of policies on climate change and desertification by the Moroccan government (Maanan et al. 2019). Increases in forests were seen in the Rif and Middle Atlas mountains (similar to Maliha et al. 2008). Rudel et al. (2005) stated that 38% of the world’s countries had begun to experience sustained increases in forest cover after earlier periods of deforestation. A study done by Barakat et al. (2018) on the changes in LULC from 2001 to 2015 in the central and eastern parts of Morocco showed an increase in forest areas. Even though there has been deforestation in the country, increases in the forest areas through reforestation have caused an overall increase in the forests. Another study by Haboudane (2007) in the Middle Atlas of Morocco revealed that the total increment of the cedar and holm forests (21%) had exceeded the effects of deforestation (17%). Mohajane et al. (2018) also concluded that the forest cover in Azrou, Middle Atlas, remained stable between 1987 and 2017. The reforestation projects by the Department of Water and Forests in Morocco contributed to the increase of forests in the country. For example, regeneration of natural species such as cedar, atlas cypress, cork oak, carob, and argan was 167,500 ha in 2003 (Allaoui 2004). The protection of national parks and reserves also contributed to the increase of forest cover. For example, the department of water and forestry established 2 national parks in the 1990s and 6 other parks between 2004 and 2008 as part of their conservation strategies which covered about 750,000 ha including the ones established in the past (Taleb 2016). One hundred fifty-four sites of biological and ecological interest classified as nature reserves covering an area of 2.5 million ha were also established beginning in 1996 to support sustainable management of the forests (Taleb 2016). The department also created the sector for protection of forests in 1999 to ensure the sustainability of forest ecosystems (Saintonge et al. 2011).

The main LULC that decreased drastically from 1992 to 2015 was grassland with a change rate of 80%. All over the world, grasslands are being degraded and transformed to other LULC categories such as agriculture (Egoh et al. 2011). This is evident in the temperate grasslands which have lost more than 70% of their natural cover by 1950 (Millenium Ecosystem Assessment 2005). In Morocco, most of the grasslands are used as rangelands for grazing and for collecting medicinal and aromatic plants (Mahyou et al. 2010). The level of degradation of grasslands, their conversion to agriculture, and the overexploitation for medicinal plants are evident in Morocco (Fennane 2004; Linstädter and Baumann 2013) and in the world (Claassen 2011; Tang et al. 2019). Grassland in the southern and the northeastern parts of the country are being converted to agriculture whereas conversion to urbanization is common in the remaining parts of the country (Mahyou et al. 2010). Water bodies also increased from 1992 to 2015 because Morocco doubled its number of dams from the 1990s to 152 in total by 2015 (Gourfi et al. 2018).

Ecosystem services and LULC change

Our results reveal a link between ecosystem services and LULC change and the way ecosystem services are affected by alterations in LULCs. Figures 4 and 7 show the impact changes in LULC have on multiple ecosystem services over the years. An integrated ecosystem assessment offers enough evidence to ensure a sustainable management of natural resources.

The increase in forests and sparse vegetation from 1992 to 2015 resulted in the improvement of carbon storage and a decline in nitrogen export, phosphorus export, sediment export, and water yield (Table 1). The contribution of forests to the improvement of carbon storage has been confirmed by several authors both in Morocco (Ghazi et al. 2018; Kusi et al. 2020; Maanan et al. 2019; Oubrahim et al. 2015) and in other parts of the world (Delphin et al. 2016; Martínez et al. 2009; Ouyang et al. 2016). This is confirmed in the hotspot analysis as the northwestern parts which are dominated by forests had higher carbon storage than anywhere else in the country (Fig. 5). The watersheds in red (Fig. 5) show the areas with the most protected forests and natural reserves. Other studies concluded that protected forests offer more ecosystem services especially carbon storage (García-Nieto et al. 2013; Palomo et al. 2013). This should help reduce the threats to forest ecosystems in the country such as firewood which is the main source of energy for the rural communities in Morocco (El Moudden 2004). Felling of trees and overgrazing constitute about 80% of crimes committed in the forest ecosystems in the country (Ellatifi 2012). Guir and Ziz Rheris had the highest increase in carbon storage due to the conversion of bare areas to sparse vegetation, 2086.9 km2 and 1722.6 km2 respectively, whereas Isly had the highest decline in carbon storage due to the conversion of sparse vegetation to agriculture (649.5 km2) and a negligible increase in forests.

Soil export had a higher decrease at 7.4% from 1992 to 2015 due to the expansion of natural vegetation. Soil export was higher in the central parts of the country which are dominated by bare areas and sparse vegetation. The southern watersheds had lower values of sediment export as compared to the northern parts since the northern watersheds presented expansion of agriculture and urban areas. The expansion of agriculture and urban areas caused Isly to have the highest increase in soil export, but the increase in sparse vegetation and forests caused Guir, Ziz Rheris, and Moulouya to have the highest decline in soil export in 2015. Agriculture and less protective vegetation generally have higher rates of soil export than other LULCs (Simonneaux et al. 2015). Morocco’s Mediterranean climate facilitates soil erosion due to the influence of concentrated precipitation in certain months with a warm and dry summer (Ramos and Martínez-Casasnovas 2009). The government’s effort to increase perennials such as olives and citric fruits especially in agricultural lands contributed to the decrease in soil export (Schilling et al. 2012). Conservation agriculture and non-tillage systems which have been introduced in the arid watersheds of the country over the years also caused the reduced soil export, greater soil water conservation, and improved soil quality (Mrabet et al. 2012).

Water yield declined in 2003 and 2015 with a percentage decrease of 6.2% and 15.5% respectively, and this was due to the steady expansion of forest and sparse vegetation in these years (similar to Gao et al. 2017; Wu et al. 2018). Moulouya and Souss watersheds had the highest decline in water yield due to the high increase in forests. Forest expansion produces an increase in evapotranspiration which reduces water yield (Geneletti 2013; Zhan et al. 2015). The expansion of natural vegetation over the years helped to reduce nitrogen and phosphorus exports from 1990 to 2015. The decline of forests or sparse vegetation and the expansion of urban areas and agriculture caused the Isly watershed to have the highest increase in nitrogen and phosphorus exports.

Land management activities and ecosystem services

The eastern and southern watersheds of Morocco would need an implementation of activities to improve the state of ecosystem services for the whole country. Activities are feasible conservation strategies or measures taken to create realistic plans for watershed interventions in order to ensure sustainable management of watersheds. Most of the watersheds in Morocco presented decreases in the overall ecosystem services, and proper land management would help to enhance the ecosystem services in the country. Activities such as reforestation, protection of native forests, and pasture management were implemented in the Guir watershed, and this resulted in the improvement of ecosystem services at the watershed scale and at the national scale. The transitioned LULC map which incorporated the implementation of all these activities showed a decrease in sediment export and nitrogen and phosphorus exports but an increase in carbon storage and a negligible decrease in water yield at the watershed scale.

At the national scale, water yield declined by 0.07% as compared to the other ecosystem services and this was due to the incorporation of pasture management activities in the watershed. Zhan et al. (2015) also concluded that pasture planting is one of the ways to improve water yield. Kusi et al. (2020) found that the combination of forest and agricultural activities will help to stabilize water yield in the future. Carbon storage had the highest increase at the national scale by 0.2% which was mainly caused by the implementation of protection of native forests and reforestation activities. The decrease in sediment export was due to the reforestation, pasture management, and forest protection activities. Debolini et al. (2018) stated that forests and silvo-pastoralism facilitate the regulation of soil erosion and maintenance of soil fertility. Nitrogen export also had the highest decrease (0.17%) at the national scale due to the implemented land management activities. The land management activities implemented at the Guir watershed affected positively the overall ecosystem services at the watershed scale and the national scale. These activities could be encouraged in other parts of the country to ensure a sustainable management of ecosystem services both at the national scale and the watershed scale.

Improving multiple ecosystem services in Morocco

This study has shown that ecosystem services can be improved at the national scale when past LULC change is analyzed in order to follow the trend and be able to make proper decisions and policies that can improve ecosystem services in the future. The study also showed that ecosystem services can be well managed and improved at the national level when they are assessed at the watershed scale and when management activities are implemented at the watershed scale. It is important to analyze the hotspots of ecosystem services in order to know the areas of intervention either through protection or restoration of these services. The overall analysis of multiple ecosystem services gives a clear evidence of the state of ecosystem services and a well-balanced management that would reduce the impact of tradeoffs among different ecosystem services. Implementation of land management activities such as reforestation and pasture management at specific watersheds can help to improve the overall ecosystem services at both the national and the watershed scales.

Since agriculture has been the main focus of the Moroccan government for some years, policies should take into consideration sustainable agriculture and agricultural expansion at selected locations in order to ensure sustainable management of ecosystem services and LULC. Agricultural policies should integrate the assessment of overall ecosystem services to improve the well-being of the population. The Department of Water and Forests of Morocco should educate the local population about the conservation of forests and the protection of reforestation sites at the watershed scale since they contribute immensely to sustainable ecosystem services. Rural-urban migration which is causing urban expansion could be reduced if local farmers and shepherds benefit from programs such as the payment of ecosystem services. Sustainable ecosystem services could be achieved when there is active participation of local farmers, the private sector, and various government agencies in policy making and its implementation.

Conclusion

LULC change is a major driving force to the provision of ecosystem services at the national scale. Past LULC changes provide evidence that could help policy makers and stakeholders to decide on future LULC planning and ecosystem management. Future LULC management is required especially at the watershed scale to facilitate sustainable ecosystem services and land management. This study assessed the impacts of LULC change and land management activities on multiple ecosystem services in Morocco, North Africa. There were increases in natural vegetation and urban areas but decreases in grasslands and bare areas during the 23 years. These changes in LULC affected all the ecosystem services analyzed in this study, with carbon storage presenting the only increase in both 2003 and 2015. The remaining ecosystem services all decreased, with water yield revealing the highest decrease, followed by sediment export, nitrogen export, and phosphorus export in both years. The overall ecosystem services assessment at the national scale revealed a high level of ecosystem services in the northwestern watersheds of the country whereas the eastern and southern watersheds showed low levels of ecosystem services. The land management activities such as reforestation and pasture management implemented in one of the watersheds with low levels of ecosystem services caused an improvement in the overall ecosystem services at the watershed and national scales. Ecosystem services can be improved at the national scale when they are analyzed at the watershed scale whereas land management activities performed at selected watersheds generate an increase in ecosystem services both at the watershed and the national scales. This would also give a holistic knowledge for proper decision-making on ecosystem service management by stakeholders.