Are Cedrus atlantica forests in the Rif Mountains of Morocco heading towards local extinction?

Abstract

Cedrus atlantica (Atlas cedar) is a relict and endemic endangered species from northwestern African mountains, whose distribution range has undergone a dramatic reduction over recent decades. Long-term studies are needed for a better understanding of the development of its range as well as for assisting in the implementation of sustainable conservation measures. The multi-proxy analysis of a high-resolution fossil record of 180 cm depth allowed us to depict the final demise of an Atlas cedar population from the western Rif Mountains (Jbel Khesana), despite its high resilience during the last ~4000 years. Currently, Atlas cedar trees are not observed in Jbel Khesana but they still occur in the nearby area as scattered populations on a few mountain tops at altitudes higher than 1400 m a.s.l. Our data show an initial relatively stable period (~4000–2400 cal. yr BP) followed by a phase where both climatic and human-induced disturbances cause an alternate dominance of oaks and Atlas cedars (2400~1550 cal. yr BP). Then, the increasing aridity and human activities favoured the depletion of Atlas cedar forests (~1550–800 cal. yr BP). Our record shows that Atlas cedar forests have recovered after each deforestation event, which reveals a high resilience of the species until the mid-20th century, when they became extinct in the study area. The main driver of their local extinction may be attributed to the strong human pressure. Management measures of Atlas cedar in the Rif Mountains should aim at limiting intensive loggings and protecting the existing populations for their local regeneration.

Keywords

charcoal extinction geochemistry late Holocene Morocco non-pollen palynomorphs pollen Rif Mountains

Introduction

The Atlas cedar (Cedrus atlantica (Endl.) Manetti ex Carrière) is a relict and endemic tree species from northern African mountains of Morocco (Rif and Atlas ranges) and Algeria (Linares et al., 2011; M’hirit and Benzyane, 2006). The Rif Mountains are located in the northwestern part of Morocco and harbour 6.5% of the current range of Atlas cedar, namely, 12,138 ha (UICN, 2015). These Rifan populations have lost 49% of their area in the last three decades (Cheddadi et al., 2017; UICN, 2015). Future predictions for 2050 indicate that their distribution range will be restricted to the top of very few mountains, with expected local extinctions from areas at lower altitude (IRES, 2010; UICN, 2015). Atlas cedar populations seem to have gradually migrated upwards forced by climate change during the Holocene, like a number of tree species growing near treelines (e.g. Cheddadi et al., 2016; Heiri et al., 2006; Tinner and Theurillat, 2003; Walther et al., 2005). Model simulations indicate an upward shift of the lower limit of their range in the Rif Mountains by about 200 m (Cheddadi et al., 2017). Recent decline in these mountains, however, is considered to have been induced by human activities, mainly cropping, cutting and overgrazing (Ajbilou et al., 2003; Benabid, 1985; Coudel et al., 2016; Grovel, 1996; Hamelin and Nwankwo, 2013; Moore et al., 1998; Rejdali, 2004). Disentangling the impact of each factor on landscape is not an easy task, although essential to take the appropriate managing measures to ensure the persistence of the modern populations (Allen et al., 2010; El-Baha et al., 2010; IRES, 2010; Vegas-Vilarrúbia et al., 2011).

Palaeoecological studies from the Mediterranean region have shown that the current tree flora is composed of very resilient and relict taxa that have experienced many abrupt and intense climate changes in the past (Carrión et al., 2010; Cheddadi et al., 2009; Magri, 2012; Magri et al., 2017; Petit et al., 2005; Terrab et al., 2008), but also display local and regional extinction events in different areas (Dullinger et al., 2012; Postigo-Mijarra et al., 2010; Rubiales et al., 2012; Tinner and Ammann, 2005).

High-resolution (~20–50 years) multi-proxy records, such as in the present study, provide an accurate analysis of the vegetation changes and disturbances which lead to the current landscape. A wide range of palaeoenvironmental proxies is necessary to properly assess the landscape changes through time and to develop ad hoc future management strategies. In addition to traditional palynological studies, the identification of spores of both green algae and fungi, as well as the analyses of microcharcoals and geochemical elements contribute to more reliable environmental reconstructions (Bindler, 2006; Chambers et al., 2011, 2012; Cook et al., 2011; Cugny et al., 2010; Iglesias et al., 2015; López-Vila et al., 2014; Montoya et al., 2012; Mooney and Tinner, 2011; Silva-Sánchez et al., 2014, 2016; Van Geel, 2001).

Recent fossil records (Figure 1) collected in the Rif Mountains (Cheddadi et al., 2016, 2015; Muller et al., 2014, 2017) as well as the pioneering palaeoecological work of Reille (1977) allow the analysis of the forest dynamics in these mountains during the Holocene. The latter has been partially re-edited (Muller et al., 2014) with new ¹⁴C dates which support a better evaluation of the chronological frame of the vegetation dynamics. Cheddadi et al. (2015) provided a focus on the spatial and temporal variation of human impact, setting the decline of Atlas cedar forests during medieval times and warning that this process is still underway. Cheddadi et al. (2016) show a record from the same wetland (Bab el Karn) as this study that spans the last 9000 years. These authors state the local extinction of Cedrus atlantica at ~2000 cal. yr BP, probably because of the combined effect of increasing aridity and anthropogenic disturbances. However, they express the difficulty in assessing which driver played the dominant role in the local extinction of the Atlas cedar populations. Here, we present a new core, Bab el Karn II, which spans the last 4000 years. We have carried out a high-resolution analysis of several environmental proxies (pollen and non-pollen palynomorphs (NPPs), fire, sedimentology and elemental geochemistry) with a main focus on the development of Atlas cedar forests and their relationship with climate and human activities.

Figure 1.

Location of Bab El Karn II (BEK II) mire (red star), showing some previous palaeoecological studies cited in the text (white stars) and the current Atlas cedar distribution range in the area (blue area).

The objective of the present study is threefold: (1) to provide an accurate record of the development of Atlas cedar forests in the western Rif Mountains, specifically in the Jbel Khesana; (2) to clarify whether the long-term Atlas cedar forest changes are related to climate or anthropogenic disturbances; and (3) to help the decision-making process for future management and monitoring of the Atlas cedar forests.

Material and methods

Study area and sampling

The mire of Bab el Karn II (35°1′23″ N, 5°12′24″ W) is located in a seasonal ponded water area developed in a lithologically controlled plain on the northern slopes of Jbel Khesana (western Rif Mountains) at around 1200 m a.s.l. (Figure 1). The narrow plain area includes two hydrologically connected open mires (ca. 400 m²) that collect the runoff water of several streams. Jbel Khesana (1705 m a.s.l.) is composed of Cretaceous carbonate rocks and Neogene terrigenous rocks. The most common outcropping sediments consist of early to late Cretaceous flysch facies, alternating marls and limestones and fine terrigenous rocks, as well as Oligo-Miocene sandstones in the upper area of the mountain range (Service Géologique du Maroc, 1980). The climate in the Rif Mountains is characteristically Mediterranean, with more than 65% of the annual rainfall occurring in spring and autumn. Annual precipitation usually exceeds 400 mm, reaching more than 2000 mm in high-altitude areas of its western range (Charco, 1999). Atlas cedar grows between 1400 and 2300 m a.s.l. in areas which receive more than ca. 1400 mm annual rainfall within a wide temperature range (−5°C to 28°C) on every kind of soil (Moore at al., 1998; M’hirit and Benzyane, 2006). Current vegetation (Figure 2) in Jbel Khesana is dominated by deciduous oaks: Quercus pyrenaica in the highest elevations of the supramediterranean bioclimatic belt, and Quercus canariensis at lower altitude in the mesomediterranean one. At lower elevation, cork oaks (Quercus suber) give way to the thermomediterranean bioclimatic belt, where holm oaks (Quercus ilex) and olive trees (Olea europaea) are the main tree species. Today, Atlas cedar trees are not observed in Jbel Khesana but they still persist on nearby ranges at the highest altitudes. The vegetation by the mire is mainly composed of several genera belonging to Cyperaceae, like Carex, Cyperus or Eleocharis; and Poaceae, like Molinia caerulea or Danthonia decumbens; as well as other species like Juncus heterophyllus, Ranunculus spp., Isoetes sp. or Sphagnum sp. The surrounding forest consists of deciduous oaks (Quercus pyrenaica) and isolated stands of Quercus canariensis. Shrubs like Cytisus villosus, Erica arborea or Cistus salviifolius; herbs like Holcus lanatus, Rumex acetosella, Lamium flexuosum, Stellaria holostea and so on; and ferns like Pteridium aquilinum develop specially in the clearings.

Figure 2.

Schematic diagram showing the vegetation communities in the Jbel Khesana and the location of Bab el Karn II (white star).

A core of 180 cm depth was extracted with a Russian corer and then sub-sampled at 1 cm intervals for the multi-proxy analyses.

Chronology

Eight radiocarbon datings were performed on bulk peat samples at the Beta Analytic Radiocarbon Dating Laboratory (Table 1). Ages BP were calibrated with the IntCal13 ¹⁴C curve (Reimer et al., 2013). Confidence intervals of the calibrations and the age–depth model were calculated at 95% (2σ). An age–depth model was constructed for the fossil record (Figure 3) using calibrated ages and applying a smooth-spline solution with Clam 2.2 software (Blaauw, 2010).

Table 1.

Results of Radiocarbon (¹⁴C) dating of core samples, showing calibrated age ranges (2σ).

Depth (cm)	Lab code	¹⁴C age	Age (cal. yr BP)		Probability
Depth (cm)	Lab code	¹⁴C age	Min.	Max.	Probability
59	Beta-319029	1040 ± 30	918	989	89.1
			992	996	0.7
			1030	1049	5.1
100	Beta-319028	2000 ± 30	1881	2003	94.6
100	Beta-319028	2000 ± 30	2030	2032	0.4
127	Beta-319027	2160 ± 30	2057	2185	52.6
			2191	2207	2.2
			2229	2306	40.1
152	Beta-319022	2410 ± 30	2351	2496	81.2
			2596	2612	2.9
			2637	2684	10.9
159	Beta-319023	2470 ± 30	2380	2393	1.7
			2404	2411	0.9
			2426	2716	92.3
171	Beta-319026	3010 ± 30	3078	3094	4.2
			3107	3132	4.5
			3136	3257	74.3
			3290	3333	12
175	Beta-319024	3340 ± 30	3480	3538	23.1
			3544	3640	70.3
			3671	3678	1.4
180	Beta-319025	3590 ± 30	3834	3974	95

Figure 3.

Age–depth model and sedimentation rate development calculated for the BEK II mire.

Pollen palynomorphs, NPPs and charcoal analysis

Samples were treated following the chemical methodology proposed by Faegri and Iversen (1989). The sediment was then concentrated in Thoulet solution for densimetric separation of pollen and NPPs (Goeury and de Beaulieu, 1979). Pollen concentration was estimated by adding a Lycopodium tablet to each sample (Stockmarr, 1971). Data processing and graphic representation was performed with TILIA and TGView programs (Grimm, 1992, 2004). Local pollen assemblage zones were determined with a cluster analysis made by CONISS (Grimm, 1987). Different keys (Faegri and Iversen, 1989; Moore et al., 1991) and pollen atlases (Reille, 1999) were used to identify pollen grains as well as the reference collection of the Archaeobiology Laboratory of CSIC (Madrid). Pinus and Quercus suber pollen differentiation follows Carrión et al. (2000a) and Carrión et al. (2000b), respectively. Ferns, hydro-hygrophilous taxa and NPPs were excluded from the total pollen sum (>500 pollen grains). The identification of NPPs is based on Van Geel et al. (1981, 1989, 2003) and Van Geel and Aptroot (2006) and nomenclature follows Miola (2012).

Microcharcoal was counted in the same slides used for pollen (Finsinger and Tinner, 2005; Mooney and Tinner, 2011; Tinner and Hu, 2003) and sorted into two size classes, namely, >100 and <100 µm, indicating local and regional fires, respectively. Charcoal accumulation rate (CHAR) was calculated by dividing the concentration of microcharcoals by the sedimentation rate of each sample, obtained from the age–depth model (Long and Whitlock, 2002).

A synthetic diagram is included (Figure 4), in which some pollen and NPPs are grouped according to their ecological affinity (Table 2), with three charts showing the total pollen concentration and the CHAR on the right.

Figure 4.

Summary diagram including selected pollen types, NPPs, CHAR and Total Pollen Concentration.

Table 2.

Groups of pollen and non-pollen palynomorphs according to their ecological affinities.

Group	Pollen and NPPs
Riparian woods	Alnus, Fraxinus, Prunus, Salix, Sambucus, Ulmus, Hedera
Thermophilous shrubland	Arbutus, Buxus, Cistus, Cytisus, Daphne, Ephedra fragilis, Erica arborea, Helianthemum, Ilex, Lamiaceae, Myrtus, Phillyrea, Pistacia, Viburnum
Anthropogenic Perennial Pastures (APP)	Allium, Apiaceae, Artemisia, Brassicaceae, Campanula, Caryophyllaceae, Fabaceae, Liliaceae, Rosaceae, Saxifraga, Scrophulariaceae, Viola
Anthropogenic Nitrophilous Communities (ANC)	Anthemis, Asphodelus albus, Aster, Boraginaceae, Cardueae, Centaurea cyanus, Cichorioideae, Erodium, Malva sylvestris
Anthropozoogenic Nitrophilous Communities (AZNC)	Chenopodiaceae-Amaranthaceae, Plantago spp., Rumex spp.
Aquatic taxa	Myryophyllum, Ranunculaceae
Open water	Botryococcus, Debarya, Mougeotia, Spirogyra, Zygnema
Coprophilous fungi	Gelasinospora sp. (HdV 1), Gelasinospora cf. retispora (HdV 2), Chaetomium sp. (HdV 7A), Sordaria-type (HdV 55), Cercophora sp. (HdV 112), Sporormiella sp. (HdV 113), Coniochaeta cf. Ligniaria (HdV 172), Podospora-type (HdV 368)
Dry conditions	Pleospora sp. (HdV 3B), Trigocladium opacum (HdV 10)
Erosive phases	Glomus, Pseudoschizaea circula

Pollen dispersal and deposition at local/regional scale

The local presence of Atlas cedar forests has usually been established in a minimum percentage of 1% (Demarteau et al., 2007; Hajar et al., 2008; Nour El Bait et al., 2014; Wright, 1952). This threshold value would be likely higher in the Rif Mountains, as pollen production increases responding to a higher moisture (Khanduri and Sharma, 2002). Percentages <1% are considered by Magri and Parra (2002) as extraregional and associated with long-distance transport of pollen grains by winds. In this work, we assume that pollen percentages of Atlas cedar >1% indicate its presence in less than 200 m (Wright, 1952) around our site.

Sedimentology and geochemical analysis

The sedimentological study consisted of stratigraphical characterisation, description of the sedimentary facies and a high-resolution geochemical analysis. The core was split into two halves and imaged with a high-resolution digital camera in a core scanner.

Elemental composition of sediments was analysed on the fresh core using an Avaatech XRF core scanner at a resolution of 1 cm and under two different working conditions: (1) with an x-ray current of 800 µA, at 10 s count time and 10 kV x-ray voltage for the measurement of Al, Si, P, S, Cl, Ar, K, Ca, Ti, V, Rh, Cr, Mn and Fe; (2) with an x-ray current of 2000 µA, at 25 s count time, 30 kV x-ray voltage and using a Pd filter, for the measurement of Ni, Cu, Zn, Ga, Ge, As, Br, Rb, Sr, Y, Zr, Nb and Pb. The XRF results are expressed as counts per second (cps) and are semi-quantitative; only statistically significant chemical elements were considered. Geochemical data were analysed using multivariate statistics, in order to extract the geochemical signatures of the elemental composition of the sediments and the insights into underlying factors accounting for them. A principal component analysis (PCA; Hotelling, 1933) was applied to XRF data using SPSS software package, in correlation mode and by applying a varimax rotation. Prior to the analysis, the data were standardised (Z-scores) to avoid scaling effects and obtain average-centred distributions (Eriksson, 1999).

Results

Chronology

The age–depth model (Figure 3) is based on the data shown in Table 1 and was also used to calculate the sediment accumulation rate. It shows an overall increasing trend towards the top of the core (0.1–0.7 mm/yr), disrupted by a strong increase between ~2300–2000 cal. yr BP, reaching maximum values of 2.2 mm/yr ~2170 cal. yr BP.

Palynological record

Two main pollen zones were identified (Figure 4) which were divided into five subzones (Table 3), using a constrained cluster analysis (Grimm, 1987).

Table 3.

Description of pollen zones.

Subzone, depth (cm)/age (cal. yr BP)	Trees/shrubs	Herbs	NPPs	Pollen concentration/CHAR
LPAZ1a, 180–149 cm/~3900–2400 cal. yr BP	AP: 61–74% Quercus pyrenaica (30–39%), Quercus ilex (10%), Cedrus (15%), Riparian woods (<4%) Pinus spp. and Abies (<1%) By ~3000 cal. yr BP, retreat of oak forests (56.5–49.4%), maximum of Cedrus (22.3%) and spread of thermophilous shrubland (4.3%)	Poaceae (15–20%) APP (8%) ANC (0.2–2.8%) AZNC (0–1.8%) Cerealia (2–3%)	Coprophilous fungi (2–3%) Open Water indicators (9%) Aquatic taxa (20–24%)	Relatively high Pollen concentration (maximum ~3300 cal. yr BP) Medium levels of CHAR
LPAZ1b, 149–81 cm/~2400–1550 cal. yr BP	AP: 62–80% Initial phase with Cedrus (27–48%) dominant compared with Quercus pyrenaica/canariensis (27–21%), followed by alternating peaks between Quercus pyrenaica/canariensis and Cedrus.	Poaceae (13–22%) APP (6–3%) ANC (1–4%) AZNC (<1%) Cerealia (>6%) ~ 1755 cal. yr BP Vicia faba appears sporadically	Coprophilous fungi almost disappear Maxima of Aquatic taxa and Open Water indicators but general low levels up to the top	Relatively high Pollen concentration (maximum ~2110 cal. yr BP) Low levels of CHAR
LPAZ1c, 81–53 cm/~1550–800 cal. yr BP	AP: 53–70% Initial (16%) and final (19%) minima of Quercus pyrenaica/canariensis Alternating minima of Cedrus ~1300 and 1000 cal. yr BP Riparian woods (<3%) Wide increase of Thermophilous shrubland (4–11%)	Poaceae (29–12%) APP (4.5–5.8%) ANC (6.5–5%) AZNC (<1%) Cerealia (5.4–4.7%) Vicia faba, Linum and Cannabis appear sporadically	Coprophilous fungi in the last samples Maxima of Aquatic taxa/ Open Water indicators Maximum of Erosive Phases indicators Final occurrence of Dry Conditions indicators	Low levels of Pollen Concentration Maximum of CHAR ~1100 cal. yr BP
LPAZ2a, 53–37 cm/~800–500 cal. yr BP	AP: ca. 60–70% Quercus pyrenaica/canariensis (24–37%), Quercus ilex (8–11%) Minima of Cedrus (ca. 5–10%) Riparian woods (1–2%) Significant increase of Thermophilous shrubland (9–15%)	Poaceae (18–11%) APP (10–6.5%) ANC (4–6%) AZNC (1–2%) Cerealia (8–0.4%) Vicia faba and Cannabis appear sporadically	Coprophilous fungi in the top samples Significant values of Aquatic taxa and Open Water indicators Final peak of Dry conditions indicators	Higher levels of Pollen Concentration Initial minor peak and a final sharp one of CHAR
LPAZ2b, 37–0 cm/~500 cal. yr BP–present	AP >90% Sharp increase of Quercus pyrenaica/canariensis (38–65%). Cedrus almost disappears Riparian woods (<1%) Thermophilous shrubland (20–7%)	Poaceae (5–7%) APP (5–3%) ANC (ca. 4%) AZNC (ca. 2%) Cerealia, Vicia faba and Cannabis become exceptional	Maxima of Coprophilous fungi Open Water and Dry conditions indicators Low levels of Aquatic taxa	High levels of Pollen Concentration, Maxima of CHAR

CHAR: charcoal accumulation rate; ANC: Anthropogenic Nitrophilous Communities; AZNC: Anthropozoogenic Nitrophilous Communities; AP: arboreal pollen; APP: Anthropogenic Perennial Pastures.

Sedimentary and geochemical records

The sedimentary units (Table 4 and Figure 5) consist of organic argillaceous marl, mixed siliciclastic (Si, Al, K) and carbonate (Ca) compounds, within which the organic content decreases towards the top. Edaphised terrigenous muds, sometimes Fe-rich, with abundant root bioturbation traces, form the upper parts of these sedimentary intervals. In general, the arrangement of the sedimentary facies (SU1 to SU6) indicates a progressive shallowing tendency and more abundant desiccation intervals towards the top of the core (Table 4).

Table 4.

Description of sedimentary units.

Unit	Depth (cm)	Age	Description	Interpretation
SU6	10–0	~145 cal. yr BP to present	Moderate sedimentation rate : 0.7 mm/yr Two sedimentary facies: 1. 3–0 cm: Brownish grey and relatively less organic Fe-rich mud 2. 10–3 cm: Brownish grey organic mud that gradually becomes more organic Modern root bioturbations in the upper half	Shallow ephemeral mire. Presence of desiccation/weathering events at the top
SU5	46–10	~725–145 cal. yr BP	Increasing sedimentation rate: 0.5–0.7 mm/yr Two sedimentary facies: 1. 23–10 cm: More Fe-rich 2. 46–23 cm: Dark grey organic mud that gradually becomes less organic Fe-rich root bioturbations in the upper half	Shallow ephemeral mire. Presence of abundant desiccation/weathering intervals.
SU4	70–46	~1300–725 cal. yr BP	Quite constant sedimentation rate: 0.4–0.5 mm/yr Two sedimentary facies: 1. 57–46 cm: More Fe-rich 2. 70–57 cm: Dark grey organic mud that gradually becomes less organic Root bioturbations near the top	Relatively shallow perennial to ephemeral open water mire. Presence of desiccation events at the top, ca. 725 cal. yr BP
SU3	95–70	~1900–1300 cal. yr BP	Quite constant sedimentation rate: 0.4 mm/yr Two sedimentary facies: 1. 77–70 cm: More siliciclastic grey clays 2. 95–77 cm: Dark grey organic mud The sediments are slightly bioturbated near the top	Relatively shallow perennial open water pond. Presence of desiccation events ca. 1.9–1.7 cal. yr BP
SU2	113–95	~2170–1900 cal. yr BP	Low sedimentation rate: 0.4–0.5 mm/yr Two sedimentary facies: 1. 102–95 cm: Greenish siliciclastic clays 2. 113–102 cm: Organic mud Not bioturbated and/or edaphised sediments	Relatively shallow perennial open water pond. Absence of desiccation events
SU1	180–113	~3900–2170 cal. yr BP	Highest sediment accumulation rate: ~2.2 mm/yr Three sedimentary facies: 1. 120–110 cm: More siliciclastic mud 2. 145–120 cm: Organic muds 3. 180–145 cm: Organic-rich facies (plant fragments) Not bioturbated and/or edaphised sediments	Relatively deep and perennial open water pond. Absence of desiccation events

Figure 5.

Sedimentary sequence of BEK II mire including core image, lithostratigraphic units (SU), geochemical stratigraphy of the most significant chemical elements, accumulation rate and principal components derived from PCA analysis (PC).

The elemental composition variations are summarised in four principal components (Figures 5 and 6), which explain 82.8% of the total variance. PC1 explains 37% of the variance. Al, Rb, Zr, V, K, Ti, Sr, Y and Nb show high positive loadings (higher than 0.5); Si and Pb have a moderately positive loading; and Mn, Ca and Br have a low negative factor loading, respectively (Figure 6). The high positive and moderate loadings of the lithogenic elements are indicative of the amount of fine-grained mineral matter (clay minerals) in the core sediments. The mineral fraction content and the record of factor scores (Figure 5) indicate that the amount of clay minerals is quite homogeneous and increases gradually because of the decrease of the organic matter content towards the top of the sedimentary units. In SU3 and SU4, the upper clay-rich interval shows a relative decrease in PC1 value (fine mineral matter) that is correlated with a significant presence of Fe, probably forming iron (hydro) oxides (Figure 5). There are two intervals with relatively low mineral content: (1) the lower organic half of SU1 (180–145 cm) with abundant coarse vegetal fragments; (2) an interval that encompasses SU3 and SU4, likely related to an increase in the input of carbonate sediment to the mire that may reflect an enhanced soil erosion episode (see Ca/Ti ratio and PC4 curves in Figure 5).

Figure 6.

Loadings of the analysed elements on the principal components obtained by PCA analysis. High and moderate loadings are marked in bold.

PC2 explains 19.2% of the variance. Fe, P and Br show high positive loadings, and Pb and Mn show moderate positive loading (Figure 6). Rb has a low negative factor loading. Factor scores (Figure 5) are positive in SU5 and SU6 (40–0 cm). In SU4 (63–40 cm), they shift sharply to negative values and then gradually become less negative until the bottom of the core, where they remain around 0. Fe and Mn are redox-sensitive elements, while P and Br are related to the presence of vegetal organic matter (Leri and Myneni, 2012). The surface enrichment in Fe and Mn is probably related to vadose oxic conditions in the upper sections of the record, which involve more favourable conditions for oxidised, less mobile forms of these elements to form (Chesworth et al., 2006). Br presence is related to the processes controlling organic matter enzymatic halogenation (Biester et al., 2006; Leri and Myneni, 2012) and dehalogenation under reducing conditions within the soils (Van Pée and Unversuch, 2003).

PC3 explains 17.3% of the variance, and it reflects the inverse relationship between lithogenic Si and Al (moderate positive loading) and biophilic and organically bound elements such as S and Zn (high negative factor loadings) (Figure 6). The record of the scores can be divided into two sections: >140 and <6 cm, with positive scores (Figure 5).

PC4 explains 9.3% of the variance. Ca and Mn show high positive factor loadings (Figure 6). The origin of Ca is likely related to the presence of terrigenous carbonate sediments derived from the marly Cretaceous rocks that outcrop along Jbel Khesana slopes. The record of the scores can be divided into three sections: (1) >154 and 55–16 cm, with negative scores (except from one sample at 129 cm); (2) 154–75 cm, with scores around zero; (3) 75–55 and 15–0 cm, with positive scores (Figure 5).

Discussion

Model simulations suggest that the range of Atlas cedar in the Rif Mountains has decreased by about 75% between 1960 and 2010 (Cheddadi et al., 2017). During the coming decades, Rifan Cedrus atlantica forests are predicted to disappear because of the increasing aridity and temperatures, as they will not have enough space to migrate upwards in search of suitable climatic conditions (IRES, 2010; Thomas, 2013; UICN, 2015). For a better understanding of the long-term drivers which have led Atlas cedar populations to local extinction, we evaluate the main driving factors involved in the Atlas cedar forests’ sensitivity since the early Holocene. Ultimately, this palaeoecological study should lead to setting up management measures and conservation policies which could avoid their potential extinction in North Africa.

Background: Climate-induced expansion and decline of Atlas cedar forests from the early to the middle Holocene (~9000–4000 cal. yr BP)

Previous studies in the Rif Mountains (e.g. Cheddadi et al., 2015, 2016, 2017; Muller et al., 2014; Reille, 1977) spanning a broader period have shown the overall landscape development in the Rif Mountains (Morocco) since the early Holocene. The analysis performed in the same wetland under study (Cheddadi et al., 2016), encompassing the last 9000 years, shows for this first period humid and temperate conditions, nuanced by high seasonality (Brayshaw et al., 2011; Jalut et al., 2009; Magny et al., 2012). The forest landscape was dominated by semi deciduous oaks (Quercus canariensis, Quercus pyrenaica), with a significant presence of Cedrus, likely more widespread at higher altitudes, and Quercus ilex, probably more widespread at lower altitudes.

During the middle Holocene, ~7000 cal. yr BP, the climate became wetter and probably cooler, with a lower seasonality (Cheddadi et al., 1998; Magny et al., 2012; Vannière et al., 2011), thus promoting a marked development of Atlas cedar forests, which reached their maximum expansion ~6000 cal. yr BP. From this date onwards, the Atlas cedar forests underwent a steady decline, likely related to increasing aridity (Alba-Sánchez et al., 2015; Jalut et al., 2009; Magny et al., 2012; Tabel et al., 2016) which allowed their replacement by oaks, especially semi-deciduous ones, also favoured by local fires (Cheddadi et al., 2016). No evidence of human impact was found in this record, but minor evidence was reported in relatively nearby sites (Muller et al., 2014) at this time, which so far prevents connecting this shift to human activities in the area.

Therefore, at the end of the middle Holocene, ~4500/4000 cal. yr BP, forests were dominated by oaks, with a significant expansion of Atlas cedar forests, which would have migrated to higher altitudes in search for more humid climatic conditions.

Resilience of Rifan forests within a low human-disturbed landscape (~3900–2400 cal. yr BP)

The beginning of the late Holocene in the Jbel Khesana is characterised by a wide expansion of the forests, mainly composed of semi-deciduous oaks, with a noteworthy presence of Cedrus which may have composed an ecotone rather than a mixed Atlas cedar-oak forest.

Aquatic taxa show constant percentages throughout this phase (Figures 4 and 7), denoting some water availability during the summer season (Tabel et al., 2016). The low occurrences of green algae, considered as open water indicators, suggest temperatures too low for their growth (Cook et al., 2011). The negative values of both PC2, interpreted as an indicator of meteorisation in oxic conditions of the sediments because of the mire desiccation and subaerial exposure, and PC3 that reflects the relative input of lithogenic versus organic sediment in the mire, support our interpretation. Both principal components point to a relatively deep wetland within stable and wet climatic conditions. This is consistent with previous studies in the area describing a dry climate at lower altitudes, but more humid upslope, forcing Atlas cedar to migrate towards higher elevations (Cheddadi et al., 2016, 2017; Martín-Puertas et al., 2010).

Figure 7.

Main environmental changes over the last ~4000 years in the BEK II mire. AP: Arboreal Pollen; for the explanation of the pollen groups: Thermophilous shrubland, Coprophilous and Nitrophilous (Anthropogenic Nitrophilous Communities, ANC) go to Table 2; NAO Index from a climate proxy reconstruction from Greenland (Olsen et al., 2012); Dry versus humid climate conditions from the first eigenvector of the west Balearic-Algerian Basin 306G core (Nieto-Moreno et al., 2011); MCA and LIA from Martín-Puertas et al., 2010; Charcoal influx as the total input of microcharcoals.

Human activities are attested by the constant record of Cerealia which may be considered as locally grown (López-Sáez and López-Merino, 2005), and coprophilous fungi which indicate the presence of cattle as well as the development of grazing activities in the area. They could be likely linked to fire occurrences which affect mostly the shrub communities (Figure 7). Cereal cultivation has been detected in the fossil record since at least 7000 years ago in Northern Morocco (Ballouche and Marinval, 2003; López-Sáez et al., 2013; López-Sáez and López-Merino, 2008), but it seems to have been most developed in coastal areas and lowlands. Hence, human impact on Rifan forests seems to be quite slight in these early phases of the late Holocene (Cheddadi et al., 2016, 2015, 2017; Kaplan et al., 2009; Muller et al., 2014). In our record, the forest ecosystem remains almost unaltered, with the exception of a minor retreat of oak forests ~3000 cal. yr BP, along with a subsequent expansion of Atlas cedars, grasslands and thermophilous shrubland (Figures 4 and 7). Such landscape change may be related to specific clearing events, as a result of a scant increase of the human use during the Phoenician occupation, of the high-altitude forests (Cheddadi et al., 2015, 2016, 2017; Reille, 1977; Taiqui and Cantarino, 1997). From this date (~3000 cal. yr BP) onwards, accumulation rates show an increasing trend (0.15–0.71 mm/yr) revealing an increasing erosion process derived from these activities, specifically fire and grazing, after a period of greater stability.

Alternate dominance of oaks and Atlas cedars in Rifan forests within an increasingly disturbed landscape (~2400–1550 cal. yr BP)

Atlas cedar forests reach their maximum extent ~2400 cal. yr BP, likely favoured by cold temperatures and the weakening of human impact on the area, as indicated by the disappearance of coprophilous fungi and the lower percentages of Anthropozoogenic Nitrophilous Communities (AZNC) and Cerealia (Figures 4 and 7). In spite of this, the synchronous retreat of oak forests could have also been triggered by thinning or felling, allowing a further development of grasslands. Such forest opening allowed the pollen transport from the cedar forests located at higher altitudes (Millerón et al., 2012). The medium levels of CHAR along with the increasing sedimentation rate (0.8–1.3 mm/yr), and particularly the high values of PC1, related to fine terrigenous sediment input because of enhanced soil erosion, also point to such a forest clearance (Figure 7).

An important shift occurs ~2100 cal. yr BP, when moisture seems to increase apparently throughout the western Mediterranean region (Cheddadi et al., 1998; Damnati et al., 2016; Martín-Puertas et al., 2010), expressed by a series of peaks of oak forests and aquatic taxa, alternating with those of Atlas cedar and Poaceae, which extend until ~1550 cal. yr BP without a significant modification of the ratio arboreal pollen/non-arboreal pollen (AP/NAP) (Figure 4). Thus, deciduous oaks (Quercus pyrenaica and/or Quercus canariensis) seem to be closely related to water availability (Cheddadi et al., 2016), revealed by the coeval high values of aquatic taxa. As already commented, oak forest could decline during drier phases, reducing the canopy cover which would favour grasses and Atlas cedar pollen grains to reach the mire. Cedrus atlantica, however, appears to be more sensitive to temperatures in these mountains, since rainfall usually exceeds 700 mm (Castelló et al., 2016; Charco, 1999; Lenoir et al., 2008), especially at higher altitudes. The peaks reached by cedar get steadily lower during this period (Figures 4 and 7), because of a progressive upward shift to outcompete oak forests. PC4 and the terrigenous fraction suggest that the water table has fluctuated with several shallow phases, which gradually impacted the organic content towards the top (Figures 5 and 7). Although wet conditions are prevailing, a number of periods with shallowing mire (SU2 and SU3) spanning ~200 years can be observed. The very high sedimentation rate between ~2300 and 2000 cal. yr BP, reaching maximum values of 2.2 mm/yr at ~2170 cal. yr BP, suggests the nearly filling up of the mire accommodation space which became subjected to more frequent desiccation periods (Figure 7). This process has been enhanced by the increase of the carbonate constituent of the sediment, which derives from the weathering of the upstream tertiary marls, likely as a result of the erosion of the original soils triggered by agricultural activities (see Ca/Ti ratio and PC4 in Figures 5 and 7). This is consistent with the intensification of cereal cultivation which corresponds to the onset of the Roman Period in northern Morocco (Cheddadi et al., 2015; Taiqui and Cantarino, 1997). Mountain forests seem to remain fairly unconnected with this increasing human impact, likely because of the long distance of the site from the main inhabited areas. Cork oak forests, by contrast, show a quite significant retreat, as a probable result of initially drier climatic conditions and the increasing regional human impact, related both to cereal cultivation and the recurrent fires, pointed out by CHAR (Figures 4 and 7). This landscape feature persisted without major changes until the decline of the Roman Empire ~1550 cal. yr BP.

The fall of Atlas cedar forests in Jbel Khesana (~1550–800 cal. yr BP)

This period (~1550–800 cal. yr BP) begins with a remarkable wide deforestation event, which led to a major expansion of grasslands, likely related to a drier climate, but also to an increase of temperature, as suggested by open water indicators (Cook et al., 2011; Quamar, 2015; Van Geel, 2001). They also point to an increasing seasonality against the more continuous water availability indicated by aquatic taxa (Amami et al., 2010; Oertli et al., 2010). This corresponds to the onset of the Medieval Climate Anomaly (~1400–700 cal. yr BP; Martín-Puertas et al., 2010), a period characterised by drier and warmer climatic conditions (Esper at al., 2007; Nieto-Moreno et al., 2015, 2011). The strong deforestation of Atlas cedar forests coincides also with a political instability between the end of the Byzantine period and the arrival of the Arabs (AD ~650–900/~1300–1050 cal. yr BP), leading to the establishment of the Idrisid dynasty during the 8th century, when an increasing consumption of forest resources triggered a major break in the regional landscape history (Cheddadi et al., 2015; Taiqui and Cantarino, 1997). Furthermore, after ~1300 cal. yr BP, PC1 and PC2 indicate an increase of the relative fine terrigenous input against organic matter (Figure 7), which can be interpreted as the start of a new shallowing sequence, as a result of drier conditions.

Also striking is the disconnection between the development of oak and Atlas cedar forests (Figure 4), pointing to the end of the competition between the two species with a further upward shift of the latter. The sharp drop of oak occurrences is no longer counteracted by Atlas cedars but rather by a wide spread of thermophilous shrubland and grasslands which is consistent with an increasing aridity trend, as well as a recovery of cereal cultivation, which had declined at the end of the previous period. It could also be linked with a slightly higher occurrence of fires, shown by CHAR values, either as a consequence of a longer dry season, or as a result of clearings to open new lands for cultivation. In this sense, Atlas cedars do not seem to have a high flammability, especially when they are old (Slimani et al., 2014). Thus, the retreat of oak forests could instead have been induced by a drier climate and an intensification of agricultural activities, while Atlas cedar forests could have been exploited for timber.

Later, ~1150 cal. yr BP, a marked peak of CHAR appears to have caused a deeper impact on grasslands and thermophilous shrublands than on mountain forests (Figures 4 and 7). Furthermore, fires could also have enhanced the erosion of the catchment area, as indicated by Glomus and Pseudoschizaea circula. Enhanced erosion is also marked by positive scores in PC4 (Figure 7), likely related to the reduction of the forest cover as well as to the development of agricultural activities.

A new wet phase (~1030 cal. yr BP), indicated by a high occurrence of aquatic taxa, allowed a recovery of the forest ecosystem where both Atlas cedar and oaks almost reached their previous occurrences. However, a new regression of the Atlas cedar forests took place ~980 cal. yr BP, likely more related to human disturbances. In this sense, timber trade between Al-Andalus and the Maghreb is well dated since the 10th century AD (Rodríguez Trobajo, 2008). Then, Atlas cedars recovered again with extended populations around the studied site. At the end of this period, the significantly high proportion of shrubs (>10%) suggests a ‘matorralization’ process (Barbero et al., 1990) for the last millennium in Jbel Khesana.

Cereal cultivation remains a strong human activity during this period, except ~1080 cal. yr BP, at the end of the Idrisid dynasty (AD 790–930). Grazing activities are attested only in the top samples, from 980 cal. yr BP onwards, with very low values of coprophilous fungi.

Final extinction event of Atlas cedar forests in Jbel Khesana (~800 cal. yr BP-present)

Atlas cedar forests undergo a sharp decrease likely related to new loggings. The initial event may coincide with the political rule of the Almohad dynasty (AD ~1150–1250/~800–700 cal. yr BP), which built a great number of mosques, presumably with Rifan Cedrus atlantica wood (Park and Boum, 2005). The use of Atlas cedar timber for buildings in Al-Andalus is also reported by various sources (Rodríguez Trobajo, 2008). Lower water levels and more frequent desiccation intervals are indicated by the more pervasive vadose oxic meteorisation (PC2) of the mire sediments, the low values of aquatic taxa and the higher occurrence of Cyperaceae. In spite of these drier conditions, cropping activities reach their maximum intensity within the record. Both CHAR and coprophilous fungi show significant values, indicating a higher incidence of fires and the intensification of grazing activities. The latter reach their maxima after the final abandonment of cereal cultivation in the area, ~550 cal. yr BP. This significant change in land-use, coeval with a noteworthy climate event, allows us to carry out a detailed diachronic analysis of this final subzone, despite the lack of a radiocarbon dating. Climatic conditions probably became cooler with the onset of the ‘Little Ice Age’ (Esper et al., 2007; Martín-Puertas et al., 2010; Nieto-Moreno et al., 2015). The final minimum (~600 cal. yr BP) in the first eigenvector of the west Balearic-Algerian Basin 306G core, indicating dry climatic conditions (Nieto-Moreno et al., 2011), as well as the shift towards negative values (~500 cal. yr BP) of the NAO Index (Olsen et al., 2012), allows us to set the abandonment of cereal cropping at this date (Figure 7). The slight expansion of aquatic taxa matches this event, although a more intense seasonality seems to have been established, as indicated by the increasing values of Dry conditions indicators (Figure 4) and the positive ones of PC2, related to a greater mire exposure (Figure 7). The LIA is considered to have been drier in Morocco, especially during its first half, than in Europe (Nieto-Moreno et al., 2011; Touchan et al., 2011). Some authors (Taiqui and Cantarino, 1997) also describe a period of social crisis between 15th and 16th centuries which would have led to the prevalence of transhumance and nomadism against cropping activities. In any case, this event would have favoured the recovery of all arboreal taxa, especially oaks, while Cedrus populations retreated. Nevertheless, it triples its values from ~665–375 cal. yr BP, after a significant retreat ~510 cal. yr BP likely related to the construction of new settlements like the town of Chefchaouen (AD 1471/479 cal. yr BP). Its percentages (5–15%) indicate that the species was present or growing a few hundred metres from the site likely further upwards from the core site. Previous lower resolution studies (Cheddadi et al., 2016, 2017; Muller et al., 2014; Reille, 1977) depict the virtual demise of Cedrus from most pollen records during medieval times and even before. The present high-resolution analysis shows, in contrast, a significant resilience of Atlas cedar forests in the area. The successive loggings, carried out ~340, 280, 220 and 105 cal. yr BP, were always counteracted by subsequent recoveries up to 12%, reporting the local persistence of Atlas cedar forests close to the site. Meanwhile, the expansion of oak forests was favoured by the retreat of the thermophilous shrubland, as a result of a higher occurrence of fires, with maxima of CHAR. Fires would have also favoured the spread of grasslands in order to supply the increasing livestock density (Figure 7). These traditional agrosilvopastoral uses shaped a dehesa-like landscape which has persisted until very recent times (Ballouche, 2013; Cheddadi et al., 2015).

A major landscape change took place with the onset of the Colonial Period (AD 1912–1956) and the subsequent independence of Morocco from France and Spain. First of all, Atlas cedar forests disappeared completely ~1950 cal. yr AD from the Jbel Khesana, most probably as a consequence of loggings carried out by the Spanish government (Ajbilou et al., 2003; Araque and Garrido, 2015). A further regeneration could have been ultimately unachievable because of the high incidence of the subsequent human activities in the area, including logging, slash-and-burn farming (‘sbir’) and overgrazing (Coudel et al., 2016; Rejdali, 2004; Taiqui, 2005), which triggered a strong soil erosion (Moore et al., 1998), along with increasing temperatures and aridity. This enhanced soil erosion is also pointed out by the high levels of PC1 and PC4, related to the increase of sediments coming from the slope bedrock (marls) weathering (Figure 7).

Thus, the local extinction of Atlas cedar forests in Jbel Khesana seems to have been mainly induced by an increasing human pressure, particularly during the 20th century, rather than by a climatic constraint, although recent warming and increasing aridity have increased their vulnerability. Both current temperature and rainfall amount still remain appropriate for the development of these forests in the area (Ajbilou et al., 2003; M’hirit and Benzyane, 2006). In this sense, Rifan Cedrus atlantica forests show a distinct pattern from the rest of their distribution range in the north of Africa (Allen et al., 2010; Castelló et al., 2016; Cheddadi et al., 2017; Linares et al., 2013; Slimani et al., 2014; UICN, 2015).

Conclusion

In the present study, we show that high-resolution multi-proxy analyses are essential for evaluating local extinctions of endangered forest species.

Over the last 4000 years, with a reduced human influence, the climate seasonality and increasing aridity were the main drivers for the upslope migration of the Atlas cedar towards more humid conditions. Then, the increased aridity and human pressure over time favoured the spread of a thermophilous shrubland and the depletion of Atlas cedar forests, in a process that lasted until ~800 cal. yr BP. Sequential deforestation events in relation with timber exploitation took place over recent centuries with subsequent recoveries of the forest, until the final local extinction of the species in the middle of the 20th century.

In summary, human pressure, either directly, by intensive loggings, or indirectly, strongly limits the species regeneration through overgrazing and the use of fire. Our data show that the increasing human pressure has been the main driver of the extinction of Atlas cedars in the studied area. Our findings have significant implications for current and future management and conservation of Atlas cedars in the Rif Mountains, beyond the unpromising forecasts linked to climate change. Conservation measures in Atlas cedar forests growing at altitudes about 2000 m a.s.l. could ensure the future of this endangered species (IRES, 2010). There should be a high priority for preserving their regeneration as well as for limiting the intensive loggings. Cropping activities, especially related to Cannabis cultivation, have become a major threat to Atlas cedar populations as main drivers of deforestation and soil erosion. More sustainable activities are required. In this sense, low-intensity levels of livestock grazing are especially advised to avoid an excessive fuel accumulation which could lead to widespread wildfires.

Footnotes

Acknowledgements

The authors thank the two anonymous reviewers for improving the final version with their constructive comments.

Funding

This work was funded by the Relic-Flora project RNM-7033 (Excellence Research Projects Program from the Andalusian Government).

ORCID iDs

Daniel Abel-Schaad

Eneko Iriarte

Francisca Alba-Sánchez

References

Ajbilou

Marañón

Arroyo

(2003) Distribución de clases diamétricas y conservación de bosques en el norte de Marruecos. Investigación Agraria, Sistemas de Recursos Forestales 12: 111–123.

Alba-Sánchez

López-Sáez

Nieto-Lugilde

et al . (2015) Long-term climate forcings to assess vulnerability in North Africa dry argan woodlands. Applied Vegetation Science 18(2): 283–296.

Allen

Macalady

Chenchouni

et al . (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management 259(4): 660–684.

Amami

Muller

Rhazi

et al . (2010) Modern pollen-vegetation relationships within a small Mediterranean temporary pool (western Morocco). Review of Palaeobotany and Palynology 162(2): 213–225.

Araque

Garrido

(2015) El Plan forestal del Protectorado español del norte de Marruecos (1944). Revista de Estudios Regionales 102: 251–300.

Ballouche

(2013) Contribution à l’histoire récente de la végétation du Bas-Loukkos (province de Larache, Maroc). Physio-Géo. Géographie Physique et Environnement 7: 67–82.

Ballouche

Marinval

(2003) Données palynologiques et carpologiques sur la domestication des plantes et l’agriculture dans le Néolithique ancien du Maroc septentrional (site de Kaf Taht El-Ghar). Revue d’Archéométrie 27(1): 49–54.

Barbero

Bonin

Loisel

et al . (1990) Changes and disturbances of forest ecosystems caused by human activities in the western part of the Mediterranean basin. Vegetatio 87(2): 151–173.

Benabid

(1985) Les écosystemes forestiers, préforestiers et présteppiques du Maroc. Forêt méditerranéenne 7(1): 53–64.

10.

Biester

Cortizas

Keppler

(2006) Occurrence and fate of halogens in mires. Developments in Earth Surface Processes 9: 449–464.

11.

Bindler

(2006) Mired in the past-looking to the future: Geochemistry of peat and the analysis of past environmental changes. Global and Planetary Change 53(4): 209–221.

12.

Blaauw

(2010) Methods and code for ‘classical’ age-modelling of radiocarbon sequences. Quaternary Geochronology 5(5): 512–518.

13.

Brayshaw

Rambeau

Smith

(2011) Changes in Mediterranean climate during the Holocene: Insights from global and regional climate modelling. The Holocene 21(1): 15–31.

14.

Carrión

Fernández

González-Sampériz

et al . (2010) Expected trends and surprises in the Lateglacial and Holocene vegetation history of the Iberian Peninsula and Balearic Islands. Review of Palaeobotany and Palynology 162: 458–475.

15.

Carrión

Navarro

et al . (2000a) The distribution of cluster pine (Pinus pinaster) in Spain as derived from palaeoecological data, relationships with phytosociological classification. The Holocene 10: 243–252.

16.

Carrión

Parra

Navarro

et al . (2000b) Past distribution and ecology of the cork oak (Quercus suber) in the Iberian Peninsula: A pollen-analytical approach. Diversity and Distributions 6: 29–44.

17.

Castelló

Cale

D’Angelo

et al . (2016) Baseline mortality analysis reveals legacy of contrasting land use practices on the structural sustainability of endangered Moroccan and Spanish Mountain Forests. Forests 7(8): 172.

18.

Chambers

Booth

De Vleeschouwer

et al . (2012) Development and refinement of proxy-climate indicators from peats. Quaternary International 268: 21–33.

19.

Charco

(1999) El Bosque Mediterráneo en el Norte de África: Biodiversidad y lucha contra la desertización. Madrid: Agencia Española de Cooperación.

20.

Cheddadi

Bouaissa

Rhoujjati

et al . (2016) Environmental changes in the Moroccan western RIF mountains over the last 9,000 years. Quaternaire 27(1): 15–25.

21.

Cheddadi

Fady

François

et al . (2009) Putative glacial refugia of Cedrus atlantica deduced from Quaternary pollen records and modern genetic diversity. Journal of Biogeography 36: 1361–1371.

22.

Cheddadi

Henrot

François

et al . (2017) Microrefugia, Climate Change, and Conservation of Cedrus atlantica in the Rif Mountains, Morocco. Frontiers in Ecology and Evolution 5: 114.

23.

Cheddadi

Lamb

Guiot

et al . (1998) Holocene climatic change in Morocco: A quantitative reconstruction from pollen data. Climate Dynamics 14(12): 883–890.

24.

Cheddadi

Nour

Bait

Bouaissa

et al . (2015) A history of human impact on Moroccan mountain landscapes. African Archaeological Review 3(2): 233–248.

25.

Chesworth

Martínez Cortizas

García-Rodeja

(2006) Redox-pH approach to the geochemistry of the Earth’s land surface, with application to peatlands. In: Martini

Martínez Cortizas

Chesworth

(eds) Peatlands: Evolution and Records of Environmental and Climate Changes. Amsterdam: Elsevier, pp. 175–196.

26.

Cook

van Geel

van der Kaars

et al . (2011) A review of the use of non-pollen palynomorphs in palaeoecology with examples from Australia. Palynology 35(2): 155–178.

27.

Coudel

Aubert

Aderghal

et al . (2016) Pastoral and woodcutting activities drive Cedrus atlantica Mediterranean forest structure in the Moroccan Middle Atlas. Ecological Applications 26(2): 574–586.

28.

Cugny

Mazier

Galop

(2010) Modern and fossil non-pollen palynomorphs from the Basque mountains (western Pyrenees, France): The use of coprophilous fungi to reconstruct pastoral activity. Vegetation History and Archaeobotany 19(5–6): 391–408.

29.

Damnati

Etebaai

Benjilani

et al . (2016) Sedimentology and geochemistry of lacustrine terraces of three Middle Atlas lakes: Paleohydrological changes for the last 2300 cal BP in Morocco (western Mediterranean region). Quaternary International 404: 163–173.

30.

Demarteau

François

Cheddadi

et al . (2007) Réponses de Cedrus atlantica aux changements climatiques passés et futurs. Geo-Eco-Trop: Revue Internationale de Géologie, de Géographie et d’Ecologie Tropicales 31: 105–146.

31.

Dullinger

Gattringer

Thuiller

et al . (2012) Extinction debt of high-mountain plants under twenty-first-century climate change. Nature Climate Change 2(8): 619–622.

32.

El-Baha

Omran

Regato

et al . (2010) Forests and Climate Change in the Near East Region. Rome: FAO.

33.

Eriksson

(1999) Introduction to Multi-and Megavariate Data Analysis using Projection Methods (PCA and PLS). Umeå: Umetrics AB.

34.

Esper

Frank

Büntgen

et al . (2007) Long-term drought severity variations in Morocco. Geophysical Research Letters 34: 17.

35.

Faegri

Iversen

(1989) Textbook of Pollen Analysis. Chichester: John Wiley and Sons.

36.

Finsinger

Tinner

(2005) Minimum count sums for charcoal concentration estimates in pollen slides: Accuracy and potential errors. The Holocene 15(2): 293–297.

37.

Goeury

de Beaulieu

(1979) A propos de la concentration du pollen à l’aide de la liqueur de Thoulet dans le sédiments minéraux. Pollen et Spores 21: 239–251.

38.

Grimm

(1987) CONISS: A FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers and Geosciences 13: 13–35.

39.

Grimm

(1992) Tilia. Version 2.0. Springfield, IL: Illinois State Museum.

40.

Grimm

(2004) TGView. Springfield, IL: Illinois State Museum.

41.

Grovel

(1996) La préservation des forêts du Rif centro-occidental: un enjeu de développement de la montagne rifaine/Preservation of forests in the central western Rif massif: A key element in the development of the Rif Mountains. Revue de Géographie Alpine 84(4): 75–94.

42.

Hajar

Khate

Cheddadi

(2008) Vegetation changes during the late Pleistocene and Holocene in Lebanon: A pollen record from the Bekaa Valley. The Holocene 18(7): 1089–1099.

43.

Hamelin

Nwankwo

(2013) Managing the environment, people and herds: Sustainability of the Moroccan cedar forest. World Journal of Science, Technology and Sustainable Development 10(4): 260–277.

44.

Heiri

Bugmann

Tinner

et al . (2006) A model-based reconstruction of Holocene treeline dynamics in the Central Swiss Alps. Journal of Ecology 94(1): 206–216.

45.

Hotelling

(1933) Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology 24(6): 417–441.

46.

Iglesias

Yospin

Whitlock

(2015) Reconstruction of fire regimes through integrated paleoecological proxy data and ecological modeling. Frontiers in Plant Science 5: 785.

47.

IRES (2010) Les écosystèmes forestiers face au changement climatique: situation et perspectives d’adaptation au Maroc. Royaume du Maroc, Rabat: IRES.

48.

Jalut

Dedoubat

Fontugne

et al . (2009) Holocene circum-Mediterranean vegetation changes: Climate forcing and human impact. Quaternary International 200(1): 4–18.

49.

Kaplan

Krumhardt

Zimmermann

(2009) The prehistoric and preindustrial deforestation of Europe. Quaternary Science Reviews 28(27): 3016–3034.

50.

Khanduri

Sharma

(2002) Pollen production, microsporangium dehiscence and pollen flow in Himalayan cedar (Cedrus deodara Roxb. ex D. Don). Annals of Botany 89(5): 587–593.

51.

Lenoir

Gégout

Marquet

et al . (2008) A significant upward shift in plant species optimum elevation during the 20th century. Science 320(5884): 1768–1771.

52.

Leri

Myneni

SCB

(2012) Natural organobromine in terrestrial ecosystems. Geochimica et Cosmochimica Acta 77: 1–10.

53.

Linares

Taïqui

Sangüesa-Barreda

et al . (2013) Age-related drought sensitivity of Atlas cedar (Cedrus atlantica) in the Moroccan Middle Atlas forests. Dendrochronologia 31(2): 88–96.

54.

Linares

Tíscar

Camarero

et al . (2011) Tree growth decline on relict Western-Mediterranean mountain forests: Causes and impacts. In: Jenkins

(ed.) Forest Decline: Causes and Impacts. Hauppauge: Nova Science Publishers, Inc.

55.

Long

Whitlock

(2002) Fire and vegetation history from the coastal rain forest of the western Oregon Coast Range. Quaternary Research 58(3): 215–225.

56.

López-Sáez

Abel

Bokbot

et al . (2013) Paisajes neolíticos del noroeste de Marruecos: Análisis arqueopalinológico de la Cueva de Boussaria. In: Goncalves

Dinz

Sousa

(eds) 5° Congresso do Neolítico Peninsular. Lisboa: Universidade de Lisboa, pp. 92–97.

57.

López-Sáez

López-Merino

(2005) Precisiones metodológicas acerca de los indicios paleopalinológicos de agricultura en la Prehistoria de la Península Ibérica. Portugalia 26: 53–64.

58.

López-Sáez

López-Merino

(2008) Antropización y neolitización durante el Holoceno en Marruecos: Una aproximación paleopalinológica. In: Hernández

Soler

López

(eds) Actas IV Congreso del Neolítico Peninsular. Tomo I. Alicante: Museo Arqueológico de Alicante, pp. 438–444.

59.

López-Vila

Montoya

Cañellas-Boltà

et al . (2014) Modern non-pollen palynomorphs sedimentation along an elevational gradient in the south-central Pyrenees (southwestern Europe) as a tool for Holocene paleoecological reconstruction. The Holocene 24(3): 327–345.

60.

Magny

Peyron

Sadori

et al . (2012) Contrasting patterns of precipitation seasonality during the Holocene in the south- and north-central Mediterranean. Journal of Quaternary Science 27(3): 290–296.

61.

Magri

(2012) Quaternary history of Cedrus in Southern Europe. Annali di Botanica 2: 57–66.

62.

Magri

Di Rita

Aranbarri

et al . (2017) Quaternary disappearance of tree taxa from Southern Europe: Timing and trends. Quaternary Science Reviews 163: 23–55.

63.

Magri

Parra

(2002) Late Quaternary western Mediterranean pollen records and African winds. Earth and Planetary Science Letters 200: 401–408.

64.

Martín-Puertas

Jiménez-Espejo

Martínez-Ruiz

et al . (2010) Late-Holocene climate variability in the southwestern Mediterranean region: An integrated marine and terrestrial geochemical approach. Climate of the Past 6(6): 807–816.

65.

M’hirit

Benzyane

(eds) (2006) Le Cèdre de l’Atlas: Mémoire du temps. Casablanca: Editions Mardaga.

66.

Millerón

De Heredia

Lorenzo

et al . (2012) Effect of canopy closure on pollen dispersal in a wind-pollinated species (Fagus sylvatica L.). Plant Ecology 213(11): 1715–1728.

67.

Miola

(2012) Tools for Non-Pollen Palynomorphs (NPPs) analysis: A list of Quaternary NPP types and reference literature in English language (1972–2011). Review of Palaeobotany and Palynology 186: 142–161.

68.

Montoya

Rull

Vegas-Vilarrúbia

(2012) Non-pollen palynomorph studies in the Neotropics: The case of Venezuela. Review of Palaeobotany and Palynology 186: 102–130.

69.

Mooney

Tinner

(2011) The analysis of charcoal in peat and organic sediments. Mires and Peat 7(9): 1–18.

70.

Moore

Fox

Harrouni

et al . (1998) Environmental challenges in the Rif mountains, northern Morocco. Environmental Conservation 25(4): 354–365.

71.

Moore

Webb

Collinson

(1991) Pollen Analysis. Oxford: Blackwell.

72.

Muller

Rhazi

Andrieux

et al . (2014) Vegetation history of the western Rif Mountains (NW Morocco): Origin, late-Holocene dynamics and human impact. Vegetation History and Archaeobotany 24: 487–501.

73.

Nieto-Moreno

Martínez-Ruiz

Gallego-Torres

et al . (2015) Palaeoclimate and palaeoceanographic conditions in the westernmost Mediterranean over the last millennium: An integrated organic and inorganic approach. Journal of the Geological Society 172(2): 264–271.

74.

Nieto-Moreno

Martínez-Ruiz

Giralt

et al . (2011) Tracking climate variability in the western Mediterranean during the Late-Holocene: A multiproxy approach. Climate of the Past 7(4): 1395–1414.

75.

Nour El Bait

Rhoujjati

Eynaud

et al . (2014) An 18000-year pollen and sedimentary record from the cedar forests of the Middle Atlas, Morocco. Journal of Quaternary Science 29(5): 423–432.

76.

Oertli

Biggs

Céréghino

et al . (eds) (2010) Pond Conservation in Europe. Dordrecht: Springer.

77.

Olsen

Anderson

Knudsen

(2012) Variability of the North Atlantic Oscillation over the past 5,200 years. Nature Geoscience 5(11): 808–812.

78.

Park

Boum

(2005) Historical Dictionary of Morocco. Lanham, MD: Scarecrow Press.

79.

Petit

Hampe

Cheddadi

(2005) Climate changes and tree phylogeography in the Mediterranean. Taxon 54(4): 877–885.

80.

Postigo-Mijarra

Morla

Barrón

et al . (2010) Patterns of extinction and persistence of Arctotertiary flora in Iberia during the Quaternary. Review of Palaeobotany and Palynology 162: 416–426.

81.

Quamar

(2015) Non-pollen palynomorphs from the late Quaternary sediments of southwestern Madhya Pradesh (India) and their palaeoenvironmental implications. Historical Biology 27(8): 1070–1078.

82.

Reille

(1977) Contribution pollen analytique a l’histoire Holocene de la vegetation des montagnes du Rif (Maroc septentrional). Bulletin de l’Association Francaise pour l’Etude du Quaternaire 50: 53–76.

83.

Reille

(1999) Pollen et Spores d’Europe et d’Afrique du Nord. 2nd Edition. Marseille: Laboratoire de Botanique Historique et Palynologie.

84.

Reimer

Bard

Bayliss

et al . (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55(4): 1869–1887.

85.

Rejdali

(2004) Forest cover changes in the Maghreb countries with special reference to Morocco. In: Mazzoleni

Di Pasquale

Mulligan

et al . (eds) Recent Dynamics of the Mediterranean Vegetation and Landscape. Chichester: John Wiley and Sons, pp. 21–31.

86.

Rodríguez Trobajo

(2008) Procedencia y uso de madera de pino silvestre y pino laricio en edificios históricos de Castilla y Andalucía. Arqueología de la Arquitectura 5: 33–53.

87.

Rubiales

Morales-Molino

Álvarez

et al . (2012) Negative responses of highland pines to anthropogenic activities in inland Spain: A palaeoecological perspective. Vegetation History and Archaeobotany 21(4–5): 397–412.

88.

Service Géologique du Maroc (1980) Carte Géologique de la Chaîne Rifaine. Echelle 1:500.000. Royaume du Maroc, Rabat: Ministère de L’Énergie et des Mines.

89.

Silva-Sánchez

Martínez-Cortizas

Abel-Schaad

et al . (2016) Influence of climate change and human activities on the organic and inorganic composition of peat during the ‘Little Ice Age’ (El Payo mire, W Spain). The Holocene 26(8): 1290–1303.

90.

Silva-Sánchez

Martínez-Cortizas

López-Merino

(2014) Linking forest cover, soil erosion and mire hydrology to late-Holocene human activity and climate in NW Spain. The Holocene 24(6): 714–725.

91.

Slimani

Touchan

Derridj

et al . (2014) Fire history of Atlas cedar (Cedrus atlantica Manetti) in Mount Chélia, northern Algeria. Journal of Arid Environments 104: 116–123.

92.

Stockmarr

(1971) Tablets with spores used in absolute pollen analysis. Pollen et Spores 13: 614–621.

93.

Tabel

Khater

Rhoujjati

et al . (2016) Environmental changes over the past 25000 years in the southern Middle Atlas, Morocco. Journal of Quaternary Science 31: 93–102.

94.

Taiqui

(2005) Évolution récente de la structure du paysage du basin de Chefchaouen (1958-1986) In: Groupe de Recherches Geographiques sur le Rif: Mutations des milieux ruraux dans les montagnes rifaines (Maroc). Tétouan: Faculté des Lettres & Sciences Humaines de Tétouan.

95.

Taiqui

Cantarino

(1997) Eléments historiques d’analyse écologique des paysages montagneux du Rif Occidental (Maroc). Mediterranea. Serie de Estudios Biológicos 16: 23–35.

96.

Terrab

Hampe

Lepais

et al . (2008) Phylogeography of North African Atlas cedar (Cedrus atlantica, Pinaceae): Combined molecular and fossil data reveal a complex Quaternary history. American Journal of Botany 95(10): 1262–1269.

97.

Thomas

(2013) Cedrus atlantica. The IUCN Red List of Threatened Species 2013. Available at: https://dx-doi-org.web.bisu.edu.cn/10.2305/IUCN.UK.2013-1.RLTS.T42303A2970716.en (accessed 16 February 2017).

98.

Tinner

Ammann

(2005) Long-term responses of mountain ecosystems to environmental changes: Resilience, adjustment, and vulnerability. In: Huber

Bugmann

Reasoner

(eds) Global Change and Mountain Regions. Dordrecht: Springer Science and Business Media, pp. 133–143.

99.

Tinner

(2003) Size parameters, size-class distribution and area-number relationship of microscopic charcoal: Relevance for fire reconstruction. The Holocene 13(4): 499–505.

100.

Tinner

Theurillat

(2003) Uppermost limit, extent, and fluctuations of the timberline and treeline ecocline in the Swiss Central Alps during the past 11,500 years. Arctic, Antarctic, and Alpine Research 35(2): 158–169.

101.

Touchan

Anchukaiti

Meko

et al . (2011) Spatiotemporal drought variability in northwestern Africa over the last nine centuries. Climate Dynamics 37(1–2): 237–252.

102.

UICN (2015) Libro rojo de los ecosistemas forestales de Marruecos. Gland y Málaga, España: UICN.

103.

Van Geel

(2001) Non-pollen palynomorphs. In: Smol

Last

Birks

(eds) Tracking Environmental Change Using Lake Sediments, vol. 3. Dordrecht: Springer Netherlands, pp. 99–119.

104.

Van Geel

Aptroot

(2006) Fossil ascomycetes in Quaternary deposits. Nova Hedwigia 82(3): 313–329.

105.

Van Geel

Bohncke

SJP

Dee

(1981) A palaeoecological study of an upper late glacial and Holocene sequence from ‘De Borchert’, the Netherlands. Review of Palaeobotany and Palynology 31: 367–448.

106.

Van Geel

Buurman

Brinkkemper

et al . (2003) Environmental reconstruction of a Roman Period settlement site in Uitgeest (the Netherlands), with special reference to coprophilous fungi. Journal of Archaeological Science 30(7): 873–883.

107.

Van Geel

Coope

GRR

van Der Hammen

(1989) Palaeoecology and stratigraphy of the late glacial type section at Usselo (the Netherlands). Review of Palaeobotany and Palynology 60(1–2): 25–129.

108.

Van Pée

Unversuch

(2003) Biological dehalogenation and halogenation reactions. Chemosphere 52: 299–312.

109.

Vannière

Power

Roberts

et al . (2011) Circum-Mediterranean fire activity and climate changes during the mid-Holocene environmental transition (8500-2500 cal. BP). The Holocene 21(1): 53–73.

110.

Vegas-Vilarrúbia

Rull

Montoya

et al . (2011) Quaternary palaeoecology and nature conservation: A general review with examples from the neotropics. Quaternary Science Reviews 30(19): 2361–2388.

111.

Walther

Beißner

Burga

(2005) Trends in the upward shift of alpine plants. Journal of Vegetation Science 16(5): 541–548.

112.

Wright

(1952) Pollen Dispersion of Some Forest Trees. Station Paper NE-46. Upper Darby, PA: U.S. Department of Agriculture, Forest Service, Northeast Forest Experiment Station.